|
THE RENAUD SOCIETY MEMBER BULLETIN BOARD
|
|
| Subject: A Controversial New Paper (reviewed) |
Author: R. Curtis Ellison, MD
Date: 12/12/2008 4:07 pm MST
|
Review of Recent Publication December 12, 2008
Article: Britton A, Marmot MG, Shipley M.
Who benefits most from the cardioprotective properties of alcohol consumption¡ªhealth freaks or couch potatoes?
J Epidemiol Community Health 2008;62:905¨C908. (doi:10.1136/jech.2007.072173).
Summary: The authors state that the cardioprotective properties of moderate alcohol consumption, compared with abstinence or heavy drinking, are widely reported, but whether the benefits are experienced equally by all moderate drinkers is less well known. Their aim was to examine the association between average alcohol intake per week and the incidence of fatal and non-fatal myocardial infarction during 17 years of followup for 9655 men and women without prevalent disease in the general population; and to test whether the level of cardioprotection differs according to subjects¡¯ other health behaviours (healthy, moderately healthy, unhealthy) at entry to the study. They carried out longitudinal evaluation of a British civil service-based cohort study, with baseline in 1985¨C8.
A significant benefit of moderate drinking compared with abstinence or heavy drinking was found among those with poor health behaviours (little exercise, poor diet and smokers). No additional benefit from alcohol was found among those with the healthiest behaviour -profile (>3 hours of vigorous exercise per week, daily fruit -or vegetable consumption and non-smokers). The authors conclude that the cardioprotective benefit from moderate drinking does not apply equally to all drinkers, and this variability should be emphasised in public health messages.
Comments: Other than a rather unusual description of subjects ("health freaks or couch potatoes") in the title, there are a number of questions about this paper. While the investigators have collected repeated data from multiple questionnaires and examinations in this prospective study, they only used data collected at the baseline examination of the study, although follow up extended for a median of 17 years. The effects of potential changes in habits over this long period of follow up were thus not evaluated and could have led to misclassification of exposure that may have reduced the effects among healthier subjects. Further, given that the majority of subjects were relatively young, and that coronary heart disease (CHD) occurs primarily in older people, the number of healthier subjects may have been inadequate to test for an effect.
The authors do not present detailed data on each of the potential modifiers of alcohol intake that they included in their analyses: smoking was classified only as non, current, or ex-smoker at baseline; diet was based only on daily consumption of any fruits and vegetables; and for exercise, the "healthy" group was classified as ¡Ý 3 hours/week of moderate or vigorous exercise. They only classify subjects as having all 3, 2, or 0 - 1 such habits. Further, they did not take into account other major risk factors (such as obesity, diabetes, hyperlipidemia, or hypertension).
There were 407 subjects in the "healthiest" group who did not consume alcohol, of whom only 8 had a cardiac event. It is not possible from the data presented to know what percentage of these were women, what their social status was, or what the other characteristics of these
subjects were. Since one of the key inferences of the paper is that "healthy" people do not have
lower rates of cardiac disease associated with alcohol, it would be useful to have further description of these subjects, especially their subsequent habits over the long follow-up period. (This could be important, as several follow-up studies have shown that subjects who stop
smoking or begin to drink in mid-life show considerably lower rates of subsequent cardiovascular disease.) Further, since the subjects were in general quite young and the occurrence of CHD prior to age 60 is uncommon, inadequate power is a possible reason why they did not see any effect in the healthiest groups; stratified analysis by age groups would have been very informative. (It is interesting that the authors focus on the subjects with the healthiest lifestyle who were non-drinkers, saying that their risk of CHD was lower than that of moderate drinkers, but did not comment on the fact that those in the same group who were heavier drinkers also had a lower risk of CHD.)
In the original report on alcohol and coronary disease from the Framingham Study in 1974, the percentage reduction in CHD risk from moderate alcohol intake was the same for heavy smokers as it was for non-smokers. Furthermore, most studies of alcohol and risk factors have shown that, in comparison with non-drinkers, people at both the lowest and the highest risk of CHD tend to show a similar reduction in relative risk from moderate drinking. The present study suggests a different effect according to other health behaviors.
