24-Hour Ambulatory Blood Pressure - Relation to the Insulin Resistance Syndrome and Cardiovascular Disease

! "# $ % ! & ' &

$( )&'*'+" $,-&).+

"*" +'/0&'*"*' %".'0+' %%"." 11 KRISTINA BJÖRKLUND

Dissertation for the Degree of Doctor of Philosophy (Faculty of Medicine) in Geriatrics presented at Uppsala University in 2002 Abstract Björklund, K. 2002. 24-hour Ambulatory Blood Pressure - Relation to the Insulin Resistance Syndrome and Cardiovascular Disease. Acta Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1199. 62 pp. Uppsala. ISBN 91-554-5451-8.

Hypertension constitutes part of the insulin resistance syndrome, and is a common and powerful risk factor for cardiovascular disease in elderly. Blood pressure (BP) measured with 24-hour ambulatory monitoring gives however more detailed information and may be a better estimate of the true BP than conventional office BP. This study examined relationships between 24-hour ambulatory BP and components of the insulin resistance syndrome, and investigated the prognostic significance of 24-hour BP for cardiovascular morbidity in a longitudinal population-based study of 70-year-old men. The findings indicated, that a reduced nocturnal BP fall, nondipping, was a marker of increased risk primarily in subjects with diabetes. A low body mass index and a more favourable serum fatty acid composition at age 50 predicted the development of white-coat as opposed to sustained hypertension over 20 years. Furthermore, cross-sectionally determined hypertensive organ damage at age 70 was detected in sustained hypertensive but not in white-coat hypertensive subjects. In a prospective analysis, 24- hour ambulatory pulse pressure and systolic BP variability at age 70 were strong predictors of subsequent cardiovascular morbidity, independently of office BP and other established risk factors. Isolated ambulatory hypertension, defined as having a normal office BP but increased daytime ambulatory BP, was associated with a significantly increased incidence of cardiovascular events during follow-up. In summary, these data provide further knowledge of 24-hour ambulatory BP and associated metabolic risk profile, and suggest that the prognostic value of 24- hour ambulatory BP is superior to conventional BP in an elderly population.

Key words: hypertension, ambulatory blood pressure, insulin, fatty acids, prognosis, morbidity Kristina Björklund, Department of Public Health and Caring Sciences, Section of Geriatrics, Box 609, SE-751 25, Uppsala, Sweden

2 24-HOUR AMBULATORY BP

3 KRISTINA BJÖRKLUND Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals:

I. Björklund K, Lind L, Lithell H. Twenty-four hour blood pressure in a population of elderly men. J Internal Medicine. 2000; 248: 501-510.*

II. Björklund K, Lind L, Andrén B, Lithell H. The majority of nondipping men do not have increased cardiovascular risk: a population-based study. Journal of Hypertension 2002;20:1501-1506.†

III. Björklund K, Lind L, Vessby B, Andrén B, Lithell H. Different metabolic predictors of white-coat and sustained hypertension over a 20-year follow-up period: a population-based study of elderly men. Circulation. 2002;106:63-68.†

IV. Björklund K, Lind L, Zethelius B, Berglund L, Lithell H. Prognostic significance of 24-hour ambulatory blood pressure characteristics for cardiovascular morbidity in a population of elderly men. (In manuscript)

V. Björklund K, Lind L, Zethelius B, Andrén B, Lithell H. Isolated ambulatory hypertension predicts cardiovascular morbidity in elderly men. (Submitted)

Reprints were made with permission from the publishers; * © Blackwell Science Ltd, and † © Lippincott Williams & Wilkins.

4 24-HOUR AMBULATORY BP Table of contents

Abbreviations...... 6 Introduction ...... 7 Hypertension in the elderly ...... 8 Hypertension and insulin resistance ...... 10 Circadian BP variation ...... 11 Aims of the study ...... 15 Methods ...... 16 Subjects ...... 16 Investigations ...... 19 Definitions ...... 23 Statistical analyses ...... 24 Discussion of Methods ...... 25 Results ...... 28 Paper I ...... 28 Paper II...... 29 Paper III ...... 31 Paper IV ...... 33 Paper V ...... 34 Discussion...... 36 Ambulatory BP normality - definition and consequences...... 36 Nondipping as cardiovascular risk indicator ...... 38 White-coat hypertension and the insulin resistance syndrome ...... 39 Prognostic value of 24-hour ambulatory BP ...... 42 Predictive value of isolated ambulatory and white-coat hypertension ...... 44 Clinical relevance ...... 47 Strengths and limitations of the study ...... 49 Future perspectives ...... 50 Conclusions ...... 52 Acknowledgements ...... 53 References ...... 55

5 KRISTINA BJÖRKLUND Abbreviations

ANOVA analysis of variance ACE angiotensin-converting enzyme BMI body mass index BP blood pressure CE cholesterol ester CI confidence interval CV coefficient of variation DBP diastolic blood pressure HDL high-density lipoprotein HR Cox proportional hazard ratio ICD International Classification of Diseases LDL low-density lipoprotein MAP mean arterial pressure PAI-1 plasminogen activator inhibitor-1 PYAR person-years at risk SD standard deviation SBP systolic blood pressure UAER urinary albumin excretion rate ULSAM Uppsala Longitudinal Study of Adult Men

6 24-HOUR AMBULATORY BP Introduction

Hypertension is a common and powerful risk factor for cardiovascular disease.[1, 2] High blood pressure (BP) often co-exists with dyslipidemia, abdominal obesity and abnormal glucose metabolism,[3, 4] and such associated metabolic risk factors further contribute to an adverse outcome in hypertension. In industrialised countries, the BP increases progressively with age, and it has been estimated that nearly 60% of the population in the USA at the age of 70 have hypertension.[5] In the elderly, hypertension is commonly characterised by an isolated increase in systolic BP (SBP), in contrast to the predominantly elevated diastolic BP (DBP) component observed in younger hypertensive subjects. Isolated systolic hypertension, for many years disregarded by clinicians as a harmless physiological phenomenon, has emerged as an important cardiovascular risk factor in elderly.[6] With an increasing number of elderly people in our society, hypertensive disease and subsequent organ damage constitute a major challenge for the health sector in the nearest future. The BP level in an individual is not constant, but continuously fluctuates over time, depending on both internal and external factors. BP is most commonly measured by a nurse or physician at the clinic. The clinic environment may in some patients cause an alerting reaction and a sudden BP rise, the so-called “white-coat effect”, which can obscure the clinical judgement, and cause misclassification of the hypertensive status of the patient. Fur- ther, occasional measurements of conventional, or office BP, reflect the BP level at a limited time point of the day. Twenty-four-hour ambulatory BP monitoring, on the other hand, eliminates the white-coat effect and provides more comprehensive information about the “true” BP level during both day- and night-time. Previous findings have indicated that hypertensive target organ damage is closer related to mean 24-hour BP level than office BP,[7, 8] and that the predictive value of 24-hour ambulatory BP for cardiovascular morbidity and mortality may be superior to office BP.[9-12] The present study aimed to evaluate relationships between 24-hour ambulatory BP and components of the insulin resistance syndrome, and further, to investigate the prognostic significance of 24-hour ambulatory BP for cardiovascular morbidity in a longitudinal population-based cohort of elderly men.

7 KRISTINA BJÖRKLUND

Hypertension in the elderly

The ascendancy of systolic over diastolic BP - a paradigm shift The view on hypertension as a risk factor in the elderly has changed during the past 15-20 years. For many years, DBP was considered the most important measure of diagnosis and target for intervention, but recently, attention has focused more on the prognostic value of the SBP component, particularly in the elderly.[13, 14] An important factor that has influenced the new approach to risk assessment in the elderly, is the emerging evidence from observational studies and clinical trials, showing that SBP, as risk indicator, is superior to DBP. While the definition of hypertension until the beginning of the 1990’s mainly took into account the DBP, new treatment recommendations in 1993[15, 16] added SBP to the diagnostic criterion, since a number of important clinical trials published around that time had reported highly beneficial effects of antihypertensive treatment in elderly hypertensive patients.[17-19] Hypertension in the elderly is typically characterised by an elevated SBP, but normal DBP, a hypertensive form which had previously not been comprised by the treatment guidelines. According to the most recent World Health Organization-International Society of Hypertension (WHO-ISH) guidelines from 1999, the upper limit of normal BP is considered similar, 140/90 mmHg, for all age categories.[20] Isolated systolic hypertension is the most common hypertensive subtype in elderly individuals,[21, 22] and constitutes a powerful predictor of cardiovascular disease.[6, 23-25] Furthermore, a wide pulse pressure has been identified as an independent risk factor for myocardial infarction,[26, 27] and may in fact be a stronger predictor of cardiovascular morbidity than mean arterial pressure in elderly hypertensive subjects.[28, 29] During recent years, the results of a number of randomised controlled trials have shown significant reductions of cardiovascular morbidity and mortality in elderly with isolated systolic hypertension.[18, 30, 31] Despite these documented benefits of antihypertensive treatment, BP, and in particular SBP, is however still inadequately controlled in a large proportion of treated hypertensive patients.[32, 33]

Pathophysiological effects of aging on the vascular system With advancing age, there is a progressive increase in SBP while the level of DBP plateaus, or even decreases,[5] resulting in a wider pulse pressure, i.e. difference between SBP and DBP. The age-related change in DBP may in part result from early cardiovascular death in subjects with high DBP levels, but is in industrialized societies also reflected by structural changes of the large capacitance vessels due to arteriosclerosis. Simultaneously, the elastic properties of the large arteries decrease, as the amount of elastin is reduced

8 24-HOUR AMBULATORY BP while the proportion of collagen in the arterial wall is increased. The conse- quence of these degenerative changes is thickening, hardening and dilata- tion of the aorta, carotids and other large elastic arteries, whereas the smaller peripheral arteries contain less elastin, and therefore are more durable. The structural changes of the large arteries lead to a reduced distensibili- ty, and an increased velocity by which the pulse wave is transmitted from the left ventricle through the arterial tree, causing an early backward return of the pulse wave from the periphery. This premature pulse-wave reflection causes an augmentation of the pressure during late systole, and a subsequent increase in SBP and decrease in DBP.[34, 35]

Hypertensive target organ damage Hypertensive disease has detrimental effects on structure and function of the vascular system and target organs. Hypertensive target organ damage may be regarded as intermediate end-points, as opposed to cardiovascular morbidity and mortality, termed “hard end-points”. The degree or presence of left ventricular hypertrophy, carotid intima-media thickening, microalbuminuria, increased serum creatinine and retinopathy, as measures of target organ damage, are important determinants of prognosis in hypertensive patients.[20, 36, 37] On the other hand, regression of these parameters may be used as markers for beneficial effect of antihypertensive treatment. The decision to treat a hypertensive patient should be based on evaluation of the total risk profile, including target organ damage and other prognostic factors such as smoking, diabetes, lipid profile and family history of cardiovascular disease.[20]

Hypertension as a risk factor for cardiovascular disease The major hazards of hypertension are the increased risk of coronary heart disease and stroke. The associations between BP level and these outcomes are strong, continuous, and independent, although the DBP component is the major risk determinant in younger populations,[2] and the SBP component in the elderly.[25] Thus, there does not seem to be an obvious thresh- old, below which the risk related to BP is negligible. The consequences of this apparently linear fashion of association between BP and cardiovascular risk was recently illustrated by observational data from the Framingham Heart Study, showing that a BP below the upper limits of normal, i.e. SBP 130-139 and/or DBP 85-89 mmHg, is a predictor of cardiovascular disease, particularly in elderly individuals.[38] A meta-analysis of previous intervention trials evaluating the effect of antihypertensive drug therapy, has reported that by lowering DBP 5-6 mmHg, 42% of strokes, but only 14% of coronary heart disease can be prevent- ed.[39] One of the reasons for the dissimilarity may be that the early clinical trials used solely an elevated DBP as inclusion criterion. This might have

9 KRISTINA BJÖRKLUND introduced a bias in the interpretation of the results, since subjects with low DBP and high SBP, who are at high risk for coronary heart disease due to a reduced coronary perfusion, were not included in those trials.[40] The “par- adox of coronary heart disease in hypertension” has also been attributed to other concomitant coronary risk factors, such as metabolic disturbances and increased sympathetic activity, which may not be reduced by antihypertensive treatment, but continue to contribute to an adverse outcome despite BP lowering.[41]

Hypertension and insulin resistance The clustering of hypertension with metabolic aberrations has been termed the insulin resistance syndrome, or metabolic syndrome,[42, 43] and con- tributes to the increased risk of hypertensive target organ damage[44-46] and cardiovascular morbidity[47] in hypertension. There are several slightly different working definitions of the insulin resistance syndrome,[48-50] of which the most recent suggests that at least three of the following criteria should be satisfied: Abdominal obesity, hypertriglyceridemia, low HDL- cholesterol level, hypertension and hyperglycemia.[50] It has recently been reported,[51] that the prevalence of the insulin resistance syndrome, using the above definition, approximates 22% of the population in the United States, implying that metabolic abnormalities are highly prevalent, with important implications for cardiovascular health in the general population. The etiology of hypertension is complex, and involves genetic as well as environmental factors. Among others, the interrelated factors obesity, hyperinsulinemia and insulin resistance to glucose uptake, have been identified as possible predictors of hypertension.[52, 53] Increased sympathetic nervous activity, which is particularly evident in the early phases of hypertension development, has been proposed as a possible common link between many of the metabolic abnormalities characteristic of the insulin resistance syndrome.[41] Moreover, dietary factors are associated with BP,[54] and the quantity as well as the quality of the dietary fat seems to be of importance. The dietary fat quality has previously been related to insulin sensitivity[55] and shown to predict the development of type 2 diabetes[56] in the presently studied ULSAM population. The Dietary Approaches to Stop Hypertension (DASH) Trial, showed that a combination diet rich in fruits and vegetables, and low in total and saturated fat had a significant BP- lowering effect during 24 hours.[57] However, the possible long-term longitudinal relationship between dietary fat intake and hypertension still re- mains to be elucidated.

10 24-HOUR AMBULATORY BP

Circadian BP variation When biological systems exhibit prominent and reproducible patterns oc- curring regularly over a 24-hour period, they are referred to as circadian. During recent years, there has been an increasing interest in the circadian BP variability, and its impact on cardiovascular risk. The occurrence of continuous fluctuations in BP was first described by Stephen Hales in 1733, who measured BP by inserting a brass tube into the carotid artery of a horse. He concluded from these pioneering experiments, that BP was “never to be exactly the same, any two minutes, throughout the whole life of an ani- mal”.[58] It is now known that a number of factors involved in atherothrombotic formation; hemodynamical, neural, hormonal and hematological, undergo a regular variation during day and night. The incidence of cardiovascular events such as myocardial infarction,[59] stroke[60] and sudden cardiac death[61] also display a circadian rhythm, with a peak incidence in the morning when BP, heart rate and platelet aggregation are increased and fibrinolysis is low. Therefore, it has been suggested that these dynamical, physiological mechanisms may contribute to, or maybe even trigger, the onset of acute cardiovascular events. Fluctuations in BP are influenced by external factors in daily life (Table 1) as well as by an internal biological rhythm. Plasma cortisol reaches a peak

Table 1. Effect of different daily activities on BP level. The changes shown are relative to BP while relaxing. (Adapted from Campbell [190] )

Effect on blood pressure (mmHg) Activity Systolic BP Diastolic BP Attending a meeting + 20 + 15 Walking + 12 + 6 Talking on telephone + 10 + 7 Reading + 2 + 2 at the time of awakening,[62] with subsequent increases in plasma catecho- lamines during the late morning hours. The a-sympathetic vasoconstrictor activity shows a circadian variation, with an increased vascular tone in the morning which may contribute to the increased BP observed at this time.[63] Furthermore, autonomic cardiovascular mechanisms, in particular the arterial baroreflexes, are important determinants of the degree of BP variability.[64] Variability of BP can be assessed over shorter or longer time-periods, and may be described as morning BP rise, night-day BP difference, or as the standard deviation of the average BPs obtained during the day or night. An

11 KRISTINA BJÖRKLUND increased BP variability over 24 hours is cross-sectionally associated with target organ damage in hypertensive patients[65, 66] and in the general population,[67] and also seems to carry prognostic information for the development of target organ damage.[68, 69] Longitudinal findings, although currently sparse, further suggest that BP variability may predict cardiovascular mortality, independently of BP level.[70]

Twenty-four-hour ambulatory BP monitoring BP is commonly measured in the clinic, and repeated clinic, or office BPs, are used as a basis for treatment decisions and evaluation of BP control. However, office BP monitoring has several limitations. Potential problems that may affect the accuracy of the measurement include observer bias such as terminal digit preference and measurement error. There is also a possibility that BP measured at the office is not representative of the “true” BP in daily life, and that some subjects experience an excessive emotional response to the clinic environment, termed “white-coat hypertension”. Twenty-four hour ambulatory BP monitoring, on the other hand, offers a reading devoid of the white-coat effect, and more accurately reflects the actual 24-hour BP load imposed on the vascular bed. In addition, ambulatory BP monitoring provides valuable information beyond that obtained with conventional BP readings, since entities such as BP variability and night- time BP level, which seem to have prognostic significance, can only be measured with 24-hour ambulatory monitoring. Reference values of normal 24-hour BP have recently been proposed,[71] and ambulatory BP monitoring is still used to a limited extent in clinical practice. According to current guidelines, ambulatory BP monitoring is es- pecially indicated in certain subgroups, such as patients with borderline hypertension, pregnant women, patients with treatment-resistant hypertension, in patients with symptoms suggesting episodes of hypotension and in elderly individuals.[20, 72] There are several reasons why the elderly may benefit from ambulatory BP monitoring. Elderly subjects more often than younger display an increased BP variability, which might obscure office BP readings. In addition, elderly have a higher prevalence of hypotensive states due to autonomic dysfunction, and the symptoms may worsen by a high susceptibility to the adverse effects of antihypertensive drugs. Misclassification of hypertension leading to unnecessary medication is therefore particularly harmful in older patients. Closer relationships have previously been found between hypertensive target organ damage and ambulatory BP than office BP,[7, 8] and prospective studies have indicated that 24-hour ambulatory BP is an independent predictor of cardiovascular disease,[9-12] as shown in Table 2. Information

12 24-HOUR AMBULATORY BP regarding the prognostic role of 24-hour ambulatory BP in elderly subjects from the general population, is however limited.

