Prevalence of Overweight and Obesity and Perception of Healthy and Desirable Body Size in Urban, Ghanaian Women
Total Page:16
File Type:pdf, Size:1020Kb
June 2012 Volume 46, Number 2 GHANA MEDICAL JOURNAL PREVALENCE OF OVERWEIGHT AND OBESITY AND PERCEPTION OF HEALTHY AND DESIRABLE BODY SIZE IN URBAN, GHANAIAN WOMEN R. M. BENKESER1, R. BIRITWUM2 and A. G. HILL1 1Harvard School of Public Health, Boston, MA, USA 2Department. of Community Medicine, University of Ghana Medical School, Accra, Ghana Author for Correspondence: Rachel Benkeser E-mail: [email protected] Conflict of Interest: None declared SUMMARY parity >2 are associated with higher risk of overweight Objective: To determine if the perceived ideal body or obesity. Education level is not associated with over- size of urban women in Accra, Ghana influences cur- weight status. Overall, 72.2% (1,936) of women were rent body size. dissatisfied with their current body size and 41.8% of Methods: The sample (n=2,814) consisted of Ghanaian women preferred a smaller figure. Overweight and women aged 18 years and older, living in the Accra obese women were significantly more likely than nor- Metropolitan Area. Data were collected between Octo- mal weight women to desire weight loss (OR: 10.12, ber 2008 and June 2009 in the Women’s Health Study CI: 8.04 – 12.72). of Accra, Wave II household survey. Height (m), Conclusion: Ideal body image in urban, Ghanaian weight (kg), waist (cm) and hip (cm) measurements women should not be viewed as an inhibitor to healthy were used to calculate body mass index (BMI, meas- weight maintenance. Interventions should be designed ured in kg/m2), waist circumference, and waist-to-hip to help women achieve their ideal and healthy weights. ratio. In addition to reported socio-demographic char- acteristics, medical history, and exercise and diet histo- Keywords: Obesity, body mass index (BMI), anthro- ry, the standardized Stunkard Figure Rating Scale pometric measures, body images, Ghana. (FRS) was used to identify silhouettes of varying body sizes that most closely resemble a woman’s current INTRODUCTION body image (CBI), the body image she sees as the ideal The prevalence of obesity in West Africa rapidly in- for Ghanaian women (IBI), and the body image she creased during the last two decades of the 20th century sees as healthiest (HBI). The validity of using the and continues to increase in the 21st century. A recent standardized Stunkard FRS to assess body image in a review of obesity prevalence in West Africa indicates population of Ghanaian women was assessed using that between 2000 and 2004, 10% of West African Receiver Operating Curve (ROC) analysis. Household adults were obese.2 In this time period, according to wealth, a proxy for income, was assessed using an in- the same study, half of the urban population in general dex derived by principal components analysis from a and 60% of urban, West African women were either selection of 20 variables measuring quality of the overweight or obese. As in earlier reviews of socio- dwelling (e.g water source, sewage arrangements, type economic status and obesity in developing countries,3 of housing, roof etc.) and possession of household du- this review also found a direct relationship between rable goods (e.g. television, radio, iron, refrigerator and obesity prevalence and increased socio-economic sta- car).1 Logistic and linear regression was used to test the tus. In Ghana, the DHS surveys indicate that the per- association between various socio-demographic char- centage of women aged 15-49 overweight or obese acteristics, BMI and body image satisfaction. grew from 25% to 30% between 2003 and 2008 with Results: Using the WHO criteria for BMI, 3.6% (95) the highest values among urban women.4 Wave 1 of the of women were underweight, 31.5% (828) normal Women’s Health Study of Accra (WHSA-1) found weight, 27.8% (730) overweight and 37.1% (973) 62.2% of 1,237 non-pregnant women living in Accra obese. In total, 64.9% of women sampled were either were overweight or obese; the same study found no overweight or obese. A WC of greater than 88 cm re- relationship between obesity prevalence and socio- vealed 78.7% (2,113) obesity. A WHR cut-off for obe- economic status.5 sity of 0.8 revealed 78.9% (2,119) obesity. ROC analy- sis confirmed the validity of using the standardized Though the public health and medical consequences of Stunkard FRS to assess body image in a population of the rise in obesity in West Africa are evident, the rea- Ghanaian women. Age, relative wealth, being married, sons for the rapid increase in body size and body mass having grown up in an urban environment and having remain speculative. 