Background. The relation between high blood pres- sure (BP) and mortality risk in older individuals. (above 65 years of age) is still debated. Some data suggest ...
Original Article doi: 10.1111/joim.12284
Relation between blood pressure and mortality risk in an older population: role of chronological and biological age G. Post Hospers1, Y. M. Smulders1, A. B. Maier1, D. J. Deeg2 & M. Muller3 From the Departments of 1 Internal Medicine; 2 Epidemiology, VU University, Amsterdam; and 3Department of Gerontology and Geriatrics, Leiden University Medical Center, Leiden, The Netherlands
Abstract. Post Hospers G, Smulders YM, Maier AB, Deeg DJ, Muller M (VU University, Amsterdam; and Leiden University Medical Center, Leiden, The Netherlands). Relation between blood pressure and mortality risk in an older population: role of chronological and biological age. J Intern Med 2015; 277:488–497. Background. The relation between high blood pressure (BP) and mortality risk in older individuals (above 65 years of age) is still debated. Some data suggest that this relation is inverted in certain subgroups of (biologically) older individuals. We therefore investigated whether the association between BP and mortality is dependent on chronological age and on physical and cognitive function as indicators of biological age. Methods. The relationship between BP and all-cause mortality was investigated in 1466 older participants (aged 65 years and older; mean age 75.8 years) of the Longitudinal Ageing Study Amsterdam using multiple adjusted Cox proportional hazard models. Analyses were stratified for
Introduction Blood pressure (BP) levels increase with age, and the prevalence of hypertension is highest in the oldest old [1, 2]. Recent guidelines for the management of high BP recommend more liberal BP targets (≥150/90 mmHg) in individuals of 60 years and older, compared to those below this age [3]. However, it is still unclear whether high BP is a risk factor for increased mortality in older populations and therefore whether high BP levels should be treated in all individuals to reduce mortality risk. Several population-based studies have shown that low rather than high BP has the greatest mortality risk in the ‘oldest old’ (individuals aged >80 years) [4, 5]. These results suggest that the current directive of ‘lower is better’ may not apply to BP
488
ª 2014 The Association for the Publication of the Journal of Internal Medicine
age, gait speed and mini mental state examination score. Results. A total of 1008 participants died after a median (range) follow-up of 10.6 (0.2; 15.9) years. Low diastolic blood pressure (DBP) was associated with an increased all-cause mortality risk: hazard ratio (HR) of low DBP (≤70 mmHg) compared to normal DBP (71–90 mmHg) was 1.36 [95% confidence interval (CI) 1.15; 1.61]. This relation was particularly strong in the oldest old (individuals aged >80 years) and in those who had lower levels of both physical and cognitive functioning: HRs (95% CIs) of low versus normal DBP were 1.58 (1.26; 1.98) and 1.45 (1.18; 1.77), respectively. Conclusion. In a large population-based cohort of older adults, low DBP was associated with an increased all-cause mortality risk, especially in the oldest old and in biologically old individuals. Keywords: ageing, blood pressure, cognition, frailty, mortality.
levels and regulation in the entire population of older people. Based on the results of a recent review, it was concluded that biological age, or frailty, of older individuals should be considered in the treatment of high BP [6]. Odden et al. [7] suggested that walking speed (measured over 6 m), as a proxy for frailty or biological age, could be used to understand the complex relation between BP and mortality risk and to categorize those older individuals who are likely to benefit from antihypertensive treatment. The authors found that high BP was associated with an increased all-cause mortality risk amongst fast walkers, whereas the inverse relation was observed in individuals who were not able to complete the walking test: high BP was
G. Post Hospers et al.
associated with reduced mortality risk. In addition, another study in the oldest old showed that the relation between BP and stroke was depended on physical and cognitive functioning; in subjects with physical and cognitive impairment, high BP was associated with lower risk of stroke [8]. Nevertheless, the role of cognitive functioning in the relation between BP and mortality remains unclear. Therefore, the aim of the present study was to investigate the relation between BP and mortality risk in a large population-based study of older individuals, and determine whether this relation is dependent on chronological age and/or on markers of biological age such as physical or cognitive performance. Methods Longitudinal Ageing Study Amsterdam Data from the Longitudinal Ageing Study Amsterdam (LASA), an ongoing prospective populationbased cohort study in older individuals (≥55 years of age), were used in the present study. The aim of LASA was to investigate the determinants of, changes in and consequences of physical, cognitive, emotional and social functioning in relation to ageing [9]. A series of measurements was carried out at baseline (1992–1993) and every 3 years thereafter. The Medical Ethics Committee of the VU University Medical Center approved the study, and informed consent was obtained from all participants. Because BP was included as part of the second set of measurements (1995–1996) in participants aged 65 years and older, this set was used as the baseline. Of the 1470 participants with valid BP measurements, mortality data were missing for only four individuals; therefore, the present study sample consisted of 1466 participants. Mortality All-cause mortality was determined by collecting data from the municipal registers. The follow-up period was defined as the time between baseline measurements (1995–1996) and death from any cause, loss to follow-up or 1 September 2011. Blood pressure Systolic (SBP) and diastolic blood pressure (DBP) were automatically measured (Omron HEM-706; Omron Corporation, Tokyo, Japan) at the upper left arm after 5 min of rest with subjects in a seated
Blood pressure and mortality
position. Pulse pressure was calculated by subtracting DBP from SBP. Biological age Physical function was assessed using a 6-m walking test, with a turn of 180° at 3 m. Gait speed was categorized as fast (≥0.8 m s 1) and slow (140 mmHg for SBP; and ≤70, 71–90 and >90 mmHg for DBP, respectively. Differences in baseline categories were tested using ANOVA for continuous variables and chi-squared test for categorical variables. Cox proportional hazards analyses were performed to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between BP (SBP and DBP; per 10 mmHg increase) and mortality risk. In addition, analyses were performed using BP categories; normal SBP and DBP were set as the reference categories. P-values for trend were calculated based on these three groups of BP. Regression analyses were ª 2014 The Association for the Publication of the Journal of Internal Medicine Journal of Internal Medicine, 2015, 277; 488–497
