Abstract P185: Alcohol Consumption and Cardiovascular Disease Mortality in Adults With Hypertension Following the 2017 Blood Pressure Guidelines

Abstract

Introduction: Approximately half of adults have hypertension (HTN) under the 2017 blood pressure (BP) guidelines. The guidelines recommend lifestyle factors, such as limiting alcohol consumption, to manage BP, although safe levels of alcohol consumption have not been investigated for this new population with HTN. Objective: To examine the dose-response associations of alcohol consumption and cardiovascular disease (CVD) mortality in 29,499 adults (45±10 years) with HTN defined as systolic/diastolic BP ≥130/80 mmHg or physician diagnosis. Methods: Participants received a preventive medical examination during 1974-2003 and were without CVD, cancer, or abnormal electrocardiogram at baseline. Participants self-reported the number of standard alcoholic beverages they consumed per week on a medical history questionnaire. Participants were classified into 5 groups: non-drinkers and sex-stratified quartiles of weekly drinks based on no significant interaction by sex. Non-drinkers with a previous history of alcohol use problems were excluded to minimize a potential selection bias. Mortality follow-up was through 2003 using the National Death Index. Cox regression models included baseline age, sex, examination year, smoking status, body mass index, meeting the aerobic physical activity guidelines, and parental CVD. Models were further adjusted for cardiorespiratory fitness (CRF; in METs). Results: During a mean follow-up of 14.5 years, 672 CVD deaths occurred. Quartiles of alcohol consumption were 1-3, 4-7, 8-14, ≥15 drinks per week for men and 1-2, 3-4, 5-9, ≥10 drinks per week for women. Compared with non-drinkers (23%; 6,756 of 29,499), the hazard ratios (95% confidence intervals) for CVD mortality among quartiles of drinking were 0.71 (0.54-0.92), 0.73 (0.58-0.92), 0.76 (0.60-0.98), and 0.84 (0.67-1.05) after adjusting for potential confounders. After additional adjustment for CRF, the HRs were 0.75 (0.58-0.98), 0.79 (0.62-0.996), 0.82 (0.65-1.05) and 0.90 (0.72-1.13) for CVD mortality compared with non-drinkers. We found similar trends for CVD mortality after adjusting for fasting glucose, resting systolic and diastolic BP, total cholesterol, diabetes, and hyperlipidemia. We also found similar results in subgroups of men and women, young (<60 years) and old (≥60 years), and normal weight (<25 kg/m 2 ) and overweight/obese (≥25 kg/m 2 ) individuals. Similar trends were observed for all-cause mortality, although the associations were attenuated in all quartiles of drinking both before and after adjusting for CRF. Conclusions: The risk of CVD mortality was lower in light-to-moderate (quartiles 1-3) alcohol drinking in adults with HTN although these associations became weaker after adjusting for CRF. These results suggest a potential confounding effect of CRF on the association between alcohol consumption and CVD mortality that should be considered for future studies.