66 June 2012 R. M. Benkeser et al Overwieght and obesity in urban Ghanaian women Overweight and obesity are well understood to be the some evidence that women in general preferred a nor- result of an energy imbalance – consuming more calo- mal weight figure, even though some obese women ries than are equivalently expended in physical activity. wanted to gain weight in order to achieve their ideal Two popularly cited theories for the West African image. Additionally, some evidence from these studies change in energy balance include increases in urbaniza- suggests that age may be an important determinant of tion and westernization. In theory, urbanization and both a woman’s weight preference and her willingness westernization lead to decreased physical activity and to lose weight, older women being less willing to lose increased food supply, including access to high caloric weight for health. At least one study suggests that the fast foods and sugar sweetened beverages.5 Such envi- majority of women would be willing to lose weight if ronmental changes, however, do not completely ex- requested by her husband or to improve her health.13 plain why women, in particular women of high socio- economic status, are disproportionally affected by the Taking this history into account, in this study we exam- obesity epidemic in West Africa. ined a large sample of urban, Ghanaian women (n=2,814) to determine if indeed an urban, West Afri- West African social desirability for overweight women can woman’s ideal body size influences her current is frequently cited as cause for the continued direct weight. relationship between obesity and socio-economic status in the region.6-11 Indeed, it is well documented that METHODS some ethnic groups in Africa historically preferred The Women’s Health Study of Accra overweight women and embraced cultural practices The Women’s Health Study of Accra, Wave II that encouraged female obesity (as in the pre-marital (WHSA-II) conducted in 2008-2009 is the second 12 “fattening rooms” of Nigeria). This argument sug- cross-sectional, community-based population study of gests that the recent increase in overweight and obesity a longitudinal assessment of health in a sample of in West Africa may not be a consequence of changes in women age 18 and older residing in Ghana’s Accra the environment but rather is the result of society-wide Metropolitan Area. Older women were over-sampled to intentional weight gain enabled by the increased avail- provide enough elderly cases for analysis but other- ability of food concurrent with decreased need for wise, the sample, after stratification for socio-economic physical exertion. status, is representative of all adult women in Accra. WHSA-II was conducted by the joint partnership be- Because behavioural interventions are only successful tween the University of Ghana Institute for Statistical, if individuals agree to participate in the intervention, Social, and Economic Research and the Harvard understanding this preference for body size is crucial to School of Public Health. understanding the approach of future overweight and obesity interventions. The WHSA-II sample consists primarily of participants from Wave I of the study (WHSA-I), conducted in Although understanding preference for body image is 2003. Participants to WHSA-I were selected by a two- critical to crafting public health obesity interventions, stage cluster probability sample stratified by socioeco- studies examining the relationship between socially nomic status based on the 2000 Ghanaian census data, desirable body size and overweight/obesity are rare. In and older women were progressively over-sampled. reviewing the literature restricted to West Africa, ten When a participant from WHSA-I could not be located, articles address the relationship between body image a woman of similar age, socioeconomic status and geo- and obesity. graphic location as the initial participant in WHSA-I replaced the woman in the sample. A total of 995 re- In the largest study considered here, Duda et al used placement women were interviewed for a total of two culturally specific silhouettes and surveyed urban Gha- thousand eight hundred and fourteen (2,814) WHSA-II naian women on whether improved health outcomes participants interviewed between October 2008 and would incentivize women to change their current body June 2009. The complete methodology and sampling of 13 image. Of the 305 participants, 86% of 214 over- WHSA-I and WHSA II are described elsewhere.14,15 weight or obese women stated they would be willing to decrease their body weight if it meant leading a health- All women interviewed in WHSA-II were eligible for ier life. Women over the age of 50, however, were sig- this study (n=2,814). Following WHO protocol, wom- nificantly less willing to decrease their body weight. en who were pregnant or less than three months post partum (n=130) at the time of interview were excluded These small sample size studies on urban, West Afri- from the analysis of body weight and body size. can women suggest a general dissatisfaction with body image irrespective of body weight. These studies offer 67 June 2012 Volume 46, Number 2 GHANA MEDICAL JOURNAL In this analysis, the WHSA-II sample population is not WC alone in the measurement of obesity risk, it is re- weighted nor are the findings standardized.