Outdoor temperature, blood pressure, and cardiovascular disease mortality among 23 000 individuals with diagnosed cardiovascular diseases from China

Abstract

Introduction: Blood pressure is a major cause of cardiovascular disease (CVD) and both may increase as outdoor temperatures fall. However, there are still limited data about seasonal variation in blood pressure and CVD mortality among patients with prior-CVD.

Methods: We analysed data on 23 000 individuals with prior-CVD who were recruited from 10 diverse regions into the China Kadoorie Biobank during 2004-8. After 7 years of follow-up, 1484 CVD deaths were recorded. Baseline survey data were used to assess seasonal variation in systolic blood pressure (SBP) and its association with outdoor temperature. Cox regression was used to examine the association of usual SBP with subsequent CVD mortality, and seasonal variation in CVD mortality was assessed by Poisson regression. All analyses were adjusted for age, sex, and region.

Results: Mean SBP was significantly higher in winter than in summer (145 vs. 136 mmHg, P < 0.001), especially among those without central heating. Above 5°C, each 10°C lower outdoor temperature was associated with 6.2 mmHg higher SBP. Systolic blood pressure predicted subsequent CVD mortality, with each 10 mmHg higher usual SBP associated with 21% (95% confidence interval: 16-27%) increased risk. Cardiovascular disease mortality varied by season, with 41% (21-63%) higher risk in winter compared with summer.

Conclusion: Among adult Chinese with prior-CVD, there is both increased blood pressure and CVD mortality in winter. Careful monitoring and more aggressive blood pressure lowering treatment in the cold months are needed to help reduce the winter excess CVD mortality in high-risk individuals.

Keywords: Blood pressure; Cardiovascular disease; China; Cohort study; Hypertension; Seasonal variation.

Figure 1

Locations of the recruitment centres in China Kadoorie Biobank. The solid circles denote urban areas, and the open circles denote rural areas.

Figure 2

Monthly variation in blood pressure and outdoor temperature in people with prior-cardiovascular disease in (A) men and (B) women. The horizontal placement of each month combined over the full 4 years of recruitment represents the mean number of days since the first participant was recruited for participants recruited in average over years for that month. J/F = January and February combined (recruitment dropped in January and February due to the Chinese New Year). For both blood pressure and temperature, the mean monthly values are the mean for all participants whose baseline survey happened during that month (regardless of the year). Means of blood pressure were adjusted for sex and age. The winter months are placed centrally to display the winter peak in blood pressure.

Figure 3

Monthly variation in systolic blood pressure and outdoor temperature among people with prior-cardiovascular disease, Other regions together vs. Harbin. Conventions as in Figure 1.

Figure 4

Seasonal variation in systolic blood pressure and outdoor temperature among people with prior-cardiovascular disease by various subgroups, with the temperature range at least >5°C. The analysis was adjusted for age, sex, and area (where appropriate). Each closed square represents a change in systolic blood pressure per 10°C lower outdoor temperatures. The dotted vertical line indicates the overall change in systolic blood pressure; the open diamond indicates it and its 95% CI.

Figure 5

Hazard ratios for cardiovascular disease mortality vs. usual blood pressure among people with prior-cardiovascular disease. Analyses were stratified by region, age, gender, and blood pressure lowering treatment status, and adjusted for education, smoking, alcohol drinking, and body mass index. The hazard ratios are plotted on a floating absolute scale. Each square has an area inversely proportional to the standard error of the log risk. Vertical lines indicate the corresponding 95% confidence intervals. Numbers above confidence intervals are of hazard ratios and those below are the numbers of cardiovascular disease deaths.

Figure 6

Seasonal variation in cardiovascular disease mortality rates among people with prior-cardiovascular disease, between 2004 and 2013. Deaths and person-days at risk for a given months are totalled across the follow-up period. The analysis is adjusted for age group, study site, and sex, but not for year of follow-up. To make the curve smoother, two calendar months are combined and winter months are again placed centrally, as in Figure 1. Vertical lines indicate the corresponding 95% confidence intervals. Numbers above confidence intervals are of mortality rates (per 1000 person-years) and those below are the numbers of deaths.