Elsevier

The Lancet

Volume 352, Issue 9143, 5 December 1998, Pages 1801-1807
The Lancet

Articles
Blood pressure, cholesterol, and stroke in eastern Asia

https://doi.org/10.1016/S0140-6736(98)03454-0Get rights and content

Summary

Background

Stroke is a major cause of death and disability in most populations of eastern Asia, and the incidence, particularly of haemorrhagic stroke, is generally higher than in western populations. This study aimed to assess the contributions of blood pressure and blood cholesterol concentrations to stroke risk in populations from eastern Asia.

Methods

The project included 13 cohorts from the People's Republic of China and five from Japan (124774 participants, 837 214 person-years of observation). All 18 cohorts provided data on blood pressure and 12 (69767 participants) provided data on cholesterol concentrations. Parametric and non-parametric analyses were done, with adjustments for several potential confounding factors. Analyses were based on estimated usual diastolic blood pressure and cholesterol concentration during follow-up, rather than baseline measurements, to avoid regression dilution bias.

Findings

Overall mean blood pressure was 124/78 mm hg and mean cholesterol concentration was 4·5 mmol/L. 1798 strokes occurred; 751 (42%) were classified as haemorrhagic and 707 (39%) were confirmed by computed tomography or necropsy. Each 5 mm hg lower usual diastolic blood pressure was associated with lower risk of non-haemorrhagic stroke (odds ratio 0·61 [95% CI 0·57–0·66]) and lower risk of haemorrhagic stroke (0·54-[0·50–0·58]). With decreasing cholesterol concentrations there were trends towards a decrease in risk of nonhaemorrhagic stroke (odds ratio for 0·6 mmol/L decrease, 0·77 [0·57–1·06]) and an increase in risk of haemorrhagic stroke (1·27 [0·84–1·91]). Overall, there was no clear evidence of any interaction between cholesterol and diastolic blood pressure.

Interpretation

Blood pressure is an important determinant of stroke risk in eastern Asian populations, whereas cholesterol concentration is less important, affecting the proportions of stroke subtypes more than overall stroke numbers. The association between blood pressure and stroke seems stronger than in western populations; a population-wide reduction of 3 mm hg in diastolic blood pressure should eventually decrease the number of strokes by about a third.

Introduction

For the past few decades, stroke has been a principal cause of death in the People's Republic of China, Japan, and other eastern Asian countries. The burden of stroke in this region is predicted to increase, both in absolute terms and as a proportion of total disease burden.1 Throughout much of eastern Asia, the pattern of stroke incidence is different from that in western Europe and North America—the incidence of both ischaemic and haemorrhagic strokes is higher, and a greater proportion of strokes are due to cerebral haemorrhage,2 the outcome of which is poor. The reasons for the greater burden of stroke in eastern Asian populations remain unclear, and direct reliable evidence about the determinants of stroke in this region is therefore needed. Such evidence is essential for prioritisation of regional stroke-prevention strategies.

Blood pressure is a key determinant of stroke risk in western populations, although few studies have investigated the separate associations of blood pressure with the risks of ischaemic and haemorrhagic stroke. Overviews of observational studies3, 4 have shown a strong, direct association between usual blood pressure and total stroke risk, and have provided precise estimates of the size of potential benefits of blood-pressure lowering in western populations. There are few data on the associations of blood cholesterol concentrations with stroke risks. In general, little association of total cholesterol with total stroke risk has been reported, but this could be the net result of a positive association with cerebral infarction and a negative association with cerebral haemorrhage.4 However, most individual studies have been too small to assess reliably the size and shape of any such associations, particularly at lower cholesterol concentrations such as those typically seen in populations from eastern Asia.

The Eastern Stroke and Coronary Heart Disease Collaborative Project was initiated to provide evidence about the associations of diastolic blood pressure and total cholesterol concentration with the risks of major cardiovascular disease in populations from eastern Asia. We describe here the results for stroke.

Section snippets

Identification of studies and collection of data

The minimum criteria for the inclusion of studies in this collaborative project were: a prospective, observational study of more than 500 individuals; a study population resident in eastern Asia; measurement of diastolic blood pressure at entry; and follow-up data on stroke deaths. Potentially eligible studies were identified from computerised searches, publication reference lists, conference proceedings, and discussion with researchers in the region. Collaborators provided standard tabular

Characteristics of study populations

18 cohort studies contributed data to this project (10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 and personal communications from H Horibe and the Shirakawa Study Group, X Li, and S Fang). 124 774 participants were included and 837 214 person-years were observed, representing average followup of 7 years (table 1). The mean values of diastolic blood pressure and cholesterol concentrations for the subgroups classified according to these variables are given in table 2, together with follow-up values

Discussion

We found strong, continuous associations between usual diastolic blood pressure and the risks of both haemorrhagic and non-haemorrhagic stroke in these populations from eastern Asia. Across a range of cholesterol concentrations, which correspond to the lower two-thirds of the distribution in most western populations, there were trends towards a positive association of cholesterol with non-haemorrhagic stroke and a negative association with haemorrhagic stroke. However, no value of usual

References (37)

  • RJ Carroll et al.

    Measurement error, instrumental variables and corrections for attenuation with applications to meta-analyses

    Statist Med

    (1994)
  • H Okada et al.

    A prospective study of cerebrovascular disease in Japanese rural communities, Akabane and Asahi—part 1, evaluation of risk factors in the occurrence of cerebral haemorrhage and thrombosis

    Stroke

    (1976)
  • A Tamakoshi et al.

    Serum cholesterol and cancer mortality in Japanese civil service workers: findings from a nested case-control study

    J Epidemiol

    (1993)
  • R Hashimoto et al.

    Association of hyperinsulinemia and serum free fatty acids with serum high density lipoproteincholesterol

    J Atheroscler Thromb

    (1995)
  • Y Ito et al.

    Relationship between serum carotenoid levels and cancer death rates in the residents living in a rural area of Hokkaido, Japan

    J Epidemiol

    (1997)
  • L Liu et al.

    A five year follow-up study of hypertension in 10 450 steel workers

    Chin Med J

    (1979)
  • Z Li et al.

    Preliminary analysis on mortality and risk factors in a Chinese population

    Asian Pacific Congress on Vascular Disease Prevention

    (1996)
  • B Zhou et al.

    Associations between dietary pattern, clustering of risk factors, and incidence of cardiovascular diseases in Chinese populations

    Asian Pacific Congress on Vascular Disease Prevention

    (1996)
  • Cited by (526)

    View all citing articles on Scopus

    Study organisation given at end of paper

    View full text