Critique 012: Effects on mortality of alcohol drinking in the elderly

McCaul KA, Almeida OP, Hankey GJ, Jamrozik K, Byles JE, Flicker L.   Alcohol use and mortality in older men and women.  Addicition 2010.  On-line prior to publication:  doi:10.1111/j.1360-0443.2010.02972.x

Authors’ Abstract

Aims.  To compare the effect of alcohol intake on 10-year mortality for men and women over the age of 65 years.

Design, setting and participants.  Two prospective cohorts of community-dwelling men aged 65–79 years at baseline in 1996 (n = 11 727) and women aged 70–75 years in 1996 (n = 12 432). Measurements.  Alcohol was assessed according to frequency of use (number of days alcohol was consumed per week) and quantity consumed per day.  Cox proportional hazards models were compared for men and women for all-cause and cause-specific mortality. 

Findings.  Compared with older adults who did not consume alcohol every week, the risk of all-cause mortality was reduced in men reporting up to four standard drinks per day and in women who consumed one or two drinks per day.  One or two alcohol-free days per week reduced this risk further in men, but not in women.  Similar results were observed for deaths due to cardiovascular disease.

Conclusions.  In people over the age of 65 years, alcohol intake of four standard drinks per day for men and two standard drinks per day for women was associated with lower mortality risk. For men, the risk was reduced further if accompanied with 1 or 2 alcohol-free days per week.


Forum Comments

Background:  As pointed out by the authors, the effects of alcohol consumption in the elderly may be modified by a decreased ability to metabolize alcohol, an altered volume of distribution due to reduced lean body mass and total body water, and an increased prevalence of co-morbid conditions.  These factors make this study of the net effects of drinking among a large number of community-dwelling elderly men and women in Australia especially important. 

Specifics of the present study:  The questions regarding drinking asked of men and women were slightly different, making it difficult for direct comparisons according to gender.  Among the two-thirds of men and slightly less than one-third of women who reported consuming alcohol on at least a weekly basis, 60% of men and 38% of women reported drinking on a daily basis.  A standard drink in Australia contains 10 grams of alcohol (slightly lower than the 12 grams assumed for typical drinks in the USA).

The male cohort lived in the urban area of Perth; this restricts the generalizability of the findings somewhat with respect to men.  Further, the two cohorts appear different from the population as a whole, which raises the possibility of some unspecified confounder.  There might be a “sick quitter” problem with the male reference group. 

Comments on reported results of study:  In both sexes, total mortality risk of drinkers of 1-2 drinks/day was significantly lower than for subjects reporting no alcohol consumption in a usual week.  For drinking at this level, there was little effect of the frequency of drinking on total mortality: a 20-30% lower risk for both men and women whether they reported drinking on 1-2 days/week or 7 days/week. 

Risk of mortality increased for men with heavier drinking (especially 9+ drinks/day), but for women the risk of all-cause mortality remained lower than that of non-drinkers for all frequency categories.  Co-morbidities showed the expected associations: for weekly drinkers, a tendency for less diabetes, stroke, and heart disease, and the same or greater occurrence of hypertension and bronchitis/emphysema, when compared with never drinkers or rarely drinkers.

Should the elderly have lower limits for drinking than middle-aged adults?  As the authors state, “The argument for lower limits of regular alcohol use for older people has been based largely on theoretical concern that specific amounts of alcohol result in higher blood alcohol concentration in older people, and also that regular medication use is common in older people.”  They conclude that, in their study, “These concerns do not seem to have been translated into all-cause mortality or accidental deaths.”  One Forum member noted: “The concerns expressed regarding the fragility of older individuals in general vis-a-vis drinking are mostly unsubstantiated theory and patronizing – that’s at best – or a masked portion of a prohibitionary agenda.”           

Are 1 to 2 alcohol-free days per week healthier?  The data presented in the paper do not provide support for the statement that daily drinking was less protective than consuming alcohol less frequently.  As shown below in data copied from Table 6 in the paper, within categories of daily intake the risk of mortality was largely unaffected by frequency of drinking.  For example, men consuming 9+ drinks/day had higher mortality than non-drinkers, but the hazard ratio was 1.51 for drinking on 1-2 days/week and 1.45 for daily drinking. 