Table 2. Previous longitudinal studies investigating prognostic value of ambulatory BP.

Mean Sntudy Study populatio follow-up E)nd-point Risk ratio (95 % CI (years)

n=1542, general population, Cardiovascular 1.047 (1.018-1.076) † Ohkubo [11] 5.1 mean age ~62 yrs mortality (/1mmHg SBP)

n=86, refractory hypertension, Cardiovascular 6.20 (1.38-28.1) † Redon [191] 4.1 mean age 53 yrs morbidity (third vs first tertile DBP)

n=2010, hypertension, Cardiovascular 1.23 (1.12-1.35) † Verdecchia [167] 3.8 mean age 52 yrs morbidity (/10mmHg SBP)

n=688, hypertension, Cardiovascular 1.14 (1.06-1.23) Khattar [166] 9.2 mean age 51 yrs morbidity (/10mmHg SBP)

n=808, isolated systolic 4.4 Cardiovascular 1.17 (1.01-1.35) † Staessen [12] hypertension, mean age 70 yrs (median) morbidity (/10mmHg SBP) † indicates that the risk is independent of office BP level

Nondipping The BP normally decreases during night-time. A reduced nocturnal BP reduction, or nondipping BP pattern, was first reported in conditions such as autonomic failure[73] and diabetes,[74] but nocturnal BP fall has later been shown to be normally distributed in the general population.[75] A blunted reduction of nocturnal BP has been associated with increased left ventricular mass,[66] cerebrovascular disease,[76] and renal dysfunction.[77] How- ever, due to the cross-sectional design of these studies, it can not be concluded whether the organ damage is a cause or effect of the disturbed night- day BP pattern. Furthermore, other investigators have reported that nondippers and dippers do not differ regarding hypertensive target organ damage,[78, 79] suggesting that the clinical implication of the nondipping phenomenon needs further clarification. Several previous studies investigating the cardiovascular risk associated with nondipping have been performed in selected populations of hypertensive subjects, and some have not adjusted for diabetes in their analyses. The question therefore persists whether nondipping is a marker of risk, independently of diabetes, in a population setting.

White-coat hypertension and Isolated ambulatory hypertension As illustrated in Figure 1, four different states are possible to identify according to office and ambulatory BP monitoring: first, normotension at both recordings; second, white-coat hypertension, when office BP is elevated but

13 KRISTINA BJÖRKLUND daytime ambulatory BP is normal; third, isolated ambulatory hypertension, when office BP is normal but ambulatory daytime BP elevated; and fourth, sustained hypertension, defined as an elevation of both office and ambulatory daytime BP. White-coat hypertension is a condition characterized by a persistently raised office BP in combination with a normal daytime ambulatory BP.[72] The white-coat effect, on the other hand, is defined as a transient rise in BP from before to during the visit at the clinic,[80] most reliably measured

Isolated Fig 1. Subgroups possible to identify Sustained according to office and ambulatory ambulatory hypertension hypertension BP measurement.

White-coat Normal Ambulatory BP hypertension

Office BP with a continuous beat-to-beat BP monitoring technique. When there is a discrepancy between office and ambulatory pressure, as in white-coat hypertension, the prognosis has been suggested to be more closely related to ambulatory BP.[81] However, the clinical relevance of white-coat hypertension is still being disputed. An association between white-coat hypertension and hypertensive target organ damage has been indicated by some previous investigators,[82-84] although this could not be supported by others.[85, 86] Furthermore, an increased resting heart rate and metabolic aberrations have been observed to be similar in white-coat and sustained borderline hypertensives, implying that white-coat hypertension is not an entirely innocent phenomenon.[87] Prognostic data suggest that white-coat hypertension may not be associated with a higher cardiovascular morbidity than normotension,[88, 89] but additional studies are required to confirm those findings. Moreover, by gaining information about predictors of white-coat hypertension we may increase the understanding of the mechanisms behind this condition, as opposed to sustained hypertension. White-coat hypertension has been quite extensively studied over the past years. However, little is known about the prevalence and prognostic role of the converse phenomenon, isolated ambulatory hypertension, characterized by an increased BP during daytime but normal office BP. Recently, cross- sectional data indicated that isolated ambulatory hypertension may be associated with metabolic risk factors and target organ abnormalities to a similar extent as sustained hypertension, a state with elevated office and ambulatory BP.[90, 91] The prognostic significance of this condition is thus far unknown. 14 24-HOUR AMBULATORY BP Aims of the study

I. To describe ambulatory and office BP characteristics and establish reference values of normal ambulatory BP in a longitudinal population-based study of 70-year-old men, and further, to evaluate the prevalence of hypertension and BP control in treated elderly hypertensive subjects in this population. II. To cross-sectionally evaluate whether nondipping and diabetes are independently related to components of the insulin resistance syndrome and prevalence of hypertensive target organ damage, using metabolic and echocardiographically determined variables measured in 70-year-old men.

III. To examine whether metabolic status and biomarkers of dietary fat quality measured in men at age 50 may predict the development of sustained and white-coat hypertension over the following 20-year-period, and further, to cross-sectionally compare subjects with sustained and white-coat hypertension at age 70 regarding metabolic risk factors and hypertensive target organ damage.

IV. To investigate the prognostic significance of 24-hour ambulatory BP and BP variability for cardiovascular morbidity in 70-year-old men from the general population, over a mean follow-up of 6.6 years.

V. To cross-sectionally examine the metabolic and cardiac risk profile in, and investigate the prognostic value of isolated ambulatory hypertension for cardiovascular morbidity in 70 year-old men.

15 KRISTINA BJÖRKLUND Methods

Subjects

ULSAM The papers in this thesis are based on a cohort of men born in 1920-24, the Uppsala Longitudinal Study of Adult Men (ULSAM). This cohort originat- ed in 1970-73, when all 50-year-old men living in the municipality of Uppsala were invited to participate in a health examination, which aimed to detect risk factors for cardiovascular disease and identify high-risk individuals for intervention.[92] Of the 2841 subjects invited, 2322 men participated in the investigation at age 50 years (participation rate 82%). There- after, the participants have been invited to three re-investigations, at the approximate ages of 60, 70 and 77. In this thesis, data are used from the investigations at age 50 and 70 years (Figure 2). Between the baseline investigation in the 1970s and 1991, 422 subjects had died and 219 had moved out of the Uppsala region. In 1991-95, 1221 of the 1681 men who were

Figure 2. The ULSAM population.

1970-73 1991-95

Age 50 Age 70

2841 2322 1681 1221 invited participants invited participants

217 died (2000-12-31)

422 died

219 moved

16 24-HOUR AMBULATORY BP alive and still living in Uppsala, took part in the re-investigation (participation rate 73%). At this re-investigation at age 70 years, a 24-hour ambulatory BP monitoring was performed. The study was approved by the Ethics Committee of Uppsala University, and all participants gave written informed consent. All investigations were done after an overnight fast. A reproducibility study was performed in 22 subjects approximately 1 month after the examination at age 70 years, from which the intra-individual coefficients of variation (CVs) are presented.

Study Populations A schematic description of the study populations is provided in Figure 3.

Paper I. The study population in Paper I consisted of the 1060 70-year- old-men with technically satisfactory recordings from the ambulatory BP monitoring, and non-missing information regarding antihypertensive treatment at the investigation at age 70 years. Paper II. Of the 1060 70-year-old subjects who constituted study population in Paper I, 1057 men with information on anti-diabetic treatment were also included in Paper II. These 1057 men were the basis for the study populations in Paper III-V. Measures of hypertensive target organ damage, represented by urinary albumin excretion rate (UAER) and echocardiographic data, were analysed in those subgroups of men with available information on these respective variables. Paper III. From the 1057 70-year-old men defined in Paper II, 373 subjects with antihypertensive treatment were excluded, as were also 82 subjects with elevated ambulatory but normal office BP, leaving 602 men for the analysis. In Paper III, two separate analyses were performed, one cross- sectional at age 70 years, and one longitudinal, where data from the investigation at age 50 years were examined in subjects who had been divided in subgroups on the basis of BP readings at age 70 years. In the cross-sectional analysis, data regarding urinary albumin excretion rate (UAER) and echocardiographic measures, were analysed in subgroups of men with available information on these respective variables.

Paper IV. Men from the investigation at age 70 years, who fulfilled the inclusion criteria for Paper II (n=1057), and who had non-missing information regarding smoking habits were included in Paper IV. Moreover, subjects who had previously been hospitalized due to coronary heart disease (Inter- national Classification of Diseases (ICD) -9 codes 410-414; n=125) and cerebrovascular disease (ICD-9 codes 431-436; n=42) were excluded. The remaining 872 subjects formed study population in Paper IV.

17 KRISTINA BJÖRKLUND

Figure 3. Description of study populations in Papers I-V. BP, blood pressure; UAER, urinary albumin excretion rate.

Age 70 years 108 missing data office or ambulatory BP I 1221 1060 53 missing data antihypertensive treatment

Age 70 years 3 missing data antidiabetic 1017 treatment with UAER II 1060 1057 426 with echocardiography

Age 70 years 373+82 584 excluded with UAER III 1057 602

Age 50 years 288 with echocardiography 602

Age 70 years 167 excluded 1057 872 IV 18 missing data smoking habits 234 with echocardiography Age 70 years 373+106 excluded 60 excluded V 1057 578 512

6 missing data smoking habits

18 24-HOUR AMBULATORY BP

Paper V. The 1057 men with satisfactory information regarding ambulatory BP and treatment for hypertension and diabetes at the investigation at age 70 years, constituted the basis for Paper V. We excluded 373 subjects who were taking antihypertensive medication, as well as 106 individuals with normal daytime ambulatory, but elevated office BP, leaving 578 men to constitute study population. Prior to the investigation at age 70 years, 40 subjects had been hospitalized due to coronary heart disease (ICD-9 codes 410-414) and 20 due to stroke (ICD-9 codes 431-436), and these 60 subjects were excluded from the survival analysis, as well as 6 subjects with missing information on smoking habits.

Antihypertensive treatment Of all 2322 50-year-old men who participated in the investigation in 1970- 73, 98 subjects were treated with antihypertensive agents. In the present thesis, data from the investigation at age 50 were used in 602 men (Paper III), and these 602 subjects were free from antihypertensive treatment at age 50 years. At the investigation at age 70 years, 285 (27%) of the 1060 men with technically satisfactory ambulatory BP recordings (Paper I) were under regular treatment with antihypertensive medication, whereas 9% were taking drugs with antihypertensive properties for other reasons than hypertension (post-myocardial infarction, angina etc). Of the 285 70-year-old men with treated hypertension, 155 were on monotherapy, 104 had two antihypertensive drugs and 26 had three or more drugs. The most frequently used antihypertensive drug was a beta-blocker, used by 162 subjects, whereas a calcium antagonist was used by 110 subjects, diuretics by 103 subjects, ACE-inhibitors by 55 subjects and alpha- blockers by 13 subjects.

Investigations

Measurements at age 50 years Anthropometry. Height (without shoes) was measured to the nearest whole cm and weight (in undershorts) to the nearest whole kg. Body mass index (BMI) was calculated as weight (in kg) divided by height (in meters) squared. Office BP. BP in the right arm was measured after 10 minutes’ rest in the recumbent position, using a mercury manometer (Kifa Ercameter, wall-model). The BP cuff was 12.5 cm wide and 35 cm long. SBP and DBP were read to the nearest 5 mmHg mark. The radial pulse rate was counted after 10 minutes’ rest before the BP measurement.

19 KRISTINA BJÖRKLUND

Glucose and insulin. Whole blood glucose concentration was measured by using the glucose oxidase method. An intravenous glucose tolerance test was performed by injection of a 50% solution of glucose at a dose of 0.5 g/kg body weight into an antecubital vein during 2.5 minutes. Blood glucose concentrations were determined before and 6, 20, 30, 40, 50 and 60 minutes after the start of the glucose injection, and the levels between 20 and 60 minutes were used for calculation of glucose tolerance. This was expressed as the K-value calculated from the formula: K=ln2·100/T1/2 where T1/2 is the time in minutes required for the concentration to be reduced by half its value. Specific insulin concentrations were determined using the Access Immunoassay System (Beckman-Coulter), which uses a chemilumi- nescent immunoenzymatic assay. These analyses were carried out between 1995 and 1998 in Cambridge, UK, using plasma samples that had been stored frozen (in -70°C) since sampling in 1970-73.

Serum lipids. Determinations of serum cholesterol and triglyceride concentrations were performed on a Technicon Auto Analyzer type II[93] in 1981- 82 on serum samples that had been stored in liquid nitrogen since 1970-73. HDL cholesterol was assayed in the supernatant after precipitation with a heparin/manganese-chloride solution. LDL cholesterol was calculated using Friedewald’s formula: LDL=serum cholesterol-HDL-(0.42·serum triglycerides). The fatty acid composition in serum cholesterol esters (CE) was determined using gas-liquid chromatography, as previously described in detail.[56] The CE fatty acids are presented as the relative percentage of the sum of the fatty acids analysed. Questionnaire. Data regarding medical treatment and heredity for hypertension were collected through a self-administered questionnaire. Informa- tion on smoking habits (smoking, non-smoking) was retrieved through an interview by a physician.

Measurements at age 70 years Anthropometry. Height was measured to the nearest whole cm, and body weight to the nearest 0.1 kg. The waist and hip circumferences were measured in the supine position, midway between the lowest rib and the iliac crest and over the widest part of the hip, respectively. The waist/hip ratio was calculated.

Office BP. Office BP was measured in the right arm with a sphygmomanometer. The cuff size was 12.35 cm or 15.45 cm depending on the arm cir- cumference. The recordings were made to the nearest 2 mmHg twice after 10 minutes supine rest, and the mean of the two measurements were used for the analyses. SBP and DBP were defined as Korotkoff´s phases I and V, 20 24-HOUR AMBULATORY BP respectively. Mean arterial pressure (MAP ) was calculated according to the formula MAP = DBP + 1/3 (SBP - DBP). We estimated pulse pressure as the difference between SBP and DBP. The resting heart rate was measured as supine pulse rate during one minute. The CVs for office SBP and DBP were 7.2% and 4.9% respectively.

Ambulatory BP. Ambulatory BP was recorded using the Accutracker 2 equip- ment (Suntech Medical Instruments Inc, Raleigh NC, USA). The device was attached to the subject´s non-dominant arm by a skilled lab technician, and BP was measured during 24 hours starting at 11 a.m. BP recordings were made every 20th minute during the whole 24-hour period. Prior to November 1993, BP was measured every 30th minute during daytime (06.00- 23.00) and every hour during night-time (23.00-06.00). Data were edited by omitting all readings of zero, all heart rate readings <30, DBP readings >170 mmHg, SBP readings >270 and <80 mmHg, and all readings where the difference between SBP and DBP was less than 10 mmHg. Subjects with ≥4/10 hours of recorded and technically satisfactory BP during daytime, and respectively, ≥3/6 hours of recorded BP during night-time, were included in the study. SBP and DBP were given by the auscultatory device. Mean arterial pressure was calculated according to the formula MAP = DBP + 1/3 (SBP - DBP). We estimated pulse pressure as the difference between SBP and DBP. BP variability was defined as the standard deviation (SD) of the hourly mean BP values during daytime and night-time for SBP and DBP respectively. The CVs for 24-hour SBP and DBP were 6.8% and 5.5% respectively. Glucose and Insulin. Plasma glucose was measured by the glucose dehydro- genase method (Gluc-DH, Merck). An oral glucose tolerance test was performed where the subjects ingested 75 g glucose dissolved in 300 ml of water, and blood samples for plasma glucose and insulin were drawn imme- diately before, and 30, 60, 90, and 120 min after ingestion of glucose. Plasma insulin was assayed using an enzymatic-immunological assay (En- zymmun, Boehringer Mannheim) performed in an ES300 automatic analyser (Boehringer Mannheim). Intact proinsulin and 32-33 split proinsulin concentrations were analysed using the two-site immunometric assay technique.[94] Specific insulin concentrations were determined using the Ac- cess Immunoassay System (Beckman-Coulter). Insulin sensitivity was assessed by using the euglycaemic hyperinsulinae- mic clamp technique according to DeFronzo et al, but slightly modified, whereby insulin (Actrapid Human(r), Novo) was infused at a rate of 56 (instead of 40[95]) mU/min per body surface area (m2). Glucose uptake (M) was calculated as the amount of glucose (milligrams) infused per minute per body weight (kilograms). The insulin sensitivity index (M/I) was calculated by dividing M by the mean insulin concentration during the same

21 KRISTINA BJÖRKLUND period of the clamp and represents the amount of glucose metabolised per unit of plasma insulin (mg×min-1×kg-1/(100mU/L)).

Serum lipids. Cholesterol and triglyceride concentrations in serum were assayed by enzymatic techniques (Instrumentation Laboratories) in a Mon- arch 2000 centrifugal analyser. HDL particles were separated by precipitation with magnesium chloride/phosphotungstate. LDL cholesterol was calculated using Friedewald’s formula (above). Nonesterified fatty acids were measured by an enzymatic colorimetric method (Wako Chemical GmbH).

