Metabolic Syndrome

THE INFLUENCE OF WORK PATTERNS ON LIFESTYLE BEHAVIOURS AND CARDIOVASCULAR RISK IN FEMALE HOSPITAL WORKERS.

Megan Kirk

A thesis submitted to the School of Nursing

In conformity with the requirements for

the degree of Master’s of Science.

Queen’s University

Kingston, Ontario, Canada

(September, 2009)

Library and Archives Bibliothèque et Canada Archives Canada

Published Heritage Direction du Branch Patrimoine de l’édition

395 Wellington Street 395, rue Wellington Ottawa ON K1A 0N4 Ottawa ON K1A 0N4 Canada Canada

Your file Votre référence ISBN: 978-0-494-65207-7 Our file Notre référence ISBN: 978-0-494-65207-7

NOTICE: AVIS:

The author has granted a non- L’auteur a accordé une licence non exclusive exclusive license allowing Library and permettant à la Bibliothèque et Archives Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par télécommunication ou par l’Internet, prêter, telecommunication or on the Internet, distribuer et vendre des thèses partout dans le loan, distribute and sell theses monde, à des fins commerciales ou autres, sur worldwide, for commercial or non- support microforme, papier, électronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats. . The author retains copyright L’auteur conserve la propriété du droit d’auteur ownership and moral rights in this et des droits moraux qui protège cette thèse. Ni thesis. Neither the thesis nor la thèse ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent être imprimés ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author’s permission.

In compliance with the Canadian Conformément à la loi canadienne sur la Privacy Act some supporting forms protection de la vie privée, quelques may have been removed from this formulaires secondaires ont été enlevés de thesis. cette thèse.

While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n’y aura aucun contenu removal does not represent any loss manquant. of content from the thesis.

Abstract

BACKGROUND: The prevalence and burden of cardiovascular disease (CVD) is a

concern. While CVD events will occur later in a woman’s life, modifiable risk factors

for CVD occur earlier during adult years. While, there is strong evidence linking modifiable risk factors to CVD, the influence of the work environment on CVD risk is poorly understood.

OBJECTIVES: The study objectives were to: 1) determine the prevalence of

cardiovascular risk indicators; 2) determine the relationships between work patterns and

lifestyle behaviours in female hospital workers; 3) determine the relationships between

work patterns and cardiovascular risk indicators; and 4) determine the relationships

between work patterns, lifestyle behaviours and cardiovascular risk while controlling for

covariates.

METHODS: Participants were female hospital workers (N= 466) from 2 hospital sites in

Southeastern Ontario. Cardiovascular risk data were obtained through anthropometric

measurements, blood sampling and self-report. Work pattern data were collected through

self-report and linked with hospital administrative work data. Lifestyle behaviour data

were obtained through self-report using validated questionnaires. Metabolic syndrome

was classified in accordance with the National Cholesterol Education Program Adult

Treatment Panel (NCEP ATP) (III) guidelines.

RESULTS: Approximately 1 in 4 female participants had the metabolic syndrome, with

elevated waist circumference being the most common CVD risk factor. After

adjustments, the multivariate analysis found a few key significant associations between

irregular work patterns, specifically extended shifts and elevated systolic and diastolic ii blood pressure. Consistent with the literature, the bivariate analyses revealed that after 6 or more years of shift work, female workers were more likely to develop the metabolic syndrome (OR 1.9, 95% CI 1.12, 3.17) and abdominally obesity (OR = 2.0, 95% CI,

1.31, 3.11).

CONCLUSIONS: The findings from this study suggest that generally work patterns do not influence the development of unhealthy behaviours and cardiovascular risk when controlling for age and other known factors. Further research is needed to elucidate the mechanisms linking harmful and protective work pattern characteristics to CVD risk, as the findings suggest relationships between work characteristics and risk. Given the prevalence of abdominal obesity and overall CVD risk, hospital decision-makers should consider cardiovascular health promotion within their healthy workplace initiatives.

iii Co-Authorship

This thesis presents the research of Megan Kirk in collaboration with her thesis

supervisor Dr. Joan Tranmer and thesis committee Dr. Ian Janssen and Dr. Elizabeth Van

Den Kerkhof.

Manuscript 1: The relationships between work patterns and modifiable lifestyle behaviours in female hospital workers. The original idea to explore the influence of work patterns on lifestyle behaviours in female hospital workers was M. Kirk’s. Dr.

Tranmer had identified the need to study the influence of work environment, inclusive of

work schedules on cardiovascular (CV) health in female hospital workers. I. Janssen and

E. Van Den Kerkhof provided content, methodological and statistical advice. The larger

study, entitled, Health and Work Study (HWS), was conducted by a research team led by

Dr. Tranmer. The HWS team contacted participants, administered the questionnaire

package and received administrative work data from Human Resources (HR) at both

hospital sites. M. Kirk carried out the statistical analyses with guidance from Dr. Tranmer

and Wilma Hopman, a research facilitator based at the Clinical Research Centre,

Kingston General Hospital. Composing of the manuscript and interpretation of the

results was completed by M. Kirk with modifications by thesis committee.

Manuscript 2: The influence of work patterns on cardiovascular risk factors in female

hospital workers. M. Kirk was responsible for the idea for this manuscript topic. Dr.

Tranmer had previously identified the need to study the influence of work environments,

including the influence of work schedules on female hospital worker’s cardiovascular iv health and designed the original cohort study. The HWS team met with participants, completed anthropometric measurements and provided blood requisitions for fasting blood work. M. Kirk completed primary data collection to gather work history data from the original cohort. I. Janssen and E. Van Den Kerkhof provided content, methodological and statistical advice. The larger study, entitled, Health and Work Study (HWS), was conducted by a research team led by Dr. Tranmer. The HWS team contacted participants, administered the questionnaire package and received administrative work data from Human Resources (HR) at both hospital sites. M. Kirk carried out the statistical analyses with guidance from Dr. Tranmer and Wilma Hopman, a research facilitator based at the Clinical Research Centre, Kingston General Hospital. Composing of the manuscript and interpretation of the results was completed by M. Kirk with modifications by thesis committee.

All other sections of this thesis (Introduction, Literature Review, General

Discussion, and Appendices) were written by M. Kirk with editorial changes provided by thesis committee.

v Acknowledgements

Firstly, I would like to extend my up-most gratitude to my thesis supervisor Dr.

Tranmer who provided me with continuous support and encouragement. Dr. Tranmer provided me with an opportunity to research a topic in which I am greatly interested in and inspired me as a researcher. From the beginning, her expertise and range of research projects provided me a breadth of opportunity and respect for nursing research. Dr.

Tranmer is an icon for nursing research, and a pleasure to learn from. I am truly honoured to have had this opportunity, and I not only grew as a researcher, but as a person as well.

Next, I would like to thank my thesis committee, Dr. Ian Janssen and Dr.

Elizabeth Van Den Kerkhof, for their individual contributions and support. They have helped to improve my study in many ways and I greatly appreciate their time and efforts.

I would like to acknowledge the Heart and Stroke Foundation Canada for the

Master’s Studentship Award and Dr. Tranmer for providing me with a research assistantship.

To my classmates I call friends, supporter, advocates, and colleagues. I need to take this moment to thank each and every one of you. We laughed together, cried together, and supported one another. This network of friends enriched my experience and I hope to work with you all in future endeavors.

vi Lastly, I would like to extend my endless appreciation and love to my family. My

father’s support and encouragement has provided me the strength to believe in myself.

My father listens with patience and interest, and I appreciate this support. To my partner,

you always believe in me and provide continuous support. To my family, you are my

foundation and strength; I dedicate this thesis to you.

vii Table of Contents

Abstract...... ii Co-Authorship ...... iv Acknowledgements...... vi Table of Contents...... viii List of Figures...... xi List of Tables ...... xii Chapter 1 Introduction ...... 1 1.1 General introduction ...... 1 1.2 Conceptual framework...... 2 1.3 Objectives ...... 5 1.4 Thesis organization and outline ...... 6 1.5 Reference list ...... 7 Chapter 2 Background Literature Review ...... 9 2.1 Introduction...... 9 2.2 The Health Issue: Cardiovascular Disease...... 9 2.2.1 Cardiovascular Disease (CVD)...... 9 2.2.2 Cardiovascular Disease Risk...... 10 2.2.3 The Metabolic Syndrome (MS) ...... 10 2.2.4 Measurement of the metabolic syndrome ...... 11 2.3 Relationships between Hospital Work Patterns and Modifiable Lifestyle Behaviours ...... 12 2.3.1 Work Patterns and Smoking ...... 13 2.3.2 Work Patterns and Alcohol Consumption ...... 15 2.3.3 Work Patterns and Diet...... 17 2.3.4 Work Patterns and Physical Activity ...... 18 2.4 Relationships between Work Patterns and Cardiovascular Disease Risk...... 21 2.4.1 Shift work and Cardiovascular Disease Risk...... 21 2.4.2 Extended Shifts and Cardiovascular Disease Risk...... 23 2.4.3 Overtime Hours and Cardiovascular Disease Risk ...... 25 2.5 Summary...... 27 2.6 References...... 28 Chapter 3...... 37

viii Manuscript 1 ...... 37 3.1 Abstract...... 38 3.2 Introduction...... 40 3.3 Study Population and Methods ...... 42 3.3.1 Study Population...... 42 3.3.2 Study Measures...... 43 3.3.3 Exposure Variables: Work Patterns ...... 43 3.3.4 Outcome Variables: Lifestyle Behaviours ...... 43 3.4 Statistical Analysis...... 45 3.5 Results...... 46 3.6 Discussion...... 47 3.7 References...... 53 Chapter 4...... 64 Manuscript 2 ...... 64 4.1 Abstract...... 65 4.2 Introduction...... 67 4.3 Study Population and Methods ...... 69 4.3.1 Study Population...... 69 4.3.2 Study Measures...... 70 4.3.3 Exposure Variables: Work Patterns ...... 70 4.3.4 Outcomes Variables: Cardiovascular Risk Factors...... 71 4.3.5 Covariates ...... 72 4.4 Statistical Analysis...... 72 4.5 Results...... 72 4.6 Discussion...... 74 4.7 References...... 79 Chapter 5 Discussion Chapter...... 91 5.1 Summary of key findings...... 91 5.2 Revisiting the objectives...... 92 5.3 Limitations ...... 95 5.4 Strengths ...... 95 5.5 Practice and policy implications ...... 96 Appendix A Additional Findings...... 99 ix Appendix B International Physical Activity Questionnaire (IPAQ)...... 114 Appendix C Rapid Eating Assessment for Patients (REAP) ...... 118

x List of Figures

Figure 1 - The adapted Knutsson & Boggild (2000) model linking irregular work patterns to CVD……………………………………………………………………………………4

Figure 2 - An organizing model of relationships between exposure, mediating and outcome variables…………………………………………………………………………5

Figure 3 - The number of participants who consented (n=547), complete self-report data (n=509) and complete administrative data (n=466)……………………………………...59

Figure 4 - The number of participants who consented (n=547), complete self-report data (n=509), complete administrative data (n=466) and complete work history data (n=359)…………………………………………………………………………………...83

1 xi List of Tables

Table 1. Descriptive Characteristics of Sample (N=466) ...... 60

Table 2. Description of lifestyle behaviours in female hospital workers (N=466) ...... 61

Table 3. Associations between personal and work patterns characteristics with smoking and excess alcohol intake...... 62

Table 4. Associations between personal and work pattern characteristics with total and leisure-time physical activity ...... 63

Table 5. Descriptive Characteristics of Sample (N= 466) ...... 84

Table 6. Prevalence of cardiovascular risk indicators among female hospital workers

(N=466)...... 86

Table 7. Associations between personal and work pattern characteristics and metabolic syndrome in female hospital workers...... 87

Table 8. Associations between personal and work patterns characteristics and waist circumference and systolic blood pressure in female hospital workers...... 89

Table 9. Associations between personal characteristics, work patterns and lifestyle behaviours with the metabolic syndrome in female hospital workers...... 100

Table 10. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated waist circumference in female hospital workers...... 102

Table 11. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated systolic blood pressure in female hospital workers...... 104

Table 12. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated diastolic blood pressure in female hospital workers...... 106

xii 1

Table 13. Associations between personal characteristics, work patterns and lifestyle behaviours with high fasting blood glucose in female hospital workers...... 108

Table 14. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated triglycerides in female hospital workers...... 110

Table 15. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated high density lipoproteins in female hospital workers...... 112

xiii 2

Chapter 1 Introduction

1.1 General introduction The prevalence and burden of cardiovascular disease (CVD) is a concern.

Historically, CVD related deaths were more prevalent in men compared to women, but recent statistics show that the number of CVD deaths is now equal in women and men

(Heart and Stroke Foundation, 2007). As women are living longer, the number of CVD related deaths in women will likely surpass that of men (Heart and Stroke Foundation,

2003). While CVD occurs later in a woman’s life, the risk for CVD and behaviours potentially leading to CVD occur earlier during working years.

There are well known modifiable and non-modifiable risk factors related to the development of CVD. The metabolic syndrome is a cluster of risk factors: high fasting blood glucose, abdominal obesity, elevated triglycerides, high blood pressure, and low high density lipoprotein cholesterol (Isomaa, et al. 2001; Lakka, et al. 2002; Sattar, et al.

2003) that are associated with increased cardiovascular events and death. The prevalence of metabolic syndrome is increasing to epidemic proportions within North America

(Cornier, et al. 2008). Metabolic syndrome is associated with a 2-fold increase in CVD risk (Cornier, et al. 2008). While there is strong evidence linking poor lifestyle behaviours, such as physical inactivity, poor dietary patterns, smoking and excess alcohol to CVD risk, the influence of contextual factors such as the work environment on lifestyle behaviours and CVD risk is poorly understood as available findings are limited and inconsistent. Presently, there are no reported Canadian studies exploring the influence of

1 work patterns on lifestyle behaviours and/or cardiovascular risk indicators in female workers.

The overall goal of this thesis study was to determine, in female hospital workers:

1) the associations between work patterns, where work patterns are defined within the context of worked hours, specifically, shift length, rotation, overtime and total hours worked, and potentially modifiable lifestyle behaviours, specifically, physical activity, dietary patterns, smoking and alcohol use; 2) the associations between work patterns and

CVD risk, specifically indicators of metabolic syndrome, and 3) the associations between work patterns, lifestyle behaviours and CVD risk indicators. It is anticipated that findings from this research will increase our understanding of the influences of work patterns on heart risk, and provide knowledge which could be used to inform the planning of workplace policy to promote the health of female hospital workers.

1.2 Conceptual framework Irregular work patterns are commonplace within healthcare settings, with shift work being the most common irregular work pattern. The negative cardiovascular implications of irregular work patterns are poorly understood in female workers, as the majority of past research has focused on the influence of shift work only and predominantly with male workers.

Based on previous research, Knutsson & Boggild (2000) described three main pathways linking shift work to cardiovascular disease: biological, behavioural and social pathways. According to this model, the mechanisms linking shift work to CVD are likely linked to disruption in three different etiological pathways: (A) Physiological disruption leading to altered circadian rhythms, internal desynchronization and metabolic changes.

Disturbed circadian rhythm may result in sleep deprivation; sleep deprivation or disruption potentially influences all pathways. (B) Behavioural changes leading to physical inactivity, poor dietary habits, increased smoking and alcohol use; and (C)

Socio- temporal effects leading to increased levels of stress. While shift work is a common work pattern characteristic in hospitals and other settings, other irregular work pattern characteristics, such as extended hours and overtime, are becoming more prevalent. In this thesis we expanded on the original model to include other irregular work patterns, and hypothesized that irregular work patterns, as a whole, will influence cardiovascular risk through similar mechanisms.

Irregular Work Patterns*

Disturbed socio- Mismatch of circadian temporal patterns rhythms

Sleep/wake disturbances Stress (work/home) Behavioural changes Coping e.g. physical activity, Internal diet, smoking, alcohol desynchronization

*Irregular Work Patterns Overtime, extended (12 h) shifts, rotating shift Cardiovascular Disease Risk work

Figure 1 - The adapted Knutsson & Boggild (2000) model linking irregular work patterns to CVD.

In this thesis study we specifically focused on the behavioural pathway as illustrated in Figure 2. Our null hypothesis was that there would be no relationships between irregular work patterns and indicators of cardiovascular risk in female hospital workers, when controlling for known confounders.

Behavioural Changes Family History of CVD

Irregular Metabolic Work Syndrome Patterns Age

*Irregular Menopausal Hormone

Work Patterns Status Use OT hours, extended (12 h) shifts, rotating shifts work

Figure 2 - An organizing model of relationships between exposure, mediating and outcome variables.

1.3 Objectives The goal of this thesis was to describe the independent and combined effects of work patterns on lifestyle behaviours and cardiovascular risk indicators in female hospital workers. The specific objectives for this thesis project were to:

1.3.1 Determine the prevalence of cardiovascular risk indicators, specifically

indicators of the metabolic syndrome, in female hospital workers.

1.3.2 Determine the relationships between work patterns and modifiable lifestyle

behaviours.

1.3.3 Determine the relationships between work patterns and cardiovascular risk

indicators.

1.3.4 Determine the relationships between work patterns, lifestyle factors and

cardiovascular risk while controlling for known confounders, e.g. age, hormone use,

menopause status and family history of CVD.

1.4 Thesis organization and outline This thesis is organized according to the Manuscript Form of Theses as specified by the School of Graduate Studies and Research at Queen’s University. Chapter 2 will provide a review of the scientific literature of the current evidence exploring the associations between work patterns and lifestyle behaviours, and work patterns and cardiovascular risk, with a focus on working women. Chapter 3 is the first manuscript entitled: The relationships between work patterns and modifiable lifestyle behaviours in female hospital workers, to be submitted to the Scandinavian Journal of Work,

Environment and Health. Chapter 4 presents the second manuscript entitled: The influence of work patterns on cardiovascular risk factors in female hospital workers, which will be submitted to the Scandinavian Journal of Work, Environment and Health.

Chapter 5 provides a discussion, summary and conclusions from the thesis work, detailing strengths and limitations and implications for future policy and research.

1.5 Reference list

Cornier, M., Dabelea, D., Hernandex, T., Lindstrom, R., Steig, A., Stob, N., et al. (2008).

The metabolic syndrome. Endocrine Reviews, 29(7), 777-822.

Heart and Stroke Foundation. (2007). Report cards on health – time to bridge the gender

gap. Ottawa, Ontario. Retrieved December 7, 2008 from

http://ww2.heartandstroke.ca/Page.asp?PageID=33&ArticleID=5911&Src=news&

From

Heart and Stroke Foundation. (2003). The growing burden of heart disease and stroke in

Canada in 2003. Ottawa, Ontario. Retrieved October 22, 2008 from

http://www.cvdinfobase.ca/cvdbook/CVD_En03.pdf

Isomaa, B., Almgren, P., Tuomi, T., Forsen, B., Lahti, K., Nissen, M., et al. (2001).

Cardiovascular morbidity and mortality associated with the metabolic syndrome.

Diabetes Care, 2, 683-389.

Knutsson, A., Boggild, H. (2000). Shift work and cardiovascular disease: review of

disease mechanisms. Reviews on Environmental Health, 15(4), 359-372.

Lakka, H.M., Laaksonen, D.E., Lakka, T.A., Niskanen, L.K., Kumpusalo, E., Tuomilehto,

J., et al. (2004). The metabolic syndrome and total and cardiovascular disease

mortality in middle-aged men. Journal of the American Medical Association, ,

288, 2709-2716.

Sattar, N., Gaw, A., Scherbakova, O., Ford, I., O’Reilly, D.S., Haffner, S.M., et al.

(2003). Metabolic syndrome with and without C-reactive protein as a predictor of

coronary heart disease and diabetes in the West of Scotland Prevention Study.

Circulation, 108, 414-419.

School of Graduate Studies and Research. General Forms of Theses. Kingston, Ontario,

Canada: Queen’s University, 2007.

Chapter 2 Background Literature Review

2.1 Introduction The following review is organized according to the objectives of this thesis. The literature review provides an overview of the problem of cardiovascular disease (CVD) in

Canada and a summary of the literature in regard to the proposed associations between work patterns, lifestyle factors, and CVD risk (with a focus on women). The last section of the review provides a summary and rationale for the study.

