The prevalence of hypertension has increased over the past decade in many developed and developing countries, including China. This increase may be associated with changes in lifestyle, including dietary patterns. We evaluated the association of dietary patterns with blood pressure (BP) by using data from a large, population-based cohort study of middle-aged and elderly Chinese men, the Shanghai Men's Health Study. The present cross-sectional analysis includes 39 252 men who reported no prior history of hypertension, diabetes, CHD, or stroke nor use of antihypertensive drugs at study enrolment. Three dietary patterns, ‘vegetable’, ‘fruit and milk’ and ‘meat’, were derived using factor analysis. The fruit and milk diet was inversely associated with both systolic and diastolic BP (Ptrend < 0·001). The adjusted mean systolic BP was 2·9 mmHg lower (95 % CI − 3·4, − 2·4), and diastolic BP was 1·7 mmHg lower (95 % CI − 2·0, − 1·4) for men in the highest quintile of the ‘fruit and milk’ pattern compared with men in the lowest quintile. This inverse association was more evident among heavy drinkers; the highest quintile of the ‘fruit and milk’ pattern was associated with a 4·1 mmHg reduction in systolic BP v. a 2·0 mmHg reduction among non-drinkers (Pinteraction = 0·003) compared to the lowest quintile. The corresponding reductions in diastolic BP were 2·0  v. 1·3 mmHg (Pinteraction = 0·011). The ‘fruit and milk’ pattern was associated with a lower prevalence of both pre-hypertension and hypertension, and the associations appeared to be stronger among drinkers. Results of the present study suggest an important role for diet in the prevention of hypertension.

The purpose of the present study was to investigate the effects of Se restriction on the excretion of Se in men who had consumed high levels of this element during their entire lives. With the use of stable isotopes of Se as selenite, the excretion of methylated Se in urine was investigated in Chinese men (n 10) who had habitual chronic high intakes of this element. The relationship between either urine Se or trimethylselenonium (TMSe) to the estimated long-term Se intake was not linear over the entire range of intake, which was also true for the infusion of labelled selenite. A non-linear relationship was also found between urine TMSe and urine Se both for TMSe arising from catabolism of endogenous body Se and that from infused selenite. The data suggest a close precursor–product relationship of urine Se and its TMSe component based on the nearly identical specific activities for these two selenocompounds. Although dimethylselenide in breath was not measured in the present study, combining urinary TMSe with this breath test may be more useful in the assessment of long-term Se status.