PAI-1-activity. Plasminogen activator inhibitor-1 (PAI-1) activity was analysed with a commercial two-step indirect enzymatic assay (Spectrolyse/pL PAI, Biopool AB) as described previously.[96, 97] Echocardiography. M-mode, two-dimensional and Doppler echocardiographic examinations were performed in the first 583 consecutive subjects in the original study population at age 70,[98] leaving 488 (Paper II), 288 (Paper III) and 234 (Paper IV) subjects with echocardiographic data in the respective Papers of the present thesis. Left ventricular mass was determined by using the M-mode formula of Troy according to the recommendations of the American Society of Echocardiography,[99] and was further divided by body surface area to obtain left ventricular mass index. Relative wall thickness was calculated as (Intraventricular septal thickness + posterior wall thickness/left ventricular end diastolic diameter). The CVs were 12.5% and 6.9% for left ventricular mass index and relative wall thickness respectively. Urinary albumin excretion. Albumin was measured in urine (Albumin RIA100, Pharmacia, Sweden) and the urinary albumin excretion rate (UAER) was calculated on the amount of urine collected during the night together with the first sample of urine after rising. Microalbuminuria was defined as a UAER of 20-200µg/min.

Questionnaire. Data concerning ongoing medical treatment were collected with a self-administered questionnaire. Leisure time physical activity was assessed using the four questions: 1) Do you spend most of your time reading, watching TV, going to the cinema or engaging in other, mostly sedentary activities?, 2) Do you often go walking or cycling for pleasure?, 3) Do you engage in any active sport or heavy gardening for at least 3 hours every week?, 4) Do you regularly engage in hard physical training or competitive sport? Based on these questions, four physical activity categories were con- structed: Sedentary, Moderate, Regular and Athletic. Information on smoking habits (smoking, non-smoking) was retrieved through interview reports. A dietary assessment was performed using a 7-day precoded food record; prepared and validated by the National Food Administration (NFA).[100]

22 24-HOUR AMBULATORY BP

Follow-up The cohort was followed for up to 9.5 years (Paper IV) since the investigation at age 70, with a mean follow-up period of 6.6±2.1 years. Outcome events were assessed using data from the Swedish Hospital Discharge and Cause of Death Registries (censor dates December 31, 1999 (Paper V) and December 31, 2000 (Paper IV) respectively). Cardiovascular morbidity was defined as a composite end-point, including death from coronary heart disease (ICD-9 codes 410-414, or ICD-10 codes I20-I25) and stroke (ICD-9 codes 431-436, or ICD-10 codes I61- I66), as well as first hospitalisation for non-fatal coronary heart disease and stroke (Paper IV). In Paper V, we also included death from peripheral vascular disease (ICD-9 codes 440-444, or ICD-10 codes I70-I74) in the composite end-point. During 9.5 years of follow-up, contributing to 5784 person-years at risk (PYAR), a total of 172/872 cardiovascular events (2.97/ 100 PYAR) occurred, including 115 coronary events, of which 27 were fatal, and 57 strokes, of which 7 were fatal. Cardiovascular morbidity, defined by combining data from the Hospital Discharge and Cause of Death Registries, is an efficient, validated alternative to revised hospital discharge notes and death certificates.[101, 102]

Definitions In Papers I and II, hypertension was defined as either taking regular antihypertensive treatment, or having an office SBP ≥140 and/or DBP ≥90 mmHg, according to the guidelines of JNC VI[103] and WHO-ISH[20]. For the 24-hour ambulatory BPs, we used short fixed-clocktime intervals and defined daytime as 10 a.m. to 8 p.m. and nighttime as midnight to 6 a.m.[75, 104] Besides the diary method, considered most reliable, fixed-time methods where the morning and evening phases are excluded increases the accuracy in estimating the actual sleep and awake periods.[105] An elevated daytime ambulatory BP was defined as ≥135/85 mmHg.[72, 103] In Paper III, where all subjects were untreated, we distinguished between white-coat hypertension, defined as office BP ≥140/90 and daytime BP <135/85 mmHg, and sustained hypertension, defined as office BP ≥140/90 and daytime BP ≥135/85 mmHg. Subjects with normal office BP (<140/90 mmHg) and daytime ambulatory BP ≥135/85 mmHg were entitled isolated ambulatory hypertensive in Paper V. The ratio between night SBP and day SBP was used to define the dipping status (Paper II). A night-day SBP ratio of ≥1.0 was considered nondipping, whereas a ratio of <0.8 was regarded extreme dipping. Dipping was defined as a night-day SBP ratio <1.0, except for in a separate analysis of extreme dipping, where dippers were subdivided into extreme dippers (defined above) 23 KRISTINA BJÖRKLUND and moderate dippers (>0.8 - <1.0). In Paper I, the night-day ratio was multiplied by 100, expressing nighttime BP as a percentage of the daytime level. Diabetes (Paper II-V) and impaired glucose tolerance (Paper II) at age 70 years were defined according to the 1985 WHO criteria.[106] A serum cholesterol level >6.5 mmol/L and/or use of lipid-lowering medications was characterized as hyperlipidemia (Paper IV).

Statistical analyses The statistical software packages Stata 6.0 (Stata Corporation; Papers I-V), JMP and SAS 8.2 (SAS Institute, Cary, NC; Papers I and IV respectively) were used for the analyses. Distributions were tested for normality by Sha- piro-Wilk´s W test, and when skewed, variables were logarithmically transformed to reach normal distribution. Two-tailed significance values were given with p<0.05 regarded as significant. We used ANOVA to calculate differences in means, and comparisons between subgroups were carried out if the overall F-test was significant. Bonferroni correction was performed in post-hoc analyses of more than two groups. In Paper I, the correspondence between clinic and ambulatory BP was analysed by linear regression analysis. In Paper II, a number of metabolic variables and variables indicating target organ damage were considered outcome variables. For each outcome variable the relation to nondipping and diabetes were estimated from a model in which these factors and their interaction were analysed. For outcome variables where the interaction term was non-significant a limited model was used, excluding this term. Proportional differences between the groups were calculated by the chi-square test. The impact of potential confounders was taken into account by using analysis of covariance (Papers II-III). Cox Proportional Hazard Regression was used to calculate crude and multivariate-adjusted hazard ratios of cardiovascular morbidity and their 95% confidence intervals (CI) over the time of follow-up since the investigation at age 70 years (Papers IV-V). In Paper IV, longitudinal relationships between BP and cardiovascular morbidity were assessed using standardized continuous BP variables, so that the hazard ratio from the Cox Regression reflected the predictive value of a 1 SD increase in a BP component. In multivariate Cox models, adjustment was made for potential confounders, treating categorical variables as dichotomous (1/0), and continuous variables as standardized variables (Papers IV-V). In Paper V, the hazard ratio represented the risk associated with the hypertensive categories defined at age 70 years. To test the significance of the additional prognostic information obtained by 24-hour pulse pressure in Paper IV, likelihood ratio tests

24 24-HOUR AMBULATORY BP were performed between pairs of maximum likelihood models, giving a χ2 statistic which was twice the difference between the log-likelihoods between models including and excluding 24-hour pulse pressure. To test the null hypothesis of equal hazard ratios for two BP variables in different models, in which a set of other variables were the same, we used a bootstrap method (Paper IV). The bootstrap method was based on 10 000 replicated samples.[107]

Discussion of Methods

Methodology of BP measurements It is known that the BP in clinical studies tends to be higher at the first visit compared with later visits. According to international standards, a diagnosis of hypertension should be based on repeated BP measurements at separate clinic visits, the number of office measurements being dependent on the BP level and the overall cardiovascular risk profile.[108] Office BP in the UL- SAM cohort, like in many population studies, was the mean of two readings at a single visit, a simplified screening procedure which may have induced measurement error, and possibly overrated the number of subjects classified as hypertensive according to office BP. Office BP was measured in the supine position both when the subjects were 50 and 70 years, in agreement with Swedish clinical practice. Supine measurements tend to cause slightly lower DBP than when BP is measured sitting, however this difference decreases with age.[109] At the investigation at age 50 years, the office BP was either taken by one registered nurse, or by one physician (Hans Hedstrand). The mean BP obtained by the physician in 231 subjects was 131.5/83.0 mmHg. The corre- sponding BP obtained by the nurse in 216 men was 131.8/83.7 mmHg, suggesting that the systematic error induced by different observers was very small.[110] In the 70-year-old men, office BP was measured by a nurse in all subjects. The ambulatory BP recordings were performed by a skilled lab technician. Subjects were asked to try to hold the arm still when the cuff was inflated, to facilitate the recording. The BP measurement procedure changed during the time of investigation, in that BP during night-time was measured once every hour between August 1991 and October 1993, and once every 20 minutes between November 1993 and May 1995. This increased frequency of readings may have increased the precision of the mean BP calculated from all intermittent readings in the subjects investigated during the latter part of the investigation period. Compared with intra-arterial continuous ambulatory BP recordings, which is considered the golden standard method

25 KRISTINA BJÖRKLUND of 24-hour BP measurement, the now widely used non-invasive techniques of intermittent BP recordings provide accurate assessment of true average BP.[111] The Accutracker II device has recently been validated against protocols from the British Hypertension Society (BHS) and the American Association for the Advancement of Medical Instruments (AAMI).[112] Accutracker II fulfilled the AAMI protocol, but failed to pass the BHS protocol due to an insufficient agreement with mercury standard regarding DBP.[113] Howev- er, the Accutracker II device showed satisfactory accuracy and precision according to available documentation at the time of investigation in the beginning of the 1990´s.[114, 115] In the present study, the reproducibility of ambulatory BP was high, in agreement with previous investigations of elderly subjects.[116]

Dichotomization of a continuous variable - classification of subgroups Twenty-four hour ambulatory BP monitoring has made it possible to identify subjects with elevated night-time BP, blunted nocturnal BP reduction, increased BP variability and discrepancies between office and daytime BP elevation. BP data can be presented and analysed as either continuous or categorical variables. The use of continuous variables may be preferable, since this approach takes into account the fact that BP is related to cardiovascular disease risk in an approximately linear fashion,[2] meaning that the risk starts to increase with an increase in BP even at levels below the cut-off for what is regarded as hypertension.[38] However, clinicians need guidance by limits of normal BP in order to make decisions regarding diagnosis and treatment. The term hypertension as well as the concepts of white-coat hypertension, isolated ambulatory hypertension and nondipping, inevitably implies a classification of subjects into subgroups on the basis of cut-off limits of ambulatory and office BP. This assumes an arbitrary dichotomization of continuous variables, and until reference values have been agreed upon, caution should be taken regarding various definitions of the phenomena. Risk assessment of white-coat hypertension and nondipping will be highly dependent on the definition used.

Selection bias The 161 subjects who were excluded from the initial 1221 subjects investigated at age 70 (Paper I), were evenly distributed across the birth-span 1921- 24, with the exception of a slightly larger amount of the men born in 1920 being excluded. Since during the first 2 years of the investigation, night- time BP was measured but once every hour, and the men born in 1920 were

26 24-HOUR AMBULATORY BP the first to be examined, there may have been an over-representation, among them, of a lower frequency of ambulatory BP readings whereby a missing value would have greater impact on the total number of readings during the night. This may be one possible explanation to the higher degree of exclu- sion in this group. More importantly, the subjects who were excluded did not differ from the study population regarding office BP level, prevalence of diabetes or history of myocardial infarction. The subsample of the population with echocardiographic data, was constituted by the first 583 consecutive subjects in the original study population, as previously described,[98] and not subject to selection. In Papers III and V, subjects treated with antihypertensive agents were excluded. This may have induced some selection bias, since untreated individuals can be assumed to be more healthy from a cardiovascular perspec- tive. Therefore, the differences and associations that were found in the analyses, may have been somewhat underrated.

27 KRISTINA BJÖRKLUND Results

Paper I

Reference values for 24-hour ABPM in 70-year-old men - Distribution vs Correspondence criteria Average 24-hour BP in the population was 133±16/75±8, and daytime BP 140±16/80±9 mmHg, the respective values in untreated subjects being 131±16/74±7 and 139±16/79±8 mmHg. Two different approaches were used to derive an upper limit of normal ambulatory BP in the population. First, in subjects identified as normotensive according to office BP, the 95th percentiles of the 24-hour and daytime ambulatory BP distributions were 142/80 and 153/85 mmHg respectively. Second, an office recording of 140/ 90 mmHg corresponded to the ambulatory pressure 130/78 (24-h) and 137/83 mmHg (daytime) in untreated subjects. When the ambulatory BP levels were compared regarding ability to predict office hypertension, the number of correctly classified hypertensives was higher using the upper lim-

Table 3. Agreement between office BP and ambulatory BP in a classification of untreated subjects as hypertensive using different methods to derive upper limits of 24-hour ambulatory BP. Numbers are n (%) correctly classified as office normotensive or hypertensive according to the criteria used.

Office Office Hypertension Normotension

Distribution criteria 24-h ABP ≥142/80 171 (41) 24 (9) 24-h ABP <142/80 244 (59) 246 (91)

Correspondence criteria 24-h ABP ≥130/78 289 (70) 69 (26) 24-h ABP <130/78 126 (30) 201 (74) Total n (%) 415 (100) 270 (100)

28 24-HOUR AMBULATORY BP it of ambulatory BP derived from correspondence criteria, compared with using the 95th percentile of ambulatory BP in office normotensives (Table 3).

Hypertension prevalence and BP control The prevalence of hypertension according to office BP was 66%. The mag- nitude of BP control among treated hypertensives according to different treatment goals, is shown in Figure 4. In treated 70-year-old men, 14% had a BP below 140 mmHg systolic and 90 mmHg diastolic. Twenty-eight percent displayed a 24-h ambulatory BP <130/78 mmHg, a proposed upper limit of normality in this study. According to guidelines at the time when the study was carried out (160/90 mmHg), 39% fulfilled the treatment goals. When only the DBP was considered, around 50% of treated subjects reached a normal BP, using any of the three limits of normal BP.

(%)

80 Figure 4. Percentage of DBP 70 treated hypertensives with SBP and DBP diastolic (full bar) and systolic 60 as well as diastolic (light grey) 50 BP control according to 40 different BP thresholds. 30 20 10 0 Office BP 24-h ABP Office BP <140/90 <130/78 <160/90

Paper II We identified 66 nondippers and 991 dippers in the population of 70-year- old men. The SBP level in the two groups during 24 hours is illustrated in Figure 5. However, diabetes was more common in nondippers (26%) than in dippers (14%). Extreme dipping occurred in 32% of the population, but extreme dippers showed no differences in metabolic status, cardiac structure or function compared with moderate dippers, and these groups together were regarded as dippers in the following analyses.

29 KRISTINA BJÖRKLUND

160

140

) )

g g H

Fig 5. SBP level in dippers and H m

m 120

m m (

nondippers. The proportion of (

individuals with hypertension P B

B 100 S was not significantly different S Dippers in the nondipper group (74%) 80 compared with the dipper Nondippers group (65%). 60 10 14 18 22 2 6

Influence of diabetes Time (hour) on the relationship between nondipping and metabolic risk profile at age 70 years To investigate the effect of diabetes and nondipping, respectively, on metabolic status, and the possibility of an interaction between them, the dipper and nondipper groups were divided into groups according to diabetes status. The most pronounced abnormalities in glucose and lipid metabolism were restricted to diabetic nondippers. A significant interaction was demonstrated between diabetes and nondipping with regard to BMI, triglyceride and fasting glucose levels, whereby nondipping was associated to these metabolic aberrations only when diabetes was present (Figure 6). As a conse- quence, we identified a large group of nondippers in this population who lacked major metabolic disturbances.

Figure 6. Serum triglyceride and fasting plasma glucose levels in nondippers with diabetes (n=17), dippers with diabetes (n=136), nondippers without diabetes (n=49), and dippers without diabetes (n=855).

Serum Triglycerides (mmol/l) Fasting Plasma Glucose (mmol/l)

3 10

2,5 8 2 6 1,5 1 4 0,5 2 Diabetes Diabetes 0 0 Nondipping No Diabetes Nondipping No Diabetes Dipping Dipping

30 24-HOUR AMBULATORY BP

Influence of diabetes on the relationship between nondipping and hypertensive target organ damage at age 70 years Measures of left ventricular geometry and urinary albumin excretion were used as indices of hypertensive target organ damage. Target organ damage did not differ between nondippers and dippers in the whole population, but a significant interaction between nondipping and diabetes contributed to an increased left ventricular mass in diabetic nondippers (Figure 7). The urinary albumin excretion rate was independently related to diabetes, and no interaction between diabetes and nondipping was present.

LVMI (g/m2)

Figure 7. Left ventricular mass in 160 nondippers with diabetes (n=11), 150 dippers with diabetes (n=47), nondippers without diabetes (n=26), and 140 dippers without diabetes (n=342) 130 120 110 100 90 Diabetes 80 Nondipping No Diabetes Dipping Paper III

Metabolic characteristics at age 50 that predict the development of white-coat and sustained hypertension at age 70 Individuals were identified as normotensive (n=188), white-coat hypertensive (n=106) or sustained hypertensive (n=308) according to office and ambulatory BP at age 70 years. Of- fice BP determined at age 50 years

) 140

g * * was significantly and similarly ele-

m 120 vated in subjects with sustained and

( white-coat hypertension determined

r at age 70 years (Figure 8). As illus- 100

e * * r 80

d Figure 8. Systolic (white bars) and diastolic

l 60 (grey bars) BP at age 50 years. * p<0.05 vs

B normotensives. NT, normotensives; WC, white-coat hypertensives; SHT, sustained 40 hypertensives. NT WC SHT

31 KRISTINA BJÖRKLUND trated in Figure 9, a number of similarities regarding an abnormal glucose metabolism were observed in white-coat and sustained hypertensives at age 50. However, a lower BMI and a more favourable serum CE fatty acid composition distinguished subjects who twenty years later were identified as white-coat hypertensives from sustained hypertensives.

Age 50 years Age 70 years

WC SHT WC SHT

Office heart rate ↑ ↑ ↑ ↑ Body mass index −(↓) ↑ −(↓) ↑ 1-hour B-glucose ↑ ↑ (IVGTT) K-value (IVGTT) ↓ ↓ 2-hour P-glucose ↑ ↑ (OGTT) Insulin sensitivity index ↓ ↓

16:0 (Palmitic acid) −(↓) − 18:2 ω-6 (Linoleic acid) −(↑) ↓ Dietary fat intake (↓)

Figure 9. Metabolic variables determined at age 50 and 70 years. ↑ and ↓ indicates difference vs NT, - no difference v s NT, and (↑) or (↓) difference vs SHT. NT, normotensives; WC, white-coat hyperten sives; SHT, sustained hyperten sives.