2.2 The Health Issue: Cardiovascular Disease

2.2.1 Cardiovascular Disease (CVD) In Canada, CVD, including myocardial infarction, ischemic heart disease, valvular heart disease, peripheral vascular disease, arrhythmias, hypertension and stroke, are the leading causes of morbidity and mortality for both men and women (Manuel, Leung,

Nguyen, Tanuseputro & Johansen, 2003; Heart and Stroke Foundation, 2003). The mortality rates for CVD have declined in the 1990’s, but the actual number of CVD related deaths has remained relatively stable. Currently, the number of CVD deaths is equal in Canadian men and women (Heart and Stroke Foundation, 2007), however, the number of CVD deaths in women is expected to increase beyond that of men as women are living longer (Heart and Stroke Foundation, 2003). Heart disease is the number one killer of women, yet many women fail to consider themselves at risk (Anderson, 2002).

Women often ignore their symptoms of heart disease and wait too long to seek medical

9 attention as symptoms tend to be vague (Health Canada, 1999). Eight times as many women die from heart disease and stroke than from breast cancer, and in total 40% of all

Canadian women's deaths are due to heart disease and stroke (Health Canada, 1999). The prevention of CVD in women is an important healthcare issue.

2.2.2 Cardiovascular Disease Risk Despite years of accumulating evidence linking modifiable risk factors, heart disease and other chronic illnesses, the prevalence of cardiovascular disease risk factors remains a concern. While CVD occurs later in a woman’s life, the modifiable risk factors which influence CVD risk occur earlier during adult years. Few women and younger adults consider CVD a personal health threat (Giardina, 2000). There is strong evidence linking modifiable risk factors, such as physical inactivity, smoking and obesity to CVD

There is a poorer understanding of the influence of contextual factors such as the work environment on CVD risk as findings are inconsistent. For all Canadians, physical inactivity, obesity, diabetes and hypertension rates continue to increase and are creating substantial detriments to health and subsequent burden on the healthcare system (Manuel, et al. 2003; Tanuseputro, Manuel, Leung, Nguyen & Johansen, 2003). It believed that

CVD is largely preventable by engaging in healthy lifestyle behaviours, such as a well balanced diet, adequate exercise, smoking cessation, and maintaining a healthy weight.

Health Canada reported that Canadians have a poor awareness of CVD risk factors, for example, 30% cannot identify one major risk (Health Canada, 1999).

2.2.3 The Metabolic Syndrome (MS) The primary outcome of interest in this thesis study is the metabolic syndrome. The metabolic syndrome, previously referred to as the insulin resistance syndrome, is a cluster 10 of risk factors: abdominal obesity, hypertension, high fasting blood glucose, elevated triglycerides, and low high density lipoprotein cholesterol that are associated with increased cardiovascular events and death (Isomaa, et al. 2001; Lakka, et al. 2002; Sattar, et al. 2003). Prevalence of the metabolic syndrome ranges between 13% and 25%, depending on the defining criteria and populations studied (Bosma, Stansfeld & Marmot,

1998), and is increasing to epidemic proportions within North America (Cornier, et al.

2008).

2.2.4 Measurement of the metabolic syndrome In 2001, The United States National Cholesterol Education Program’s Adult

Treatment Panel III (NCEP ATP III) provided evidence-based recommendations outlining diagnostic criteria for the metabolic syndrome. The NCEP ATP III diagnostic criteria for metabolic syndrome in women are: high fasting blood glucose (> 6.1 mmol/L), abdominal obesity (waist circumference > 88 cm), blood pressure > 140/90 mmHg, elevated triglycerides (> 1.69 mmol/L), and low HDL cholesterol (<1.29 mmol/L) (2001). At least three of these risk factors are to be present for classification of the metabolic syndrome. We applied the US NCEP ATP III guidelines to assess the prevalence of the metabolic syndrome disease, as it is the most frequently used definition in the current literature (Kahn, Buse, Ferrannini & Stern, 2005). Other definitions of the metabolic syndrome are available from the World Health Organization (WHO) and

American Heart Association (AHA). WHO provided the first definition of the metabolic syndrome in 1998, classifying this syndrome as: BMI >30 kg/m2 , triglycerides > 150 mg/dL, HDL <35 mg/dL in men and <39 mg/dL in women, blood pressure > 140/90 mmHg, and increased urinary albumin secretion (Alberti & Zimmet, 1998). The

11 definition provided by AHA outlines the syndrome as: waist circumference > 102 cm in men and >88 cm in women, triglycerides > 150 mg/dL, HDL <40 mg/dL in men and <50 mg/dL in women, blood pressure >130/85 mmHg, glucose >110 mg/dL (Grundy, Brewer,

Cleeman, Smith & Lenfant, 2004). The WHO and NCEP ATP III have similar defining criteria for the metabolic syndrome. The key difference is that NCEP ATP III guidelines use waist circumference as a measure for obesity, whereas WHO criteria assesses body mass index.

There are strong associations between the metabolic syndrome and increased cardiovascular morbidity and mortality (Isomma, et al. 2001; Lakka, et al. 2002). A longitudinal epidemiological study exploring the association between metabolic syndrome and CVD using the NCEP and

WHO guidelines found men with metabolic syndrome were at increased risk of CVD and all- cause mortality (Lakka, et al. 2002). Using the NCEP criteria, men with the metabolic syndrome were 2.1 (95% CI, 0.93-4.65) times more likely to die from CVD, whereas, when using the WHO criteria, men were 2.5 (95% CI, 1.33-4.80) times more likely, compared to men without CVD risk factors, when controlling for age (Lakka, et al. 2002). Regardless of the criteria used, it is generally felt that individuals with metabolic syndrome have a two-fold increase in morbidity and mortality related to CVD (Bonora, Kiechl, Willeit, Oberhollenzer, Egger & Bonadonna, 2003;

Resnick, et al. 2003; Rutter, Meigs, Sullivan, D’Agostino & Wilson, 2004; Hunt, Resendez,

Williams, Steve & Stern, 2004; Ford, 2005; McNeill, et al. 2005).

2.3 Relationships between Hospital Work Patterns and Modifiable Lifestyle Behaviours Hospital work environments can be characterized as demanding on many fronts.

Hospital employees are exposed to long working hours, shift work, family/work imbalance, high work pace, and high emotional demands. Emerging evidence shows that

CVD risk is associated with 1) work stress (Steptoe, Cropley & Joekes, 1999; Steptoe,

Siegrist, Kirschbaum & Marmot, 2004; Vrijkotte, Van Doornen & de Geus, 1999,

Vrijkotte, Van Doornen & de Geus, 2000, Kristensen, 2005), 2) shift work (Knuttson &

Boggild, 2000), and 3) sleep disturbances (Harma, 2006). Evidence supports links between healthy lifestyle behaviours and cardiovascular health, which include health enhancing physical activity levels, balanced dietary patterns, specifically patterns that meet the recommended Canada Food Guidelines, smoking abstinence, and limited excessive alcohol intake (De Backer, et al. 2003). Healthy lifestyle behaviours may be affected by irregular work patterns due to increased fatigue, increased time at work and disrupted normal social and recreational activities. Irregular work patterns may result in varied coping strategies for employees that can range from advantageous to detrimental on health (Costa, 1995; Kivimaki, Paivi, Virtanen & Elovainio, 2001). The mechanisms of irregular work patterns leading to poorer health remain unclear, although unhealthy coping strategies, such as increased smoking and alcohol consumption have been identified (Burch, et al. 2009).

2.3.1 Work Patterns and Smoking Multiple studies have linked irregular work patterns to smoking (Kawachi, et al.

1995, Trinkoff & Storr, 1998, Kroenke, et al. 2007, Shields, 1999). Researchers have hypothesized that smoking, specifically nicotine, acts as stimulant during night shifts. As well, ‘smoke-breaks’ are commonly viewed as a coping mechanism as they provide time for the worker to leave the work unit, interrupting extended shifts within the stressful healthcare work environment. There is strong evidence supporting higher rates of smoking among male and female shift workers when compared to day workers (Di

Lorenzo, et al. 2003; Suwazono, et al. 2008; van Amelsvoort, Schouten & Kok, 2004,

Knutsson, Akerstedt & Jonsson, 1988). Knutsson, Hallquist, Reuterwall, Theorell and

Akerstedt (1999), in a case-control study explored the relationships between shift work and coronary heart disease, found that smoking, both current and past, was higher among female shift workers compared to female day workers. In a study, of 227 shift workers and 150 day workers starting new jobs, smoking increased during the first year for shift workers and not day workers (van Amelsvoort, et al. 2004). Van Amelsvoort et al.

(2004) hypothesized increased cigarette smoking among shift workers would increase their relative cardiovascular risk by 10% if the increased rate of smoking remained constant over 20 years.

Two studies exploring the influence of work schedules on nurses’ health found that shift work and long working hours resulted in increased smoking rates. The USA

Nurses’ Health Study (NHS) reported a higher prevalence of smoking within shift work nurses, compared to day work nurses (19.2% versus 17.4%) (no p value), and with longer shift work durations (1-2 years – 17.3%, 15 or more years – 24.2%)(no p value)

(Kawachi, et al. 1995). A study exploring the influence of work schedules on substance use in 3917 nurses (95% female) found that smoking increased with shift work duration and extended shifts (Trinkoff & Storr, 1998). Similarly, the NHS II found that as working hours increased, smoking increased. Highest rates of smoking were found among employees working 41-60 and > 61 hours/week (Kroenke, et al. 2007). A report from

Canada found that female employees changing from regular to longer shifts were more likely to smoke (Shields, 1999).

Therefore, shift workers, and more generally, employees working irregular work patterns, may be more likely to engage in unhealthy smoking behaviours.

2.3.2 Work Patterns and Alcohol Consumption Similar to smoking behaviours, alcohol intake patterns may be influenced by irregular work patterns. Past findings are inconsistent. A recent cross-sectional study exploring the impact of shift work on the health of healthcare workers (n=376); 90% of sample were female workers reported no difference in alcohol intake between shift and day workers (study did not show data) (Burch, et al. 2009). Other studies, however, have reported associations between shift work and increased alcohol intake (van Amelsvoort, et al. 2004; Suwazono, et al. 2008; Sookoian, et al. 2007). A 14-year historical cohort study exploring the effects of shift work on blood pressure in 8251 male steel workers found that shift workers in comparison to day workers had a significantly higher alcohol intake (46.7% versus 36.5%, p<0.001) (Suwazono, et al. 2008). Shields (1999) reported that women working longer hours (>41 hours/week), compared to those who continued to work standard hours (35 -40 hours/week) had higher alcohol consumption. The prospective NHS II followed 62,574 registered nurses and these investigators also found that alcohol consumption increased with consecutive hours worked per week (Kroenke, et al. 2007). However, when investigators examined the relationship between shift work and CHD within the same female nurses cohort (NHS) they found an inverse relationship between shift work duration and alcohol intake; as shift work duration increased, alcohol consumption decreased (Kawachi, et al. 1995). In a study exploring the influence of work schedules on substance use in 3917 nurses (95% female), results showed associations between night work (OR=1.29, 95% CI 0.99, 1.67), rotational shift work

(OR=1.54, 95% CI 1.21, 1.97, p<0.01), extended shifts (OR=1.44, 95% CI 1.20, 1.72, p<0.01) and alcohol consumption (Trinkoff & Storr, 1998).

Determining healthy or unhealthy alcohol intake patterns is complex. It is believed that low to moderate alcohol intake may reduce cardiovascular related mortality

(Nanchahal, Ashton & Wood, 2000). Moderate alcohol consumption seems to have two protective effects: increase in HDL and anti-thrombotic effects (Nanchahal et al. 2000).

A cross-sectional study of CVD risk factors in 14,077 female employees in the UK found a linear decrease in LDL cholesterol with increasing alcohol intake (p<0.001), and an increase in HDL cholesterol with increasing alcohol consumption up to 15-21 units/week

(p<0.001) (Nanchahal et al.). A threshold is important to identify to balance the risks

(hemorrhagic stroke, breast cancer, hypertension, non-cardiovascular events) and benefits of moderate alcohol consumption. Nanchahal et al. found that women when consuming

1-7 units/week had the greatest CHD risk reduction (OR=0.79, 95% CI 0.72, 0.87, p<0.001), when compared to non-drinkers. A decrease in the prevalence of hypertension was found up to 14 units/week of alcohol, after this threshold, an increase in hypertension occurred (Nanchahal et al.). The reduction in CHD risk for women when consuming 1-7 units/week of alcohol was 21%, with an additional 13% reduction when consuming 8-14 units/week (p<0.001), when compared to non-drinkers (Nanchahal et al.). Thus there seems to be support for associations between low to moderate (1-14 units/week) alcohol intake and reduction in CHD in women (Nanchahal et al.). There is no consistent evidence supporting relationships between irregular work patterns and unhealthy alcohol consumption.

2.3.3 Work Patterns and Diet A few studies have examined the influence of irregular work patterns; predominantly shift work, on dietary habits, more specifically the frequency of meals, timing of meals, and nutrient content (van Amelsvoort, et al. 2004; Knutson, Andersson

& Berglund, 1990; Costa, 1996; Geliebter, Gluck, Tanowitz, Aronoff & Zammit, 2000).

In a one-year follow-up study, female and male shift workers were assessed for changes in cardiovascular risk factors (van Amelsvoort, et al. 2004). At baseline, significantly more shift workers had a higher percentage of energy intake from fat than day workers

(40.3% vs. 38.1%, p<0.002). At one year follow-up, both groups had decreases in energy intake but only day workers demonstrated significant decreases in the amount of energy from fat and cholesterol intake (-22.4, p<0.05) (van Amelsvoort, et al. 2004). Knutson et al. (1990) conducted a prospective cohort study designed to examine serum lipoproteins changes in day and shift workers before employment and six months post. Nutrient intake was assessed using a four day diary and found that shift workers significantly reduced their intake of dietary fiber (p<0.03) through a decreased intake of vegetables and potatoes (Knutson et al.). Shift workers were found to have an increase in sucrose which was linked to an increase in soft drink consumption (p<0.07) (Knutson et al.). Although this study only included a small group of participants (n = 25) and is dated, the findings do suggest that irregular work patterns (shift work) may influence CVD risk.

Costa (1996) explains that even though calorie intake often remains unaltered, the quality of food consumed and the timing of meals often changes in shift workers. For instance, when on night shifts, quick meals are consumed, consisting of pre-packed food and increased intake of caffeinated beverages, such as coffee and tea (Costa, 1996) and fats and refined carbohydrates (de Castro, 2004; Wardle, Steptoe & Oliver, 2000). 17

Geliebter et al. (2000) found that shift workers self-reported eating fewer larger meals

(1.9 vs. 2.5, p = 0.002) and ate later in the day (p < 0.001) since beginning shift work which resulted in weight gain. Contrary to these findings, a study exploring the impact of rotational schedules on eating habits in male industrial workers found no changes in energy intake, nutrients or caffeinated beverages in day and shift workers, except again that shift workers ate at varied times (Lennernas, Hambraeus & Akerstedt, 1995).

Two studies have explored the associations between overtime and dietary patterns.

A descriptive study of male Japanese white-collar non-management workers determined the differences between overtime and dinner time, breakfast consumption, snacking and preference for rich, fatty foods. Overtime was associated with later dinner times

(r=0.436, p<0.0001), but no other variables (Nakamura, et al. 1998). These investigators hypothesized that working overtime leads to eating dinner chronically late which may lead to weight gain (Nakamura, et al. 1998). A more recent longitudinal cohort study

(N=1266) examining the links between overtime and the development of diabetes in men found no significant difference between hours of work per day, breakfast, vegetable, and fruit consumption (Nakanishi, et al. 2001).

Thus, the evidence in regard to relationships between irregular work patterns and unhealthy dietary patterns is inconsistent.

2.3.4 Work Patterns and Physical Activity Similar to other lifestyle behaviours, the relationships between work patterns and attainment of healthy physical activity levels is complex, and the findings from studies are inconsistent.

Shift work seems to adversely influence physical activity levels. A cohort study exploring the impact of one year of shift work on male and female participants found physical activity levels decreased in shift workers and not in day workers (van

Amelsvoort, et al. 2004). Nakamura et al. (1997) also found a significant difference in frequency of non-exercise between shift workers and day workers (69% vs. 50%, p <

.05). A recent 14-year prospective cohort study with Japanese male workers found that alternating shift workers were less likely to exercise routinely (46.7%) when compared to day workers (42.6%) (p<0.001) (Suwazono, et al. 2008). These findings are supported by

Burch, Tom, Zhai, Criswell, Leo & Ogoussan (2009) in which night workers reported lower physical activity levels.

A cross-sectional study including female nurses, reported inverse correlations between physical activity and shift work duration (r=0.14, p<0.05), which may have been partially explained by the correlation between physical inactivity and age (r=0.32, p<0.05) (Ha & Park, 2005). The Nurses’ Health Study was the only prospective cohort study to explore the associations between shift work duration and physical activity in female nurses. Longer durations of shift work were associated with a higher age-adjusted prevalence of physical activity level (Kawachi, et al. 1995). Kivimaki, et al. (2001) found that sedentary lifestyles were more prevalent in day workers than shift workers. It was also found that sedentary lifestyles significantly increased by age in both day and shift workers (Kivimaki, et al. 2001).

A recent review by Atkinson, Fullick, Grindey & Maclaren (2008) examined the current literature on energy expenditure and energy intake of shift workers. This review suggests that shift workers often have more difficulty maintaining exercise routines.

Although shift workers often report lower levels of physical activity due to lack of time and general fatigue, Atkinson et al. (2008) explain how physical activity actually improves sleep as it decreases anxiety, depression and aids in circadian phase shifting.

Other irregular work patterns, such as extended shifts and long work hours may impact an individual’s capacity to engage in health enhancing physical activity. A few studies have explored the effect of other work pattern characteristics on physical activity

(Kroenke, et al. 2007, Nakamura, et al. 1998, Sookoian, et al. 2007, Park, et al. 2001). A

Canadian longitudinal cohort study examining long working hours and health, followed

1649 women and found modest decreases in exercise levels with increases in hours of work per day, although associations were not significant (Shields, 1999). The Nurses’

Health Study II followed 62 574 young to middle-aged women for 6 years to better understand associations between work characteristics and the development of type 2 diabetes (Kroenke, et al. 2007). Kroenke et al. found that women working 40 hours or more a week tended to engage in higher amounts of leisure-time physical activity.

Whereas, in two cross-sectional studies of Japanese male workers, lower physical activity levels were found to be associated with longer shift length > 11 hours/day (p=0.04)

(Nakanishi, et al. 2001), but not overtimes hours (Nakamura, et al. 1998). Sookoian et al.

(2007) reported similar findings. Shift workers working 12 hours were less active in comparison to 8 hour day workers (physical activity duration 1.4 h/week + 0.3 vs.1.9 h/week + 0.2) (p=0.001). Park et al. (2001), found no significant difference between number of working hours and regular exercise.

It is clear from the literature reviewed, that there are no consistent relationships between different work pattern characteristics and physical activity attainment.

2.4 Relationships between Work Patterns and Cardiovascular Disease Risk

2.4.1 Shift work and Cardiovascular Disease Risk Shift work and exposure to night work is a common work pattern within the hospital setting linked to increased cardiovascular risk (Knutsson & Boggild, 2000;

Kawachi, et al. 1995; Karlsson, Knutsson, Lindahl & Alfredsson, 2003). Approximately

20% of the labour workforce of most developed countries are employed in varying rotational shift work schedules (Angersbach, et al. 1980). This percentage is not likely to decrease, thus, long-term effects of shift work ought to be of particular interest to researchers and policy makers, alike. It has been estimated that working rotating shifts increases risk of cardiovascular disease by 30-40%, after adjustments are made for social class and other traditional risk factors (Harma, 2006). A number of studies have explored the effects of shift work on female employee’s health (Kawachi, et al. 1995; Van

Amelsvoort, et al. 2004; Ha & Park, 2005; Kivimaki, et al. 2001). Again, the prominent

Nurses Health Study (NHS) based in the United States found that coronary heart disease risk increased in women after 6 or more years of shift work (RR=1.51, 95% CI, 1.12,

2.03), compared to women who had never been involved in shift work before (Kawachi, et al. 1995). Overall, the NHS found that having ever worked shift work increased a women’s risk of coronary heart disease 1.4 fold and as duration of shift work increased, so did relative risk for coronary heart disease (Kawachi, et al. 1995). The outcome of interest was coronary heart disease (i.e., myocardial infarction, mortality) and not cardiovascular risk indicators as of interest in this study. Since cardiac events occur later in life, the contribution of shift work to the development of CVD risk factors in women remains unclear.