489
G. Post Hospers et al.
adjusted for age and sex (model 1), and additionally for educational level, BMI, smoking, alcohol consumption, total cholesterol, CVD, diabetes and use of antihypertensive drugs (model 2). To assess the possible modifying effects of chronological and/or biological age on the relation between BP and mortality, regression analyses were stratified for age (three equal-sized groups of ≤72, 73–80 and >80 years), gait speed categories (fast, slow and did not complete), MMSE score (≤26, 27–28 and >28) and the combination score (0–1 and 2–4). Interaction terms between BP (SBP or DBP) and measures of chronological or biological age (age, gait speed, MMSE score and combination score) were added to the regression models. Survival curves for the categories of BP combined with measures of chronological or biological age were constructed. To minimize the possibility that underlying comorbidity may explain the relation between BP and mortality risk, the analyses were repeated within strata of follow-up years (0–5 and 6–16 years), assuming that longer follow-up is more suggestive of a causal relation between BP and mortality. To determine whether the results differed in participants with the highest BP values, additional analyses were performed with four categories of SBP (≤120, 121–140, 141–160 and >160 mmHg). Analyses with four groups of DBP were not possible because of small numbers of subjects in the highest DBP group. Data were analysed using SPSS version 20.0 (Chicago, IL, USA). Results Of the total population of 1466 subjects aged 65 years and older, those with higher SBP and DBP were younger, had a lower level of education, were less likely to use antihypertensive drugs and had higher BMI values, higher cholesterol levels and lower rates of CVD (Table 1). In addition, higher DBP was associated with higher MMSE scores. Pulse pressure was almost twofold higher in participants with high SBP compared to those with low SBP. A total of 1008 participants died after a median (range) follow-up of 10.6 (0.2; 15.9) years. In the total population, SBP was not related to all-cause mortality risk: HR (95% CI) per 10 mmHg increase in SBP was 1.01 (0.99; 1.04; Table 2). However, lower DBP was related to an increased all-cause mortality risk independent of age and sex: HR (95% CI) per 10 mmHg increase in DBP was 0.93 (0.88; 0.97; P < 0.05). After further adjustment for CVD 490
ª 2014 The Association for the Publication of the Journal of Internal Medicine Journal of Internal Medicine, 2015, 277; 488–497
Blood pressure and mortality
risk factors, both low DBP and high DBP were associated with an increased mortality risk (Table 3, model 2). BP and mortality: effect of chronological age Following stratification for categories of chronological age, we found that SBP and DBP were not associated with all-cause mortality risk in the youngest age group (≤72 years; Tables 2 and 3). By contrast, in the oldest age group, SBP and DBP were both inversely associated with mortality risk. However, the association between SBP and mortality was attenuated after adjustment for CVD risk factors (Table 2, model 2). The results indicate that age significantly modified the relation between DBP and mortality risk (interaction between SBP and age: P = 0.10; interaction between DBP and age: P = 0.01). BP and mortality: effect of biological age When stratifying for categories of gait speed or MMSE score, or the combination of these measures, we found that the relation between SBP and mortality risk did not depend on either physical or cognitive function alone or on their combination (Table 2, model 1 and Fig. 1). Effect estimates did not change after further adjustment for CVD risk factors (Table 2, model 2). P-values for the interaction between SBP and gait speed, MMSE score and the combination score were 0.70, 0.74 and 0.39, respectively. However, lower DBP was associated with an increased mortality risk in participants with physical dysfunction (gait speed slow or did not complete), cognitive dysfunction (MMSE score ≤26) or a combination of these variables (combination score 2–4), independent of age (Table 3, model 1). After further adjustment for CVD risk factors, high DBP was also associated with increased mortality, but only in participants without cognitive dysfunction or without a combination of physical and cognitive dysfunction (Table 3, model 2). P-values for the interaction between DBP and gait speed, MMSE score or the combination score were 0.70, 0.64 and 0.04, respectively, suggesting that the combination of physical and cognitive function in particular modified the relation between DBP and mortality risk (Fig. 2). When the analyses were stratified according to follow-up time (0–5 years, n = 1466 vs. 6–16 years,
59
33
Low
Use of
drugs, %
DBP (mmHg)
)
2 a
)
1 a
1 a
28 (26; 29)
0.8 (0.3)
3 (2; 6)
14
46
8
5.5 (1.0)
25.8 (4.2)
50
28 (25; 29)
0.8 (0.3)
3 (2; 4)
12
40
6
5.6 (1.1)
26.0 (4.2)
75
48
18
55 (50; 61)
76 (70; 81)
132 (127; 136)
32
59
29
12
75.7 (7.0)
27 (26; 29)
0.8 (0.3)
2 (1; 4)
13
34
9
5.7 (1.0)
27.2 (4.0)
74
45
19
78 (68; 89)
87 (80; 95)
163 (152; 180)
32
64
25
12
75.9 (6.4)
49
968 (66)
>140
61 (50; 74)