<0.05 <0.05 20 150 ns <0.05 ns <0.05 140 18 130 16 ) 2 120 14 110 12 100 10 90 8 LVMI (g/m LVMI

80 (µg/min) UAER 6 70 4 60 2 50 0 NTWC SHT NT WC SHT

Figure 10. Measures of target organ damage in the different subgroups at age 70. LVMI, left ventricular mass index; UAER, Urinary albumin excretion rate; NT, normotensives; WC, White-coat hypertensives; SHT, sustained hypertensives.

32 24-HOUR AMBULATORY BP

Metabolic characteristics and target organ damage at age 70 in subjects with white-coat and sustained hypertension At age 70, both white-coat and sustained hypertensives had an impaired insulin sensitivity, increased plasma glucose and heart rate, however, body mass index was lower in white-coat hypertensive subjects (Figure 9). Seven- day dietary records showed that white-coat hypertensives at age 70 had a lower fat intake and a higher carbohydrate intake than sustained hypertensives. Furthermore, left ventricular mass and urinary albumin excretion rate, as indicators of hypertensive target organ damage, were only increased in sustained hypertensives (Figure 10).

Paper IV

Predictive value of ambulatory vs office BP levels A total of 172/872 cardiovascular morbid events (2.97/100 PYAR), 34 of which were fatal, occurred during the follow-up period of 9.5 years since baseline investigation at age 70 years. In a Cox regression model, adjusting for antihypertensive treatment at baseline, SBP and pulse pressure, both office and ambulatory, were significant predictors of cardiovascular morbidity, whereas DBP did not have any prognostic value. Figure 11 shows the results from a multivariate Cox regression analysis,

0,6 0,8 1 1,2 1,4 1,6 1,8 OSBP 24-h SBP Day SBP Night SBP ODBP 24-h DBP Day DBP Night DBP OPP 24-h PP Day PP Night PP 0,6 0,8 1 1,2 1,4 1,6 1,8

Figure 11. Adjusted hazard ratios and 95% confidence intervals for cardiovascular morbidity during follow-up. The hazard ratio reflects the risk associated with 1 SD increase of a BP variable measured at age 70 years. PP, pulse pressure. 33 KRISTINA BJÖRKLUND where smoking, diabetes, hyperlipidemia, antihypertensive treatment and body mass index were included as covariates in the model assessing the risk associated with a 1 SD increase of a BP variable at age 70 years. Twenty-four hour ambulatory pulse pressure remained the most powerful predictor of cardiovascular morbidity (hazard ratio (HR) for 1 SD increase in BP 1.32, 95% CI 1.15-1.52) and the association was independent of office pulse pressure level. Daytime, night-time and office pulse pressure, and all measures of SBP were significant, but weaker predictors of risk in the multivari-

ate analysis.

) )

R R A

A Figure 12. Cardiovascular event rate Y

Y 5 P

P according to tertiles of 24-hour

0 0 0

0 4 ambulatory SBP and daytime SBP

1 1

/ / (

( variability

e 3 t

t at baseline. PYAR, Person-years at

a a

r r

2 risk.

e e

c c n

n 1

e e

d d i

i III c

c 0 n

n II I I III I II I 24-hour SBP 24-hour SBP variability

Predictive Value of BP Variability Variability of SBP, reflected by the standard deviation (SD) of the daytime and night-time SBP was a predictor of cardiovascular morbidity, and the relationship remained after adjustment for other cardiovascular risk factors. The variability of daytime (HR 1.19, 95% CI 1.04-1.37), but not night- time SBP (HR 1.14, 95% CI 0.99-1.31), provided prognostic information independently of the 24-hour SBP level. As shown in Figure 12, the incidence of cardiovascular events rose with increasing daytime SBP variability within all tertiles of 24-hour SBP.

Paper V

Metabolic and cardiac risk profile at baseline in 70-year-old subjects with isolated ambulatory hypertension Subjects with elevated ambulatory but normal office BP, isolated ambulatory hypertension, at age 70 years, were more obese, showed higher plasma glucose levels during the oral glucose tolerance test, and lower insulin sensi-

34 24-HOUR AMBULATORY BP tivity compared with subjects with normal BP according to both office and ambulatory readings. The prevalence of smoking did not differ between the groups. Echocardiographically determined relative wall thickness was increased both in subjects with isolated ambulatory and sustained hypertension (Figure 13), whereas only sustained hypertensives showed an increased left ventricular mass. * * 0,45 Figure 13. Echocardiographically determined relative wall thickness at baseline 0,4 investigation, age 70 years. * p<0.05 vs NT. NT, normotensives; IAH, isolated ambulatory hypertensives; SHT, sustained hyperten- 0,35 sives. 0,3

0,25 Relativethickness wall

) 100

% 0,2

(

a NT IAH SHT

s 75 SHT

r IAH

t Figure 15. Kaplan-Meier survival curves for

e probability of event-free survival during follow-

E 50 up. NT, normotensives; IAH, isolated ambula- 0 2 4 6 8 tory hypertensives; SHT, sustained hyperten- Years sives.

Prognostic significance of isolated ambulatory and sustained hypertension for cardiovascular morbidity A total of 72/512 cardiovascular morbid events (2.37/100 PYAR), of which 48 were coronary events, 20 strokes and 4 peripheral vascular deaths, occurred during the mean follow-up of 6.6 years. The probability of cardiovascular events was increased in both hypertensive subgroups, as shown in Figure 15. The event rate (/100 PYAR) was 3.14 in sustained hypertensives, 2.74 in isolated ambulatory hypertensives and 0.99 in normotensives. In multivariate Cox Proportional hazard models, adjusting for the possible con- founding effect of serum cholesterol, smoking and diabetes, both isolated ambulatory and sustained hypertension remained significant predictors of cardiovascular morbidity, the adjusted hazard ratios being 2.77 (95% CI, 1.15-6.68) and 2.94 (95% CI, 1.49-5.82 ) for subjects with isolated and sustained hypertension respectively.

35 KRISTINA BJÖRKLUND Discussion

Ambulatory BP normality - definition and consequences Normal reference values of ambulatory BP may be defined using different approaches, among which the most common are 1) to evaluate the 24-hour BP distribution in normal populations, 2) to identify the 24-hour BP level that by regression analysis corresponds to the upper limit of normal office BP, or 3) to determine the relationship of ambulatory BP to prognostic outcome of morbidity and mortality. In Paper I, based on a large cohort of elderly men from the general population, we suggested the reference values of 24-hour ambulatory BP derived from correspondence criteria in untreated subjects, 130/78 mmHg, as they resulted in a greater number of correctly classified office hypertensives than thresholds derived from distribution criteria. The upper limit of normal 24-hour ambulatory BP in our population of 70-year-old men was slightly higher than in the Italian PAMELA study of elderly,[117] but in agreement with that which has been proposed based on other population studies,[118-120] and closely resembling the cut-off values for 24-hour ambulatory BP that best predicted cardiovascular mortality in a study that used a prognostic criterion to define normality.[121] Given the progressive increase in office SBP that is seen with age, the similarities between our study population and others, comprised of younger subjects, were a little surprising. Interestingly though, a tendency for SBP to increase less with age according to ambulatory BP, has been observed in other populations,[117, 122] and the present findings suggest that similar definitions as in younger populations may be applicable also in elderly subjects.

Indeed, reference values derived from large population studies during recent years have been consistent, and the consistencies have made it possible to reach consensus regarding a working definition of normal 24-hour ambulatory BP, 130/80 mmHg, and daytime ambulatory BP, 135/85 mmHg.[71, 123] These figures for normality were also applied by the Sixth Joint National Committee on Prevention, Detection, Evaluation and Treat-

36 24-HOUR AMBULATORY BP ment of High BP (JNC-VI).[103] In contrast, in the 1999 WHO-ISH guidelines,[20] the reference values for ambulatory BP have been quite strictly defined, with 120/80 mmHg considered as an upper limit of normal daytime ambulatory BP. These recommendations have been subject to criti- cism,[124] due to the discrepancy between such strict thresholds and the consistent findings from a large body of population data. Caution has been raised that application of the WHO guidelines may lead to over-classification of hypertension, and that treatment costs may have serious implications for the health economy. Whether we use the reference values recommended by the WHO or consensus documents, the current thresholds for normal ambulatory BP may still be viewed as preliminary, and eventually need to be verified in prospective outcome studies.

Hypertension - prevalence and control in an elderly population The fact that the present study was conducted in an elderly population, has important implications for the results. As described in Paper I, the prevalence of hypertension in the ULSAM population at age 70 years was 66%, similar to what has been shown in health surveys of elderly populations in the United States,[5] and Great Britain.[125] Despite consistent evidence showing that adequate treatment of hypertension in elderly reduces cardiovascular morbidity and mortality,[17, 126] the present findings (Paper I) corroborate others,[32, 127, 128] in that BP is insufficiently controlled in hypertensive populations. This applies in particular to the SBP component in elderly, which has started to gain interest as a more important prognostic factor than the DBP.[29] Although Paper I was not primarily designed to investigate BP control, the results suggest that more effort is needed to im- plement current guidelines and treatment goals in clinical practice.

Relationship between office and 24-hour ambulatory BP In agreement with previous population studies,[117, 129, 130] office and ambulatory BPs in Paper I were significantly correlated, with correlation coefficients of about 0.60 for SBP and DBP. Like our study, most previous investigations have found office BP to be higher than ambulatory BP.[129, 130] The discrepancy between office and ambulatory BP has been attributed to various factors, such as office BP level, age and sex,[131, 132] but also to whether the office BP is measured by a doctor or a nurse.[133] BP measurement in the office may induce an alerting reaction, the so-called white- coat effect, reflected by an increase in BP and heart rate.[80] The office-daytime BP difference has sometimes been used as a surrogate measure of the white-coat effect, but since it is not associated with a systematic office-daytime difference in heart rate, and has shown limited repro-

37 KRISTINA BJÖRKLUND ducibility,[134] the difference between office and daytime BP does not seem to be a reliable measure of this phenomenon. Furthermore, the office- daytime BP difference is less than 30% of the increase in non-invasive finger BP measured continuously before and during the clinic visit.[135] Another factor believed to influence the difference between office and daytime BP, is the “regression towards the mean”, when comparing office BP measured at a single visit with the mean of 24-hour ambulatory measurements.[132] Repeated office measurements may be an approach to reduce the measurement error, and the mean of standardised repeated office BP measurements on several occasions is likely to more closely correspond to 24-hour ambulatory BP.[136]

Nondipping as cardiovascular risk indicator We found a high prevalence of diabetes among nondipper subjects in the population of 70-year-old men (Paper II). Impaired diurnal BP patterns have previously been reported in both normotensive and hypertensive diabetic subjects.[74, 137] An increased sympathetic nervous activity and impaired glucose tolerance has been found in hypertensive subjects with a nondipping BP pattern,[138] and an altered autonomic nervous function, which may be associated with both diabetes and hypertension, has been proposed as a possible pathophysiological explanation for a blunted nocturnal BP decline.[139] The possible interrelation between diabetes and nondipping was further elucidated in Paper II, where an interaction was demonstrated between diabetes and nondipping with regard to several important metabolic risk factors. Similarly, a significant interaction between diabetes and nondipping contributed to an increased left ventricular mass in diabetic nondippers, implying that signs of organ damage and the most pronounced abnormalities in glucose and lipid metabolism were restricted to diabetic nondippers. The findings suggest that nondipping, at least in an elderly population, is a marker of risk mainly when diabetes is present. The results in Paper II support the hypothesis that the pathophysiologic mechanisms behind an impaired nocturnal BP reduction may, in part, differ in diabetic and nondiabetic subjects.[140] In diabetic subjects, the dys- functional autonomic cardiovascular regulation may be the most important factor contributing directly to nondipping, whereas it in hypertensive nondiabetic subjects may be attributed to hypertensive damage to the vascular wall, and subsequent persistent elevation of BP during night-time. Other factors suggested to be responsible for a disturbed nocturnal BP pattern, include a high sodium sensitivity,[141] and sleep deprivation.[142] Our findings that left ventricular mass was not increased in nondiabetic nondippers, compared with dippers, agree with some previous studies[78,

38 24-HOUR AMBULATORY BP

79] but disagree with others,[66, 143] although the latter did not report adjusting for diabetes in their analyses. Further, the studies are difficult to compare due to several differences in study design and populations. The conflicting findings may partially depend on the use of varying definitions of nondipping. Another general problem is methodological, represented by limited reproducibility of diurnal BP variation and dipping profile.[144, 145] In a recent study, a blunted nocturnal BP reduction confirmed by two separate ambulatory BP recordings, and thus reproducible over time, was associated with increased left ventricular mass and intima media thickness,[146] suggesting that a more careful identification of nondippers may increase the possibility to find subjects at high risk.

White-coat hypertension and the insulin resistance syndrome Hypertension is a well known feature of the insulin resistance syndrome, and metabolic risk factors related to hypertension contribute to the development of target organ damage and eventually, atherosclerotic diseases. In Paper III, not only sustained hypertensive subjects, but in addition, the 70- year-old men with white-coat hypertension showed impaired insulin sensitivity, elevated blood glucose and serum insulin compared with normotensives in the cross-sectional analysis. Although clamp insulin sensitivity has previously not been documented in a similar study of this size, metabolic aberrations have been reported in white-coat hypertensives by some,[87] but not all[84] previous investigators. In a study by Julius et al,[87] white- coat hypertension was preceded by elevated BP levels since childhood, suggesting that predictors of this condition may be possible to identify long before the office hypertension is discovered. In the ULSAM study, we found that subjects characterised as white-coat hypertensives at the age of 70, had higher office BP and heart rate, a marker of increased sympathetic activity, and an impaired glucose tolerance, to a similar degree as sustained hypertensives at the investigation twenty years earlier. However, white-coat hypertensives differed from sustained hypertensives at age 50 by a lower BMI and a more favourable fatty acid composition in the serum cholesterol esters. The findings that an increased heart rate and disturbances in glucose metabolism were consistent over time in both white-coat and sustained hypertensives are important, since they support the hypothesis that both of these conditions may originate from a state of increased sympathetic drive.[147] Sympathetic overactivity is, by some, considered to be a driving force in the development of insulin resistance and an important feature of borderline hypertension in its early stage.[148]

39 KRISTINA BJÖRKLUND

It is important to remember that most previous studies have used office BP to assess hypertension. This is also applicable for several studies where an increased heart rate has predicted the development of hypertension.[149] Indeed, in the present study an increased heart rate predicted both states where an elevated office BP was observed at age 70, that is, if not ambulatory BP monitoring would have been performed, both of these groups would have been classified as “hypertensive”. However, in our population, the hypertension that persisted throughout the day was the state characterised by the highest risk of target organ damage. Although the two hypertensive groups in our study of elderly men showed similarities regarding metabolic risk profile over time, the severity of target organ damage, reflected by left ventricular mass and urinary albumin excretion rate, was greater only in subjects with sustained hypertension in the cross-sectional analysis (Paper III). Therefore, it might be interesting to spec- ulate about possible mechanisms whereby one of the hypertensive groups may have been protected from developing sustained hypertension and organ damage over 20 years of follow-up, but not against becoming insulin resistant. At age 50, the serum cholesterol fatty acid composition, although more favourable, operated in an environment of increased sympathetic tone in the white-coat hypertensive subjects. The metabolic risk profile at baseline was otherwise similar in white-coat and sustained hypertensives, with the exception that body mass index was not higher in white-coat hypertensives than in normotensives. This was also observed twenty years later, when white- coat hypertensives were characterized by impaired insulin sensitivity, but sustained hypertensives were slightly more affected by the entire insulin resistance syndrome, with increased abdominal obesity, body mass index and serum proinsulin levels. In our study (Paper III), it seemed as if an increased sympathetic drive, in order to constitute a predictor of sustained hypertension and more pronounced metabolic disturbances, had to act in combination with a higher body mass and less favourable fatty acid profile. The absence of these additional factors, as seen in the white-coat subjects, may have reduced the risk, and resulted in a “moderately severe” state of the insulin resistance syndrome, devoid of a persistently increased BP and target organ damage. We can only hypothesize whether the lower body mass index observed in the white-coat hypertensives mainly was a marker of life style factors, or possibly also an indicator of a different genetic predisposition which together with the fatty acid profile put these subjects at another metabolic and cardiovascular risk level.

40 24-HOUR AMBULATORY BP

Fatty acids and BP Fatty acids in serum may influence the vascular system and BP level through several possible mechanisms. A previous experimental study,[150] showed that sympathetic nervous activity in humans was stimulated by an increase in plasma fatty acid concentration. Furthermore, the circulating free fatty acid levels seem to be associated with both the endothelial function[151] and insulin mediated vasodilation.[152] The fatty acid composition of serum cholesterol esters partly reflects the dietary fat quality over the previous weeks,[153] but also depends on genetically determined differences in fatty acid uptake and metabolisation.[56] A randomised controlled study previously showed that compared with a control diet, a diet rich in polyunsaturated fat induced a significant increase in the proportion of linoleic acid, and decrease in saturated fatty acids in the serum cholesterol esters, suggesting that the serum fatty acid profile can be modified by dietary changes.[154] A high-fat diet has been found to impair endothelial function,[155, 156] and the fatty acid composition of serum lipids is related to endothelium- dependent vasodilation,[157] as well as BP level.[158, 159]

In Paper III, subjects who twenty years later were identified as white-coat hypertensives, showed a significantly lower proportion of palmitic acid (16:0), and higher proportions of linoleic acid (18:2 ω-6), indicating a lower intake of foods containing saturated fat and higher relative intake of vegetable fat, compared with those men who later developed sustained hypertension. This fatty acid composition can be regarded as more favourable, since saturated fatty acids are associated with insulin resistance[55] and adversely affects the endothelium-dependent vasodilation,[157] while unsaturated fatty acids, such as linoleic acid, have positive effects on cell membrane fluidity and presumably, function.[160] The lower proportions of long-chain ω-3 fatty acids, commonly regarded as favourable, in white-coat hypertensive subjects at age 50, may be ex- plained by the fact that a high proportion of linoleic acid in serum can reduce the levels of eicosapentaenoic and docosahexaenoic acid through a decreased conversion of α-linolenic acid to long-chain ω-3 fatty acids, since these series are competitors for the same enzymatic systems.[153] Previous data have shown, that the serum fatty acid composition of 70-year old men reflects that 20 years earlier,[161] implying that subjects with a more favourable dietary fat intake at age 50 maintain these habits. In Paper III, this was supported by the observation that white-coat hypertensive subjects had a more favourable fatty acid pattern in serum lipids at age 50, a lower self- reported total fat intake at age 70 and a lower BMI that consisted over twenty years. In the ULSAM study, longitudinal relationships have previously been found

41 KRISTINA BJÖRKLUND between the fatty acid composition in serum cholesterol esters and left ventricular hypertrophy.[162] Although there is no established pathophysiological explanation to this observed relationship, hypothetical mechanisms may be suggested. Since endothelial dysfunction is related to left ventricular hypertrophy,[163] and the serum fatty acid profile is associated with endothelium dependent vasodilation,[157] this may represent a possible link between the fatty acid profile, sustained hypertension and left ventricular hypertrophy.