In regard to shift work, phase shifts in circadian rhythm between day and night work can lead to hormonal and metabolic abnormalities. Van Amelsvoort et al. (2004) investigated changes in anthropometric measurements, plasma cholesterol and blood pressure after one year of shift work in male and female workers (n = 377) starting new positions as hospital nurses, occupational health or waste incineration plant employees. A higher prevalence of smoking in shift workers was the only identified risk factor. Given the follow-up period was one year, this was likely too short of a period for the development of cardiovascular risk factors. More recently, Ha & Park (2005) examined the relationship between duration of shift work and metabolic risk factors. The study included 226 female nurses (average age 28.5 years, range: 19-49 years) and 134 male plant workers (average age 29.1 years, range: 25-44 years) who completed questionnaires pertaining to shift work duration, total years of work, lifestyle habits and past medical history (Ha & Park, 2005). In female workers, duration of shift work was inversely associated with increased cholesterol (r2=-2.82, p<0.05) over the age of 30 and increased diastolic blood pressure (DBP) (r2=-1.01, p<0.05) under the age of 30, when controlling for physical activity, job strain, smoking and alcohol intake. Waist-hip ratio in female nurses 30 years and older, was significantly associated with duration of shift work

(r2=0.004, p<0.05) after adjustments. Shift work duration was broadly defined as the total number of months during which the person had engaged in shift work in his or her lifetime, with the mean being 2.8 years (1 mos-10yrs) for nurses (Ha & Park, 2005).

Within our sample population, and more generally in Canada, the mean length of years worked for nurses is greater than that noted in the previous study. In a Finnish cross- sectional 10 hospitals cohort study, investigators compared 506 shift work and 183 day

22 work female nurses (Kivimaki, et al. 2001). Shift workers were more likely to be overweight, when defined as BMI >25 kg/m2 than day workers (OR=1.54, 95% CI, 1.06,

2.25). Shift workers over 45 years of age had on an average one unit higher body mass index than day workers, a difference associated with a 4-5% increase in coronary heart disease mortality (Kivimaki, et al. 2001). A study of 12 male workers reported that male night workers had elevated glucose (p<0.01) and insulin (p<0.01) levels likely related insulin insensitivity at night (Lund, Arendt, Hampton, English & Morgan, 2001). A retrospective, longitudinal study followed a cohort of 488 male municipal workers to assess the development of metabolic syndrome risk factors among night workers (Biggi,

Consonni, Galluzzo, Sogliani & Costa, 2008). A catabolic response was found among workers at night as triglyceride and total cholesterol levels were elevated (Biggi, et al.

2008).

Based on the epidemiological evidence the emerging evidence supports a relationship between shift work, and in particular long durations of shift work, and the development of cardiovascular risk factors. The mechanisms linking shift work to CVD risk are unclear, as not all shift workers develop CVD. As well, other important work patterns, such as extended shifts and overtime hours that may or may not impact CVD risk have not been explored in these studies.

2.4.2 Extended Shifts and Cardiovascular Disease Risk Traditional 8 hour shifts for hospital nurses have changed over the past few decades to predominantly include extended 12 hour shift patterns (Rogers, Hwang, Scott,

Aiken & Dinges, 2004). Most studies on shift work have only assessed the effects of 8 hour shifts on cardiovascular risk, leaving us with a poor understanding of the influence

23 of extended shifts. A recent cross-sectional study in Argentina, explored the effects of rotating shift work on metabolic syndrome risk factors in 877, 8 hour day employees and

474, 12 hour male shift workers (Sookoain, et al. 2007). Rotating shift workers working extended hours were found to be at greater risk of developing metabolic syndrome

(OR=1.51, 95% CI, 1.01-2.25, p<0.04) compared to day workers working 8 hours, after controlling for age and physical activity. The simultaneous presence of three or more metabolic risk factors were significantly more common in rotating shift workers (17.2%) when compared to day workers (10.7%, p<0.005). Extended 12 hour shift workers when compared to 8 hour day workers, showed the following characteristics: BMI (27.1+ 0.3;

26.3 + 0.2, p<0.02); waist-to-hip ratio (WHR) (0.95 + 0.01; 0.93 + 0.01, p<0.01), DBP

(78 mmHg+ 1; 76 mmHg+ 1, p<0.04), fasting insulin (65.5 pmol/L + 2.9; 55.9 pmol/L +

1.9, p< 0.02) and triglycerides (1.71 mmol/L + 0.1; 1.5 mmol/L + 0.1, p< 0.003

(Sookoain, et al. 2007).

A dated, retrospective longitudinal cohort study completed by Angersbach, et al.

(1980) in Germany explored the influence of extended shifts on cardiovascular disease and sick leave. The health insurance records of 210 shift and 142 male day workers were analyzed over an 11 year span and found that 32% of the shift workers involved in extended, 12 hour shifts and 7% of day workers, 8 hour shifts, left original employment for medical reasons (Angersbach, et al. 1980). Although, no statistical difference between shift and day workers could be detected, 16.8% shift workers and 14.8% of day workers were diagnosed with a cardiovascular disease (Angersbach, et al. 1980).

These studies are limited as only male participants were included, again, leaving us with a gap in understanding the effects of extended, 12 hours, shifts on female workers’ health.

2.4.3 Overtime Hours and Cardiovascular Disease Risk In addition to extended shifts, the influence of overtime hours on cardiovascular health is of concern within the healthcare system. With the growing shortage of nurses, there is a steady demand for overtime work. The term “karoshi,” meaning death from overwork, is referred to frequently in Japanese articles discussing the effects of increasing hours of work on employee’s health (Kawakami, et al. 1999). Five studies were available, two assessing the association between overtime and anthropometric measurements (Nakamura, et al. 1998, Wada, et al. 2006) and three studies exploring the influence of overtime and the development of diabetes (Kawakami, et al. 1999,

Nakanishi, et al. 2001, Kroenke, et al. 2007), both risk factors for the metabolic syndrome. Unfortunately, these studies have conflicting results and only the USA based study included female participants (Kroenke, et al. 2007).

Nakamura et al. (1998) reported that working overtime was associated with increased BMI (r=0.21, p<0.0017) and waist circumference (r=0.22, p=0.0091) after a three year period. Whereas, Wada et al. 2006, found that age-adjusted incidence rates of developing hypertension and increasing BMI to be significantly lower among participants whose mean overtime was >50 hours/month than <50 hours/month (p<0.05). Both were longitudinal studies, but opposing results may be due to differences in the defining of hours worked per week and participants. For instance, Nakamura et al. (1998) compared non-management white collar workers at a computer manufacturing company working

>40 hours/week to <40 hours/week. Wada, et al. (2006) compared manual workers working >50 hours/week to <50 hours/week.

A USA based prospective cohort study of 62, 574 female participants, aged 29-46 years from the NHS II explored work pattern characteristics and development of diabetes in female workers (Kroenke, et al. 2007). Results revealed 365 cases of type 2 diabetes developed over a 6 year follow-up. Women were at increased risk of diabetes when working 41-60 hours/week (RR=1.57, 95% CI, 1.26, 1.94) and >61 hours/week

(RR=1.49, 95% CI, 0.85, 2.63), p<0.001, compared to women working 21-40hr/week after controlling for age. Also, the risk of diabetes increased for women with duration of shift work: 2-<5 years (RR=1.04, 95% CI, 0.76, 1.39), 5-<10 years (RR=1.59, 95% CI,

1.15, 2.16), >10 years (RR=1.64, 95% CI, 1.11, 2.37), p<0.001.

In an 8-year prospective cohort study of male participants, longer overtime was found to be a risk factor for non-insulin dependent diabetes mellitus (NIDDM)

(Kawakami, et al. 1999). Kawakami et al. (1999) found that male workers who worked over 50 hours of overtime a month had a 3.7 times higher risk of NIDDM (95% CI, 1.41,

9.90). Nakanishi et al. (2001) conducted a 5 year cohort study on 1266 Japanese male office workers, finding that longer overtime was a negative risk factor for the development of impaired fasting glucose (IFG) and type 2 diabetes mellitus. The relative risk of IFG or type 2 DM with increasing hours of work were: 8.0-8.9 h/day (RR=0.82,

95% CI, 0.54, 1.26), 9.0-9.9 h/day (RR=0.69, 95% CI, 0.38, 1.26), 10.0-10.9 h/day

(RR=0.63, 95% CI, 0.37, 1.09), >11.0 h/day (RR=0.50, 95% CI, 0.25, 0.98), when compared to those who worked <8.0 hours a day (Nakanishi, et al. 2001).

Again, a gap exists in understanding the influence of work patterns, especially overtime hours, on the cardiovascular risk of female employees.

2.5 Summary The considerable variation in the literature regarding the influence of work patterns on lifestyle behaviours and cardiovascular risk in both male and female employees supports the need to identify and understand the mechanisms linking work patterns to cardiovascular risk. It is suggested in the available literature that work patterns may play a role in impacting on the ability of workers to develop routine healthy lifestyle behaviours. Few studies explore the influence of independent and combined irregular work pattern characteristics on health enhancing lifestyle behaviours and cardiovascular risk of female employees, and none are currently available that address this issue for Canadian female employees. This thesis project aimed to address this gap by improving our current understanding of the influence of work pattern characteristics of female hospital workers on behaviours associated with, and indicators of cardiovascular health.

2.6 References

Alberti, K.G., & Zimmet, P.Z. (1998). Definition, diagnosis, and classification of

diabetes mellitus and its complications. Part 1: Diagnosis, and classification of

diabetes mellitus. Provisional report of a WHO consultation. Diabetic Medicine,

15, 539-553.

Anderson, R.N. (2002). Deaths: Leading Causes for 2000. National vital statistics

reports, 50(16), 1-85.

Angersbach, D., Knauth, P., Loskant, H., Karvonen, M.J., Undeutsch, K., & Rutenfranz,

J. (1980). A retrospective cohort study comparing complaints and diseases in day

and shift workers. International Archives of Occupational and Environmental

Health, 45(2), 127.

Atkinson, G., Fullick, S., Grindey, C., & Maclaren, D. (2008). Exercise, energy balance

and the shift worker. Sports Medicine, 38(8), 671-685.

Biggi, N., Consonni, D., Galluzzo, V., Sogliani, M., & Costa, G. (2008). Metabolic

syndrome in permanent night workers. Chronobiology International, 25, 443-454.

Bonora, E., Kiechl, S., Willeit, J., Oberhollenzer, F., Egger, G., & Bonadonna, R. C.

(2003). Carotid atherosclerosis and coronary heart disease in the metabolic

syndrome. Diabetes Care, 26(4), 1251.

Bosma, H., Stansfeld, S., & Marmot, M. (1998). Job control, personal characteristics, and

heart disease. Journal of Occupational Health and Psychology, 3, 402-409.

Burch, J.B., Tom, J., Zhai, Y., Criswell, L., Leo, E., & Ogoussan, K. (2009). Shiftwork

impacts and adaptation among health care workers. Occupational Medicine, 10,

1093-2001.

Cornier, M., Dabelea, D., Hernandex, T., Lindstrom, R., Steig, A., Stob, N., et al. (2008).

The metabolic syndrome. Endocrine Reviews, 29(7), 777-822.

Costa, G. (1995). Factors influencing health of workers and tolerance to shift work.

Theoretical Issues in Ergonomics Science, 4, 263-288.

Costa, G. (1996). The impact of shift and night work on health. Applied Ergonomics,

27(1), 9-16.

De Backer, G., Ambrosioni, E., Borch-Johnsen, K., Brotons, C., Cifkova, R.,

Dallongeville, J., et al. (2003). European guidelines on cardiovascular disease

prevention in clinical practice: Third joint task force of European and other

societies on cardiovascular disease prevention in clinical practice. European Heart

Journal, 24(17), 1601-1610.

De Castro, J. (2004). The time of day of food intake influences overall intake in humans.

Journal of Nutrition, 134, 104-111.

Di Lorenzo, L., De Pergola, G., Zocchetti, C., L’Abbate, N., Basso, A., Pannacciulli, N.,

et al. (2003). Effect of shift work on body mass index : results of a study

preformed in 319 glucose-tolerant men working in a Southern Italian industry.

International Journal of Obesity, 27, 1353-1358.

Executive Summary of the Third Report of the National Cholesterol Education Program

(NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood

Cholesterol in Adults (Adult Treatment Panel III). (2001). Journal of the

American Medical Association, 285, 2486-2497.

Ford, E. S. (2005). Risks for all-cause mortality, cardiovascular disease, and diabetes

associated with the metabolic syndrome: A summary of the evidence. Diabetes

Care, 28(7), 1769-1778.

Geliebter, A., Gluck, M., Tanowitz, M., Aronoff, N., & Zammit, G. (2000). Work-shift

period and weight change. Nutrition, 16(1), 27-29.

Giardina, E.G. Heart disease in women. (2000). International Journal of Fertility and

Women’s Medicine, 45, 350-357.

Grundy, S.M., Brewer, B., Cleeman, J.I., Smith Jr., S.C., & Lenfant, S.C. (2004).

Definition of metabolic syndrome. Report of the National Heart, Lung, and Blood

Institute/American Heart Association Conference on scientific issues relating to

definition. Circulation, 109, 433–438.

Ha, M., & Park, J. (2005). Shift work and metabolic risk factors of cardiovascular

disease. Journal of Occupational Health, 47, 89-95.

Harma, M. (2006). Workhours in relation to work stress, recovery and health.

Scandinavian Journal of Work, Environment and Health, 32(6), 502-514.

Health Canada. (1999). Women and Heart Health. Retrieved February 6, 2009 from

http://www.hc-sc.gc.ca/hl-vs/pubs/women-femmes/heart-cardiovasculaire-eng.php

Heart and Stroke Foundation. The growing burden of heart disease and stroke in Canada

in 2003. (2003). Ottawa, Ontario.

Heart and Stroke Foundation. (2007). Report cards on health – time to bridge the gender

gap. Ottawa, Ontario. Retrieved December 7, 2008 from

http://ww2.heartandstroke.ca/Page.asp?PageID=33&ArticleID=5911&Src=news&

From

Hunt, K. J., Resendez, R. G., Williams, K. H., Steve M., & Stern, M. P. (2004). National

cholesterol education program versus world health organization metabolic

syndrome in relation to all-cause and cardiovascular mortality in the San Antonio

heart study. Circulation, 110(10), 1251-1257.

Isomaa, B., Almgren, P., Tuomi, T., Forsen, B., Lahti, K., Nissen, M., et al. (2001).

Cardiovascular morbidity and mortality associated with the metabolic syndrome.

Diabetes Care, 2, 683-389.

Janzon, E., Engstrom, G., Lindstrom, M., Berglund, G., Hedblad, B., & Janzon, L.

(2005). Who are the “quitters”? A cross-sectional study of circumstances

associated with women giving up smoking. Scandinavian Journal of Public

Health, 33(3), 175-182.

Kahn, R.A., Buse, J., Ferrannini, E., & Stern, M. (2005). The metabolic syndrome: time

for critical appraisal. Diebetologia, 48, 1684-1699.

Karlsson, B., Knutsson, A., Lindahl, B., & Alfredsson, L. (2003). Metabolic disturbances

in male workers with rotating three-shift work. Results of the WOLF study.

International Archives of Environmental and Occupational Health, 76, 424-430.

Kawachi, I., Colditz, G., Stampfer, M., Willett, W., Manson, J., Speizer, et al. (1995).

Prospective study of shift work and risk of coronary heart disease in women.

Circulation, 92, 3178-3182.

Kawakami, N., Araki, S., Takatsuka, N., Shimizu, H., & Ishibashi, H. (1999). Overtime,

psychosocial working conditions, and occurrence of non- insulin dependent

diabetes mellitus in Japanese men. Journal of Epidemiology and Community

Health, 53(6), 359-363.

Kivimaki, M., Paivi, K., Virtanen, M., & Elovainio, M. (2001). Does shift work lead to

poorer health habits? A comparison between women who had always does shift

work with those who had never done shift work. Work Stress, 15, 3-13.

Knutsson, A., Akerstedt, T., & Jonsson, B. (1988). Prevalence of risk factors for coronary

artery disease among day and shift workers. Scandinavian Journal of Work,

Environment and Health, 14, 317-321.

Knutsson, A., Hallquist, J., Reuterwall, C., Theorell, T., & Akerstedt, T. (1999).

Shiftwork and myocardial infarction: a case-control study. Occupational and

Environmental Medicine, 56(1), 46-50.

Knutsson, A., & Boggild, H. (2000). Shift work and cardiovascular disease: review of

disease mechanisms. Reviews on Environmental Health, 15(4), 359-372.

Knutson, A., Andersson, H., & Berglund, U. (1990). Serum lipoproteins in day and shift

workers: A prospective study. British Journal of Industrial Medicine, 47(2), 132-

134.

Kristensen, T.S. The changing nature of work and its implications for cardiovascular

disease. (2005). Fourth International Conference on Work Environment and

Cardiovascular Diseases.

Kroenke, C.H., Spiegelman, D., Manson, J., Schernhammer, E.S., Colditz, A., &

Kawachi, I. (2007). Work characteristics and incidence of type 2 diabetes in

women. American Journal of Epidemiology, 165(2), 175-183.

Lennernas, M., Hambraeus, L., & Akerstedt, T. (1995). Shift related dietary intake in day

and shift workers. Appetite, 25(3), 352-166.

Lakka, H.M., Laaksonen, D.E., Lakka, T.A., Niskanen, L.K., Kumpusalo, E., Tuomilehto,

J., et al. (2004). The metabolic syndrome and total and cardiovascular disease

mortality in middle-aged men. Journal of the American Medical Association, 288,

2709-2716.

Lund, J., Arendt, J., Hampton, S., English, J., & Morgan, L. (2001). Postprandial

hormone and metabolic responses amongst shift workers in Antarctica. Journal of

Endocrinology, 171(3), 557-564.

Manuel, D.G., Leung, M., Nguyen, K., Tanuseputro, P., & Johansen, H. (2003). Burden

of cardiovascular disease in Canada. Canadian Journal of Cardiology, 19, 997-

1004.

McNeill, A. M., Rosamond, W. D., Girman, C. J., Golden, S. H., Schmidt, M. I., & East,

H. E. (2005). The metabolic syndrome and 11-year risk of incident cardiovascular

disease in the atherosclerosis risk in communities study Diabetes Care, 28(2),

385.

Nakamura, K., Shimai, S., Kikuchi, S., Takahashi, H., Tanaka, M., Nakano, S., et al.

(1998). Increases in body mass index and waist circumference as outcomes of

working overtime. Occupational Medicine, 48(3), 169-173.

Nakamura, K., Shimai, S., Kikuchi, S., Tominaga, K., Takahashi, H., Tanaka, M., et al.

(1997). Shift work and risk factors for coronary heart disease in Japanese blue-

collar workers: Serum lipids and anthropometric characteristics. Occupational

Medicine, 47(3), 142-146.

Nakanishi, N., Nishina, K., Yoshida, H., Matsuo, Y., Nagano, K., Nakamura, et al.

(2001). Hours of work and the risk of developing impaired fasting glucose or type

2 diabetes mellitus in Japanese male office workers. Journal of Occupational and

Environmental Medicine, 58, 569-574.

Nanchahal, K., Ashton, W.D., & Wood, D.A. (2000). Alcohol consumption, metabolic

cardiovascular risk factors and hypertension in women. International Journal of

Epidemiology, 29, 57-64.

Park, J., Kim, Y., Cho, Y., Woo, K., Chung, H., Iwasaki, K., et al. (2001). Regular

overtime and cardiovascular functions. Industrial Health, 39, 244-249.

Resnick, H. E., Jones, K., Ruotolo, G., Jain, A. K., Henderson, J., Lu, W., et al. (2003).

Insulin resistance, the metabolic syndrome, and risk of incident cardiovascular

disease in nondiabetic American Indians. Diabetes Care, 26(3), 861.