Derived from Table 6:  Risks of all-cause mortality by quantity (drinks per day) and frequency (days per week) relative to no alcohol consumption in a usual week
  HR (95% CI) 
Quantity Frequency Mena Womenb
1-2 drinks 1-2 days 0.81    (0.73-0.89) 0.76    (0.65-0.90)
  3-4 days 0.71    (0.62-0.83) 0.71    (0.58-0.86)
  5-6 days 0.69    (0.55-0.86) 0.78    (0.63-0.96)
  7 days 0.78    (0.72-0.85) 0.78    (0.68-0.90)
3-4 drinks 1-2 days 0.86    (0.75-0.99) 0.84    (0.66-1.07)
  3-4 days 0.76    (0.65-0.90) 0.78    (0.60-1.00)
  5-6 days 0.74    (0.58-0.93) 0.85    (0.65-1.12)
  7 days 0.83    (0.74-0.93) 0.86    (0.70-1.04)
5-8 drinks 1-2 days 0.96    (0.81-1.15) 0.48    (0.24-0.93)
  3-4 days 0.86    (0.70-1.04) 0.44    (0.22-0.87)
  5-6 days 0.83    (0.64-1.07) 0.48    (0.24-0.97)
  7 days 0.93    (0.80-1.09) 0.49    (0.25-0.94)
9+ drinks 1-2 days 1.51    (1.09-2.08)  
  3-4 days 1.34    (0.95-1.88)  
  5-6 days 1.29    (0.88-1.89)  
  7 days 1.45    (1.06-2.00)  

a.  Adjusted for age in 5-year age groups and smoking (current smoker, former smoker, never smoked).  b. Adjusted for smoking (current smoker, former smoker, never smoked).

Further, in a figure in the paper showing cause-specific mortality, the point estimates for death from cardiovascular disease, cerebrovascular disease, and dementia were essentially identical for consumers of 1-2 drinks/week and for daily drinkers.  The risk of accidental death showed a tendency toward being higher for men and lower for women for daily drinking in comparison with less frequent drinking. 

One Forum member stated “I think I can see why the authors made the statement that their study supported 1-2 alcohol-free days per week but I suspect it is a result of not testing their results with appropriate statistics.”  Another member stated that it is surprising that the authors conclude that “for men the risk was further reduced if accompanied with 1 or 2 alcohol-free days per week.  The conclusion is not in accordance with their own data, and to my knowledge the benefit of one or several alcohol-free days per week has never been substantiated with solid evidence from any large epidemiological study.  The art of drinking is to limit your daily consumption to moderate amounts.”

Another Forum member described: “At one end of the weekly spectrum of drinking, those who drink every day derive cardiovascular benefit while, at the other, those who drink on one day a week appear at increased risk.  We can argue about the contribution of the amount drunk to cardio-protection in those who drink daily, but the unfortunate ‘binge-drinkers’ who drink a lot on one or two days a week seem to be placing themselves in harm’s way.  At what point in the weekly spectrum of consumption does the ‘binge-drinking’ effect give way to the ‘constant moderate-drinking’ effect?  The suggestion by the authors of this paper that people who drink on five or six days a week are better off than the constant imbibers is probably an artifact caused by the way in which they have analyzed their data.”

Conclusions of Forum review:  Our Forum review concludes that in terms of total mortality risk, recommendations regarding drinking among the elderly need not be overly restrictive, as this study showed lower total mortality for up to 4 drinks/day for men and 2 drinks/day for women.  Further, this study does not provide support for the authors’ conclusion that “For men, the risk was reduced further if accompanied with 1 or 2 alcohol-free days per week.”

 *                     *                      *

Contributions to these comments from the International Scientific Forum on Alcohol Research were from the following:

R. Curtis Ellison, MD, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA

Alun Evans, MD, Centre for Public Health, The Queen’s University of Belfast, Belfast, UK

Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA

Ross McCormick PhD, MSC, MBChB, Associate Dean, Faculty of Medical and Health Sciences, The University of Auckland, Auckland, New Zealand

Francesco Orlandi, MD, Dept. of Gastroenterology, Università degli Studi di Ancona. Italy

Erik Skovenborg, MD, Scandinavian Medical Alcohol Board, Practitioner, Aarhus, Denmark  

Gordon Troup, MSc, DSc, School of Physics, Monash University, Victoria, Australia