A prospective study of chronic disease and risk factors in an urban Chinese population

The relationships of serum cholesterol, blood pressure and cigarette smoking with certain chronic diseases were investigated in a prospective study among more than 9,000 middle-aged adults in urban Shanghai. At baseline, the mean serum cholesterol was 4.2 mmol/l, 14 per cent of the participants had definite hypertension, and 61 per cent of males and 7 per cent of females were regular smokers. During 8-13 years of follow-up, 620 deaths were recorded. 231 (37%) of the deaths were ascribed to cardiovascular disease, including 44 (7%) from CHD and 152 (25%) from stroke. Cancer caused 274 deaths (44%), of which 66 deaths (11%) were from lung cancer, 63 (10%) from stomach cancer and 54 deaths (9%) from liver cancer. Other causes accounted for 115 deaths (19%), 29 (5%) of which were from chronic liver disease, and 31 (5%) from chronic obstructive pulmonary disease. In this study, there was a strong positive and apparently independent relationship of serum cholesterol level to CHD death (z=3.47, 2P<0.001). Within the range of usual serum cholesterol studied (about 3.8-4.7 mmol/l), there was no evidence of any apparent "threshold". After appropriate adjustment for the "regression dilution" bias, a 4% difference in usual cholesterol was associated with a 21% (95% confidence interval 9-35%) difference in the risk of CHD death. There was no significant relationship of serum cholesterol with total stroke mortality, or with total cancer mortality. The 79 deaths due to liver cancer or other chronic liver diseases were inversely related to cholesterol concentration at baseline. This inverse association appears to be secondary to prolonged hepatitis B virus infection, which accounts for most of the deaths from liver disease in China and which chronically lowers blood cholesterol. There was a strong positive relationship between blood pressure and risk of death from stroke and CHD. Within the range of usual blood pressure studied (SBP: 117-161 mmHg; DBP:75-101 mmHg), there was no evidence of any apparent threshold. After appropriate adjustment for the "regression dilution" bias, a 10 mmHg difference in usual SBP was associated with a 67% (95% Cl 52-83%) difference in the risk of stroke deaths, and with a 44% (95% confidence interval 21- 73%) difference in the risk of CHD death; a 7 mmHg difference in usual DBP was associated with a 124% (95% Cl 96-155%) difference in the risk of stroke deaths, and with a 58% (95% Cl 22-105%) difference in the risk of CHD deaths. Cigarette smoking was significantly associated with deaths from any disease. There was a strong positive relationship between cigarette smoking and risk of all cancer deaths, and specifically cancer of the lung and cancer of the upper aerodigestive tract. The relative risk of lung cancer for a current smoker was 3.5 (95% Cl 1.8-7.0; 2P<0.001), and among the male population 63% of lung cancers were directly attributed to the smoking. The relative risk of upper aero-digestive cancer death for regular smokers was 3.4 (95% Cl 1.1-10.5; 2P<0.05). The risk of chronic obstructive lung disease was also significantly related to smoking, with a relative risk in a smoker of 2.2 (95% Cl 1.1-4.4; 2P<0.05). In the present population, smokers had a 60% excess risk of deaths from total stroke compared with nonsmokers (z=2.40, 2P<0.05).