A recent previous paper on this topic by Mukamal et al1 concluded that even in the presence of other components of a healthy lifestyle (not smoking, not obest, exercising, eating a healthy diet), moderate alcohol intake was associated with a significantly lower risk of myocardial infarction. That previous study used updated exposure information rather than limiting their analyses to baseline data; they used dietary records to judge quality of diet and had more detailed data on smoking history. And, unlike the present paper by Britton et al, that study showed an inverse association with alcohol for all subjects, whether they had 0-1, 2-3, or 4 other "healthy lifestyle factors, ".
For subjects with all 4 healthy characteristics (non-smokers, lean, good diet, getting exercise), the consumption of 15-29.9 g/day of alcohol was associated with 69% further reduction in risk of a MI. Mukamal et al conclude: "In this prospective analysis of men with favorable levels of 4 healthy lifestyle factors, moderate alcohol consumption was associated with
a lower risk for MI, with the lowest risk in men who drank consumed 5 to 30 g of alcohol per day (approximately one half to 2 drinks)."
In the discussion of the present paper, however, the authors seem to misrepresent the findings of the Mukamal paper. Britton et al state: "Mukamal and colleagues found that, among 8867 healthy men, free from major disease and with healthy lifestyle behaviours, followed for 16 years, those consuming alcohol in moderate amounts did not have a statistically significant lower
risk of myocardial infarction than abstainers." In the paper by Mukamal et al, subjects who had all four other healthy factors, the risk ratios were 0.62 (95% CI = 0.43, 0.90) for those reporting 5.0 - 14.9 g of alcohol per day and 0.54 (95% CI = 0.34, 0.86) for those reporting 15.0 - 29.9 g/day, in comparison with non-drinkers.
The Britton et al paper concludes: "The cardioprotective benefit from moderate drinking does not apply equally to all drinkers, and this variability should be emphasised in public health messages." It has been well demonstrated that individuals vary markedly in their response to alcohol, from genetic and behavioral factors as well as other environmental influences. The problems with the analysis in the present study make us question the definitive statements of the authors that healthier subjects do not show a reduction in their risk from drinking. Their findings are not supported by most other studies.
Lay Summary: This observational study from the UK suggests that people with other aspects of a healthy lifestyle (non-smoking, regular exercise, good diet) do not show protection against coronary heart disease (CHD) from alcohol. There are a number of problems with the paper that make us question that conclusion. The main weaknesses are that the investigators had a rather limited and somewhat unusual definition of a ¡°healthy lifestyle,¡± did not evaluate changes in alcohol consumption over the 17 year follow-up period, probably had an inadequate sample size (only 8 events in the healthiest group) to make conclusions among these relatively young subjects, and misrepresented previous research on this topic. While undoubtedly people vary in the degree of protection against CHD from alcohol, the data presented in this paper should not be used as a basis for public health announcements saying that only ¡°unhealthy¡± people may benefit from moderate alcohol consumption.
R. Curtis Ellison, MD
Yuqing Zhang, MD, DSc
Luc Djouss¨¦, MD, DSc
Institute on Lifestyle & Health
Boston University School of Medicine
1 Mukamal KJ, Chiuve SE, Rimm EB. Alcohol consumption and risk for coronary heart disease in men with healthy lifestyles. Arch Intern Med 2006;166:2145-2150
|
|
| Topics |
Author |
Date |
|
|
|
|
R. Curtis Ellison, MD |
11/23/2008 7:29 am MST |
|
|
|
|
R. Curtis Ellison, MD |
11/23/2008 7:30 am MST |
|
|
|
|
R. Curtis Ellison, MD |
11/23/2008 7:32 am MST |
|
|
|
   |
A Controversial New Paper (reviewed) |
|
|
R. Curtis Ellison, MD |
12/12/2008 4:07 pm MST |
|
|
|
|
R. Curtis Ellison, MD |
1/11/2009 7:27 am MST |
|
|
|
|
R. Curtis Ellison, MD |
1/11/2009 7:13 am MST |
|
|
|
|
