Author + information
- Received March 21, 2000
- Revision received August 9, 2000
- Accepted September 26, 2000
- Published online January 1, 2001.
- Athanase Benetos, MD, PhD∗,†,* (, )
- Frederique Thomas, PhD∗,
- Michel E Safar, MD†,
- Kathryn E Bean, MA, MPH∗ and
- Louis Guize, MD∗
- ↵*Reprint requests and correspondence: Dr. Athanase Benetos, Centre d’Investigations Préventives et Cliniques (IPC), 6-14 rue de la Pérouse, 75116 Paris, France
The goal of this study was to evaluate the role of diastolic blood pressure (DBP) in cardiovascular mortality for different systolic blood pressure (SBP) levels in middle-aged men and women.
In middle-aged subjects it is unclear whether DBP, in addition to SBP, should be considered for risk evaluation.
Subjects (77,023 men; 48,480 women) aged 40 to 70 years old, had no major cardiovascular disease, no antihypertensive treatment and were examined at the Centre d’Investigations Préventives et Cliniques between 1972 and 1988. Mortality was assessed for an 8- to 12-year period.
In both genders, cardiovascular mortality increased with the SBP level. In men and women with normal SBP levels, DBP did not influence cardiovascular mortality after adjustment for age and SBP. In men with systolic hypertension, a U-shaped curve relationship between cardiovascular mortality and DBP was observed, with the lowest mortality rates in the group with DBP 90 to 99 mm Hg. Compared with this group, age- and SBP-adjusted cardiovascular mortality was higher by 73% (p < 0.02) in the group with DBP <90 mm Hg and by 65% (p < 0.001) in the group with DBP ≥110 mm Hg. In women with systolic hypertension, however, DBP was positively correlated with cardiovascular mortality.
In middle-aged subjects, classification of cardiovascular risk according to DBP levels should take into account gender, especially when SBP levels are elevated. Men with systolic hypertension are at higher risk when their DBP is “normal” than when they present a mild to moderate increase in DBP. In women of the same age, however, systolic-diastolic hypertension represents a higher risk than isolated systolic hypertension.
In various epidemiological studies, a strong positive relationship between cardiovascular disease and blood pressure (BP) has been reported (1,2). Both systolic blood pressure (SBP) and diastolic blood pressure (DBP) have been shown to be associated with cardiovascular risk. In the 1970s and early 1980s, most international medical authorities suggested that DBP was the main determinant of cardiovascular risk. More recently, the importance of SBP has also been pointed out, especially in older subjects (3,4). The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (5)and the recent report of the World Health Organization/International Society of Hypertension (6)classified BP into stages based on either SBP or DBP levels. Both classifications define the lowest SBP and DBP levels as being the optimal levels and the highest levels of both SBP and DBP as being the levels with the highest cardiovascular morbidity and mortality. During the last years however, a large number of studies have suggested that in several populations, pulse pressure (PP = SBP-DBP) is a major indicator, if not the best, of cardiovascular risk (7–9). These observations suggest that more attention should be given to SBP and PP since both of them may play a greater role in determining cardiovascular risk than DBP (10–12). Isolated systolic hypertension is identified as a particular form of hypertension that requires treatment, but it is generally thought that isolated systolic hypertension is a risk factor only in older subjects. For middle-aged subjects it is unclear whether or not for a given level of SBP, cardiovascular risk is proportional to the level of DBP. The majority of clinicians think that having both an elevated SBP and DBP is a more important risk factor than having an isolated increase in SBP. It is also unclear whether or not the role of DBP is influenced by gender.
The aim of this study was to evaluate the role of DBP in cardiovascular mortality for different SBP levels in middle-aged men and women. This study was conducted in a large French population of 77,023 men and 48,480 women examined at the Center d’Investigations Préventives et Cliniques (IPC) (7,13).
Subjects were examined at the IPC Center, a medical center that is subsidized by the French national health care system (Sécurité Sociale [CNAM]) and provides free medical examinations for all working and retired persons and their families. It is one of the largest medical centers of this kind in France, having carried out approximately 20,000 to 25,000 examinations per year since 1970 for people living in the Paris area. Between January 1972 and December 1988, 139,315 individuals 40 to 70 years of age (84,228 men and 55,087 women) had a free health check-up at the IPC Center. Subjects who indicated that they had previously been hospitalized for a heart attack, cerebrovascular disease or other cardiac disease, as well as those who presented clinical or electric signs of myocardial infarction, heart failure stage III or IV or other major cardiovascular diseases were excluded from the study (1,584 men and 1,606 women). Subjects with antihypertensive treatment were also excluded from this analysis (4,564 men and 4,160 women). An additional 1,848 subjects (1,057 men and 841 women) were not included for a variety of reasons, such as missing clinical or mortality data or their refusal to participate. All other subjects aged 40 to 70 years were included in the present analysis.
