Alcohol increases the risk of most chronic diseases with the exception of a reported cardio-protective role from moderate alcohol intake,1 although this has been debated.2 The data that led to the familiar ‘J-shaped’ curve for heart disease were first derived from cohorts recruited in the period 1940–1980.3,4,5 We are concerned that the dramatic rise in overweight and obesity among adults since the late 1980s may change the assumptions about the protective effects of moderate alcohol intake in the general population.
Using data from the Framingham Heart Study we separated more adipose adults from less adipose adults using Body Mass Index (BMI >27.5 kg/m2 versus BMI <25 kg/m2) and waist circumference (WC >102 cm for men and >88 cm for women, versus <94cm for men and <80 cm for women). We eliminated records indicating pre-existing heart disease. Death due to coronary heart disease was noted over the following 50 years. Study details can be found at http://www.framinghamheartstudy.org.
Figure 1 shows the relative risks for CHD death for different levels of alcohol intake compared with abstinence, for the 2,603 adults with available BMI data. The smoothed curves (polynomial second order) indicate the expected dip in relative risk for those adults with lower BMIs, at moderate levels of alcohol intake. However, the dip appears to be reduced for higher BMI adults, with lowest risk at intake levels between 10 mL and 30 mL alcohol per week (about one large glass of wine or one pint of lager). Point estimates show a significant difference between the BMI groups at levels of alcohol intake of 80–120 mL per week, and suggest no health benefit for overweight adults from drinking more than three glasses of wine or pints of lager per week. For waist circumference, we found the expected dip at low levels of intake for adults with low WC, but the dip was entirely absent for adults with high WCs (figure not shown).
These preliminary findings need to be adjusted for possible confounding factors, and need to be validated using other longitudinal cohorts. However, they indicate that health authorities and the alcohol industry may need to review health messages about ‘safe’ levels of consumption, and in particular may need to ensure that overweight people (currently the majority of middle-aged and older adults in Australia, the UK and the US) are aware that any cardio-protective benefit from moderate alcohol intake is unlikely to apply to them.