Re: Alcohol consumption and mortality: modelling risks for men and women at different ages
In their recently published paper, White, Altmann and Nanchahal, BMJ
2002 (1) reported the existence of a direct dose-response relationship
between alcohol consumption and risk of death in women aged 16-54 and in
men aged 16-34, whereas at older ages the relation is U shaped.
However a significant methodological issue related to the assessment
of lifetime alcohol consumption that may affect the conclusions was not
discussed in the paper.
The non-drinkers group could be mainly composed of ex-heavy drinkers (2)
(3), who became abstainers. And this is more likely to be true with the
"old ages" group, where due to alcohol related health issues ex-drinkers
may become "non-drinkers" (3).
If the data used allows it, the "non-drinking" group should be separated
into "ex-heavy drinkers", "ex-moderated drinkers", and "lifetime
abstainers" or any other classification assessing former drinking
patterns.
If the data does not allow assessing former drinking behavior in the "non-
drinking" group, the authors should discuss this as a limitation of they
study.
1. White IR, Altmann DR, Nanchahal K. Alcohol consumption and
mortality: modelling risks for men and women at different ages. British
Medical Journal 2002: 325: 191.
2. Shaper AG. Alcohol and mortality: a review of prospective studies.
Br J Addict 1990;85:837-47.
3. Sarr M. Re: "Alcohol consumption and coronary heart disease
morbidity and mortality." (Letter). Am J Epidemiol 1999;149:682-3.
Rapid Response:
Re: Alcohol consumption and mortality: modelling risks for men and women at different ages
In their recently published paper, White, Altmann and Nanchahal, BMJ
2002 (1) reported the existence of a direct dose-response relationship
between alcohol consumption and risk of death in women aged 16-54 and in
men aged 16-34, whereas at older ages the relation is U shaped.
However a significant methodological issue related to the assessment
of lifetime alcohol consumption that may affect the conclusions was not
discussed in the paper.
The non-drinkers group could be mainly composed of ex-heavy drinkers (2)
(3), who became abstainers. And this is more likely to be true with the
"old ages" group, where due to alcohol related health issues ex-drinkers
may become "non-drinkers" (3).
If the data used allows it, the "non-drinking" group should be separated
into "ex-heavy drinkers", "ex-moderated drinkers", and "lifetime
abstainers" or any other classification assessing former drinking
patterns.
If the data does not allow assessing former drinking behavior in the "non-
drinking" group, the authors should discuss this as a limitation of they
study.
1. White IR, Altmann DR, Nanchahal K. Alcohol consumption and
mortality: modelling risks for men and women at different ages. British
Medical Journal 2002: 325: 191.
2. Shaper AG. Alcohol and mortality: a review of prospective studies.
Br J Addict 1990;85:837-47.
3. Sarr M. Re: "Alcohol consumption and coronary heart disease
morbidity and mortality." (Letter). Am J Epidemiol 1999;149:682-3.
Competing interests: No competing interests