In this study the data describe a population composed of 125,503 individuals (77,023 men and 48,480 women). The IPC Center received authorization from the “Comité National d’Informatique et des Libertés” to conduct these analyses. All subjects gave their informed consent at the time of the examination. Based on the national statistics of mortality, our cohort presented a 20% lower mortality rate than the general French population. This finding can be explained by the assumption that people who came for the health check-up were apparently healthy and motivated to be followed up. Interestingly, compared with the national data, the distribution of the different causes of mortality in our cohort was identical to that of the general population.
Supine BP was measured in the right arm using a manual mercury sphygmomanometer, after a 10-min rest period. The first and the fifth Korotkoff phases were used to define systolic and diastolic pressures. Biological parameters were measured under fasting conditions, and electrocardiogram measurements were also recorded.
For all screened subjects, vital status was obtained for an 8- to 12-year period (mean 11.6 + 0.2), which extended from the time of inclusion through December 1996. These data were obtained from the mortality records of the “Institut National de Statistiques et d’Etudes Economiques” (INSEE) after a previously described procedure (7). Causes of mortality, taken from the death certificates, were provided by INSERM’s Department of Mortality Studies (Unit SC 8). Causes of death were codified according to the International Classification of Disease (8th revision until 1978 and 9th revision thereafter). Cardiovascular-related deaths were coded 390 to 459 and 798 (sudden death). Permission to obtain mortality data was given by the “Comité National de l’Informatique et des Libertés.”
Men and women were studied separately. They were divided into three groups of SBP. Inside each SBP group, men were divided into four DBP groups, whereas women were divided into three DBP groups. Different classifications were necessary for men and women due to the lower number of subjects and lower mortality rates in women.
Differences in cardiovascular mortality among SBP groups were calculated after adjustment for age by using a Cox proportional hazards regression analysis. Cardiovascular deaths among the DBP groups, within each SBP group, were compared with the use of a chi-square test. Differences in survival probability for cardiovascular mortality, from one DBP group to the other, were tested by using a Cox proportional hazards regression analysis with adjustments for age and SBP levels. Complementary adjustments for other major cardiovascular risk factors (total cholesterol, diabetes, and tobacco consumption) did not significantly modify the results.
For cardiovascular mortality, the role of DBP within each SBP group was also assessed when DBP was considered as a continuous quantitative parameter. Risk ratios (RR) and 95% confidence intervals related to the increase of one standard deviation of DBP were calculated using a Cox proportional hazards regression analysis, after adjustment for age and SBP or age, SBP and other risk factors (total cholesterol, diabetes, tobacco consumption).
The level of statistical significance was 5%. All statistical analyses were carried out using the SAS statistical software package.
The general characteristics of the studied population are shown in Table 1. During the follow-up period, a total of 4,055 men and 1,112 women died. Among them, 849 men and 154 women died from cardiovascular-related diseases.
Characteristics of SBP/DBP groups
The different subgroups according to SBP and DBP in men and women, respectively, are shown in Tables 2 and 3. ⇓⇓Within each SBP group, the increase in mean arterial pressure in the higher DBP groups was not only due to higher diastolic levels but also to higher SBP levels. As the increase in SBP was less pronounced than DBP, PP levels were lower in the higher DBP groups. Total cholesterol, glycemia, triglycerides and heart rate did not differ greatly among the DBP groups (data not shown). In both genders, mortality clearly increased in the higher SBP groups. Compared with the men with normal SBP (<140 mm Hg), age-adjusted RR (95% confidence intervals) for cardiovascular mortality was 1.57 (1.30 to 1.90) in men with borderline systolic hypertension (SBP ≥140 to 149 mm Hg) and 2.99 (2.54 to 3.52) in men with SBP ≥150 mm Hg. In women, similar results were observed, and the respective relative risks were 2.75 (1.82 to 4.14) and 2.94 (1.96 to 4.42).
Mortality rates according to DBP for different levels of SBP
Normal SBP group (<140 mm Hg)
In men, unadjusted cardiovascular mortality rates (Table 2, upper panel) were more elevated in the higher DBP groups (trend test: p < 0.05). However, after adjustment for age and SBP, cardiovascular mortality was the same in the different DBP subgroups. A similar result was observed when DBP was considered as a continuous quantitative variable: age- and SBP-adjusted RR for one standard deviation of DBP was 1.16 (0.92 to 1.45). In women with normal SBP, cardiovascular mortality rates were very low in all DBP subgroups without any significant differences among them (Table 3, upper panel).