Rogers, A. E., Hwang, W., Scott, L. D., Aiken, L. H., & Dinges, D. F. (2004). The

working hours of hospital staff nurses and patient safety. Health Affairs, 23(4),

202-212.

Rutter, M. K., Meigs, J. B., Sullivan, L. M., D'Agostino, R. B. S., & Wilson, P. W. F.

(2004). C-reactive protein, the metabolic syndrome, and prediction of

cardiovascular events in the framingham offspring study. Circulation, 110(4),

380-385.

Sattar, N., Gaw, A., Scherbakova, O., Ford, I., O’Reilly, D.S., Haffner, S.M., et al.

(2003). Metabolic syndrome with and without C-reactive protein as a predictor of

coronary heart disease and diabetes in the West of Scotland Prevention Study.

Circulation, 108, 414-419.

Shields, M. (1999). Long working hours and health. Health Reports, 11(2), 33-48.

Sookoian, S., Gemma, C., Gianotti, T., Burgueno, A., Alvarez, A., Gonzalez, C., et al.

(2007). Effects of rotating shift work on biomarkers of metabolic syndrome and

inflammation. Journal of Internal Medicine, 261, 285-292.

Steptoe, A., Cropley, M., & Joekes, K. (1999). Job strain, blood pressure and response to

uncontrollable stress. Journal of Hypertension, 17, 193-200.

Steptoe, A., Siegrist, J., Kirschbaum, C., & Marmot, M. (2004). Effort-reward imbalance,

overcommitment, and measures of cortisol and blood pressure over the working

day. Psychosomatic Medicine, 66, 323-329.

Suwazono, Y., Dochi, M., Sakata, K., Okubo, Y., Oishi, M., Tanaka, K., et al. (2008).

Shift work is a risk factor for increased blood pressure in Japanese men: a 14-year

historical cohort study. Hypertension, 52, 581-586.

Tanuseputro, P., Manuel, D.G., Leung, M., Nguyen, K., & Johansen, H. (2003). Risk

factors for cardiovascular disease in Canada. Canadian Journal of Cardiology, 19,

1249-1259.

Trinkoff, A.M., & Storr, C.L. (1998). Work schedule characteristics and substance use in

nurses. American Journal of Industrial Medicine, 34, 266-271. van Amelsvoort, L,G., Jansen, N.W., & Kant, I. (2006). Smoking among shift workers:

more than a confounding factor. Chronobiology International, 23(6), 1105-1113. van Amelsvoort, L.G., Schouten, E.G., & Kok, F.J. (2004). Impact of one year of shift

work on cardiovascular disease risk factors. Journal of Occupational and

Environmental Medicine, 46(7), 699-706.

Vrijkotte, T.G., Van Doornen, L.J., & de Geus, E.J. (1999). Work stress and metabolic

and hemostatic risk factors. Psychosomatic Medicine, 61, 796-805.

Vrijkotte, T.G., Van Doornen, L.J., & de Geus, E.J. (2000). Effects of work stress on

ambulatory blood pressure, heart rate, and heart rate variability. Hypertension, 35,

880-886.

Wada, K., Katoh, N., Aratake, Y., Furukawa, Y., Hayashi, T., Satoh, E., et al. (2006).

Effects of overtime work on blood pressure and body mass index in Japanese male

workers. Occupational Medicine, 56(8), 578-580.

Wardle, J., Steptoe, A., & Oliver, G. (2000). Stress, dietary restraint and food intake.

Journal of Psychosomatic Research, 48, 195-202.

Chapter 3

Manuscript 1

THE RELATIONSHIPS BETWEEN WORK PATTERNS AND MODIFIABLE LIFESTYLE BEHAVIOURS IN FEMALE HOSPITAL WORKERS Kirk, M., MSc (c.) 1., Janssen, I., PhD 3,2, Van Den Kerkhof, E., DrPH 1,4, Tranmer, J., PhD1,2

1 School of Nursing, Queen’s University 2 Department of Community Health and Epidemiology 3 School of Kinesiology and Health Studies 4 Department of Anesthesiology Queen’s University, Kingston

Contact Author: Megan Kirk, RN, BScN., MSc (student) c/o J Tranmer, RN, PhD. Associate Professor School of Nursing Queen's University Cataraqui Building, 92 Barrie Street Kingston, Ontario K7L 3N6 613-549-6666 Ext. 4952

3.1 Abstract Objectives: The purpose of this study was to determine the relationships between work pattern characteristics and modifiable lifestyle behaviours, specifically, physical activity, smoking, alcohol and dietary intake, in female hospital workers.

Methods: Participants were female hospital workers (n = 466) from 2 hospital sites in

Southeastern Ontario. Lifestyle behaviour data were obtained through self-report using the International Physical Activity Questionnaire (IPAQ); a dietary pattern assessment tool based on Canadian Food Guidelines, and selected questions in regard to smoking and alcohol use. Work pattern data were collected through self report and hospital administrative work data. Irregular work patterns were defined as: extended 12 hour shifts, shift work and working overtime. Poor lifestyle behaviours were defined as: physical inactivity, an unbalanced diet, excessive alcohol consumption and cigarette use.

Results: Overall, the prevalence of smoking and excess alcohol intake was low. However

93.8% (n = 437) reported that they did not usually meet recommended Canadian guidelines for whole grain, fruit and vegetable intake, and only 53.6 % (n =250) met recommended levels of health-enhancing (high) physical activity. Multivariate analysis findings showed that participants working 12 hrs extended shifts were more likely to smoke ( OR 2.5, 95% CI, 1.01, 6.16); working overtime was associated with meeting recommendations for health-enhancing (high) physical activity levels (OR 1.7, 95% CI,

1.04, 2.88); and working part-time was associated with less leisure-time activity (OR

0.46, 95% CI, 0.28, 0.77). No significant associations were found between work patterns and excess alcohol and dietary patterns.

Conclusions: The findings from this study suggest that different work pattern characteristics are not significantly associated with unhealthy lifestyle behaviours, with a 38 few key exceptions. Women working overtime or extended shifts may be more likely to smoke and be physically inactive. Overall, for all employees, poor dietary patterns and low levels of physical activity are concerning. Further research with objective measures of lifestyle behaviours and a larger sample is warranted; however, the findings do support the need for workplaces to engage in activities aimed to promote health and prevent cardiovascular disease.

Key Words: smoking, alcohol, diet, exercise, shift work, shift length, overtime.

3.2 Introduction The prevalence and burden of cardiovascular disease (CVD) is a concern. In

Canada, CVD is the leading cause of morbidity and mortality in both men and women

(Heart and Stroke Foundation, 2003; Manuel, Leung, Nguyen, Tanuseputro & Johansen,

2003). CVD related deaths are now equal in Canadian men and women, although as women live longer, rates in women are likely to increase beyond that of men (Heart and

Stroke Foundation, 2007). While CVD events occur later in a woman’s life, the modifiable behaviours which influence CVD risk occur earlier during adult years. There is strong evidence linking modifiable behaviours, such as physical inactivity and smoking to CVD (Health Canada, 1999), but the influence of contextual factors such as the work environment on these risk factors is poorly understood as available findings are limited and inconsistent. Presently, no studies have examined the independent and combined effects of work patterns on lifestyle behaviours in Canadian female workers.

Irregular work patterns are an important and prevalent characteristic of hospital work environments. Work patterns consisting of long working hours (i.e., extended 12 hours shifts), shift work, and overtime may influence cardiovascular health through a number of behavioural pathways (Knutsson & Boggild, 2000). Heart healthy lifestyle behaviours such as the ability to maintain physical activity, consume a balanced diet, avoidance of smoking and a moderate alcohol intake may be impacted by different types of work patterns.

Most of our understanding about the impact of work patterns has emerged from studies exploring the impact of shift work, one work pattern characteristic, on health (van

Amelsvoort, Schouten & Kok, 2004; Knutsson & Boggild, 2000). The available literature exploring work patterns and CVD has predominantly focused on male workers, although 40 one study suggests that men and women shift workers are at similar risk for CVD (van

Amelsvoort, et al. 2004). The Nurses’ Health Study (NHS), a large prospective epidemiological study (n=79109) found that longer durations of shift work were associated with a higher age-adjusted prevalence of physical activity level and smoking

(Kawachi, et al. 1995). The associations between shift work and alcohol intake are inconsistent, with some studies reporting an association (van Amelsvoort, et al. 2004;

Suwazono, et al. 2008; Sookoian, et al. 2007), and others not (Burch, et al. 2009).

Irregular work patterns may also negatively influence health through changes in dietary behaviours. A longitudinal study exploring cardiovascular risk in day and shift workers (n=25) found that shift workers significantly reduced their intake of dietary fiber

(p<0.03) through a decreased intake of vegetables and potatoes over the 6 month follow- up (Knutson, Anderson & Berglund, 1990). In a review, Costa (1996) explains that even though calorie intake often remains unaltered, the quality of food consumed by shift workers frequently changes. Researchers have reported changes in dietary intake in shift workers (Knutson, et al. 1990; Costa, 1996; Nakamura, et al. 1998). Metabolic disruptions may play a role in increased CVD risk (Lund, Arendt, Hampton, English &

Morgan, 2001; Scheer, Hilton, Mantzoros & Shea, 2009; Ribeiro, Hampton, Morgan,

Deacon & Arendt, 1998). Therefore, the time of intake of different nutrients may be as important as the nutritional content (Knutsson & Boggild, 2000).

While shift work is an important work pattern characteristic, other work patterns such as extended shifts and overtime may also influence the ability for women to engage in healthy behaviours in a similar manner. Therefore, the purpose of this study was to determine the relationships between work patterns (overtime, shift length, rotation and

41 total hours worked) and modifiable lifestyle factors (smoking, alcohol, physical activity and dietary intake) in female hospital workers.

3.3 Study Population and Methods

3.3.1 Study Population

This cross-sectional study was conducted at two university affiliated acute care teaching hospital sites in South Eastern Ontario between 2007 and 2008. Data were obtained from a larger cohort study, designed to determine the relationships between work and home environments and cardiovascular risk in female hospital workers. Ethics approval was obtained through the Queen’s University Health Sciences Research Ethics

Board.

Female hospital employees who were employed for at least one year prior to the beginning of the study were invited to participate through local hospital newsletters and postings. Interested participants provided consent, met with the study coordinator and completed an interview and questionnaire. As well, participants consented to having the investigators obtain a detailed human resources administrative report of worked and unworked hours for the year prior to study enrollment.

Figure 3 outlines the number of participants consented (n=547), and those who had complete self-report data (n=509) and administrative data (n=466). This study reported on those participants which had complete administrative data and who worked more than 390 hrs in the past year (i.e., at least one shift per week).

3.3.2 Study Measures

3.3.3 Exposure Variables: Work Patterns

<1600 hrs/years and >1600 hrs/year. We categorized overtime as a Yes (any amount) or

No variable. Participants documented their usual shift length, rotation, work status

(full/part-time), and position. Response categories were dichotomized for: a) shift length

(8 hr vs. other, which consisted of 12 hour and mixed shift lengths), and b) Rotation

(regular days vs. rotational, which consisted of mixed, night, and evening shifts).

3.3.4 Outcome Variables: Lifestyle Behaviours

Physical Activity: Physical activity levels were assessed with the long version of the International Physical Activity Questionnaire (IPAQ) (International Physical Activity

Questionnaire [IPAQ], 2005). The IPAQ collects detailed information on domains of household, occupational, active transportation and leisure-time activity. Participants reported the minutes per week within each activity category which was used to estimate total weekly physical activity by calculating the total metabolic equivalent (MET) energy expenditure using IPAQ data processing guidelines (IPAQ, 2005). We focused on two physical activity outcomes: total physical activity and leisure-time physical activity. For total physical activity, participants were categorized into three levels of physical activity which are based on the energy expended in activities involved in all four domains. Level

1, the lowest physical activity describes persons with no physical activity or those who do not meet requirements of level 2 or 3. Level 2, moderate physical activity, describes persons who have 3 or more days/week of vigorous activity for at least 20 minutes/day, or at least 5 days/week of moderate intensity for at least 30 minutes/day and/or at least 5 days/week any intensity achieving a minimum of 600 MET-minutes/week. Level 3, high physical activity describes persons who have vigorous exercise at least 3 days/week and

1500 MET-minutes/week or 7 days/week of any combination of intensities, accumulating at least 3000 MET-minutes/week. Calculation of IPAQ scores followed the IPAQ procedural guidelines (IPAQ, 2005). Leisure-time activity was based on energy expended in both the active transportation and leisure-time activity domains, and reflected the amount of potentially modifiable activity levels for participants. It is recommended that adults expend 500-1000 METS-minutes/week in physical activity to receive substantial health benefits (U.S. Department of Health & Human Services, 2008). Leisure activity was dichotomized at 500 METS-minutes/week. Thus, <500 METS- minutes/week of activity did not meet recommendations and >500 METS-minutes/week met recommendations. The IPAQ demonstrates good reliability correlations (median of

.80) and validity correlations of .30 across different populations and is a reliable tool for measuring population levels of physical activity (Craig, et al. 2003).

Dietary Patterns: Participants completed the Rapid Eating and Activity

Assessment for Patients (REAP) (Pearson, et al. 2001). The REAP was developed to rapidly assess diet and physical activity (Gans, et al. 2003). The tool contains questions assessing intake frequency of various foods, snacking, condiment use and questions related to food preparation. The REAP questionnaire was modified to assess dietary

44 patterns in correspondence with Canadian Food Guidelines (CFG) (Gans, et al. 2003;

Gans, et al. 2006). To assess dietary patterns, we determined those who met the recommended daily food group (whole grains, fruit, vegetables, dairy, meat/protein) servings. Dietary patterns were rated with a 3 category response option (usually, sometimes, or rarely met selected guideline). The response ‘usually’ was considered to have met the guideline. Meat and dairy were excluded as the questionnaire did not measure extremes in consumption, which could have negative cardiovascular results.

Whole grain, fruit, and vegetable categories were analyzed. Because so few participants met all 3 guidelines (6.9%), a cut-off of > 2 or more (17.0%) food groups was created for the analysis.

Smoking and Alcohol Patterns: A general health questionnaire was administered to participants to determine smoking status (current, former, never) and frequency of excess alcohol intake (>5 drinks on one occasion; response options of never, monthly, weekly, and > once a week). Current smokers were compared to former and non- smokers. Excess Alcohol was also dichotomized as participants who never consumed >5 drinks on one occasion were compared to all other patterns of excess alcohol consumption. In addition, general demographic data such as age, marital status, income and education were also collected.

3.4 Statistical Analysis

Data were analyzed using version 16.0 of Statistical Package from Social Science

(SPSS). Descriptive statistics were used to describe the sample characteristics. Bivariate analyses (chi-square test and independent sample t-tests) were used to evaluate the associations between the demographic, work pattern and lifestyle data. Logistic

45 regression analysis was completed to determine relationships between exposure and outcome variables, while controlling for potentially significant confounders (i.e. age).

Variables that met p < 0.15 in the bivariate analysis were entered into the multivariate logistic regression models. The 0.05 level of statistical significance was accepted.

3.5 Results

Demographic and work pattern characteristics of our sample of female hospital workers (n=466) are described in Table 1. The age range of this female cohort was 22 to

67 years with an average age of 45.6 years. The average length in a current position was

9.5 years. Approximately 45% of the participants provided direct patient care within the hospital setting, while the remaining worked in non-direct care, in such positions as managers, coordinators and/or administrators.

Lifestyle behaviours are described in Table 2. Only 6.4% of the participants were current smokers, and 60.3% did not consume alcohol in excess at any time in the past year. Overall, dietary patterns were poor as approximately 70% of the sample met one or none of the recommended food guidelines for whole grains, fruits and vegetables.

Bivariate and multivariate analyses are presented in Tables 3-4. Bivariate analyses illustrated that participants with higher education were less likely to smoke

(OR=0.31, 95% CI, 0.11, 0.91). The multivariate analysis showed that participants working extended 12 hour shifts were more likely to smoke (OR=2.5, 95% CI, 1.01,

6.16), when controlling for age and education. Bivariate analyses found associations between extended shifts and excess alcohol intake (OR=1.63, 95% CI, 1.09, 2.42) and younger workers (OR=0.94, 95% CI, 0.92, 0.96). When controlling for age within the multivariate analyses, no work pattern variables were associated with excess alcohol

46 intake in that. In regards to total physical activity, the bivariate analysis revealed that female workers with higher education (OR=0.64, 95% CI, 0.41, 1.00) were less likely to meet recommended guidelines. Rotational shift work (OR=1.69, 95% CI, 1.04, 2.74) and working overtime (OR=1.88, 95% CI, 1.22, 2.89) were significantly associated with achieving recommended total physical activity guidelines. When controlling for age, female employees working overtime were more likely to reach recommended total physical activity levels (OR=1.7, 95% OR, 1.04, 2.88). The factors associated with leisure-time activity levels were seemingly different. Female participants working part- time were less likely to achieve recommended levels of leisure-time activity in the bivariate (OR=0.46, 95% CI, 0.28, 0.77) and multivariate analyses (OR=0.50, 95% CI,

0.27, 0.92). Full-time workers consisted of 70.3% day workers (p<0.001) and 74.1% of 8 hour shift length workers (p<0.001). When controlling for age, no work pattern variables were significantly associated with leisure-time activity. There were no significant associations for dietary patterns (data not shown). Thus, dietary patterns do not appear to be influenced by varying work patterns in this study.

3.6 Discussion

Overall, few female employees in this sample were smokers or consumed excess alcohol However, dietary patterns appeared poor as almost 70% of the sample only met one or none of the recommended daily food group guidelines, indicating an unbalanced diet. As well, just over half of the sample met the recommended health-enhancing physical activity guidelines. Thus generally, as a whole, women seemed to have more difficulty mainting a healthy diet and physical activity levels, The influence of work

47 pattern characteristics on women’s ability to engage in healthy behaviours appears complex and multifaceted.

In general, findings suggest that work patterns do not influence lifestyle behaviours of female workers, with a few key exceptions. Working extended shifts was associated with increased rates of smoking, a finding that is consistent with other research linking shift work, overtime and extended shifts to increased rates of smoking (Kawachi, et al. 1995; Trinkoff, & Storr, 1998; Kroenke, et al. 2007; Di Lorenzo, et al. 2003;

Suwazono, et al. 2008; van Amelsvoort, et al. 2004, Knutsson, Akerstedt, & Jonsson,

1988). Smoking is commonly viewed as a coping mechanism for those working extended shifts in stressful work environments in that “smoke breaks” may relax workers and nicotine may stimulate workers during the night. When controlling for age, the influence of extended 12 hours shifts became insignificant. Our bivariate findings do suggest an association, although our study may not have had adequate power to detect associations within the multivariate regression model. Available literature supports the association between participants working irregular work patterns and increased alcohol intake (van

Amelsvoort, et al. 2004; Suwazono, et al. 2008; Sookoian, et al. 2007). Again, alcohol consumption could be a coping method for females in positions requiring extended shifts, often direct-care providers. Although, a recent study found no difference in alcohol consumption when comparing shift and day workers, they did not specify whether excess alcohol intake was considered (Burch, Tom, Zhai, Criswell, Leo & Ogoussan, 2009).

Higher levels of total physical activity were found in women who worked rotational shift work and overtime hours. Once we controlled for age, only overtime hours were significantly associated with total physical activity. In our study, overall

48 physical activity was a composite score based on the individual’s perception of the amount of time and intensity of physical activity in work, household, active transportation and leisure-time activities. Women who work irregular work patterns were more likely to be direct care workers either as nurses, environmental assistants, porters and so forth. It makes sense for these women perceived their work as more strenuous and demanding, although it is likely of a low intensity physical activity. Recommendations indicate moderate-to-vigorous intensity physical activity, 20- 30 minutes daily, can positively impact cardiovascular health and reduce risk of premature death (U.S. Department of

Health & Human Services, 2008). It is unclear, however, whether low-intensity physical activity accumulated in the health care working environment has protective effects on female cardiovascular health (Kodama, Miao, Yamada & Sone, 2006).

No work pattern variables were significantly associated with leisure-time activity.