Prognostic value of 24-hour ambulatory BP

The role of ambulatory SBP and pulse pressure in risk assessment In agreement with previous studies,[9-12] findings from the ULSAM population (Paper IV) indicate that 24-hour ambulatory BP is an important predictor of cardiovascular morbidity independent of office BP and other established cardiovascular risk factors. More specifically, 24-hour ambulatory SBP and pulse pressure showed the strongest predictive power, while the DBP component did not contribute with any prognostic information (Paper IV). The fact that these results were achieved from a longitudinal analysis of an elderly population has important implications for the results. The prognostic value of SBP is known to increase with age, and SBP is an important predictor of cardiovascular morbidity and mortality in elderly.[6, 23] Epidemiological data suggest that BP components may influence cardiovascular risk differently at different ages,[29, 164, 165] shifting from DBP to be of importance below 50 years of age, to SBP and pulse pressure above age 60. In recent observational studies, using office[27, 28] or ambulatory[166, 167] recordings, pulse pressure has emerged as a major determinant of cardiovascular risk. In contrast with findings from the Syst-Eur Study,[12] daytime, but not night-time SBP was associated with a significantly increased cardiovascular risk in the ULSAM population (Paper IV). Furthermore, the night-day SBP ratio provided prognostic information independent of the 24-hour BP level in the Syst-Eur Study, whereas no such association was detected in ULSAM. The explanation to these discrepancies is not quite simple, but one possible reason may be found in the different characteristics of the populations studied. While Syst-Eur consisted solely of elderly patients with isolated systolic hypertension, thus being at high risk for cardiovascular events, we studied elderly men from the general population in UL- SAM. Data from Paper II further indicated that target organ damage was only present in diabetic subjects with nondipping, suggesting that in a population setting, a reduced nocturnal BP decline may primarily be a risk marker in subjects with diabetes.

42 24-HOUR AMBULATORY BP

Aortic stiffness - pathogenetical mechanisms behind an increased pulse pressure BP consists of a steady component, expressed as the product of total peripheral resistance and the cardiac output, and a pulsatile component, determined by ventricular ejection and arterial stiffness.[168] DBP and SBP represent the extremes of the fluctuations during the cardiac cycle, and pulse pressure is the difference between them. Whereas DBP reflects the mean, steady BP component more closely than SBP, the latter has a stronger association with pulse pressure. Arterial compliance is defined as the relationship between changes in distending pressure inside the artery, and the concomitant change in radius, or volume. The relationship is non-linear, and depends on BP level and the composition of the arterial wall. The arterial wall is a mixture of smooth muscle cells, collagen and elastin fibres, and as BP increases, the tension on elastin fibres, which are highly distensible, is transferred to collagen fibres, which are less distensible, leading to a stiffer arterial wall with increasing BP. Arterial compliance constitutes an important physiological mechanism, through the so called “Wind-kessel function”, whereby pulsatile flow in the large arteries is transformed into a steady flow in the peripheral tissues to maintain organ perfusion.[168] With aging, the elastic properties of the arteries decrease, a development that may be accelerated by hypertension and arteriosclerosis. As the large arteries stiffen, they become less distensible, causing a quicker return of the backward pressure wave towards the heart,[34] and a subsequently higher SBP and lower DBP. These age-related changes may have deleterious consequences for the heart, since the coronary perfusion is reduced while the increased SBP causes a simultaneously increased myocardial oxygen demand, predisposing to left ventricular hypertrophy and myocardial ischemia.[35] The increased SBP and pulse pressure, being a result of as well as causing a further accelerating arterial damage, creates a vicious cycle.

There are different methods of measuring aortic stiffness. The speed of wave travel in the aorta, pulse wave velocity, can be indirectly non-invasive- ly estimated by applanation tonometry, whereby the pulse wave is recorded at the radial or carotid artery, and a computerised process is used to generate an aortic pressure wave-form.[34] Aortic stiffness assessed by pulse wave analysis has recently been shown to predict cardiovascular morbidity[169] and mortality[170, 171] in population studies. Due to a difference between the pulse pressure in the aorta and the upper limb arteries when measured clinically, the adverse effects of aortic stiffening are underestimat- ed by measurement of brachial artery pressure by sphygmomanometer.[35] However, this difference is most evident in younger subjects, and may even be irrelevant in elderly.[168, 172] Thus, the use of brachial arterial pulse

43 KRISTINA BJÖRKLUND pressure as a surrogate measure of arterial stiffness seems appropriate in elderly. In addition, our study of 70-year-old men as well as previous longitudinal studies,[27, 28, 167] suggest that brachial arterial pulse pressure represents a more sensitive cardiovascular risk marker than SBP and DBP in elderly, indicating that this measure provides useful prognostic value in the population at large.

BP variability and prognosis In Paper IV, a high daytime SBP variability measured at baseline was a significant predictor of cardiovascular morbidity in 70-year-old men after 9 years of follow-up, independent of other established cardiovascular risk factors and 24-hour ambulatory BP level. A similarly positive longitudinal association between daytime systolic, but not diastolic, BP variability and cardiovascular mortality was previously found in a Japanese population,[70] whereas the relationship to an adverse outcome did not remain significant in the multivariate analysis in an earlier Italian study.[173] An increased BP variability has been directly associated with age and BP level.[174, 175] Autonomic cardiovascular regulatory mechanisms, in particular the arterial baroreflex function, are important determinants of fluctuations in BP, and an impaired baroreflex sensitivity is more common in elderly and in hypertensive subjects. The sensitivity of the arterial baroreflex is inversely related to the degree of BP variability and directly related to the heart rate variability. The baroreflex sensitivity, a marker of the capability to reflexly increase vagal activity and decrease sympathetic activity in response to an increase in BP, was recently shown to predict cardiac mortality in patients after a myocardial infarction.[176] An imbalance in the autonomic nervous system, with relatively high sympathetic and low vagal activity may thus represent one explanation for the adverse outcome related to increased BP variability in our study (Paper IV). Previous studies have indicated that target organ damage, evaluated as left ventricular hypertrophy, was greater with increasing BP variability in hypertensive subjects, independent of 24-hour BP level.[65, 66] However, from these cross-sectional investigations it was difficult to determine whether the BP variability was the cause or effect of the organ damage. In the ULSAM study, we assessed the long-term BP variability as the standard deviation of hourly BPs over day and night respectively. The intermittent measurements derived from non-invasive ambulatory BP monitoring may be a limitation, as they only roughly estimate long-term BP changes. In addition to intra-arterial BP monitoring, the more recently developed noninvasive continuous beat-to-beat finger BP recording devices might provide a more precise evaluation of different BP frequency components.[177, 178]

44 24-HOUR AMBULATORY BP

Predictive value of isolated ambulatory and white-coat hypertension

Isolated ambulatory hypertension When there is a discrepancy between office and ambulatory BP, the prognosis has been suggested to be more closely related to ambulatory BP.[81] Of the two situations where office and ambulatory BP may be discordant, the state with normal office but elevated ambulatory BP probably represents the greatest challenge. This is partly due to the fact that the knowledge of this condition is very sparse, and partly because it is not detected by the established hypertension screening instrument; office BP measurement. Two previous studies have shown that normal office but elevated ambulatory BP, or as we prefer to call it, “isolated ambulatory hypertension”, has a noticea- ble prevalence in the general population, and is associated with cardiac and arterial organ damage to a similar degree as sustained hypertension.[90, 91] A higher daytime than office BP was recently observed at baseline in a large proportion of elderly hypertensive subjects in the Second Australian National BP Study, a randomised-outcome trial.[179] Moreover, a retrospective analysis of treated hypertensive subjects in the SHEAF (Self-Meas- urement of BP at Home in the Elderly: Assessment and Follow-up) study, showed that cardiovascular risk profile and history of cardiovascular disease in patients with “isolated home hypertension” resembled those in patients with uncontrolled hypertension.[180] Although these cross-sectional and retrospective studies suggest that isolated ambulatory hypertension is associated with an increased cardiovascular risk, longitudinal data regarding cardiovascular outcome are thus far lacking. In Paper V, we present evidence that isolated ambulatory hypertension is a predictor of cardiovascular morbidity, independent of other established cardiovascular risk factors. Our study was performed in 70-year-old men from a population-based cohort, who had no antihypertensive treatment at baseline, and were followed for 8.4 years. At baseline, subjects with isolated ambulatory hypertension had more abdominal obesity, and higher plasma glucose levels than normotensives. Further, a subanalysis of echocardiographic data showed, in agreement with previous studies,[90, 91] signs of cardiac hypertrophy in subjects with isolated ambulatory hypertension. The possible mechanisms behind an elevated ambulatory daytime BP, but normal office BP, are not fully understood, but subjects with this BP pattern have previously been characterised as older,[90, 181] more often male,[90, 180, 181] and former smokers.[180, 181] Smoking elicits an acutely increased BP which becomes apparent during daytime BP monitoring, while individuals are told to refrain from smoking before a visit to the clinic.[182] In the ULSAM population, we found no differences in smok-

45 KRISTINA BJÖRKLUND ing prevalence between the subgroups defined in Paper V, whereas gender and age differences were not possible to investigate due to the study design. Other factors that may influence BP level during daytime include physical activity, and working conditions. In an analysis of self-reported leisure-time physical activity at baseline, we found no differences between the subgroups. Further, since the ULSAM population at age 70 had passed the age-limit for retirement, we did not expect to find working condition as an important contributor to differences in daytime BP patterns. Subjects in ULSAM were identified as normotensive or hypertensive based on the mean of two BPs taken at a single visit. As for white-coat hypertension, the number of correctly classified office hypertensives may increase with an increased number of BP readings. However, according to a previous study, the accuracy of office BP measurements in isolated ambulatory hypertension was not improved by repeated clinic visits.[181] Thus, although Paper V leaves some unanswered questions to be further investigated, the findings add to the evidence that office BP measurement as a method has shortcomings, and that the prognostic implications may be serious.

White-coat hypertension and prognosis Although much attention has focused on white-coat hypertension during recent years, there are very limited data regarding the prognostic significance of this presumably innocent condition. Verdecchia et al showed, that white-coat hypertensive subjects had a risk of cardiovascular events not different from that in normotensive controls,[10] however the study was limited by a short follow-up time. In a later extension of that Italian study, the outcome in white-coat hypertension, either similar to normotensives or to sustained hypertensives, was dependent on the definition used to identify white-coat hypertensive subjects.[88] In another study,[89] using intra-arterial BP monitoring, patients with white-coat hypertension was associated with substantially reduced risk of cardiovascular events over 10 years of follow-up, compared with patients whose BP was sustained throughout 24 hours. Figure 16 illustrates the office and ambulatory BPs in untreated subjects from ULSAM, and the four subgroups possible to identify according to these monitoring methods. Ninety-six cardiovascular morbid events occurred over 9.4 years of follow-up since the baseline investigation at age 70 years. Preliminary results (unpublished data, 2002) from a multivariate Cox Pro- portinal Hazard regression, adjusting for smoking, diabetes and serum cholesterol levels, showed that the hazard ratio for cardiovascular morbidity was highest in sustained hypertension (HR 2.79, 95% CI 1.49-5.21) followed by isolated ambulatory hypertension (HR 2.44, 95% CI 1.05-5.66). White- coat hypertension was not a significant predictor of cardiovascular disease in the ULSAM population (HR 1.79, 95% CI 0.82-3.87), thus corroborating 46 24-HOUR AMBULATORY BP the two previous prospective studies available today.[10, 89] The results are also in line with previous findings in ULSAM (Paper III), where signs of hypertensive cardiac and renal damage were similar in white-coat hypertensives and normotensives.

220 HR 2.44 HR 2.79 200

180

160

140

120

100 HR 1.00 HR 1.79 80 80 100 120 140 160 180 200 220 Office SBP (mmHg)

Fig 15. Distribution of office and daytime ambulatory SBP in untreated subjects (n=684) of the ULSAM population at age 70 years. HR indicates adjusted Cox proportional hazard ratios for cardiovascular morbidity in subgroups according to normal or abnormal office and daytime BP. All groups compared with nomotensives (HR 1.00).

Clinical relevance The present findings, in a population of elderly men, support previously published data suggesting that 24-hour ambulatory BP is a stronger predictor of cardiovascular morbid events than office BP, and speak in favour of a more wide-spread clinical use of 24-hour ambulatory BP monitoring. Despite the advantages of the method, ambulatory BP monitoring is used to a limited extent in the clinic, and in clinical trials, and it is still mainly regarded as a research tool.[20] Reasons for this might be practical issues such as inconvenience to the patient caused by disturbed sleep at night, and cost aspects associated with the monitoring procedure. Furthermore, there has been a lack of knowledge regarding interpretation of the measurements and limited data about prognostic value of the method. Not until recently, reference values for normal 24-hour BP were suggested,[71] and with increasing prognostic information, these recommendations may need revision. However, the reference values of ambulatory BP based on the present study of an elderly population (Paper I), are close to the current working defini-

47 KRISTINA BJÖRKLUND tions, suggesting that these may be appropriate to use also in the elderly. It has been proposed, that a large number of repeated, standardised office BP recordings may provide a mean BP close to the “true” BP level obtained by 24-hour ambulatory BP monitoring. This may be partially accurate, but we, and others[70] have shown that 24-hour ambulatory BP monitoring provides additional prognostically important information, represented by BP variability (Paper IV), that can not be obtained by office BP. Findings from the present study also suggest that office BP measurement, may in some subjects, fail to identify daytime ambulatory hypertension (Paper V), a condition with a fairly large prevalence in the present study population of 70-year-old men, and associated with an increased incidence of cardiovascular events. We have, in agreement with others,[12] demonstrated that ambulatory SBP is a stronger predictor of cardiovascular events than office SBP. The Syst-Eur Study, conducted in an elderly hypertensive population, showed that SBP in particular may be overestimated by office readings.[12] Elderly are more prone to postural and postprandial hypotension, and the symptoms may worsen by a high susceptibility to the adverse effects of antihypertensive drugs. Misclassification of hypertension leading to unnecessary medication is therefore particularly harmful in older patients. Furthermore, an issue that previously has been raised,[183, 184] concerns the potential overtreatment of hypertension in society, due to inaccurate treatment decisions in subjects with white-coat hypertension. The hypothesis that white-coat hypertension is a harmless condition, with an associated cardiovascular risk similar to that in normotension, is now supported by two longitudinal studies[10, 89] and preliminary results (unpublished data, 2002) from the UL- SAM population. Furthermore, the only study that has investigated the effects of antihypertensive teatment on prognosis in white-coat hypertension, showed no benefit of treatment.[185] To correctly diagnose, or to treat white-coat hypertension due to igno- rance, is a matter of benefit for the patient, and cost for society, represented by on one hand expenses for drugs, and on the other, for an ambulatory monitoring procedure. It was recently shown, that by using ambulatory BP monitoring instead of office BP measurement as a basis for drug prescrip- tion, significantly less drug treatment was prescribed, and the cost for ambulatory BP measurement were offset by the savings from less drug prescrip- tion and fewer clinic visits.[186] In summary, the clinical relevance of the present findings is important. Considering the strong prognostic value of ambulatory BP, it might be rea- sonable to suggest the performance of at least one ambulatory BP monitoring session in a hypertensive patient, at some point of time, to gain more detailed information on the total risk profile. In non-diabetic subjects, a BP monitoring during 12 hours may be enough to identify subjects with white- 48 24-HOUR AMBULATORY BP coat hypertension, who are not at increased cardiovascular risk, and in whom treatment is not necessary. In contrast, the present findings propose, that a full 24-hour ambulatory BP monitoring should be recommended in patients with diabetes, since the night-day BP difference may be a marker of increased risk primarily in this patient group. Elderly subjects, in particular, might benefit from a health check-up, where daytime ambulatory BP monitoring could be included to identify subjects with white-coat, as well as isolated ambulatory hypertension. Thus, given the independent predictive value of the mean daytime ambulatory BP in the general population, different approaches to the use of ambulatory BP monitoring in different patient groups may reduce the unnecessary inconvenience to the patient, represented by night-time BP monitoring. A final important clinical implication of the present thesis does not con- cern any new findings, but the well known fact that hypertensive patients are insufficiently treated, as observed in Paper I. To improve hypertension control is a main future goal, and 24-hour ambulatory BP monitoring may be a helpful tool in that effort. Further, many elderly with moderately or borderline elevated BP still do not receive any treatment, as demonstrated in Paper V, although this condition is associated with a substantially increased risk in the elderly.