Borderline systolic hypertension group (140 to 149 mm Hg)
As shown in Table 2(middle panel), in men with borderline systolic hypertension, the highest cardiovascular mortality rates were observed in those with the lowest DBP (<80 mm Hg) and the highest DBP levels (≥100 mm Hg), whereas the lowest cardiovascular mortality rates were recorded in subjects with DBP ≥90 to 99 mm Hg (p < 0.05 vs. DBP < 80 or vs. DBP ≤ 100 mm Hg). However, after adjustment for age and SBP, the influence of DBP on cardiovascular mortality was not significant in this group; the RR for one standard deviation of DBP was 1.05 (0.75 to 1.43).
Women with DBP ≥90 mm Hg had cardiovascular mortality rates twice as high as those with low DBP (Table 3, middle panel) (p < 0.01). The difference among groups was significant even after adjustment for age and SBP (p < 0.01). When DBP was considered as a continuous variable, age- and SBP-adjusted RR for one standard deviation of DBP was 1.54 (0.97 to 4.46).
High SBP group (≥150 mm Hg)
In men, the highest unadjusted cardiovascular mortality rates were observed in those with the lowest DBP levels and in those with the highest DBP levels (Table 2, lower panel). The lowest cardiovascular mortality rates were recorded in subjects with DBP between 90 and 99 mm Hg. Compared with this group, age- and SBP-adjusted cardiovascular mortality was higher by 73% (p < 0.02) in the isolated systolic hypertension group, by 8% (NS) in the DBP = 100 to 109 mm Hg group and by 62% (p < 0.001) in the DBP ≥ 110 mm Hg group (Fig. 1, left panel). When DBP was considered as a continuous variable, age- and SBP-adjusted RR of cardiovascular mortality for one standard deviation of DBP was 0.99 (0.89 to 1.10). On the contrary, in women, the highest DBP levels were associated with the highest cardiovascular mortality rates (Table 3). This difference among groups was significant even after adjustment for age and SBP (trend test: p < 0.01) (Fig. 1, right panel). When DBP was considered as a continuous variable, age- and SBP-adjusted RR of cardiovascular mortality for one standard deviation of DBP was 1.39 (1.03 to 1.87).
In men, PP was a significant determinant of cardiovascular mortality. Relative risk for one standard deviation of PP was 1.24 (1.16 to 1.33). In women, however, PP was not significantly associated with cardiovascular mortality (RR: 0.93 [0.74 to 1.17]).
In a large middle-aged French population, we observed that the role of DBP in cardiovascular mortality varies according to gender and SBP level.
In both men and women with normal SBP (<140 mm Hg), DBP was not a significant determinant of cardiovascular mortality, after adjustment for age and SBP.
In men with systolic hypertension, a U-shaped curve relationship between cardiovascular mortality and DBP was observed. The highest cardiovascular death rates were recorded in men with DBP values less than 90 mm Hg (isolated systolic hypertension) and in those with DBP values greater than 110 mm Hg.
In women with systolic hypertension, however, DBP was positively correlated with cardiovascular mortality; women with isolated systolic hypertension were at a much lower risk than those with systolic-diastolic hypertension.
SBP/DBP and risk determination
In various epidemiological studies both SBP and DBP have been shown to be associated with cardiovascular risk (1–4). A predominant role of SBP has been pointed out, especially in older subjects (4,14). A number of studies have also shown that, even in middle-aged populations, SBP is a stronger predictor of coronary heart disease than DBP (15,16). Based on the analyses of the Multiple Risk Factor Intervention Trial (MRFIT) database, Neaton et al. (15)showed that, in white men aged 35 to 57, differences in coronary mortality among DBP categories for each SBP category were small. On the contrary, a strong risk gradient was evident for SBP for each DBP category. They also showed that the highest coronary age-adjusted mortality rates were observed in the group with the higher SBP (>160 mm Hg) and the lower DBP (<70 mm Hg). The authors suggested that this group might include men with pre-existing heart disease. A recent analysis of the Framingham population by Franklin et al. (16)showed that, in men and women (studied together) aged 50 to 79 without pre-existing heart disease, coronary risk strongly increased with higher SBP. This risk also increased with lower DBP for any given level of SBP. In our analysis, all subjects with pre-existing cardiovascular disease were excluded. In this population (which was younger than the one studied by Franklin et al. ), we also found a higher risk of cardiovascular mortality with lower DBP, but only in men and especially in those with high SBP.