Work status was the only variable associated with leisure-time activity, within the bivariate and multivariate analyses. Women who worked full-time reported higher leisure-time activity levels. Full time workers were more likely to be 8 hour day workers with sedentary positions providing non-direct care (i.e., desk work). These findings are consistent with a recent review conducted by Atkinson, Fullick, Grindey, & Maclaren

(2008) which reported that shift work generally decreases opportunities and participation in physical activity. Working regular and predictable shifts may allow working women to incorporate leisure time exercise into their normal schedule, whereas, this seems to be more problematic for women working irregular shifts. Furthermore, responses to activity are potentially altered if the exercise is at an unusual time of day and/or if the shift worker

49 is sleep deprived which may impact the longer-term adherence to an activity (Atkinson, et al. 2008).

One of the major limitations in studies exploring the relationships between physical activity behaviours and CVD risk is the reliance on self-report measures.

Measurements of physical activity using accelerometers or another objective measure would have enhanced the validity of our physical assessment measure. Our study has several other limitations. Research exploring irregular work patterns need to consider potential bias due to selection out of the study. This drop-out is of particular importance as it may lead to under-estimations of associations. This bias applies to those who drop out of irregular work because they are less healthy than others, leaving the healthiest workers within the irregular work cohort and unhealthiest in day work. This is known as the ‘healthy worker effect’ (Karlsson, Knutsson & Lindahl, 2001; Kivimaki, Kuisma,

Virtanen & Elovaino, 2001). Often, poor health is associated with poor health habits.

Thus, the healthy worker effect may mask the long-term effects related to irregular shift schedules, implying that poor outcomes in shift workers are likely to underestimate of the problem if it exists. Also, this cross-sectional study was conducted at two acute care teaching hospitals and participants volunteered for study participation. Thus, causality cannot be determined, biases may have existed that would have resulted in an underestimation of the associations, and the findings may not be generalizable to other work place settings.

To the best of our knowledge, this is the first study performed in Canada examining the influence of irregular work patterns in female workers. Female hospital employees working in acute care settings are exposed to highly demanding work

50 environments. Our findings suggest that work patterns are generally not influencing the lifestyle behaviours of female workers. Although, the few exceptions exist, for instance extended shifts linked with increased smoking and overtime hours linked with higher total physical activity. Opportunities exist in the workplace to promote health by promoting and facilitating healthy lifestyle behaviours. Further research is required to better elucidate the pathways and mechanisms linking work environment characteristics to unhealthy behaviours, and to delineate inconsistent findings. In general, by improving our understanding of the influence of work environments on lifestyle behaviours we would be better positioned to improve the cardiovascular health of female workers.

Acknowledgements The authors gratefully acknowledge the women who participated in this study; the

Nursing Research Team who carried out the research study; Julia Nijboer for her assistance with data collection; The Health and Work Study investigators (Tranmer, J.,

McGillis-Hall, L, Katzmarzyk, P., Parry, M., Rivoire, E., Day, A., O’Callaghan, C.);

CIHR for funding the larger cohort study; and the Heart and Stroke Foundation of Canada

(Master’s Studentship Award) for their financial support for the first author.

3.7 References Al-Naimi, S., Hampton, S. M., Richard, P., Tzung, C., & Morgan, L.M. (2004).

Postprandial Metabolic Profiles Following Meals and Snacks Eaten during

Simulated Night and Day Shift Work. Chronobiology International, 21(6), 937 –

947.

Atkinson, G., Fullick, S., Grindey, C., & Maclaren, D. (2008). Exercise, energy balance

and the shift worker. Sports Medicine, 38(8), 671-685.

Boggild, H., Suadicani, P., Ole Hein, H., & Gyntelberg, F. (1999). Shift work, social

class, and ischemic heart disease in middle aged and elderly men; a 22 year follow

up in the Copenhagen male study. Occupational and Environmental Medicine,

56(9), 640-645.

Burch, J. B., Tom, J., Zhai, Y., Criswell, L., Leo, E., & Ogoussan, K. (2009). Shiftwork

impacts and adaptation among health care workers. Occupational Medicine, 59(3),

159-166.

Costa, G. (1996). The impact of shift and night work on health. Applied Ergonomics,

27(1), 9-16.

Craig, C.L., Marshall, A.L., Sjostrom, M., Bauman, A.E., Booth, M.L., Ainsworth, B.E.,

et al. (2003). International physical activity questionnaire: 12-country reliability

and validity. Medicine and Science in Sports and Exercise, 35(8), 1381-1395.

Di Lorenzo, L., De Pergola, G., Zocchetti, C., L’Abbate, N., Basso, A., Pannacciulli, N.,

et al. (2003). Effect of shift work on body mass index : results of a study

preformed in 319 glucose-tolerant men working in a Southern Italian industry.

International Journal of Obesity, 27, 1353-1358.

Executive Summary of The Third Report of The National Cholesterol Education Program

(NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood

Cholesterol in Adults (Adult Treatment Panel III). JAMA, 2001, 285, 2486-2497.

Gans, K.M., Ross, E., Barner, C.W., Wylie-Rosett, J., McMurray, J., & Eaton, C. (2003).

REAP and WAVE: new tools to rapidly assess/discuss nutrition with patients.

Journal of Nutrition, 133, 556-562.

Gans, K.M., Risics, P.M., Wylie-Rosett, J., Ross, E.M., Strolla, L.O., McMurray, J., et al.

(2006). Development and evaluation of the nutrition component of the Rapid

Eating and Activity Assessment for Patients (REAP): a new tool for primary care

providers. Journal of Nutrition Education and Behavior, 38(5), 286-292.

Ha, M., & Park, J. (2005). Shiftwork and metabolic risk factors of cardiovascular disease.

Journal of Occupational Health, 47(2), 89-95.

Harma, M. (2006). Workhours in relation to work stress, recovery and health.

Scandinavian Journal of Work, Environment and Health, 32(6), 502-514.

Heart and Stroke Foundation of Canada. (2003). The growing burden of heart disease and

stroke in Canada in 2003. Ottawa, Ontario. Retrieved October 22, 2008 from

http://www.cvdinfobase.ca/cvdbook/CVD_En03.pdf

Heart and Stroke Foundation of Canada. (2007). Report cards on health – time to bridge

the gender gap. Ottawa, Ontario. Retrieved December 7, 2008 from

http://ww2.heartandstroke.ca/Page.asp?PageID=33&ArticleID=5911&Src=news&

From

Institute for Community Health Promotion. (2005). Rapid Eating and Activity

Assessment for Patients. Brown University, Providence, RI. Retrieved February

21, 2009 from http://bms.brown.edu/nutrition/acrobat/REAP%206.pdf

International Physical Activity Questionnaire (IPAQ). (2005). Guidelines for data

processing and analysis of the international physical activity questionnaire – short

and long forms. Retrieved February 18, 2009 from

http://www.ipaq.ki.se/scoring.pdf

Janssen, I., Katzmarzyk, P.T., & Ross, R. (2002). Body mass index, waist circumference,

and health risk: evidence in support of current National Institutes of Health

guidelines. Archives of Internal Medicine, 162(18), 2074-2079.

Janssen, I., Katzmarzyk, P.T., & Ross, R. (2004). Waist circumference and not body mass

index explains obesity-related health risk. American Journal of Clinical Nutrition,

79 (3), 379-384.

Kawachi, I., Colditz, G., Stampfer, M., Willett, W., Manson, J., Speizer, F., et al. (1995).

Prospective study of shift work and risk of coronary heart disease in women.

Circulation, 92(11), 3178-3182.

Karlsson, B., Knutsson, A., & Lindahl, B. (2001). Is there an association between shift

work and having a metabolic syndrome? Results from a population based study of

27 485 people. Occupational and Environmental Medicine, 58(11), 747-752.

Kivimaki, M., Kuisma, P., Virtanen, M., & Elovaino, M. (2001). Does shift work lead to

poorer health habits? A comparison between women who had always done shift

work with those who had never done shift work. Work & Stress, 15(1), 3-13.

Knutsson, A., Anderson, H., & Berglund, U. (1990). Serum lipoproteins in day and shift

workers: a prospective study. British Journal of Industrial Medicine, 47(2), 132-

134.

Knutsson, A., Akerstedt, T., & Jonsson, B. (1988). Prevalence of risk factors for coronary

artery disease among day and shift workers. Scandinavian Journal of Work,

Environment and Health, 14, 317-321.

Knutsson, A., & Boggild, H. (2000). Shift work and cardiovascular disease: review of

disease mechanisms. Reviews on Environmental Health, 15(4), 359-372.

Kodama, S., Miao, S., Yamada, N., & Sone, H. (2006). Exercise training for ameliorating

cardiovascular risk factors – focusing on exercise intensity and amount.

International Journal of Sport and Health Science, 4, 325-338.

Koller, M. (1983). Health risks related to shift work. International Archives of

Occupational and Environmental Health, 53, 59-75.

Kroenke, C., Spiegelman, D., Manson, J., Schernhammer, E., Colditz, G., & Kawachi, I.

(2007). Work characteristics and incidence of type 2 diabetes in women.

American Journal of Epidemiology, 165(2), 175-183.

Lund, J., Arendt, J., Hampton, S., English, J., & Morgan, L. (2001). Postprandial

hormone and metabolic responses amongst shift workers in antarctica. Journal of

Endocrinology, 171(3), 557-564.

Manuel, D.G., Leung, M., Nguyen, K., Tanuseputro, P., & Johansen, H. (2003). Burden

of cardiovascular disease in Canada. Canadian Journal of Cardiology, 19(9), 997-

1004.

Nakamura, K., Shimai, S., Kikuchi, S., Takahashi, H., Tanaka, M., Nakano, S., et al.

(1998). Increases in body mass index and waist circumference as outcomes of

working overtime. Occupational Medicine, 48(3), 169-173.

Nanchahal, K., Ashton, W., & Wood, D. Alcohol consumption, metabolic cardiovascular

risk factors and hypertension in women. International Journal of Epidemiology,

29(1), 57-64.

Pearson, T.A., Stone, E.J., Grundy, S.M., McBride, P.E., Van Horn, L., & Tobin, B.W.

(2001). NAA Collaborative Group. Translation of nutritional sciences into

medical education: the Nutrition Academic Award Program. .American Journal of

Clinical Nutrition, 74(2), 164-170.

Ribeiro, D., Hampton, S., Morgan, L., Deacon, S., & Arendt, J. (1998). Altered

postprandial hormone and metabolic responses in a simulated shift work

environment. Journal of Endocrinology, 158(3), 305-310.

Scheer, F. A. J. L., Hilton, M. F., Mantzoros, C. S., & Shea, S. A. (2009). Adverse

metabolic and cardiovascular consequences of circadian misalignment. Proceedings

of the National Academy of Sciences, 106(11), 4453-4458.

Sookoian, S., Gemma, C., Gianotti, T., Burgueno, A., Alvarez, A., Gonzalez, C., et al.

(2007). Effects of rotating shift work on biomarkers of metabolic syndrome and

inflammation. Journal of Internal Medicine, 261, 285-292.

Suwazono, Y., Dochi, M., Sakata, K., Okubo, Y., Oishi, M., Tanaka, K., et al. (2008).

Shift work is a risk factor for increased blood pressure in japanese men: A 14-year

historical cohort study. Hypertension, 52(3), 581-586.

Trinkoff, A.M., & Storr, C.L. (1998). Work schedule characteristics and substance use in

nurses. American Journal of Industrial Medicine, 34, 266-271.

U.S. Department of Health & Human Services. (2008). Physical activity guidelines for

Americans. Appendix 1: Translating scientific evidence about total amount and

intensity of physical activity into guidelines. Washington, DC. Retrieved July 26,

2009 from http://www.health.gov/paguidelines/guidelines/Appendix1.aspx

Van Amelsvoort, L., Schouten, E., & Kok, F. (2004). Impact of one year of shift work on

cardiovascular disease risk factors. Journal of Occupational and Environmental

Medicine, 46(7), 699-706.

Figure 3

547 Female Hospital Workers Enrolled (322 day and 187 shift workers)

509 466 Completed Self-Report Data Human Resources Data (38 incomplete data / withdrew) (31 missing HR data, 14 total hours <390hrs/year)

Figure 3 - The number of participants who consented (n=547), complete self-report data (n=509) and complete administrative data (n=466).

Table 1. Descriptive Characteristics of Sample (N=466) Characteristics Mean (SD) Demographics N (%) Marital Status Married 360 (77.3) Single 106 (22.7) Household Income($) <75 000 219 (47.0) >75 000 237 (50.9) Education < Diploma 313 (67.1) > Degree 152 (32.6) Work Status Full time (FT) 343 (73.6) Part time (PT) 123 (26.4) Work Domain Non-direct care 262 (56.2) Direct care 204 (43.8) Work Patterns Shift Length 8 hour 319 (68.5) 12 hour 81 (17.4) Other 66 (14.2) Rotation Day 300 (64.4) Rotational 166 (35.6) Overtime Hours None 199 (42.7) < 40 hours/year 130 (27.9) > 40 hours/year 137 (29.4) Total Hours Worked <1600 hours/year 233 (50.0) >1600 hours/year 233 (50.0)

Percentages may not add to 100% due to missing data. Work Domain: Direct care: an employee which provides direct patient care. Non-direct care: an employee which does not provide direct patient care. Rotation: Day: an employee working during day hours, Monday to Friday, with a set weekly schedule. Rotational: an employee working rotating work hours, 7 days a week. Shift Length: Other: any length other than 8.

Table 2. Description of lifestyle behaviours in female hospital workers (N=466) Lifestyle Behaviours N (%) Smoking Never 295 (63.3) Current 30 (6.4) Former 140 (30.0) Excess Alcohol (>5 drinks/day) Never 281 (60.3) 2 79 (14.4)

Percentages may not add to 100% due to missing data. Leisure Activity Guidelines are based on the Physical Activity Guidelines for Americans. Guidelines recommend 500-1000 METS- minutes/week of exercise. CFG – Canadian Food Guidelines. Total Physical Activity Guidelines, see text for classification of each level.

Table 3. Associations between personal and work patterns characteristics with smoking and excess alcohol intake (N=466) Characteristics Smoking Excess Alcohol Bivariate OR Multivariate OR Bivariate Multivariate OR (95% CI) (95% CI) OR(95% CI) (95% CI) Age (years) 0.98 (0.94, 1.02) 0.97 (0.93, 1.01) 0.94 (0.92, 0.96) 0.94 (0.92, 0.96) Marital Status Married 1.00 - 1.00 - Single 0.85 (0.34, 2.13) 1.25 (0.81, 1.95) Household Income($) <75 000 1.00 - 1.00 - >75 000 0.46 (0.21, 1.02) 0.78 (0.54, 1.14) Work Status Full time 1.00 - 1.00 - Part time 1.01 (0.44, 2.34) 0.93 (0.61, 1.41) Education < Diploma 1.00 1.00 1.00 - > Degree 0.31 (0.11, 0.91) 0.24 (0.08, 0.72) 0.84 (0.56, 1.25) Work Domain Non-direct care 1.00 - 1.00 - Direct Care 0.73 (0.34, 1.56) 1.30 (0.90, 1.90) Work Patterns Shift length 8 hrs 1.00 1.00 1.00 1.00 Other (12hrs) 1.99 (0.94, 4.19) 2.50 (1.01, 6.16) 1.63 (1.09, 2.42) 1.46 (0.91, 2.36) Rotation Day 1.00 1.00 1.00 1.00 Rotational 1.22 (0.57, 2.59) 0.65 (0.23, 1.83) 1.23 (0.84, 1.81) 0.91 (0.53, 1.54) Overtime Hours No 1.00 1.00 1.00 1.00 Yes 1.12 (0.53, 2.38) 0.84 (0.32, 2.18) 1.05 (0.72, 1.53) 0.88 (0.54, 1.42) Total Hours Worked <1600 hrs/year 1.00 1.00 1.00 1.00 >1600 hrs/year 0.65 (0.31, 1.39) 0.64 (0.28, 1.44) 0.81 (0.56, 1.18) 0.90 (0.60, 1.35) The reference group is the first category listed under each variable.

62 Table 4. Associations between personal and work pattern characteristics with total and leisure-time physical activity (N=466) Characteristics Total Physical Activity Leisure-Time Physical Activity Bivariate OR Multivariate OR Bivariate Multivariate OR (95% CI) (95% CI) OR(95% CI) (95% CI) Age (years) 0.99 (0.97, 1.02) 0.99 (0.97, 1.02) 0.99 (0.96, 1.01) 0.98 (0.96, 1.01) Marital Status Married 1.00 - 1.00 - Single 1.18 (0.71, 1.95) 1.45 (0.81, 2.58) Household Income($) <75 000 1.00 - 1.00 - >75 000 0.71 (0.46, 1.10) 1.29 (0.81, 2.06) Education < Diploma 1.00 - 1.00 - > Degree 0.64 (0.41, 1.00) 1.49 (0.90, 2.47) Work Status Full time 1.00 - 1.00 1.00 Part time 0.88 (0.55, 1.42) 0.46 (0.28, 0.77) 0.50 (0.27, 0.92) Work Domain Non-direct care 1.00 - 1.00 Direct Care 1.50 (0.96, 2.34) 0.82 (0.51, 1.32) - Work Patterns Shift length 8 hrs 1.00 1.00 1.00 1.00 Other (12hrs) 1.24 (0.76, 2.03) 0.94 (0.55, 1.63) 0.66 (0.40, 1.10) 0.79 (0.44, 1.41) Rotation Day 1.00 1.00 1.00 1.00 Rotational 1.69 (1.04, 2.74) 1.31 (0.72, 2.39) 0.75 (0.46, 1.23) 1.05 (0.56, 1.96) Overtime Hours No 1.00 1.00 1.00 1.00 Yes 1.88 (1.22, 2.89) 1.73 (1.04, 2.88) 0.73 (0.46, 1.17) 0.87 (0.50, 1.52) Total Hours Worked <1600 hrs/year 1.00 1.00 1.00 1.00 >1600 hrs/year 0.96 (0.63, 1.47) 1.20 (0.76, 1.89) 1.49 (0.93, 2.38) 1.03 (0.58, 1.83) The reference group is the first category listed under each variable.

63 Chapter 4

Manuscript 2

THE INFLUENCE OF WORK PATTERNS ON CARDIOVASCULAR RISK FACTORS IN FEMALE HOSPITAL WORKERS

Kirk, M., MSc (c.) 1., Janssen, I., PhD 3,2, Van Den Kerkhof, E., DrPH 1,4, Tranmer, J., PhD1,2

1School of Nursing, Queen’s University 2Department of Community Health and Epidemiology 3School of Kinesiology and Health Studies 4 Department of Anesthesiology Queen’s University, Kingston

Contact Author:

Megan Kirk, RN, BScN, MSc (student) c/o J Tranmer, RN, PhD. Associate Professor School of Nursing Queen's University Cataraqui Building, 92 Barrie Street Kingston, Ontario K7L 3N6 613-549-6666 Ext. 4952

64 4.1 Abstract Objectives: The purpose of this study was to determine the associations between work pattern characteristics and indicators of cardiovascular risk, specifically indicators of the metabolic syndrome, in female hospital workers.

Methods: Participants were female hospital workers (n = 466) from 2 hospital sites in

Southeastern Ontario. Cardiovascular risk data was obtained through anthropometric measurements, blood sampling and self-report. Work pattern data was collected through self report and linked with hospital administrative work data. Irregular work patterns were defined as: extended 12 hour shifts, shift work and overtime hours. Metabolic syndrome was classified in accordance with the NCEP ATP (III) guidelines.

Results: The prevalence of metabolic syndrome was 22.1% in this cohort of female hospital workers; elevated waist circumference (≥ 88 cm) was the most prevalent cardiovascular risk indicator (39.1%, n=182). Bivariate analyses showed that after 6 years of shift work, female workers were at increased risk of developing metabolic syndrome, compared to day workers and those with less than 6 years of shift work

(OR=1.88, 95% CI, 1.12, 3.17). Bivariate analyses revealed that shift work duration, after 6 years, was also associated with abdominal obesity (OR=2.01, 95% CI, 1.31, 3.11).

Multivariate analyses showed that female workers working extended shifts were at increased of elevated systolic blood pressure (OR=1.89, 95% CI, 1.00, 3.58), when controlling for age and menopausal status. No other work patterns were significantly associated with the metabolic syndrome in the multivariate analyses.