Strengths and limitations of the study The most obvious limitation of this study, is the lack of women in the study population. When ULSAM was initiated in the beginning of the 1970´s, one of the aims was to identify risk factors for cardiovascular disease in a high risk population, at that time typically represented by middle-aged men. It is now known, that although coronary heart disease is not as common in women as in men during middle-age, the risk increases substantially after menopause, and coronary heart disease in fact constitutes the leading cause of death in women. This age-dependent risk increase has partly been attributed a possible protective effect of estrogen in pre-menopausal women.[187] Age-dependent gender differences also apply for hypertensive disease. Data from observational population studies have indicated that women show a lower BP, both office and ambulatory, than men up to the fifth decade of age.[5, 130] This does however not reduce the role of hypertension as an important cardiovascular risk factor in elderly women. Indeed, the fact that hypertension has similar prevalence rates and has become a similarly strong cardiovascular risk factor in women as in men at the age of 70, suggest that the present findings may be applicable also in women. However, some caution should be taken regarding generalizability of these results, until they have been confirmed in elderly women.

49 KRISTINA BJÖRKLUND

Regarding the cohort, we might assume, that subjects who participated in the re-investigation at age 70, were somewhat healthier than those who did not participate, and this may have lead to some underestimation of the associations found. Another important limitation of the present study, is the lack of withdrawal of antihypertensive medication before the investigations. This is however a common drawback of cohort studies of this size. All the papers in this thesis are based on data derived from the general population, resulting in limited selection bias. The population studied was homogeneous, comprising men of similar age and from a well-defined geo- graphical area. The characteristics of ULSAM as being representative of the general population, and including both hypertensive and normotensive subjects, gave us the opportunity to investigate the prognostic significance of BP for cardiovascular disease in elderly at different levels of normal and elevated BP, and to further evaluate the prevalence, associated risk profile and predictive value of isolated ambulatory hypertension, requiring 24-hour ambulatory BP data from subjects defined as normotensive according to office BP. Hypertension is a complex state, and despite being an independent cardiovascular risk factor, the strength of association with cardiovascular disease may be related to other components of the insulin resistance syndrome, as well as lifestyle factors. The broad multifactorial approach of the present study material, including metabolic parameters, dietary information and data on physical activity, therefore makes the ULSAM study particularly valuable for an investigation into epidemiological aspects of hypertension.

Future perspectives The growing evidence from this (Paper IV), and previous studies, demon- strating that pulse pressure is a powerful predictor of cardiovascular morbidity and mortality in elderly, leads to the question whether a treatment regi- men that reduces not only mean BP, but also pulse pressure, may prevent adverse events in elderly with this hypertensive subform. This question should be answered by randomised clinical trials, designed to evaluate the specific effect of conventional as well as new antihypertensive drugs on reducing large artery stiffness, as this should be the target of treatment.[188] The inclusion criteria must be based on elevated SBP or pulse pressure, as already has been done in Syst-Eur and Syst-China,[30, 31] and the efficacy of treatment should be measured in terms of decrease in pulse pressure. Com- mon antihypertensive drugs, such as thiazide diuretics, ß-blockers and calcium channel blockers, lower SBP as well as DBP, but in subjects with a wide pulse pressure, a further decrease in DBP may be detrimental for the coronary perfusion. In contrast, low-dose nitrates have been shown to reduce

50 24-HOUR AMBULATORY BP arterial stiffness and lower pulse pressure without affecting the DBP. Anoth- er target of treatment may be to modify arterial structure and thereby reduce pulse wave velocity, possibly by inhibiting collagen accumulation, which has been achieved experimentally through blockade of angiotensin II (type1) receptors.[188] The adverse prognostic outcome associated with isolated ambulatory hypertension presently observed in the ULSAM study (Paper V) illustrates the limitations of office BP as a screening instrument, and addresses the important question of potential undertreatment of these patients. Although it seems likely that these patients would benefit from antihypertensive treatment, this should be a matter for investigation in future randomised clinical trials. Large clinical intervention trials have a huge impact on hypertension treatment policy. It is therefore essential that BP in these trials is measured correctly. Although the standardized office measurement procedures are followed, there may be differences in treatment effects during day and night, that are not revealed by a conventional office measurement. Such differences may be crucial for an accurate interpretation of the trial, as recently illustrated in a substudy of ambulatory BP in the Heart Outcomes Preven- tion Evaluation (HOPE) trial.[189] In all future clinical trials evaluating effect of antihypertensive treatment, ambulatory BP monitoring should therefore be performed in a representative subsample of the study population. Furthermore, future intervention trials should, to a larger extent, use 24- hour ambulatory BP as inclusion criteria and efficacy parameter in relation to cardiovascular outcome measures. The importance of cost-benefit related issues for the clinical implications of ambulatory BP monitoring has already been mentioned. With more focus directed towards health economy, future studies may be able to show not only clinical, but also economical benefits of an increased use of 24-hour ambulatory BP monitoring. Finally, the potential significance of a correct diagnosis and an efficient hypertension management for the quality of life of the patient, can not be overestimated.

51 KRISTINA BJÖRKLUND Conclusions

I. Reference values of ambulatory BP were established in a large cohort of 70-year-old men, and the upper limits of normal ambulatory BP in this elderly population were in agreement with those suggested in current guidelines. The study confirmed previous findings of a high prevalence of hypertension in elderly, and indicated that BP was inadequately controlled in elderly treated hypertensive subjects.

II. An interaction was demonstrated between nondipping and diabetes, regarding several metabolic parameters as well as echocardiographically determined left ventricular mass. Thus, we identified a nondipping high-risk group being diabetic, and a group of nondippers with a normal metabolic status and a lack of organ damage, implying that nondipping, at least in elderly, is a marker of increased risk primarily in the presence of diabetes.

III. Several metabolic disturbances at age 50 predicted both white-coat and sustained hypertension and were consistent over twenty years of follow-up. How- ever, the development of white-coat hypertension was preceded by lower BMI and a more favourable serum CE fatty acid composition at age 50. In a cross-sectional analysis at age 70, only subjects with sustained hypertension showed signs of organ damage. Thus, an unfavourable dietary fat intake and obesity may be important contributors to sustained hypertension and organ damage. IV. In a longitudinal analysis, SBP and pulse pressure, but not DBP, were significant predictors of cardiovascular morbidity. Twenty-four-hour and daytime ambulatory pulse pressure had the strongest predictive value, independently of office BP, antihypertensive medication, hyperlipidemia, diabetes, smoking and BMI. The results indicate that pulse pressure is an important predictor of cardiovascular morbidity in elderly, and further that ambulatory BP is superior to office BP in evaluating cardiovascular risk in an elderly population. V. Subjects with isolated ambulatory hypertension showed metabolic abnormalities and signs of cardiac hypertrophy in the cross-sectional analysis. More importantly, isolated ambulatory hypertension was a significant independent predictor of cardiovascular morbidity. These findings suggest that 24-hour ambulatory BP may disclose important prognostic information also in subjects with a normal office BP.

52 24-HOUR AMBULATORY BP Acknowledgements

I wish to express my sincerest gratitude to everyone who has contributed to this work in different ways. My special thanks go to Lars Lind, my principal supervisor, for guiding me into the field of cardiovascular research, for his unlimited enthusiasm and bright scientific mind, and for patience and support whenever needed, Hans Lithell, head of the Section of Geriatrics, my mentor and supervisor, for sharing his extensive knowledge of hypertension as well as good advice about life in general, for consistent encouragement and nice chats about life after the thesis, and without whose outstanding persuasive skill I might have stayed in Trollhättan to do AT instead of writing this thesis, Bengt Vessby, head of the Section for Clinical Nutrition, co-author and pos- sessing invaluable knowledge of fatty acids, for saving me from getting lost in the fatty acid swamp, Lars Berglund and Rawya Mohsen, for excellent statistical advice and exper- tise regarding data handling in general, and ULSAM in particular, Margareta Öhrwall, head of the Metabolic Ward and expert on treatment of obesity, for being my clinical tutor and turning theory into clinical practice, Previous or current PhD students, colleagues and friends at the Section of Geriatrics; Ann-Cristin, Anu, Arvo, Björn, Claes, Johan S, Johan Ä, Kristina D, Lena, Lennart, Liisa, Karin, Merike, Meta and Per-Erik, and at the Section of Clinical Nutrition Research; Achraf, Agneta, Anders, Anette, Annika, Brita, Cecilia, Eva, Johanna, Samar and Ulf, and the “newcomers” Bernice, Kristin, Klara, and Martin, to all for contributing to the friendly and warm atmos- phere, for good laughs and interesting conversations over numerous coffee- breaks, to those who have participated in the “doktorandklubben”; a special thanks for dynamic and constructive discussions, the lunch-team, for getting me used to extravagant long lunch-breaks - I will miss them, either we had thai food or lukewarm sushi, Johan S, Johan Ä and Ulf; for letting me join in and sometimes win the office-golf competitions, Johan Ä; for great travel-company to London, Milan and New Orleans, and for saving me from annoying males, whether my annoyance was justified or not, Jan Hall, particularly for skillful performance of the ambulatory BP monitoring, and Siv Tengblad, Eva Sejby and Barbro Simu, for outstanding labora- tory work, and professional arrangement of parties,

53 KRISTINA BJÖRKLUND

Bertil Andrén, for co-authorship, and for carrying out all the echocardiographic examinations, Birgitta and the entire staff of nurses and dieticians at the Metabolic Ward, for taking such good care of the study participants and creating a pleasant working climate,

All the brave Uppsala-men, for participating in the ULSAM study, without you there would definitely not have been a thesis,

Anna-Pia, Eva, Frida, Jenny and Lotta, for keeping me updated on the world outside Uppsala, for great friendship and late nights out with the girls, Gabriella, for being an awesome friend, and for sharing enthusiasm over Bach´s Concerto for two violins although it is really too complicated for us, Franziska, my dear friend and colleague, for always being there, and for sharing my devotion for travelling, but also for sharing unpleasant Mexican experiences of having mice in bed at night, Erik, my dear older brother, for letting me for once be the first; to complete a thesis, and together with Lotta and Sofia, for being an important part of my precious family, My dear mother and father, for being the best of parents, for all your love and generosity, for always believing in me and supporting me in everything, and Carl, without whom I might not have stayed in Uppsala to write this thesis, for making my heart beat a little faster, for making me laugh every day and for being my best friend, for endless encouragement and valuable help with thesis cover design – I love you.

This work was supported by grants from the Foundation for Geriatric Re- search, Ernfors Fund for Diabetic Research, Swedish Medical Research Coun- cil (grant no. 5446), the Swedish Association against Heart and Lung Dis- ease, the Geriatric Fund, Trygg Hansa, “Förenade Liv” Mutual Group Life Insurance Company, the Swedish Council for Planning and Coordination of Research, the Thuréus Foundation, and Uppsala University.

The cover illustrates the blood pressure during a common day of the author´s life measured by 24-hour ambulatory BP monitoring.

54 24-HOUR AMBULATORY BP

References

1. Stokes J3rd, Kannel WB, Wolf PA, D’Agostino RB, and Cupples LA. Blood pressure as a risk factor for cardiovascular disease. The Framingham Study—30 years of follow-up. Hypertension 1989; 13:I13-8. 2. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J, Dyer A, and Stamler J. Blood pressure, stroke, and coronary heart disease. Part 1, Prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet 1990; 335:765-74. 3. Pollare T, Lithell H, and Berne C. Insulin resistance is a characteristic feature of primary hypertension independent of obesity. Metabolism 1990; 39:167-74. 4. DeFronzo RA, and Ferrannini E. Insulin resistance. A multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care 1991; 14:173-94. 5. Burt VL, Whelton P, Roccella EJ, Brown C, Cutler JA, Higgins M, Horan MJ, and Labarthe D. Prevalence of hypertension in the US adult population. Results from the Third National Health and Nutrition Examination Survey, 1988-1991. Hypertension 1995; 25:305-13. 6. Staessen JA, Gasowski J, Wang JG, Thijs L, Den Hond E, Boissel JP, Coope J, Ekbom T, Gueyffier F, Liu L, Kerlikowske K, Pocock S, and Fagard RH. Risks of untreated and treated isolated systolic hypertension in the elderly: meta-analysis of outcome trials. Lancet 2000; 355:865-72. 7. Sokolow M, Werdegar D, Kain HK, and Hinman AT. Relationship between level of blood pressure measured casually and by portable recorders and severity of complications in essential hypertension. Circulation 1966; 34:279-98. 8. Pessina AC, Palatini P, Sperti G, Cordone L, Libardoni M, Mos L, Mormino P, Di Marco A, and Dal Palu C. Evaluation of hypertension and related target organ damage by average day-time blood pressure. Clin Exp Hyper- tens [A] 1985; 7:267-78. 9. Perloff D, Sokolow M, and Cowan R. The prognostic value of ambulatory blood pressures. JAMA 1983; 249:2792-8. 10. Verdecchia P, Porcellati C, Schillaci G, Borgioni C, Ciucci A, Battistelli M, Guerrieri M, Gatteschi C, Zampi I, Santucci A, and et al. Ambulatory blood pressure. An independent predictor of prognosis in essential hypertension. Hypertension 1994; 24:793-801. 11. Ohkubo T, Imai Y, Tsuji I, Nagai K, Watanabe N, Minami N, Itoh O, Bando T, Sakuma M, Fukao A, Satoh H, Hisamichi S, and Abe K. Prediction of mortality by ambulatory blood pressure monitoring versus screening blood pressure measurements: a pilot study in Ohasama. J Hypertens 1997; 15:357-64. 12. Staessen JA, Thijs L, Fagard R, O’Brien E, Clement D, de Leeuw PW, Mancia G, Nachev C, Palatini P, Parati G, Tuomilehto J, and Webster J. Predicting cardiovascular risk using conventional vs ambulatory blood pressure in older patients with systolic hypertension. Systolic Hypertension in Europe Trial Investigators. JAMA 1999; 282:539- 46. 13. Kannel WB, Gordon T, and Schwartz MJ. Systolic versus diastolic blood pressure and risk of coronary heart disease. The Framingham study. Am J Cardiol 1971; 27:335-46. 14. Black HR. The paradigm has shifted, to systolic blood pressure. Hypertension 1999; 34:386-7. 15. Summary of 1993 World Health Organisation-International Society of Hypertension guidelines for the management of mild hypertension. Subcommittee of WHO/ISH Mild Hypertension Liaison committee. BMJ 1993; 307:1541-6. 16. Läkemedelsverket. Behandling av hypertoni hos äldre (>=70 år). Rekommendationer. Läkemedelsverket. 1993; 3:5-9. 17. Dahlof B, Lindholm LH, Hansson L, Schersten B, Ekbom T, and Wester PO. Morbidity and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension). Lancet 1991; 338:1281-5. 18. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). SHEP Cooperative Research Group. JAMA 1991; 265:3255-64. 19. Medical Research Council trial of treatment of hypertension in older adults: principal results. MRC Working Party. BMJ 1992; 304:405-12. 20. 1999 World Health Organization-International Society of Hypertension Guidelines for the Management of Hypertension. Guidelines Subcommittee. J Hypertens 1999; 17:151-83. 21. Sagie A, Larson MG, and Levy D. The natural history of borderline isolated systolic hypertension. N Engl J Med 1993; 329:1912-7. 22. Owens PE, Lyons SP, Rodriguez SA, and O’ Brien E. Is elevation of clinic blood pressure in patients with white coat hypertension who have normal ambulatory blood pressure associated with target organ changes? J Hum Hypertens 1998; 12:743-8.