Gender differences in the role of DBP
No such U-shaped curve was observed in the women of our population. To the best of our knowledge, most of the studies conducted so far have evaluated the role of SBP and DBP only in male or in mixed populations, without separate analyses for women. The authors of a recent analysis of the respective roles of SBP and DBP in the Framingham population reported that gender was not a significant determinant in the relationship between DBP and coronary mortality (16). However, subjects up to 79 years of age were included in that study. We previously reported that the role of PP was not the same in men as it was in women (17). Actually, no association between cardiovascular mortality and PP was observed in either normotensive or hypertensive women. In that study, we had suggested that low mortality rates could explain this observation in normotensive, but not in hypertensive, women for whom cardiovascular mortality rates were relatively high. Low mortality rates in women could also be one of the explanations for the absence of any association between SBP, DBP or PP and cardiovascular mortality in untreated women of the Medical Research Council (MRC) trial (18). Despite low cardiovascular rates in women observed in this study, our results clearly show that, in women, cardiovascular mortality is positively correlated with DBP levels without any trend toward a U-shaped curve.
Arterial determinants of SBP, DBP and PP
Although SBP increases with an elevation in total peripheral resistance or an increase in large artery stiffness, DBP levels are influenced by arterial or arteriolar alterations in opposite ways; an increase in peripheral vascular resistance leads to an elevation in DBP, whereas stiffening of large arteries can contribute to a decrease in DBP (19,20). Therefore, normal levels of DBP may be due to the combination of these two vascular alterations, which are both independent risk factors of cardiovascular disease, especially coronary artery disease. This leads to the conclusion that levels of DBP may not necessarily reflect cardiovascular risk, especially in subjects whose SBP is disproportionally elevated, i.e., those in whom arterial stiffness is more pronounced. Our results show that the U-shaped relationship between DBP and cardiovascular mortality was observed in men with high SBP. This suggests that, among men with systolic hypertension, two high risk subgroups can be identified, one with very low DBP (<90 mm Hg) and, therefore, high PP, the main indicator of pronounced large artery stiffness, and one with very high DBP (≥110 mm Hg), where the main alteration is an elevation of peripheral resistance.
Our results suggest that arterial stiffness in women develops later in life, and, therefore, low DBP is mainly an indicator of low peripheral resistance and not, as has been observed in men, an indicator of pronounced large artery stiffness. It would be interesting to evaluate whether or not these findings are similar in older women.
Our study presents some limitations. One limitation is that BP values are based on a single visit, therefore, not enabling us to clearly identify people with permanently elevated BP. However, it has previously been reported that multiple BP measurements added little to determining the risk already predicted by a single measurement (21,22). Another limitation is the lack of information about possible antihypertensive treatment after the visit. One could assume that patients with high levels of both SBP and DBP were more likely to receive treatment after the visit than those with high SBP and low DBP. This was especially true in the 1980s when, in the majority of cases, isolated systolic hypertension was not treated. However, possible subsequent treatment cannot entirely explain the relationship between DBP and cardiovascular mortality. However, if subsequent treatment was the reason for lower mortality rates in subjects with systolic-diastolic hypertension, similar results should also be observed in women. It is interesting to note that Franklin et al. (16)reported that taking into account postbaseline antihypertensive treatment had little or no effect on the cardiovascular risk as evaluated during the pretreatment visit.
In conclusion, in middle-aged subjects, classification of cardiovascular risk according to DBP levels should take into account gender, especially when SBP levels are elevated. Men with systolic hypertension are at higher risk when their DBP is “normal” than when they present a mild to moderate increase in DBP. In women of the same age, however, systolic-diastolic hypertension represents a higher risk than isolated systolic hypertension.
☆ Supported by the National Institute of Medical Research (INSERM) and the French National Health Care System (CNAM).
- blood pressure
- Sécurité Sociale (French national health care system)
- diastolic blood pressure
- Institut National de Statistiques et d’Etudes Economiques
- National Institute of Medical Research
- IPC Center
- Centre d’Investigations Préventives et Cliniques
- pulse pressure
- risk ratio
- systolic blood pressure
- Received March 21, 2000.
- Revision received August 9, 2000.
- Accepted September 26, 2000.
- American College of Cardiology
- ↵Kannel WB, Stokes JL. Hypertension as a cardiovascular risk factor. In: Robertson JIS, editor. Handbook of hypertension epidemiology of hypertension. Vol 6. New York/Amsterdam: Elsevier Science Publishing Co Inc, 1985:15–34.
- SAVE Investigators,
- Mitchell G.F,
- Moye L.A,
- Braunwald E,
- et al.
- Benetos A,
- Zureik M,
- Morcet J,
- et al.
- Franklin S.S,
- Khan S.A,
- Wong N.D,
- Larson M.G,
- Levy D
- MRC Working Party
- ↵Nichols WV, O’Rourke MF, eds. McDonald’s Blood Flow in Arteries: Theoretic, Experimental, and Clinical Principles. 3rd ed. London, UK: Edward Arnold, 1990:77–142, 216–269, 398–411.
- Kannel W.B,
- Dawber T.R,
- McGree D.L
- Working Group on Risk and High Blood Pressure