Conclusions: In general, work patterns do not appear to influence the development of cardiovascular risk factors in female workers, with a few key exceptions. For example, participants working extended shifts were at increased risk of developing elevated

65 systolic blood pressure. Given the prevalence, and the likely underestimation, of the metabolic syndrome and in particular abdominal obesity in this sample, suggests that

further research is warranted to determine mechanistic pathways. However, our findings

do support the need for healthy workplace policy aimed at optimizing the health of

female workers in Canada.

Key Words: metabolic syndrome, abdominal obesity, shift work, shift length, overtime.

66 4.2 Introduction

The number of cardiovascular disease (CVD) related deaths is currently equal in

Canadian men and women, (Heart and Stroke Foundation, 2007) although as women are living longer, deaths in women are projected to surpass that of men (Heart and Stroke

Foundation, 2007). While CVD occurs later in a woman’s life, the modifiable risk factors which influence the development of cardiovascular disease occur earlier during the adult working years. The metabolic syndrome represents a clustering of modifiable cardiovascular risk factors (i.e., high fasting blood glucose, abdominal obesity, elevated triglycerides, and low high density lipoprotein cholesterol) that are associated with increased cardiovascular events and death (Isomaa, et al. 2001; Lakka, et al. 2002; Sattar, et al. 2003). The prevalence of metabolic syndrome is increasing in North America

(Cornier, et al. 2008), with ranges between 12% and 25%, depending on the defining criteria and populations studied (Bosma, Stansfeld & Marmot, 1998).

Past research suggests that irregular work patterns, specifically shift work, overtime hours and extended shifts, may negatively impact the cardiovascular health of workers (Knutson & Boggild, 2000; Kawachi, et al. 1995; Karlsson, Knutsson, Lindahl &

Alfredsson, 2003; Kroenke, et al. 2007; Sookoain, et al. 2007; Biggi, Consonni,

Galluzzo, Sogliani & Costa, 2008; Beerman, & Nachreiner, 1995; Boggild, Suadicani,

Ole Hein, & Gyntelberg, 1999; Ha & Park, 2005). Shift work and exposure to night work is a common work pattern within the hospital setting which has been linked to increased cardiovascular disease risk in both men and women (Knutsson & Boggild,

2000; Kawachi, et al. 1995; Karlsson, et al. 2003). The Nurses Health Study (NHS), a large epidemiological study based in the United States followed a large cohort of female nurses who reported on work characteristics and cardiovascular risk factors (Kawachi, et

67 al. 1995). Analyses revealed that coronary heart disease (CHD) risk increased in women

after 6 or more years of shift work, (RR=1.51, 95% CI, 1.12, 2.03), compared to women

who had never done shift work before (Kawachi, et al. 1995). In addition, women were

found to be at increased risk of diabetes when working 41-60 hours/week (RR=1.57, 95%

CI, 1.26, 1.94) and >61 hours/week (RR=1.49, 95% CI, 0.85, 2.63), compared to women

working 21-40 hours/week when controlling for age (Kroenke, et al. 2007). The

influence of overtime hours on cardiovascular health is of concern as the continuous

shortage of workers within the Canadian health care system creates a demand for

overtime work.

The traditional 8 hour shifts for hospital nurses have changed over the past few decades to predominantly include extended 12 hour shift patterns (Rogers, Hwang, Scott,

Aiken & Dinges, 2004). Most studies exploring the health effects of shift work have

assessed the effects of 8 hour shifts on cardiovascular risk. We were unable to find a

study assessing the influence of 12 hours shifts on the cardiovascular health of female

workers. A recent cross-sectional study in male workers, explored the effects of rotating

shift work and extended shifts on metabolic syndrome risk factors (Sookoain, et al.

2007). Rotating shift workers (n=474) working extended hours were found to be at

greater risk of developing metabolic syndrome (OR=1.51, 95% CI, 1.01, 2.25, p<0.04)

compared to day workers working 8 hours (n=877), after adjustments (Sookoain, et al.

2007).

Hospitals employ a large number of women. Female nurses make up 94% of all

Canadian nurses (Statistics Canada, 2007), and nurses make up 80% of the health care workforce (Statistics Canada, 2005). Hospital work environments are characterized by extended shifts, rotational shift work and overtime hours. Thus, an opportunity exists to

68 enhance our understanding of the influence of work patterns on the health of female

hospital workers with aim of developing heart healthy initiatives to improve female

cardiovascular health.

The influence of contextual factors such as the work environment on CVD risk is

poorly understood as findings are unavailable or inconsistent. The main objective of this

study was to describe the associations between contemporary work pattern

characteristics, such as shift work and shift length, and indicators of cardiovascular risk,

such as the metabolic syndrome, in female hospital workers.

4.3 Study Population and Methods

4.3.1 Study Population

This cross-sectional study was conducted at two university affiliated teaching hospital sites in South Eastern Ontario between 2007 and 2008. Data were obtained from a larger cohort study designed to determine the relationships between work and home environments and cardiovascular risk in female hospital workers. Ethics approval was obtained through the Queen’s University Health Sciences Research Ethics Board.

Female hospital workers were included if employed for a year prior to the initiation of the study. Interested participants provided consent, met the study coordinator and completed a clinical examination and questionnaire. During the clinical

examination, anthropometric measurements were completed and instructions and the

requisition form were provided for the fasting blood work sample. General health and

work pattern data were collected through self-report in a questionnaire. As well, participants consented to having the investigators obtain a detailed human resources

69 administrative report of worked and unworked hours for the year prior to study

enrollment.

Figure 4 outlines the number of participants consented (n=547), and those for who

there was complete self-report data (n=509), administrative data (n=466) and work

history data (n=359). This study reports on the 359 participants who had complete administrative data, and work history data.

4.3.2 Study Measures

4.3.3 Exposure Variables: Work Patterns

Work pattern data were collected through self-report and human resource administrative records. Administrative records contained data on the total worked hours and the number of overtime hours worked for the 12 month period prior to study enrollment. Participants were included within the analyses if they worked more than 390 hours in the past year, i.e., one shift per week. We categorized total worked hours as a dichotomous variable at the median, <1600 hrs/years and >1600 hrs/year. We

categorized overtime as a Yes (any amount) or No variable. Participants documented

their usual shift length, rotation, work status (full/part-time), and position. Response

categories were dichotomized for: a) shift length (8 hr vs. other, which consisted of 12

hour and mixed shift lengths), and b) Rotation (regular days vs. rotational, which

consisted of mixed, night, and evening shifts). Participants were re-contacted in the later

stages of the study to collect work history data containing: total years in shift work and/or

total years in day work. Shift work duration was dichotomized at 6 years based on a

priori evidence, suggesting an increased risk of CVD after 6 years of shift work

(Kawachi, et al. 1995; Knuttson, Akerstedt, Jonsson, & Orth-Gomer, 1986).

70 4.3.4 Outcomes Variables: Cardiovascular Risk Factors

Cardiovascular risk was assessed using metabolic syndrome risk criteria. The prevalence of metabolic syndrome was determined using the US National Cholesterol

Education Program (NCEP) Adult Treatment Panel III (ATP III) criteria. Metabolic syndrome was present in female workers if three or more of the five following risk factors existed: abdominal obesity (waist circumference > 88 cm); high triglycerides (>

1.69 mmol/L); low high-density lipoprotein (< 1.29 mmol/L); high blood pressure

(systolic > 130 mmHg or diastolic > 85 mmHg); or high fasting blood glucose (> 6.1 mmol/L). Participants previously diagnosed with high blood pressure, diabetes and/or

heart disease and/or reported taking antihypertensive, hypoglycemic, and/or cholesterol-

lowering agents were classified as having the high blood pressure, high fasting blood

glucose, and high triglyceride components of the metabolic syndrome, respectively.

Metabolic syndrome was dichotomized into absent/present. Risk factors were

dichotomized using the above US NCEP Adult Treatment Panel III (ATP III) criteria as a

cut-off.

To obtain accurate measurements of waist circumference, blood pressure and

blood samples, three trained nurses conducted the clinical examinations. The trained

nurses measured the waist circumference of participants to the nearest cm using a flexible

anthropometric tape, measured midway between the margin of the lowest ribs and the

iliac crest, at the point of minimal inspiration. Trained nurses collected blood pressure

measurements using a BP Tru device. Blood pressure measurements were taken 3 times

during a study session beginning, after interview and at end of the session after

completion of questionnaire. The mean systolic and diastolic pressure was recorded.

Blood samples were collected at hospital sites and were analyzed using standardized

71 hospital procedures and equipment to determine lipid profiles (triglycerides, HDL

cholesterol), and fasting blood glucose. Participants fasted for 12-hours prior to the

blood draw, and samples were taken on the morning of day shifts.

4.3.5 Covariates

Age, marital status, family income, family history of CVD, menopause status and

hormone use were collected through self-report.

4.4 Statistical Analysis

Data were analyzed using version 16.0 of Statistical Package from Social Science

(SPSS). Descriptive statistics were used to describe the sample characteristics. Bivariate analyses (chi-square test and independent sample t-tests) were used to evaluate the associations between demographic, work pattern and metabolic syndrome data.

Multivariate logistic regression analyses were used to determine relationships between work pattern characteristics, the metabolic syndrome, and the metabolic syndrome components while controlling for potentially significant confounders (i.e. age, hormone use). Variables that met p < 0.15 were entered into the multivariate logistic regression

models. The 0.05 level of statistical significance was accepted.

4.5 Results

Demographic and work pattern characteristics of our sample of female hospital workers (n=466) are described in Table 5. Age ranged between 22 to 67 years, with an average age of 45.6 years. Approximately 45% of the participants provided direct patient care within the hospital setting, while the remaining worked in non-direct care, in such positions as managers, coordinators and/or administrators. The prevalence of cardiovascular risk factors is outlined in Table 6. The prevalence of elevated waist

72 circumference (> 88 cm) was 39.1% (n=182) and 32.2% were classified as having high blood pressure. The prevalence of the metabolic syndrome within this cohort was 22.1%

(n=103).

Bivariate and multivariate analyses are presented in Table 7 and 8. Increasing age were significantly associated with the metabolic syndrome in the bivariate (OR=1.07,

95% CI, 1.04, 1.10) and multivariate (OR=1.06, 95% CI, 1.01, 1.10) analyses. The bivariate analyses revealed that post-menopausal workers (OR=2.39, 95% CI, 1.53, 3.73) were more likely to have the metabolic syndrome, where as, female participants with higher education (OR=0.39, 95% CI, 0.23, 0.67) were less likely to have metabolic syndrome. Bivariate analyses revealed that years in shift work was significantly associated the metabolic syndrome (OR=1.88, 95% CI, 1.12, 3.17). Therefore, after 6 years of shift work, female workers were at increased risk of having metabolic syndrome, compared to day workers and those less than 6 years of shift work. Although, when controlling for age, menopausal status and education within the multivariate analyses, shift work duration lost significance. Results suggest trends between irregular work patterns linked and increased risk of the metabolic syndrome which may be clinically relevant. We were unable to detect statistical significance likely due to insufficient power in this study, and the strong collinearity between age and shift work duration.

Older female workers (OR=1.05, 95% CI, 1.03, 1.08) and those who were post- menopausal (OR=2.14, 95% CI, 1.46, 3.16) were more likely to be abdominally obese.

Female workers with higher education were less likely to have abdominal obesity. Also, female workers who had worked 6 or more years of shift work were more likely to have abdominal obesity (OR=2.01, 95% CI, 1.31, 3.11). Seventy three percent of past and current shift workers had an obese waist circumference (p=0.04) and 61.6% of female

73 workers who had worked 6 or more years of shift work were abdominally obese

(p=0.001). Once we controlled for age, menopausal status and education within the

multivariate analysis, we were unable to detect significant associations between work

patterns and waist circumference. Again, the trend suggests that irregular work patterns may be linked with increasing abdominal obesity. The only significant multivariate analyses finding was that older female workers are more likely to have abdominal obesity

(OR=1.04, 95% CI, 1.00, 1.07).

Bivariate findings suggest that older female workers (OR=1.10, 95% CI, 1.07,

1.13) and those who are post-menopausal (OR=2.61, 95% CI, 1.70, 4.03) were more likely to have high blood pressure, where as, female workers with higher education

(OR=0.53, 95% CI, 0.33, 0.86) were less likely to have high blood pressure. The multivariate analysis showed that female participants working extended 12 hour shifts

(OR=1.89, 95% CI, 1.00, 3.58) were more likely to have high blood pressure, when controlling for age, menopausal status and education. No other significant associations were found between work patterns, metabolic syndrome, waist circumference and systolic blood pressure.

4.6 Discussion

Findings from this study suggest that generally work patterns do not influence the development of the metabolic syndrome within female workers, with a few key exceptions. Bivariate analysis showed that after 6 years of shift work, female workers were more likely to have indicators of the metabolic syndrome. Our findings are supported by similar results from the prospective Nurses Health Study, which also found an increased CVD risk after 6 years of shift work in female workers (Kawachi, et al.

74 1995). We were unable to detect statistical significance between shift work duration and

the metabolic syndrome once covariates were controlled for, leading us to believe that

our study may have lacked statistical power to detect such associations if they existed.

Multivariate analysis showed that working extended 12 hour shifts was associated

with increased risk of elevated systolic blood pressure (OR=1.89, 95% CI, 1.00, 3.58),

when controlling for age, menopausal status and education. Within our cohort, workers

working extended shifts typically were involved in irregular work pattern scheduling.

For instance, of those working extended shifts, 80.3% worked overtime, 69.4% worked irregular rotations (including shift work) and 69.0% worked 6 or more years of shift work

(p<0.001). No other work pattern variables were significantly associated with systolic blood pressure.

The most prevalent risk factors present in this cohort were abdominal obesity and high blood pressure, specifically elevated systolic blood pressure. Almost 40% of female workers were classified as abdominally obese. Evidence suggests that abdominal obesity is associated with increased cardiovascular risk independent of overall obesity (Zhang,

Rexrode, van Dam, Li & Hu, 2008; de Koning, Merchant, Pogue & Anand, 2007; Ho, et al. 2001, Janssen, Katzmarzyk & Ross, 2004). The Nurses Health Study (NHS), a prospective cohort study (n=44,636), examined associations between abdominal obesity and CVD. Waist circumference was positively associated with cardiovascular disease mortality (RR=1.99, 95% CI, 1.44, 2.73) when controlling for lifestyle behaviours and personal characteristics (Zhang, et al. 2008). In addition, that study found that women with a normal body weight (body mass index, 18.5 - 25 kg/m2) but with abdominal

obesity (waist circumference > 88 cm) were also at increased risk CVD mortality

(RR=3.02, 95% CI, 1.31, 6.99) (Zhang, et al. 2008). Cardiovascular risk is also

75 associated with elevated systolic blood pressure (Kannel, 1996; Kannel, 2000). Kannel

suggests that elevated systolic blood pressure is more of an indicator of cardiovascular

disease risk than elevated diastolic blood pressure, as it is often associated with early

atherosclerotic disease. Thus, emphasizing the importance of identifying work-related factors which may have negative influence on the cardiovascular health, specifically, abdominal obesity and elevated systolic blood pressure.

The prevalence of metabolic syndrome, 22.1%, could be partially explained by

age of the female workers within our sample (average age of 45 years) and the high

proportion of post-menopausal women. While a 22% prevalence rate is in the high range

of reported levels in the literature, this is likely an underestimate of the problem as

women volunteered for the study and women who were self-conscious about their body

weight may be less likely to volunteer. The Canadian workforce is aging and it is

estimated that over the next decade the 55-64 age group will at least double, which will

result in this cohort making up almost 50% of the working-age population (Statistics

Canada, 2004). Findings from this study suggest that older workers, irrespective of work

pattern characteristics, need to be aware of their cardiovascular health.

As with any study, our study has limitations. This study was cross-sectional,

conducted at two health science centers and participants volunteered for study

participation. Thus causality cannot be determined and if bias existed the associations are

likely underestimated. As shift work history data were collected later on in the study

protocol, we were only able to conduct the multivariate analysis on a sub-sample, thus

limiting the power. A strength of this study was the ability to address, in part, the healthy

shift worker effect as data were collected on the work history of participants. The healthy

shift worker effect refers to the bias associated with participants who drop out of irregular

76 work patterns because they are less healthy than others, therefore leaving the healthiest

workers within the irregular work cohort and unhealthiest in day work (Karlsson,

Knutsson & Lindahl, 2001; Kivimaki, Kuisma, Virtanen & Elovaino, 2001). Therefore, we were able to identify past shift workers, who were currently day workers.

To the best of our knowledge, this is the first study performed in Canada examining the influence of work patterns on cardiovascular risk factor development in female workers. Female hospital employees working in acute care settings are exposed

to highly demanding work environments. Generally, our findings suggest that work

patterns do not influence cardiovascular disease risk among female workers. In this study

we focused on a heterogeneous sample of female workers, and thus we were likely under

powered to determine significant findings within the designated work pattern pathways.

Further research is required to better elucidate the pathways and mechanisms linking

work environment characteristics to CVD risk, and to continue to build this area of

research in female health promotion. The few key exceptions do indicate that

opportunities still exist to improve the health of work environments. Further research is

required to better elucidate the pathways and mechanisms linking work environment

characteristics to CVD risk, and to continue to build this area of research in female health

promotion. By understanding the influence of the work environment on cardiovascular

health we would be better positioned to improve the cardiovascular health of female

workers.

77 Acknowledgements

The authors gratefully acknowledge the women who participated in this study; the

Nursing Research Team who carried out the research study; Julia Nijboer for her assistance with data collection; The Health and Work Study investigators (Tranmer, J.,

McGillis-Hall, L, Katzmarzyk, P., Parry, M., Rivoire, E., Day, A., O’Callaghan, C.);

CIHR for funding the larger cohort study; and the Heart and Stroke Foundation of

Canada (Master’s Studentship Award) for their financial support for the first author.

78 4.7 References

Beerman, B., & Nachreiner, F. (1995). Working shifts—different effects for women and

men? Work & Stress, 9(2&3), 289-297.

Biggi, N., Consonni, D., Galluzzo, V., Sogliani, M., & Costa, G. (2008). Metabolic

syndrome in permanent night workers. Chronobiology International, 25, 443-454.

Boggild, H., Suadicani, P., Ole Hein, H., & Gyntelberg, F. (1999). Shift work, social

class, and ischemic heart disease in middle aged and elderly men; a 22 year

follow up in the Copenhagen male study. Occupational and Environmental

Medicine, 56(9), 640-645.

Bosma, H., Stansfeld, S., & Marmot, M. (1998). Job control, personal characteristics, and

heart disease. Journal of Occupational Health and Psychology, 3, 402-409.

Cornier, M., Dabelea, D., Hernandex, T., Lindstrom, R., Steig, A., Stob, N., et al. (2008).

The metabolic syndrome. Endocrine Reviews, 29(7), 777-822. de Koning, L., Merchant, A. T., Pogue, J., & Anand, S. S. (2007). Waist circumference

and waist-to-hip ratio as predictors of cardiovascular events: Meta-regression

analysis of prospective studies. European Heart Journal, 28(7), 850-856.

Executive Summary of the Third Report of the National Cholesterol Education Program

(NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood

Cholesterol in Adults (Adult Treatment Panel III). (2001). Journal of the

American Medical Association, 285, 2486-2497.

Ha, M., & Park, J. (2005). Shiftwork and metabolic risk factors of cardiovascular disease.

Journal of Occupational Health, 47, 89-95.

79 Heart and Stroke Foundation. (2007). Report cards on health – time to bridge the gender

gap. Ottawa, Ontario. Retrieved December 7, 2008 from

http://ww2.heartandstroke.ca/Page.asp?PageID=33&ArticleID=5911&Src=news

&From

Ho, S. C., Chen, Y. M., Woo, J. L. F., Leung, S. S. F., Lam, T. H., & Janus, E. D. (2001).

Association between simple anthropometric indices and cardiovascular risk

factors. International Journal of Obesity & Related Metabolic Disorders, 25(11),

1689-1697.

Isomaa, B., Almgren, P., Tuomi, T., Forsen, B., Lahti, K., Nissen, M., et al. (2001).

Cardiovascular morbidity and mortality associated with the metabolic syndrome.

Diabetes Care, 2, 683-389.