55 KRISTINA BJÖRKLUND

23. Rutan GH, Kuller LH, Neaton JD, Wentworth DN, McDonald RH, and Smith WM. Mortality associated with diastolic hypertension and isolated systolic hypertension among men screened for the Multiple Risk Factor Interven- tion Trial. Circulation 1988; 77:504-14. 24. O’Donnell CJ, Ridker PM, Glynn RJ, Berger K, Ajani U, Manson JE, and Hennekens CH. Hypertension and borderline isolated systolic hypertension increase risks of cardiovascular disease and mortality in male physicians. Circulation 1997; 95:1132-7. 25. Alli C, Avanzini F, Bettelli G, Colombo F, Torri V, and Tognoni G. The long-term prognostic significance of repeated blood pressure measurements in the elderly: SPAA (Studio sulla Pressione Arteriosa nell’Anziano) 10-year follow-up. Arch Intern Med 1999; 159:1205-12. 26. Fang J, Madhavan S, Cohen H, and Alderman MH. Measures of blood pressure and myocardial infarction in treated hypertensive patients. J Hypertens 1995; 13:413-9. 27. Franklin SS, Khan SA, Wong ND, Larson MG, and Levy D. Is pulse pressure useful in predicting risk for coronary heart Disease? The Framingham heart study. Circulation 1999; 100:354-60. 28. Blacher J, Staessen JA, Girerd X, Gasowski J, Thijs L, Liu L, Wang JG, Fagard RH, and Safar ME. Pulse pressure not mean pressure determines cardiovascular risk in older hypertensive patients. Arch Intern Med 2000; 160:1085- 9. 29. Franklin SS, Larson MG, Khan SA, Wong ND, Leip EP, Kannel WB, and Levy D. Does the relation of blood pressure to coronary heart disease risk change with aging? The Framingham Heart Study. Circulation 2001; 103:1245-9. 30. Staessen JA, Fagard R, Thijs L, Celis H, Arabidze GG, Birkenhager WH, Bulpitt CJ, de Leeuw PW, Dollery CT, Fletcher AE, Forette F, Leonetti G, Nachev C, O’Brien E, Rosenfeld J, Rodicio JL, Tuomilehto J, and Zanchetti A. Randomised double-blind comparison of placebo and active treatment for older patients with isolated systolic hypertension. The Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Lancet 1997; 350:757-64. 31. Liu L, Wang JG, Gong L, Liu G, and Staessen JA. Comparison of active treatment and placebo in older Chinese patients with isolated systolic hypertension. Systolic Hypertension in China (Syst-China) Collaborative Group. J Hypertens 1998; 16:1823-9. 32. Berlowitz DR, Ash AS, Hickey EC, Friedman RH, Glickman M, Kader B, and Moskowitz MA. Inadequate management of blood pressure in a hypertensive population. N Engl J Med 1998; 339:1957-63. 33. Mancia G, Bombelli M, Lanzarotti A, Grassi G, Cesana G, Zanchetti A, and Sega R. Systolic vs diastolic blood pressure control in the hypertensive patients of the PAMELA population. Pressioni Arteriose Monitorate E Loro Associazioni. Arch Intern Med 2002; 162:582-6. 34. O’Rourke MF. Wave travel and reflection in the arterial system. J Hypertens 1999; 17:S45-7. 35. Black HR, Kuller LH, O’Rourke MF, Weber MA, Alderman MH, Benetos A, Burnett J, Cohn JN, Franklin SS, Mancia G, Safar M, and Zanchetti A. The first report of the Systolic and Pulse Pressure (SYPP) Working Group. J Hypertens 1999; 17:S3-14. 36. Mancia G, Lanfranchi A, Turri C, and Grassi G. Can good surrogate end-points predict the prognosis of hypertensive patients? J Hypertens Suppl 1998; 16:S3-7. 37. Cuspidi C, Ambrosioni E, Mancia G, Pessina AC, Trimarco B, and Zanchetti A. Role of echocardiography and carotid ultrasonography in stratifying risk in patients with essential hypertension: the Assessment of Prognostic Risk Observational Survey. J Hypertens 2002; 20:1307-14. 38. Vasan RS, Larson MG, Leip EP, Evans JC, O’Donnell CJ, Kannel WB, and Levy D. Impact of high-normal blood pressure on the risk of cardiovascular disease. N Engl J Med 2001; 345:1291-7. 39. Collins R, Peto R, MacMahon S, Hebert P, Fiebach NH, Eberlein KA, Godwin J, Qizilbash N, Taylor JO, and Hennekens CH. Blood pressure, stroke, and coronary heart disease. Part 2, Short-term reductions in blood pressure: overview of randomised drug trials in their epidemiological context. Lancet 1990; 335:827-38. 40. Safar ME. Epidemiological aspects of pulse pressure and arterial stiffness. J Hypertens 1999; 17:S37-40. 41. Julius S. Corcoran Lecture. Sympathetic hyperactivity and coronary risk in hypertension. Hypertension 1993; 21:886-893. 42. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988; 37:1595-607. 43. Haffner SM, Valdez RA, Hazuda HP, Mitchell BD, Morales PA, and Stern MP. Prospective analysis of the insulin-resistance syndrome (syndrome X). Diabetes 1992; 41:715-22. 44. Suzuki M, Shinozaki K, Kanazawa A, Hara Y, Hattori Y, Tsushima M, and Harano Y. Insulin resistance as an independent risk factor for carotid wall thickening. Hypertension 1996; 28:593-8. 45. Bianchi S, Bigazzi R, Nenci R, and Campese VM. Hyperinsulinemia, circadian variation of blood pressure and end-organ damage in hypertension. J Nephrol 1997; 10:325-33. 46. Lind L, Andersson PE, Andren B, Hanni A, and Lithell HO. Left ventricular hypertrophy in hypertension is associated with the insulin resistance metabolic syndrome. J Hypertens 1995; 13:433-8. 47. Pyorala M, Miettinen H, Halonen P, Laakso M, and Pyorala K. Insulin resistance syndrome predicts the risk of

56 24-HOUR AMBULATORY BP coronary heart disease and stroke in healthy middle-aged men: the 22-year follow-up results of the Helsinki Police- men Study. Arterioscler Thromb Vasc Biol 2000; 20:538-44. 48. Beck-Nielsen H. General characteristics of the insulin resistance syndrome: prevalence and heritability. Europe- an Group for the study of Insulin Resistance (EGIR). Drugs 1999; 58:7-10; discussion 75-82. 49. WHO. Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications. Report of a WHO consultation. Part 1: Diagnosis and Classification of Diabetes Mellitus. 1999. WHO/NCD/NCS/99.2 50. National Institutes of Health. Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Bethesda, Md: National Institutes of Health; 2001. NIH Publication 01-3670. 51. Ford ES, Giles WH, and Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA 2002; 287:356-9. 52. Skarfors ET, Lithell HO, and Selinus I. Risk factors for the development of hypertension: a 10-year longitudinal study in middle-aged men. J Hypertens 1991; 9:217-23. 53. Haffner SM, Ferrannini E, Hazuda HP, and Stern MP. Clustering of cardiovascular risk factors in confirmed prehypertensive individuals. Hypertension 1992; 20:38-45. 54. Beilin LJ, Burke V, Cox KL, Hodgson JM, Mori TA, and Puddey IB. Non pharmacologic therapy and lifestyle factors in hypertension. Blood Press 2001; 10:352-65. 55. Vessby B, Tengblad S, and Lithell H. Insulin sensitivity is related to the fatty acid composition of serum lipids and skeletal muscle phospholipids in 70-year-old men. Diabetologia 1994; 37:1044-50. 56. Vessby B, Aro A, Skarfors E, Berglund L, Salminen I, and Lithell H. The risk to develop NIDDM is related to the fatty acid composition of the serum cholesterol esters. Diabetes 1994; 43:1353-7. 57. Moore TJ, Vollmer WM, Appel LJ, Sacks FM, Svetkey LP, Vogt TM, Conlin PR, Simons-Morton DG, Carter- Edwards L, and Harsha DW. Effect of dietary patterns on ambulatory blood pressure : results from the Dietary Approaches to Stop Hypertension (DASH) Trial. DASH Collaborative Research Group. Hypertension 1999; 34:472-7. 58. Hales S. Statistical assays: containing haemastaticks. London: Innys, Manby & Woodward; 1733: 2. 59. Cohen MC, Rohtla KM, Lavery CE, Muller JE, and Mittleman MA. Meta-analysis of the morning excess of acute myocardial infarction and sudden cardiac death. Am J Cardiol 1997; 79:1512-6. 60. Elliott WJ. Circadian variation in the timing of stroke onset: a meta-analysis. Stroke 1998; 29:992-6. 61. Muller JE, Ludmer PL, Willich SN, Tofler GH, Aylmer G, Klangos I, and Stone PH. Circadian variation in the frequency of sudden cardiac death. Circulation 1987; 75:131-8. 62. Weitzman ED, Fukushima D, Nogeire C, Roffwarg H, Gallagher TF, and Hellman L. Twenty-four hour pattern of the episodic secretion of cortisol in normal subjects. J Clin Endocrinol Metab 1971; 33:14-22. 63. Panza JA, Epstein SE, and Quyyumi AA. Circadian variation in vascular tone and its relation to alpha-sympathetic vasoconstrictor activity. N Engl J Med 1991; 325:986-90. 64. Parati G, Di Rienzo M, Ulian L, Santucciu C, Girard A, Elghozi JL, and Mancia G. Clinical relevance of blood pressure variability. J Hypertens Suppl 1998; 16:S25-33. 65. Parati G, Pomidossi G, Albini F, Malaspina D, and Mancia G. Relationship of 24-hour blood pressure mean and variability to severity of target-organ damage in hypertension. J Hypertens 1987; 5:93-8. 66. Palatini P, Penzo M, Racioppa A, Zugno E, Guzzardi G, Anaclerio M, and Pessina AC. Clinical relevance of nighttime blood pressure and of daytime blood pressure variability. Arch Intern Med 1992; 152:1855-60. 67. Sega R, Corrao G, Bombelli M, Beltrame L, Facchetti R, Grassi G, Ferrario M, and Mancia G. Blood pressure variability and organ damage in a general population: results from the PAMELA study (Pressioni Arteriose Moni- torate E Loro Associazioni). Hypertension 2002; 39:710-4. 68. Frattola A, Parati G, Cuspidi C, Albini F, and Mancia G. Prognostic value of 24-hour blood pressure variability. J Hypertens 1993; 11:1133-7. 69. Sander D, Kukla C, Klingelhofer J, Winbeck K, and Conrad B. Relationship between circadian blood pressure patterns and progression of early carotid atherosclerosis: A 3-year follow-up study. Circulation 2000; 102:1536- 41. 70. Kikuya M, Hozawa A, Ohkubo T, Tsuji I, Michimata M, Matsubara M, Ota M, Nagai K, Araki T, Satoh H, Ito S, Hisamichi S, and Imai Y. Prognostic significance of blood pressure and heart rate variabilities: the Ohasama study. Hypertension 2000; 36:901-6. 71. O’ Brien E, and Staessen J. Normotension and hypertension defined by 24-hour ambulatory blood pressure monitoring. Blood Press 1995; 4:266-82. 72. Staessen JA, Beilin L, Parati G, Waeber B, and White W. Task force IV: Clinical use of ambulatory blood pressure monitoring. Participants of the 1999 Consensus Conference on Ambulatory Blood Pressure Monitoring. Blood Press Monit 1999; 4:319-31. 73. Mann S, Altman DG, Raftery EB, and Bannister R. Circadian variation of blood pressure in autonomic failure.

57 KRISTINA BJÖRKLUND

Circulation 1983; 68:477-83. 74. Fogari R, Zoppi A, Malamani GD, Lazzari P, Destro M, and Corradi L. Ambulatory blood pressure monitoring in normotensive and hypertensive type 2 diabetes. Prevalence of impaired diurnal blood pressure patterns. Am J Hypertens 1993; 6:1-7. 75. Staessen JA, Bieniaszewski L, O’ Brien E, Gosse P, Hayashi H, Imai Y, Kawasaki T, Otsuka K, Palatini P, Thijs L, and Fagard R. Nocturnal blood pressure fall on ambulatory monitoring in a large international database. The “Ad Hoc” Working Group. Hypertension 1997; 29:30-9. 76. Shimada K, Kawamoto A, Matsubayashi K, Nishinaga M, Kimura S, and Ozawa T. Diurnal blood pressure variations and silent cerebrovascular damage in elderly patients with hypertension. J Hypertens 1992; 10:875-8. 77. Bianchi S, Bigazzi R, Baldari G, Sgherri G, and Campese VM. Diurnal variations of blood pressure and microalbuminuria in essential hypertension. Am J Hypertens 1994; 7:23-9. 78. Roman MJ, Pickering TG, Schwartz JE, Cavallini MC, Pini R, and Devereux RB. Is the absence of a normal nocturnal fall in blood pressure (nondipping) associated with cardiovascular target organ damage? J Hypertens 1997; 15:969-78. 79. Cuspidi C, Lonati L, Sampieri L, Macca G, Valagussa L, Zaro T, Michev I, Fusi V, Leonetti G, and Zanchetti A. Impact of nocturnal fall in blood pressure on early cardiovascular changes in essential hypertension. J Hypertens 1999; 17:1339-44. 80. Mancia G, Bertinieri G, Grassi G, Parati G, Pomidossi G, Ferrari A, Gregorini L, and Zanchetti A. Effects of blood-pressure measurement by the doctor on patient’s blood pressure and heart rate. Lancet 1983; 2:695-8. 81. Pickering TG, Coats A, Mallion JM, Mancia G, and Verdecchia P. Blood Pressure Monitoring. Task force V: White-coat hypertension. Blood Press Monit 1999; 4:333-41. 82. Hoegholm A, Bang LE, Kristensen KS, Nielsen JW, and Holm J. Microalbuminuria in 411 untreated individuals with established hypertension, white coat hypertension, and normotension. Hypertension 1994; 24:101-5. 83. Palatini P, Mormino P, Santonastaso M, Mos L, Dal Follo M, Zanata G, and Pessina AC. Target-organ damage in stage I hypertensive subjects with white coat and sustained hypertension: results from the HARVEST study. Hypertension 1998; 31:57-63. 84. Muldoon MF, Nazzaro P, Sutton-Tyrrell K, and Manuck SB. White-coat hypertension and carotid artery atherosclerosis: a matching study. Arch Intern Med 2000; 160:1507-12. 85. Pierdomenico SD, Lapenna D, Guglielmi MD, Antidormi T, Schiavone C, Cuccurullo F, and Mezzetti A. Target organ status and serum lipids in patients with white coat hypertension. Hypertension 1995; 26:801-7. 86. Cavallini MC, Roman MJ, Pickering TG, Schwartz JE, Pini R, and Devereux RB. Is white coat hypertension associated with arterial disease or left ventricular hypertrophy? Hypertension 1995; 26:413-9. 87. Julius S, Mejia A, Jones K, Krause L, Schork N, van de Ven C, Johnson E, Petrin J, Sekkarie MA, Kjeldsen SE, and et al. “White coat” versus “sustained” borderline hypertension in Tecumseh, Michigan. Hypertension 1990; 16:617-23. 88. Verdecchia P, Schillaci G, Borgioni C, Ciucci A, and Porcellati C. White-coat hypertension. Lancet 1996; 348:1444-5; discussion 1445-6. 89. Khattar RS, Senior R, and Lahiri A. Cardiovascular outcome in white-coat versus sustained mild hypertension: a 10-year follow-up study. Circulation 1998; 98:1892-7. 90. Liu JE, Roman MJ, Pini R, Schwartz JE, Pickering TG, and Devereux RB. Cardiac and arterial target organ damage in adults with elevated ambulatory and normal office blood pressure. Ann Intern Med 1999; 131:564-72. 91. Sega R, Trocino G, Lanzarotti A, Carugo S, Cesana G, Schiavina R, Valagussa F, Bombelli M, Giannattasio C, Zanchetti A, and Mancia G. Alterations of cardiac structure in patients with isolated office, ambulatory, or home hypertension: Data from the general population (Pressione Arteriose Monitorate E Loro Associazioni [PAMELA] Study). Circulation 2001; 104:1385-92. 92. Hedstrand H. A study of middle-aged men with particular reference to risk factors for cardiovascular disease. Ups J Med Sci Suppl 1975; 19:1-61. 93. Rush R, Leon L, and Turrel J. Automated simultaneous cholesterol and triglyceride determination on the Auto- Analyzer II instrument. Advances in Automated Analysis 1971; 1:503-507. 94. Sobey WJ, Beer SF, Carrington CA, Clark PM, Frank BH, Gray IP, Luzio SD, Owens DR, Schneider AE, Siddle K, and et al. Sensitive and specific two-site immunoradiometric assays for human insulin, proinsulin, 65-66 split and 32-33 split proinsulins. Biochem J 1989; 260:535-41. 95. DeFronzo RA, Tobin JD, and Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979; 237:E214-23. 96. Eriksson E, Ranby M, Gyzander E, and Risberg B. Determination of plasminogen activator inhibitor in plasma using t-PA and a chromogenic single-point poly-D-lysine stimulated assay. Thromb Res 1988; 50:91-101. 97. Byberg L, Siegbahn A, Berglund L, McKeigue P, Reneland R, and Lithell H. Plasminogen activator inhibitor- 1 activity is independently related to both insulin sensitivity and serum triglycerides in 70-year-old men. Arterioscler

58 24-HOUR AMBULATORY BP

Thromb Vasc Biol 1998; 18:258-64. 98. Andrén B, Lind L, Hedenstierna G, and Lithell H. Left ventricular hypertrophy and geometry in a population sample of elderly males. Eur Heart J 1996; 17:1800-7. 99. Levy D, Savage DD, Garrison RJ, Anderson KM, Kannel WB, and Castelli WP. Echocardiographic criteria for left ventricular hypertrophy: the Framingham Heart Study. Am J Cardiol 1987; 59:956-60. 100. Becker W, Lennernäs M, Gustafsson I, Haraldsdottir J, Nydahl M, Vessby B, and Ytterfors A. Precoded food records compared with weighed food records measuring dietary habits in a population of Swedish adults. Scandina- vian Journal of Nutrition 1998; 42:145-149. 101. Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, and Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation 1994; 90:583-612. 102. Merlo J, Lindblad U, Pessah-Rasmussen H, Hedblad B, Rastam J, Isacsson SO, Janzon L, and Rastam L. Comparison of different procedures to identify probable cases of myocardial infarction and stroke in two Swedish prospective cohort studies using local and national routine registers. Eur J Epidemiol 2000; 16:235-43. 103. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med 1997; 157:2413-46. 104. van Ittersum FJ, Ijzerman RG, Stehouwer CD, and Donker AJ. Analysis of twenty-four-hour ambulatory blood pressure monitoring: what time period to assess blood pressures during waking and sleeping? J Hypertens 1995; 13:1053-8. 105. Fagard R, Brguljan J, Thijs L, and Staessen J. Prediction of the actual awake and asleep blood pressures by various methods of 24 h pressure analysis. J Hypertens 1996; 14:557-63. 106. World Health Organization. Diabetes mellitus: Report of a WHO Study Group. Technical Report Series 727. Geneva: WHO; 1985. 107. Bradley E. The Jackknife, the Bootstrap and Other Resampling Plans. Bristol: J.W. Arrowsmith Ltd., 1982. 108. Perloff D, Grim C, Flack J, Frohlich ED, Hill M, McDonald M, and Morgenstern BZ. Human blood pressure determination by sphygmomanometry. Circulation 1993; 88:2460-70. 109. Netea RT, Smits P, Lenders JW, and Thien T. Does it matter whether blood pressure measurements are taken with subjects sitting or supine? J Hypertens 1998; 16:263-8. 110. Hedstrand H. Studies in Preventive Medicine with Particular Reference to Detection and Treatment of Risk Factors for Cardiovascular Disease. A Feasibility Study in Middle-Aged Men. Uppsala University. 1975. 111. di Rienzo M, Grassi G, Pedotti A, and Mancia G. Continuous vs intermittent blood pressure measurements in estimating 24-hour average blood pressure. Hypertension 1983; 5:264-9. 112. O’Brien E, Waeber B, Parati G, Staessen J, and Myers MG. Blood pressure measuring devices: recommendations of the European Society of Hypertension. BMJ 2001; 322:531-6. 113. Taylor R, Chidley K, Goodwin J, Broeders M, and Kirby B. Accutracker II (version 30/23) ambulatory blood pressure monitor: clinical validation using the British Hypertension Society and Association for the Advancement of Medical Instrumentation standards. J Hypertens 1993; 11:1275-82. 114. White WB, Lund-Johansen P, McCabe EJ, and Omvik P. Clinical evaluation of the Accutracker II ambulatory blood pressure monitor: assessment of performance in two countries and comparison with sphygmomanometry and intra-arterial blood pressure at rest and during exercise. J Hypertens 1989; 7:967-75. 115. Appel LJ, Whelton PK, Seidler AJ, Patel AR, and Klag MJ. The accuracy and precision of the Accutracker ambulatory blood pressure monitor. Am J Epidemiol 1990; 132:343-54. 116. Fotherby MD, and Potter JF. Twenty-four-hour ambulatory blood pressure in old and very old subjects. J Hypertens 1995; 13:1742-6. 117. Sega R, Cesana G, Milesi C, Grassi G, Zanchetti A, and Mancia G. Ambulatory and home blood pressure normality in the elderly: data from the PAMELA population. Hypertension 1997; 30:1-6. 118. O’ Brien E, Murphy J, Tyndall A, Atkins N, Mee F, McCarthy G, Staessen J, Cox J, and O’Malley K. Twenty- four-hour ambulatory blood pressure in men and women aged 17 to 80 years: the Allied Irish Bank Study. J Hypertens 1991; 9:355-60. 119. Staessen JA, O’ Brien E, Amery AK, Atkins N, Baumgart P, De Cort P, Degaute JP, Dolenc P, De Gaudemaris R, Enstrom I, et al. Ambulatory blood pressure in normotensive and hypertensive subjects: results from an international database. J Hypertens Suppl 1994; 12:S1-12. 120. Staessen JA, Fagard R, Lijnen P, Thijs L, van Hulle S, Vyncke G, and Amery A. Ambulatory blood pressure and blood pressure measured at home: progress report on a population study. J Cardiovasc Pharmacol 1994; 23 Suppl 5:S5-11. 121. Ohkubo T, Imai Y, Tsuji I, Nagai K, Ito S, Satoh H, and Hisamichi S. Reference values for 24-hour ambulatory blood pressure monitoring based on a prognostic criterion: the Ohasama Study. Hypertension 1998; 32:255-9. 122. Staessen J, Bulpitt CJ, Fagard R, Mancia G, O’ Brien E, Thijs L, Vyncke G, and Amery A. Reference values for