Janssen, I., Katzmarzyk, P.T., & Ross, R. (2002). Body mass index, waist circumference,

and health risk: evidence in support of current National Institutes of Health

guidelines. Archives of Internal Medicine, 162, 2074-2079.

Janssen, I., Katzmarzyk, P.T., & Ross, R. (2004). Waist circumference and not body

mass index explains obesity-related health risk. American Journal of Clinical

Nutrition, 79, 379-384.

Kannel, W. B. (2000). Elevated systolic blood pressure as a cardiovascular risk factor.

The American Journal of Cardiology, 85(2), 251-255.

Kannel, W. B. (1996). Blood pressure as a cardiovascular risk factor: Prevention and

treatment. JAMA, 275(20), 1571-1576.

Kawachi, I., Colditz, G., Stampfer, M., Willett, W., Manson, J., Speizer, F., et al. (1995).

Prospective study of shift work and risk of coronary heart disease in women.

Circulation, 92, 3178-3182.

80 Karlsson, B., Knutsson, A., & Lindahl, B. (2001). Is there an association between shift

work and having a metabolic syndrome? Results from a population based study of

27 485 people. Occupational and Environmental Medicine, 58, 747-752.

Karlsson, B., Knutsson, A., Lindahl, B., & Alfredsson, L. (2003). Metabolic disturbances

in male workers with rotating three-shift work. Results of the WOLF study.

International Archives of Occupational and Environmental Health, 76, 424-430.

Kivimaki, M., Kuisma, P., Virtanen, M., & Elovaino, M. (2001). Does shift work lead to

poorer health habits? A comparison between women who had always done shift

work with those who had never done shift work. Work & Stress, 15(1), 3-13.

Knutsson, A., Akerstedt, T., Jonsson, B., & Orth-Gomer, K. (1986). Increased risk of

ischaemic heart disease in shift workers. The Lancet, 2, 89-92.

Knutsson, A., & Boggild, H. (2000). Shift work and cardiovascular disease: review of

disease mechanisms. Reviews on Environmental Health, 15(4), 359-372.

Knutsson, A., Hallquist, J., Reuterwall, C., Theorell, T., & Akerstedt, T. (1999).

Shiftwork and myocardial infarction: a case-control study. Occupational and

Environmental Medicine, 56(1), 46-50.

Kroenke, C., Spiegelman, D., Manson, J., Schernhammer, E., Colditz, G., & Kawachi, I.

(2007). Work characteristics and incidence of type 2 diabetes in women.

American Journal of Epidemiology, 165(2), 175-183.

Lakka, H.M., Laaksonen, D.E., Lakka, T.A., Niskanen, L.K., Kumpusalo, E.,

Tuomilehto, J., et al. (2004). The metabolic syndrome and total and

cardiovascular disease mortality in middle-aged men. Journal of the American

Medical Association, 288, 2709-2716.

81 Rogers, A. E., Hwang, W., Scott, L. D., Aiken, L. H., & Dinges, D. F. (2004). The

working hours of hospital staff nurses and patient safety. Health Affairs, 23(4),

202-212.

Sattar, N., Gaw, A., Scherbakova, O., Ford, I., O’Reilly, D.S., Haffner, S.M., et al.

(2003). Metabolic syndrome with and without C-reactive protein as a predictor of

coronary heart disease and diabetes in the West of Scotland Prevention Study.

Circulation, 108, 414-419.

Sookian, S., Gemma, C., Gianotti, T., Burgueno, A., Alvarez, A., Gonzalez, C., et al.

(2007). Effects of rotating shift work on biomarkers of metabolic syndrome and

inflammation. Journal of Internal Medicine, 261, 285-292.

Statistics Canada (2005). Labour force survey. Retrieved August 10, from

http://www.statcan.ca/cgobin/imdbp2SV.pl?Function=getSurvey&SDDS=3701&l

ang=en&db=IMDB&dbg=f&adm=8&dis=2>

Statistics Canada (2007). RN workforce by sex and Province/Territory of registration.

Retrieved August 10, from

http:///secure.cihi.ca/cihiweb/dispPage.jsp?cw_page=statistics_results_topic_nurs

es_e&cw_topic=Health%20Human%20Resources%cw_subtopic=Nurses

Zhang, C., Rexrode, K. M., van Dam, R. M., Li, T. Y., & Hu, F. B. (2008). Abdominal

obesity and the risk of all-cause, cardiovascular, and cancer mortality: Sixteen

years of follow-up in US women. Circulation, 117(13), 1658-1667.

Figure 4

547 Female Hospital Workers Enrolled (322 day and 187 shift workers)

509 466 Completed Self-Report Data Human Resources Data (38 incomplete data / withdrew) (31 missing HR data, 14 total hours <390hrs/year)

359 Work History Data (107 missing)

Figure 4 - The number of participants who consented (n=547), complete self-report data (n=509), complete administrative data (n=466) and complete work history data (n=359).

83 Table 5. Descriptive Characteristics of Sample (N= 466) Characteristics Mean (SD) Demographics N (%) Marital Status Married 360 (77.3) Single 106 (22.7) Post-Menopausal No 297 (63.7) Yes 168 (36.1) Hormone Use No 346 (74.2) Yes 119 (25.5) Family history CVD No 155 (33.3) Yes 301 (64.6) Household Income($) <$75 000 219 (47.0) >$75 000 237 (50.9) Education < Diploma 313 (67.1) > Degree 152 (32.6) Work Status Full time 343 (73.6) Part time 123 (26.4) Work Domain Non-direct care 262 (56.2) Direct care 204 (43.8) Work Patterns Shift Length 8 hour 319 (68.5) Other (includes 12 hour) 147 (31.6) Current Rotation Day 300 (64.4) Rotational 166 (35.6) Overtime Hours No 199 (42.7) Yes 267 (57.3) Total Hours Worked <1600 hours/year 233 (50.0) >1600 hours/year 233 (50.0) Duration of Shift Work 0 years 123 (26.4) 1-5 years 55 (11.4) 6-14 years 49 (10.5) 15+ years 134 (28.8) Missing 107 (23.0) Past Shift Work No 123 (26.4) Yes 238 (50.7)

84 Percentages may not add to 100% due to missing data. Work Domain: Direct care: an employee which provides direct patient care. Non-direct care: an employee which does not provide direct patient care. Rotation: Day: employee working during day hours, Monday to Friday, with a set weekly schedule. Rotational: an employee working rotating work hours, 7 days a week.

85 Table 6. Prevalence of cardiovascular risk indicators among female hospital workers (N=466)

Risk Factor Mean SD Median Risk Indicator N (%) with risk indicator or receiving treatment* Waist circumference (cm) 85.4 13.9 83.0 > 88 cm 182 (39.1%)

Systolic BP (mmHg) 120.3 17.2 118.0 > 130 mmHg 115 (24.7%)

Diastolic BP (mmHg) 76.6 10.7 76.0 > 85 mmHg 108 (23.2%)

Triglycerides (mmol/L) 1.1 0.68 0.92 > 1.69 mmol/L 55 (11.8%)

Fasting blood glucose 5.1 0.78 5.1 > 6.11 mmol/L 24 (5.2%) (mmol/L) High density lipoprotein 1.8 1.1 1.7 < 1.29 mmol/L 59 (12.7%) cholesterol (mmol/L) Metabolic syndrome - - - > 3 indicators 103 (22.1%) present

* Participants who were currently taking antihypertensive, lipid lowering agents and/or hypoglycemic agents were considered to have the risk indicator. Or if participants were previously diagnosed as hypertensive, hypercholesterolemia, or diabetic were also considered to have the risk indicator applicable.

86 Table 7. Associations between personal and work pattern characteristics and metabolic syndrome in female hospital workers Metabolic Syndrome

Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N = 466) (N = 359) Demographics Age (years) 1.07 (1.04, 1.10) 1.06 (1.01, 1.10) Post-Menopausal No 1.00 1.00 Yes 2.39 (1.53, 3.73) 1.52 (0.78, 2.98) Hormone Use No 1.00 - Yes 1.04 (0.63, 1.71) Family history CVD No 1.00 - Yes 1.01 (0.63, 1.62) Marital Status Married 1.00 - Single 1.36 (0.82, 2.24) Household Income($) <75 000 1.00 - >75 000 1.10 (0.71, 1.71) Education < Diploma 1.00 1.00 > Degree 0.39 (0.23, 0.67) 0.59 (0.31, 1.10) Work Status Full time 1.00 Part time 0.76 (0.46, 1.28) - Work Domain Non-direct care 1.00 Direct Care 1.05 (0.68, 1.64) - Work Patterns Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.04 (0.65, 1.66) 1.28 (0.67, 2.43) Current Rotation Day 1.00 1.00 Rotational 1.20 (0.77, 1.89) 1.20 (0.60, 2.42) Overtime Hours No 1.00 1.00 Yes 1.45 (0.92, 2.27) 1.44 (0.74, 2.81) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 0.97 (0.63, 1.50) 1.24 (0.72, 2.14) Years in Shift Work (SW) * <6 years 1.00 1.00 >6 years 1.88 (1.12, 3.17) 1.23 (0.65, 2.35)

The reference group is the first category listed under each variable. Work Domain: Direct care: a position which provides direct patient care. Non-direct care: a position which does not provide direct patient care. Rotation: Day: a position during day hours, 5 days a week, Monday to Friday, which a set weekly schedule. Rotational: a position which rotates working hours, 7 days a week. Shift Length: Other: any length other than 8 hours. * Years in Shift work based on a sub-sample (N=359)

Table 8. Associations between personal and work patterns characteristics and waist circumference and systolic blood pressure in female hospital workers Characteristics Waist Circumference Systolic Blood Pressure Bivariate OR Multivariate OR Bivariate OR Multivariate OR (95% CI) (95% CI) (95% CI) (95% CI) (N= 466) (N= 359) (N= 466) (N=359) Demographics Age (years) 1.05 (1.03, 1.08) 1.04 (1.00, 1.07) 1.10 (1.07, 1.13) 1.10 (1.05, 1.15) Post-Menopausal No 1.00 1.00 1.00 1.00 Yes 2.14 (1.46, 3.16) 1.38 (0.79, 2.44) 2.61 (1.70, 4.03) 1.13 (0.59, 2.16) Hormone Use No 1.00 - 1.00 - Yes 0.93 (0.60, 1.42) 1.25 (0.78, 2.00) Family history CVD No 1.00 - 1.00 - Yes 0.80 (0.54, 1.19) 1.00 (0.64, 1.57) Marital Status Married 1.00 - 1.00 - Single 1.03 (0.66, 1.61) 1.37 (0.84, 2.23) Household Income($) <75 000 1.00 - 1.00 - >75 000 0.91 (0.63, 1.33) 0.99 (0.65, 1.52) Education < Diploma 1.00 1.00 1.00 1.00 > Degree 0.54 (0.36, 0.82) 0.65 (0.39, 1.07) 0.53 (0.33, 0.86) 0.66 (0.67, 1.21) Work Status Full-time 1.00 - 1.00 - Part-time 1.10 (0.72, 1.67) 0.77 (0.47, 1.26) Work Domain Non-direct care 1.00 - 1.00 1.00 Direct Care 1.13 (0.78, 1.64) 0.71 (0.46, 1.10) 0.62 (0.32, 1.19) Work Patterns Shift length 8 hrs 1.00 1.00 1.00 1.00 Other 1.16 (0.78, 1.73) 1.39 (0.80, 2.41) 1.10 (0.70, 1.73) 1.89 (1.00, 3.58) Current Rotation Day 1.00 1.00 1.00 1.00 Rotational 1.43 (0.97, 2.10) 1.20 (0.65, 2.19) 1.06 (0.68, 1.64) 1.07 (0.54, 2.15) Overtime Hours No 1.00 1.00 1.00 1.00 Yes 1.44 (0.98, 2.10) 1.04 (0.59, 1.81) 1.21 (0.79, 1.87) 1.45 (0.76, 2.76) Total Hours Worked <1600 hrs/year 1.00 1.00 1.00 1.00 >1600 hrs/year 0.70 (0.48, 1.01) 0.74 (0.47,1.17) 1.40 (0.92, 2.14) 1.67 (0.97, 2.87)

Years in SW * <6 years 1.00 1.00 1.00 1.00 >6 years 2.01 (1.31, 3.11) 1.42 (0.84, 2.41) 1.37 (0.85, 2.21) 1.11 (0.58, 2.09) The reference group is the first category listed under each variable. * Years in Shift work based on a sub-sample (N=359)

Chapter 5 Discussion Chapter

5.1 Summary of key findings The overall purpose of this thesis project was to explore the associations between work patterns, lifestyle behaviours and cardiovascular risk (CVR) indicators among a cohort of female hospital workers. The first manuscript aimed to determine the associations between work patterns and lifestyle behaviours. The second manuscript examined the associations between work patterns and indicators of the metabolic syndrome indicators. The additional findings section in Appendix A explored the associations between lifestyle behaviours, work patterns and the metabolic syndrome, and the metabolic syndrome indicators.

The results from both manuscripts add to the current literature by increasing our understanding about the influence of the work environment on the health of female hospital employees. These two manuscripts provide information that could not be found in the current literature regarding the relationships between work pattern characteristics and cardiovascular risk in Canadian female hospital employees.

Findings indicate:

• In general, work patterns do not influence lifestyle behaviours in female

employees, with a few key exceptions. Extended shifts are associated with

an increased risk of smoking. As well, females working overtime hours are

more likely to meet total physical activity guidelines.

• The prevalence of poor dietary patterns was high, as 93.8% reported not

meeting the recommended CFG for vegetables, fruits and whole grain

intake. As well, the prevalence of physical inactivity is high, as only 53.6%

met recommended guidelines. The influence of irregular work patterns on

physical activity levels is conflicting. Participants working irregular work

patterns were more likely to meet recommended total physical activity;

whereas, participants with irregular work patterns, were less likely to meet

recommended leisure-time physical activity. However, those who engaged

in higher levels of leisure time physical activity were less likely to have the

metabolic syndrome and abdominal obesity.

• Consistent with the literature, older workers, and those who are post-

menopausal were more likely to have the metabolic syndrome. Higher

education seems to have a protective effect, in that women with higher

levels of education were less likely have metabolic syndrome indicators and

higher levels of leisure time activity.

5.2 Revisiting the objectives Objective 1: Determine the relationships between work patterns and lifestyle behaviours.

The first manuscript addressed the associations between work patterns and lifestyle behaviours. Overall, female employees in this sample engaged in some healthy behaviours, such as low rates of smoking and excess alcohol intake. In general, our findings suggest that work patterns do not influence the ability of individual workers to 92

engage in healthier lifestyle behaviours, with a few key exceptions. Working extended

shifts was associated with increased rates of smoking and drinking alcohol in excess.

More women reported poor dietary patterns and physical inactivity. Working regular day

shifts seem to allow working women to incorporate leisure time exercise into their

routine, whereas, developing an exercise routine appears to be more problematic for

women working irregular shifts. It is currently unclear whether physical activity at work

has protective effects. The influence of work pattern characteristics on the ability of

women to engage in healthy behaviours is seemingly multifaceted.

Objective 2: Determine the prevalence of cardiovascular risk indicators, specifically indicators of the metabolic syndrome, in female hospital workers. The second

manuscript of this thesis reported on the prevalence of cardiovascular risk indicators.

Results from the analysis found that abdominal obesity and elevated systolic blood

pressure were the most frequent risk indicators. Our sample of female hospital workers is

an aging cohort which is reflective of the national female hospital workforce. While

CVR indicators are likely to be more common in women as they age, the high prevalence

of abdominal obesity and high blood pressure is concerning.

Objective 3: Determine the relationships between work patterns and cardiovascular risk indicators. The second manuscript addressed the associations between work patterns and cardiovascular risk indicators, those specifically of the metabolic syndrome. This analysis showed that within this sample the most common CVD risk indicators were

abdominal obesity and elevated systolic blood pressure. The prevalence of the metabolic

syndrome was found to be 22.1%. Bivariate analyses revealed that after 6 years of shift

work, female workers were at increased risk of developing the metabolic syndrome,

abdominal obesity, and fasting blood glucose. When we control age confounders, we are

unable to detect significant associations. These findings may accurately reflect what is

taking place, or our study may have inadequate power. When controlling for

confounders, extended shift length was associated with elevated systolic blood pressure.

This recurring finding may be indicating that after 6 years of shift work, negative physiological changes may be taking place and that the chronic exposure to shift work and more generally irregular work patterns are impacting the risk of developing CVD and, perhaps, other health problems.

Objective 4: Determine the relationships between work patterns, lifestyle factors and cardiovascular risk while controlling for known confounders, e.g. age, hormone use, menopause status and family history of CVD. The additional findings section (Appendix

A) reported on the associations between work patterns, lifestyle behaviours and the metabolic syndrome and the metabolic syndrome indicators. Metabolic syndrome and all risk indicators (except HDL) were significantly associated with increasing age and post- menopausal status in women. Leisure-time activity was found to be protective and decreased risk of the metabolic syndrome and abdominal obesity, when controlling for age, menopausal status and education. As well, elevated diastolic blood pressure was associated with extended shift length, when controlling for covariates. Overall, work

patterns do not appear to be influencing cardiovascular risk factors. These additional findings demonstrate that overall, work patterns do not appear to be associated with cardiovascular risk factors within this study.

5.3 Limitations There were limitations to the study. Participants volunteered to join the study.

Female workers who may be concerned or self-conscious about their health or weight

may not have volunteered as readily. If sample bias existed, the prevalence of metabolic

syndrome would have been underestimated. This thesis was cross-sectional, thus we

were unable to determine the long-term impact of irregular work patterns on CVD risk

indicators. However, given our focus was on identification of risk and prevention of

CVD, this limitation is not as concerning. The convenience sample was collected from

two hospital sites, therefore reducing the generalizability of the findings to different work

places other than hospitals. Lifestyle behaviours were self-reported, therefore potential for error or biased responses were possible. A more objective tool to measure physical activity may be needed to more accurately assess amount of physical activity. Missing data was also a concern. Data in regard to shift work duration was collected later during the study project and we were unable to re-contact some participants. Therefore, shift work duration analyses were based on a smaller subset of participants.

5.4 Strengths To the best of our knowledge, this is the first reported study to explore work patterns, lifestyle behaviours and cardiovascular risk indicators among female healthcare workers in Canada. There is a lack of research exploring the impact of the work 95

environment on female workers health, as much of the research has focused on male

workers. Presently, the number of women in the workforce continues to increase, and we

need to understand the healthy and unhealthy influences of the work environment

influences on women’s health.

Within this study, trained nurses carried out clinical examinations to collect the anthropometric data, thus increasing our chances of accurate and reliable data.

Participants were provided instructions on the blood sampling procedure. Again, accuracy in assessing cardiovascular risk indicators was obtained. The IPAQ used to assess physical activity is a validated questionnaire and used for cross-national comparison. The work pattern data was collected through self-report, as well as administrative human resources through use of employee numbers to ensure accuracy and trustworthiness. We are confident we were able to comprehensively examine the intended study objectives.

5.5 Practice and policy implications Although findings from this study suggest that work patterns are not influencing lifestyle behaviours and cardiovascular risk in female workers, additional benefits, such as productivity and morale, are often improved through healthy workplace policy. Female employees working part-time, rotational, 12 hour shift positions, mainly direct care providers, may require more assistance or support to attain healthy lifestyle goals.

Working irregular work patterns may be creating barriers for developing leisure-time physical activity routines, thus facilitation and support may be beneficial. Given that direct-care providers also reported higher levels of work related activity, they may have 96

been too fatigued to participate in leisure-time activity. Workplace policies could support

physical activity through on-work or off-work programs. Promoting adequate amounts of

leisure-time activity will likely decrease cardiovascular disease and subsequent health

problems at later stages of life.

The low prevalence of smoking among this sample demonstrates the successful

efforts of public health initiatives aimed at smoking cessation. Findings also indicate the need of public health initiatives to focus on promoting a balanced diet and physical activity, among those working regular and irregular work patterns. Perhaps, more education is required to facilitate healthy eating habits among female workers working

shift work, overtime and extended shifts. More specifically, hospital decision makers

could facilitate the process of establishing healthy balanced diets among rotational

workers, possibly by providing access to healthy foods on night shifts.

The alarming prevalence of abdominal obesity needs to be addressed. Obesity

impacts women at an individual level, but is likely also to have a system impact. Thus, it

would seem that targeting this ever increasing health problem should be a priority for the planning of healthy workplaces. While beyond the scope of this thesis, interventions to promote physical activity and balanced dietary intake could be encouraged and promoted within workplaces.

Given that the prevalence of CVD indicators increases with age and the consistent finding that shift work duration is associated with increased CVD risk, is particularly concerning for older women who engage in shift work or other irregular work patterns.

As the healthcare workforce ages, hospital administrators perhaps need to consider other

work pattern alternatives for their employees and, as well, continue to monitor the impact

of irregular work patterns on health.

Female hospital workers, who are taking care of people in need, deserve attention themselves. It is important to address the health of hospital workers to ensure the future functioning of our healthcare system. Health management policies could further explore the work environment and its role in deterring or facilitating the health of workers. With more female employees in the workforce and the multiple roles which women engage in, consideration needs to be given to the establishment of workplace policy that supports both workplace and life demands.

Appendix A Additional Findings

The following results are based on additional analyses of the data, completed to address the stated objectives, and are not presented in Manuscript 1 and 2. These findings will be discussed in Chapter 6. Initially, bivariate analyses were run to evaluate associations between demographic, work pattern, lifestyle behaviour and metabolic syndrome data. Variables that met p < 0.15 and all work pattern variables were entered into the multivariate logistic regression models to determine relationships between work pattern characteristics, lifestyle behaviours, and the metabolic syndrome, and the metabolic syndrome components while controlling for potentially significant confounders (i.e. age, hormone use). The 0.05 level of statistical significance was accepted. The results are displayed in a tabular format with a subsequent results paragraph summarizing the key findings.

Table 9. Associations between personal characteristics, work patterns and lifestyle behaviours with the metabolic syndrome in female hospital workers. Metabolic Syndrome

Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 1.07 (1.04, 1.10) 1.06 (1.01, 1.12) Post-Menopausal No 1.00 1.00 Yes 2.39 (1.53, 3.73) 1.52 (0.71, 3.25) Hormone Use No 1.00 - Yes 1.04 (0.63, 1.71) Family history CVD No 1.00 - Yes 1.01 (0.63, 1.62) Marital Status Married 1.00 - Single 1.36 (0.82, 2.24) Household Income($) <75 000 1.00 - >75 000 1.10 (0.71, 1.71) Education < Diploma 1.00 1.00 > Degree 0.39 (0.23, 0.67) 0.56 (0.28, 1.14) Work Status Full time (FT) 1.00 - Part time (PT) 0.76 (0.46, 1.28) Work Domain Non-direct care 1.00 - Direct Care 1.05 (0.68, 1.64) Work Patterns Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.04 (0.65, 1.66) 1.08 (0.52, 2.25) Current Rotation Day 1.00 1.00 Rotational 1.20 (0.77, 1.89) 1.53 (0.70, 3.33) Overtime Hours No 1.00 1.00 Yes 1.45 (0.92, 2.27) 1.60 (0.77, 3.31) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 0.97 (0.63, 1.50) 1.38 (0.74, 2.57) Years in Shift Work <6 years 1.00 1.00 100

>6 years 1.88 (1.12, 3.17) 1.30 (0.64, 2.64) Lifestyle Behaviours Smoking Former/never 1.00 1.00 Current 1.94 (0.87, 4.30) 0.81 (0.19, 3.39) Excess Alcohol No 1.00 - Yes 0.77 (0.49, 1.22) Total Physical Activity Low/moderate 1.00 - High 0.78 (0.47, 1.28) Leisure-time Activity Guidelines Unmet 1.00 1.00 Met 0.59 (0.38, 0.91) 0.43 (0.23, 0.81) Dietary Patterns Unbalanced 1.00 - Balanced 0.65 (0.34, 1.23) The reference group is the first category listed under each variable.

Bivariate analyses showed that older women (OR=1.07, 95% CI, 1.04, 1.10) and post-menopausal women (OR=2.39, 95% CI, 1.53, 3.73) were at increased risk of developing the metabolic syndrome. Female workers with higher education (OR=0.39, 95% CI, 0.23, 0.67) and participants who met recommended levels of leisure activity (OR=0.59, 95% CI, 0.38, 0.91) were less likely to develop the metabolic syndrome. Whereas, female participants who had worked 6 or more years of shift work (OR=1.88, 95% CI, 1.12, 3.17) were at greater risk of developing the metabolic syndrome. Multivariate analyses revealed that female workers who met recommended leisure activity guidelines were less likely to develop the metabolic syndrome (OR=0.43, 95% CI, 0.23, 0.81), when controlling for age, menopausal status, education and smoking. No work pattern characteristics were significantly associated with the metabolic syndrome in the multivariate analysis.

101

Table 10. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated waist circumference in female hospital workers. Waist Circumference

Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 1.05 (1.03, 1.08) 1.05 (1.01, 1.10) Post-Menopausal No 1.00 1.00 Yes 2.14 (1.46, 3.16) 1.17 (0.62, 2.20) Hormone Use No 1.00 - Yes 0.93 (0.60, 1.42) Family history CVD No 1.00 - Yes 0.80 (0.54, 1.19) Marital Status Married 1.00 - Single 1.03 (0.66, 1.61) Household Income($) <75 000 1.00 - >75 000 0.91 (0.63, 1.33) Education < Diploma 1.00 1.00 > Degree 0.54 (0.36, 0.82) 0.70 (0.40, 1.21) Work Status Full time (FT) 1.00 - Part time (PT) 1.10 (0.72, 1.67) Work Domain Non-direct care 1.00 - Direct Care 1.13 (0.78, 1.64) Work characteristics Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.16 (0.78, 1.73) 1.16 (0.62, 2.18) Rotation Day 1.00 1.00 Rotational 1.43 (0.97, 2.10) 1.50 (0.75, 2.97) Overtime Hours No 1.00 1.00 Yes 1.44 (0.98, 2.10) 1.25 (0.68, 2.28) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 0.70 (0.48, 1.01) 0.68 (0.41, 1.14) Years in Shift work None/<6 years 1.00 1.00 102

>6 years 2.01 (1.31, 3.11) 1.40 (0.78, 2.49) Lifestyle Behaviours Smoking Former/never 1.00 1.00 Current 1.20 (0.57, 2.54) 0.82 (0.24, 2.77) Excess Alcohol No 1.00 - Yes 0.89 (0.61, 1.31) Total Physical Activity Low/moderate 1.00 - High 0.84 (0.54, 1.29) Leisure-time Activity Guidelines Unmet 1.00 1.00 Met 0.46 (0.29, 0.74) 0.48 (0.27, 0.85) Dietary Patterns Unbalanced 1.00 - Balanced 0.89 (0.54, 1.46) The reference group is the first category listed under each variable.

Bivariate analyses showed that increasing age (OR=1.05, 95% CI, 1.03, 1.08) and post-menopausal status (OR=2.14, 95% CI, 1.46, 3.16) were associated with abdominal obesity. Female workers with higher education (OR=0.54, 95% CI, 0.36, 0.82), and participants with adequate leisure activity (OR=0.46, 95% CI, 0.29, 0.74) were less likely to be abdominally obese. Females working 6 or more years of shift work were at increased risk of abdominal adiposity (OR=2.01, 95% CI, 1.31, 3.11). Multivariate analyses found that female workers who met recommended leisure activity guidelines (OR=0.48, 95% CI, 0.27, 0.85) were less likely to have abdominal obesity, when controlling for age, menopausal status, education and smoking.

103

Table 11. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated systolic blood pressure in female hospital workers. Systolic Blood Pressure Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 1.10 (1.07, 1.13) 1.09 (1.04, 1.15) Post-Menopausal No 1.00 1.00 Yes 2.61 (1.70, 4.03) 1.13 (0.56, 2.27) Hormone Use No 1.00 - Yes 1.25 (0.78, 2.00) Family history CVD No 1.00 - Yes 1.00 (0.64, 1.57) Marital Status Married 1.00 - Single 1.37 (0.84, 2.23) Household Income($) <75 000 1.00 - >75 000 0.99 (0.65, 1.52) Education < Diploma 1.00 1.00 > Degree 0.53 (0.33, 0.86) 0.64 (0.34, 1.21) Work Status Full time (FT) 1.00 - Part time (PT) 0.77 (0.47, 1.26) Work Domain Non-direct care 1.00 1.00 Direct Care 0.71 (0.46, 1.10) 0.79 (0.39, 1.58) Work characteristics Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.10 (0.70, 1.73) 1.72 (0.87, 3.42) Rotation Day 1.00 1.00 Rotational 1.06 (0.68, 1.64) 1.11 (0.52, 2.35) Overtime Hours No 1.00 1.00 Yes 1.21 (0.79, 1.87) 1.69 (0.86, 3.31) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 1.40 (0.92, 2.14) 1.63 (0.90, 2.94) Years in Shift work None/<6 years 1.00 1.00 104

>6 years 1.37 (0.85, 2.21) 1.07 (0.54, 2.10) Lifestyle Behaviours Smoking Former/never 1.00 - Current 1.56 (0.71, 3.44) Excess Alcohol No 1.00 - Yes 0.80 (0.52, 1.25) Total Physical Activity Low/moderate 1.00 - High 1.04 (0.64, 1.69) Leisure-time Activity Guidelines Unmet 1.00 1.00 Met 0.62 (0.38, 1.04) 0.63 (0.34, 1.16) Dietary Patterns Unbalanced 1.00 - Balanced 0.89 (0.50, 1.59) The reference group is the first category listed under each variable.

Bivariate analyses revealed that elevated systolic blood pressure was associated with increasing age (OR=1.10, 95% CI, 1.07, 1.13) and post-menopausal status (OR=2.61, 95% CI, 1.70, 4.03). Female workers with higher education (OR=0.53, 95% CI, 0.33, 0.86) were less likely to have elevated systolic blood pressure. Multivariate analysis revealed increasing age (OR=1.09, 95% CI, 1.04, 1.15) to be associated with elevated systolic blood pressure. No work pattern characteristics were significantly associated with the metabolic syndrome in the multivariate analysis.

105

Table 12. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated diastolic blood pressure in female hospital workers. Diastolic Blood Pressure Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 1.07 (1.04, 1.10) 1.08 (1.03, 1.14) Post-Menopausal No 1.00 1.00 Yes 1.95 (1.26, 3.02) 0.88 (0.43, 1.82) Hormone Use No 1.00 - Yes 1.19 (0.73, 1.94) Family history CVD No 1.00 - Yes 1.15 (0.72, 1.84) Marital Status Married 1.00 - Single 1.19 (0.72, 1.97) Household Income($) <75 000 1.00 - >75 000 0.94 (0.61, 1.46) Education < Diploma 1.00 1.00 > Degree 0.52 (0.31, 0.86) 0.65 (0.34, 1.24) Work Status Full time (FT) 1.00 1.00 Part time (PT) 0.61 (0.36, 1.04) 0.55 (0.24, 1.26) Work Domain Non-direct care 1.00 - Direct Care 0.78 (0.50, 1.20) Work characteristics Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.11 (0.70, 1.75) 2.02 (0.99, 4.13) Rotation Day 1.00 1.00 Rotational 0.87 (0.55, 1.38) 0.93 (0.42, 2.03) Overtime Hours No 1.00 1.00 Yes 1.16 (0.75, 1.80) 1.51 (0.76, 3.00) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 0.86 (0.56, 1.32) 0.86 (0.44, 1.67) Years in Shift work None/<6 years 1.00 1.00 106

>6 years 1.01 (0.99, 1.03) 0.63 (0.32, 1.24) Lifestyle Behaviours Smoking Former/never 1.00 1.00 Current 1.74 (0.79, 3.84) 1.31 (0.36, 4.86) Excess Alcohol No 1.00 - Yes 0.99 (0.64, 1.54) Total Physical Activity Low/moderate 1.00 - High 0.99 (0.60, 1.64) Leisure-time Activity Guidelines Unmet 1.00 1.00 Met 0.69 (0.41, 1.19) 0.68 (0.36, 1.29) Dietary Patterns Unbalanced 1.00 - Balanced 1.08 (0.61, 1.90) The reference group is the first category listed under each variable.

Bivariate analyses showed that increasing age and post-menopausal status were both significantly associated with elevated diastolic blood pressure. Multivariate analyses revealed that irregular shift length, mainly extended 12 hour shifts (OR=2.02, 95% CI, 0.99, 4.13) were significantly associated with elevated diastolic blood pressure, when controlling for age, menopausal status, education, smoking and leisure-time activity.

107

Table 13. Associations between personal characteristics, work patterns and lifestyle behaviours with high fasting blood glucose in female hospital workers. Fasting Blood Glucose

Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 1.08 (1.02, 1.14) 1.05 (0.95, 1.17) Post-Menopausal No 1.00 1.00 Yes 4.56 (1.85, 11.25) 2.88 (0.60, 13.84) Hormone Use No 1.00 - Yes 0.58 (0.19, 1.72) Family history CVD No 1.00 - Yes 2.05 (0.75, 5.60) Marital Status Married 1.00 - Single 1.40 (0.57, 3.48) Household Income($) <75 000 1.00 - >75 000 0.89 (0.39, 2.02) Education < Diploma 1.00 1.00 > Degree 0.30 (0.87, 1.02) 0.27 (0.05, 1.39) Work Status Full time (FT) 1.00 - Part time (PT) 0.58 (0.19, 1.72) Work Domain Non-direct care 1.00 - Direct Care 1.61 (0.70, 3.67) Work characteristics Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.12 (0.47, 2.68) 1.52 (0.41, 5.60) Rotation Day 1.00 1.00 Rotational 0.93 (0.39, 2.23) 0.85 (0.20, 3.68) Overtime Hours No 1.00 1.00 Yes 1.57 (0.66, 3.76) 1.13 (0.29, 4.38) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 1.15 (0.51, 2.63) 3.23 (0.88, 11.77) Years in Shift work None/<6 years 1.00 1.00 108

>6 years 3.11 (1.11, 8.76) 2.64 (0.64, 10.95) Lifestyle Behaviours Smoking Former/never 1.00 - Current 1.46 (0.32, 6.55) Excess Alcohol No 1.00 - Yes 0.77 (0.32, 1.83) Total Physical Activity Low/moderate 1.00 - High 1.40 (0.49, 4.01) Leisure-time Activity Guidelines 1.00 1.00 Unmet 0.45 (0.17, 1.21) 0.43 (0.13, 1.37) Met Dietary Patterns Unbalanced 1.00 - Balanced 0.43 (0.10, 1.85) The reference group is the first category listed under each variable.

Bivariate analyses revealed that older female workers (OR=1.08, 95% CI, 1.02, 1.14) and post-menopausal women (OR=4.56, 95% CI, 1.85, 11.25) were at increased risk of developing high fasting blood glucose. Female workers with higher education (OR=0.33, 95% CI, 0.87, 1.02) were less likely to have high fasting blood glucose, whereas, females working > 6 years of shift work (OR=3.11, 95% CI, 1.11, 8.76) were at increased risk of developing high fasting blood glucose. No significant findings were revealed in the multivariate analyses when controlling for covariates.

109

Table 14. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated triglycerides in female hospital workers. Triglycerides

Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 1.04 (1.01, 1.08) 1.00 (0.95, 1.06) Post-Menopausal No 1.00 1.00 Yes 2.52 (1.42, 4.46) 2.04 (0.79, 5.23) Hormone Use No 1.00 - Yes 1.52 (0.83, 2.78) Family history CVD No 1.00 1.00 Yes 0.64 (0.36, 1.15) 0.77 (0.35, 1.70) Marital Status Married 1.00 - Single 0.92 (0.47, 1.82) Household Income($) <75 000 1.00 - >75 000 0.96 (0.54, 1.69) Education < Diploma 1.00 1.00 > Degree 0.39 (0.19, 0.82) 0.40 (0.15, 1.05) Work Status Full time (FT) 1.00 - Part time (PT) 0.90 (0.47, 1.75) Work Domain Non-direct care 1.00 - Direct Care 0.95 (0.53, 1.68) Work characteristics Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 0.81 (0.43, 1.53) 0.91 (0.35, 2.36) Rotation Day 1.00 1.00 Rotational 0.82 (0.45, 1.50) 0.78 (0.29, 2.11) Overtime Hours No 1.00 1.00 Yes 1.28 (0.72, 2.29) 1.39 (0.57, 3.39) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 0.66 (0.38, 1.17) 0.98 (0.45, 2.10) Years in Shift work None/<6 years 1.00 1.00 110

>6 years 1.16 (0.60, 2.26) 1.36 (0.56, 3.34) Lifestyle Behaviours Smoking Former/never 1.00 - Current 1.27 (0.42, 3.82) Excess Alcohol No 1.00 - Yes 1.24 (0.70, 2.19) Total Physical Activity Low/moderate 1.00 - High 0.77 (0.41, 1.45) Leisure-time Activity Guidelines 1.00 1.00 Unmet 0.54 (0.28, 1.50) 0.53 (0.24, 1.16) Met Dietary Patterns Unbalanced 1.00 - Balanced 1.09 (0.52, 2.27) The reference group is the first category listed under each variable.

Bivariate analyses revealed that increasing age (OR=1.04, 95% CI, 1.01, 1.08) and post-menopausal status (OR=2.52, 95% CI, 1.42, 4.46) were both significantly associated with elevated triglycerides. Female hospital workers with higher education (OR=0.39, 95% CI, 0.19, 0.82) were less likely to have elevated triglycerides. No statistically significant results were found in the multivariate analyses when controlling for covariates.

111

Table 15. Associations between personal characteristics, work patterns and lifestyle behaviours with elevated high density lipoproteins in female hospital workers. High Density Lipoprotein

Characteristics Bivariate OR (95% CI) Multivariate OR (95% CI) (N=466) (N=359) Demographics Age (years) 0.99 (0.96, 1.02) 1.01 (0.96, 1.05) Post-Menopausal No 1.00 - Yes 0.87 (0.49, 1.55) Hormone Use No 1.00 - Yes 1.11 (0.60, 2.06) Family history CVD No 1.00 - Yes 0.82 (0.46, 1.47) Marital Status Married 1.00 - Single 1.29 (0.69, 2.39) Household Income($) <75 000 1.00 - >75 000 1.21 (0.69, 2.11) Education < Diploma 1.00 - > Degree 0.72 (0.38, 1.34) Work Status Full time (FT) 1.00 - Part time (PT) 1.01 (0.54, 1.89) Work Domain Non-direct care 1.00 1.00 Direct Care 1.56 (0.90, 2.69) 2.23 (0.93, 5.34) Work characteristics Shift length 8 hrs 1.00 1.00 Other (includes 12hrs) 1.27 (0.71, 2.25) 0.78 (0.30, 2.05) Rotation Day 1.00 1.00 Rotational 1.70 (0.98, 2.96) 2.25 (0.81, 6.23) Overtime Hours No 1.00 1.00 Yes 1.24 (0.71, 2.16) 0.90 (0.35, 2.34) Total Hours Worked <1600 hrs/year 1.00 1.00 >1600 hrs/year 0.73 (0.42, 1.27) 1.35 (0.63, 2.92) Years in Shift work None/<6 years 1.00 1.00 112

>6 years 1.10 (0.56, 2.15) 0.75 (0.30, 1.89) Lifestyle Behaviours Smoking Former/never 1.00 - Current 1.16 (0.39, 3.49) Excess Alcohol No 1.00 - Yes 0.99 (0.57, 1.74) Total Physical Activity Low/moderate 1.00 1.00 High 0.54 (0.30, 0.99) 0.51 (0.23, 1.14) Leisure-time Activity Guidelines 1.00 1.00 Unmet 0.54 (0.29, 1.02) 0.77 (0.33, 1.78) Met Dietary Patterns Unbalanced 1.00 1.00 Balanced 0.51 (0.21, 1.24) 0.35 (0.08, 1.58) The reference group is the first category listed under each variable.

Bivariate analyses revealed that female workers who had adequate (high) total physical activity (OR=0.45, 95% CI, 0.30, 0.99) were less likely to have low high density lipoproteins. Multivariate analyses showed no significant findings when controlling for covariates.

113

Appendix B International Physical Activity Questionnaire (IPAQ)

114

115

116

117

Appendix C Rapid Eating Assessment for Patients (REAP)

118

119

120