59 KRISTINA BJÖRKLUND the ambulatory blood pressure and the blood pressure measured at home: a population study. J Hum Hypertens 1991; 5:355-61. 123. Pickering T. Recommendations for the use of home (self) and ambulatory blood pressure monitoring. Amer- ican Society of Hypertension Ad Hoc Panel. Am J Hypertens 1996; 9:1-11. 124. O’Brien E, and Staessen JA. What is “hypertension”? Lancet 1999; 353:1541-3. 125. Colhoun HM, Dong W, and Poulter NR. Blood pressure screening, management and control in England: results from the health survey for England 1994. J Hypertens 1998; 16:747-52. 126. Hansson L, Zanchetti A, Carruthers SG, Dahlof B, Elmfeldt D, Julius S, Menard J, Rahn KH, Wedel H, and Westerling S. Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomised trial. HOT Study Group. Lancet 1998; 351:1755-62. 127. Mancia G, Sega R, Milesi C, Cesana G, and Zanchetti A. Blood-pressure control in the hypertensive population. Lancet 1997; 349:454-7. 128. Barker WH, Mullooly JP, and Linton KL. Trends in hypertension prevalence, treatment, and control: in a well- defined older population. Hypertension 1998; 31:552-9. 129. Imai Y, Nagai K, Sakuma M, Sakuma H, Nakatsuka H, Satoh H, Minami N, Munakata M, Hashimoto J, Yamagishi T, and et al. Ambulatory blood pressure of adults in Ohasama, Japan. Hypertension 1993; 22:900-12. 130. Mancia G, Sega R, Bravi C, De Vito G, Valagussa F, Cesana G, and Zanchetti A. Ambulatory blood pressure normality: results from the PAMELA study. J Hypertens 1995; 13:1377-90. 131. Aihara A, Imai Y, Sekino M, Kato J, Ito S, Ohkubo T, Tsuji I, Satoh H, Hisamichi S, and Nagai K. Discrepancy between screening blood pressure and ambulatory blood pressure: a community-based study in Ohasama. Hyper- tens Res 1998; 21:127-36. 132. Rasmussen SL, Torp-Pedersen C, Borch-Johnsen K, and Ibsen H. Normal values for ambulatory blood pressure and differences between casual blood pressure and ambulatory blood pressure: results from a Danish population survey. J Hypertens 1998; 16:1415-24. 133. Little P, Barnett J, Barnsley L, Marjoram J, Fitzgerald-Barron A, and Mant D. Comparison of agreement between different measures of blood pressure in primary care and daytime ambulatory blood pressure. BMJ 2002; 325:254. 134. Parati G, Omboni S, Staessen J, Thijs L, Fagard R, Ulian L, and Mancia G. Limitations of the difference between clinic and daytime blood pressure as a surrogate measure of the ‘white-coat’ effect. Syst-Eur investigators. J Hypertens 1998; 16:23-9. 135. Parati G, Ulian L, Santucciu C, Omboni S, and Mancia G. Difference between clinic and daytime blood pressure is not a measure of the white coat effect. Hypertension 1998; 31:1185-9. 136. Pearce KA, Grimm RH, Jr., Rao S, Svendsen K, Liebson PR, Neaton JD, and Ensrud K. Population-derived comparisons of ambulatory and office blood pressures. Implications for the determination of usual blood pressure and the concept of white coat hypertension. Arch Intern Med 1992; 152:750-6. 137. Lurbe A, Redon J, Pascual JM, Tacons J, Alvarez V, and Batlle DC. Altered blood pressure during sleep in normotensive subjects with type I diabetes. Hypertension 1993; 21:227-35. 138. Chen JW, Jen SL, Lee WL, Hsu NW, Lin SJ, Ting CT, Chang MS, and Wang PH. Differential glucose tolerance in dipper and nondipper essential hypertension: the implications of circadian blood pressure regulation on glucose tolerance in hypertension. Diabetes Care 1998; 21:1743-8. 139. Kohara K, Nishida W, Maguchi M, and Hiwada K. Autonomic nervous function in non-dipper essential hypertensive subjects. Evaluation by power spectral analysis of heart rate variability. Hypertension 1995; 26:808- 14. 140. Mancia G, and Parati G. Ambulatory blood pressure monitoring and organ damage. Hypertension 2000; 36:894-900. 141. Uzu T, Kazembe FS, Ishikawa K, Nakamura S, Inenaga T, and Kimura G. High sodium sensitivity implicates nocturnal hypertension in essential hypertension. Hypertension 1996; 28:139-42. 142. Lusardi P, Zoppi A, Preti P, Pesce RM, Piazza E, and Fogari R. Effects of insufficient sleep on blood pressure in hypertensive patients: a 24-h study. Am J Hypertens 1999; 12:63-8. 143. Verdecchia P, Schillaci G, Guerrieri M, Gatteschi C, Benemio G, Boldrini F, and Porcellati C. Circadian blood pressure changes and left ventricular hypertrophy in essential hypertension. Circulation 1990; 81:528-36. 144. James MA, Fotherby MD, and Potter JF. Reproducibility of the circadian systolic blood pressure variation in the elderly. J Hypertens 1995; 13:1097-103. 145. Dimsdale JE, and Heeren MM. How reliable is nighttime blood pressure dipping? Am J Hypertens 1998; 11:606-9. 146. Cuspidi C, Macca G, Sampieri L, Fusi V, Severgnini B, Michev I, Salerno M, Magrini F, and Zanchetti A. Target organ damage and non-dipping pattern defined by two sessions of ambulatory blood pressure monitoring in

60 24-HOUR AMBULATORY BP recently diagnosed essential hypertensive patients. J Hypertens 2001; 19:1539-45. 147. Julius S, and Jamerson K. Sympathetics, insulin resistance and coronary risk in hypertension:the “chicken-and- egg” question. J Hypertens 1994; 12:495-502. 148. Julius S, Krause L, Schork NJ, Mejia AD, Jones KA, van de Ven C, Johnson EH, Sekkarie MA, Kjeldsen SE, Petrin J, and et al. Hyperkinetic borderline hypertension in Tecumseh, Michigan. J Hypertens 1991; 9:77-84. 149. Palatini P, and Julius S. Heart rate and the cardiovascular risk. J Hypertens 1997; 15:3-17. 150. Paolisso G, Manzella D, Rizzo MR, Ragno E, Barbieri M, Varricchio G, and Varricchio M. Elevated plasma fatty acid concentrations stimulate the cardiac autonomic nervous system in healthy subjects. Am J Clin Nutr 2000; 72:723-30. 151. Steinberg HO, Tarshoby M, Monestel R, Hook G, Cronin J, Johnson A, Bayazeed B, and Baron AD. Elevated circulating free fatty acid levels impair endothelium-dependent vasodilation. J Clin Invest 1997; 100:1230-9. 152. Andersson P, Lind L, Berne C, Berglund L, and Lithell HO. Insulin-mediated vasodilation and glucose uptake are independently related to fasting serum nonesterified fatty acids in elderly men. J Intern Med 1999; 246:529- 37. 153. Nikkari T, Luukkainen P, Pietinen P, and Puska P. Fatty acid composition of serum lipid fractions in relation to gender and quality of dietary fat. Ann Med 1995; 27:491-8. 154. Vessby B, Gustafsson IB, Boberg J, Karlstrom B, Lithell H, and Werner I. Substituting polyunsaturated for saturated fat as a single change in a Swedish diet: effects on serum lipoprotein metabolism and glucose tolerance in patients with hyperlipoproteinaemia. Eur J Clin Invest 1980; 10:193-202. 155. Plotnick GD, Corretti MC, and Vogel RA. Effect of antioxidant vitamins on the transient impairment of endothelium-dependent brachial artery vasoactivity following a single high-fat meal. JAMA 1997; 278:1682-6. 156. Ong PJ, Dean TS, Hayward CS, Della Monica PL, Sanders TA, and Collins P. Effect of fat and carbohydrate consumption on endothelial function. Lancet 1999; 354:2134. 157. Sarabi M, Vessby B, Millgard J, and Lind L. Endothelium-dependent vasodilation is related to the fatty acid composition of serum lipids in healthy subjects. Atherosclerosis 2001; 156:349-55. 158. Grimsgaard S, Bonaa KH, Jacobsen BK, and Bjerve KS. Plasma saturated and linoleic fatty acids are independently associated with blood pressure. Hypertension 1999; 34:478-83. 159. Simon JA, Fong J, and Bernert JT, Jr. Serum fatty acids and blood pressure. Hypertension 1996; 27:303-7. 160. Storlien LH, Hulbert AJ, and Else PL. Polyunsaturated fatty acids, membrane function and metabolic diseases such as diabetes and obesity. Curr Opin Clin Nutr Metab Care 1998; 1:559-63. 161. Ohrvall M, Tengblad S, Gref CG, Salminen I, Aro A, and Vessby B. Serum alpha tocopherol concentrations and cholesterol ester fatty acid composition in 70-year-old men reflect those 20 years earlier. Eur J Clin Nutr 1996; 50:381-5. 162. Sundstrom J, Lind L, Vessby B, Andren B, Aro A, and Lithell H. Dyslipidemia and an Unfavorable Fatty Acid Profile Predict Left Ventricular Hypertrophy 20 Years Later. Circulation 2001; 103:836-41. 163. Perticone F, Maio R, Ceravolo R, Cosco C, Cloro C, and Mattioli PL. Relationship between left ventricular mass and endothelium-dependent vasodilation in never-treated hypertensive patients. Circulation 1999; 99:1991-6. 164. Khattar RS, Swales JD, Dore C, Senior R, and Lahiri A. Effect of aging on the prognostic significance of ambulatory systolic, diastolic, and pulse pressure in essential hypertension. Circulation 2001; 104:783-9. 165. Domanski M, Mitchell G, Pfeffer M, Neaton JD, Norman J, Svendsen K, Grimm R, Cohen J, and Stamler J. Pulse pressure and cardiovascular disease-related mortality: follow-up study of the Multiple Risk Factor Intervention Trial (MRFIT). JAMA 2002; 287:2677-83. 166. Khattar RS, Swales JD, Banfield A, Dore C, Senior R, and Lahiri A. Prediction of coronary and cerebrovascular morbidity and mortality by direct continuous ambulatory blood pressure monitoring in essential hypertension. Circulation 1999; 100:1071-6. 167. Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Pede S, and Porcellati C. Ambulatory pulse pressure: a potent predictor of total cardiovascular risk in hypertension. Hypertension 1998; 32:983-8. 168. London GM, and Guerin A. Influence of arterial pulse and reflective waves on systolic blood pressure and cardiac function. J Hypertens Suppl 1999; 17:S3-6. 169. Boutouyrie P, Tropeano AI, Asmar R, Gautier I, Benetos A, Lacolley P, and Laurent S. Aortic stiffness is an independent predictor of primary coronary events in hypertensive patients: a longitudinal study. Hypertension 2002; 39:10-5. 170. Laurent S, Boutouyrie P, Asmar R, Gautier I, Laloux B, Guize L, Ducimetiere P, and Benetos A. Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients. Hypertension 2001; 37:1236-41. 171. Meaume S, Benetos A, Henry OF, Rudnichi A, and Safar ME. Aortic pulse wave velocity predicts cardiovascular mortality in subjects >70 years of age. Arterioscler Thromb Vasc Biol 2001; 21:2046-50. 172. O’Rourke M, and Frohlich ED. Pulse pressure: Is this a clinically useful risk factor? Hypertension 1999;

61 KRISTINA BJÖRKLUND

34:372-4. 173. Verdecchia P, Borgioni C, Ciucci A, Gattobigio R, Schillaci G, Sacchi N, Santucci A, Santucci C, Reboldi G, and Porcellati C. Prognostic significance of blood pressure variability in essential hypertension. Blood Press Monit 1996; 1:3-11. 174. Mancia G, Ferrari A, Gregorini L, Parati G, Pomidossi G, Bertinieri G, Grassi G, di Rienzo M, Pedotti A, and Zanchetti A. Blood pressure and heart rate variabilities in normotensive and hypertensive human beings. Circ Res 1983; 53:96-104. 175. Imai Y, Aihara A, Ohkubo T, Nagai K, Tsuji I, Minami N, Satoh H, and Hisamichi S. Factors that affect blood pressure variability. A community-based study in Ohasama, Japan. Am J Hypertens 1997; 10:1281-9. 176. La Rovere MT, Bigger JT, Jr., Marcus FI, Mortara A, and Schwartz PJ. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet 1998; 351:478-84. 177. Omboni S, Parati G, Castiglioni P, Di Rienzo M, Imholz BP, Langewouters GJ, Wesseling KH, and Mancia G. Estimation of blood pressure variability from 24-hour ambulatory finger blood pressure. Hypertension 1998; 32:52-8. 178. Parati G, and Mancia G. Blood pressure variability as a risk factor. Blood Press Monit 2001; 6:341-7. 179. Wing LM, Brown MA, Beilin LJ, Ryan P, and Reid CM. ‘Reverse white-coat hypertension’ in older hypertensives. J Hypertens 2002; 20:639-644. 180. Bobrie G, Genes N, Vaur L, Clerson P, Vaisse B, Mallion JM, and Chatellier G. Is “isolated home” hypertension as opposed to “isolated office” hypertension a sign of greater cardiovascular risk? Arch Intern Med 2001; 161:2205- 11. 181. Selenta C, Hogan BE, and Linden W. How often do office blood pressure measurements fail to identify true hypertension? An exploration of white-coat normotension. Arch Fam Med 2000; 9:533-40. 182. Mann SJ, James GD, Wang RS, and Pickering TG. Elevation of ambulatory systolic blood pressure in hypertensive smokers. A case-control study. JAMA 1991; 265:2226-8. 183. Waeber B, Scherrer U, Petrillo A, Bidiville J, Nussberger J, Waeber G, Hofstetter JR, and Brunner HR. Are some hypertensive patients overtreated? A prospective study of ambulatory blood pressure recording. Lancet 1987; 2:732-4. 184. Myers MG, Reeves RA, Oh PI, and Joyner CD. Overtreatment of hypertension in the community? Am J Hypertens 1996; 9:419-25. 185. Fagard RH, Staessen JA, Thijs L, Gasowski J, Bulpitt CJ, Clement D, de Leeuw PW, Dobovisek J, Jaaskivi M, Leonetti G, O’Brien E, Palatini P, Parati G, Rodicio JL, Vanhanen H, and Webster J. Response to antihypertensive therapy in older patients with sustained and nonsustained systolic hypertension. Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Circulation 2000; 102:1139-44. 186. Staessen JA, Byttebier G, Buntinx F, Celis H, O´Brien E, and Fagard R. Antihypertensive treatment based on conventional or ambulatory blood pressure measurement. A randomized controlled trial. Ambulatory Blood Pres- sure Monitoring and Treatment of Hypertension Investigators. JAMA 1997; 278:1065-72. 187. Barrett-Connor E. Sex differences in coronary heart disease. Why are women so superior? The 1995 Ancel Keys Lecture. Circulation 1997; 95:252-64. 188. Safar ME, and London GM. Therapeutic studies and arterial stiffness in hypertension: recommendations of the European Society of Hypertension. The Clinical Committee of Arterial Structure and Function. Working Group on Vascular Structure and Function of the European Society of Hypertension. J Hypertens 2000; 18:1527- 35. 189. Svensson P, de Faire U, Sleight P, Yusuf S, and Ostergren J. Comparative effects of ramipril on ambulatory and office blood pressures: a HOPE Substudy. Hypertension 2001; 38:E28-32. 190. Campbell NR, and McKay DW. Accurate blood pressure measurement: why does it matter? CMAJ 1999; 161:277-8. 191. Redon J, Campos C, Narciso ML, Rodicio JL, Pascual JM, and Ruilope LM. Prognostic value of ambulatory blood pressure monitoring in refractory hypertension: a prospective study. Hypertension 1998; 31:712-8