Light-to-moderate drinking and dementia risk: The former drinkers problem re-visited

1. We thank E.J. Neafsey, who kindly provided a list of the remaining five citations (email from Neafsey to Roizen, October 21, 2011).

2. Anttila et al. wrote: “Since some studies, as well as our preliminary analyses, indicated that there was a J or U shaped relation between alcohol drinking and mild cognitive impairment and dementia, we used infrequent drinkers as our reference group” (p. 2 of 6, reprint).

3. See Orgogozo et al., Table IV, p. 188.

4. Ruitenberg et al. also added 28 participants who reported drinking less frequently than twice monthly into their nondrinker category (see Table 1, p. 283); their paper employed the expanded “nondrinker” category as reference (see p. 282).

5. Personal communication (email from Au Yeung Shiu Lun Ryan, October 29, 2011).

6. According to Virginia G. Wadley (personal communication, email January 1, 2011), the categorization of alcohol consumption was based on two questions: “1. Do you presently drink alcoholic beverages, including beer, wine, and other drinks made with hard liquor, even occasionally? Yes – current drinker[;] If no – query for drinking history.” “2. Have you ever drunk alcoholic beverages, including beer, wine, and other drinks made with hard liquor, on a regular basis? By regular, we mean at least 1 drink per month for 1 year. Yes – past drinker[;] No – never drinker.”

7. For example, as offered in Table 3 (p. 244).

8. Personal communication (email from Annie Britton, October 17, 2011).

9. Kivipelto et al. (Citation2008) described their alcohol measure as follows: “A yearly alcohol drinking habit was determined in the 1972 and 1977 questionnaires only, and in this study the analyses including alcohol data were based on the subsample of the persons who participated in 1972 or 1977…. Alcohol drinking was considered as the times person drank alcohol per year. For the current study alcohol drinking was classified into three categories as in our previous study [5]: (i) never drinkers (= persons who never drank alcohol, n = 276), (ii) infrequent drinkers (= persons who drank alcohol less than once a month, n = 391) and (iii) frequent drinkers (= persons who drank alcohol once a month or more often, n = 278). The never drinkers served as the reference category in the analyses concerning alcohol drinking.” (p. 2763)

10. See questions 156-158 at http://www.cfas.ac.uk/pages/bquestionnaires/index.html (accessed on Nov. 18, 2011). We thank Fiona Matthews for directing us to this page (personal communication, email Matthews to Roizen, November 17, 2011).

11. Orgogozo et al. described their drinking measure as follows: “A structured questionnaire was administered individually, starting with average daily water intake to reduce the emotional reaction often associated with questions about alcohol intake (Midanik, 1989). Wine, beer and liquor drinking were recorded separately. Subjects were then classified as: non drinkers if they did not drink any alcoholic beverage or no more than one drink per week; mild drinkers if they drank at least 2 drinks per week but not more than 250 ml (2 standard glasses) per day; moderate drinkers if they drank between 250 and 500 ml of alcoholic beverage (3 to 4 standard glasses) per day; heavy drinkers if they drank more than 500 ml (5 standard glasses and up) per day.” (p. 186)

12. See also Lemeshow et al.'s (Citation1998) description of the average daily approach employed in the French PAQUID study.

13. Presumably “≥3 drinks” was intended.

14. In Ruitenberg et al.: “To check whether any observed relation between alcohol intake and risk of dementia was due to selection bias in the reference group (no alcohol intake), we repeated the analyses excluding those who reported use of medication in which alcohol intake was contraindicated (anxiolytics, antidepressants, and hypnotics). Additionally, we excluded participants with a history of alcoholism (n = 4) and those with alcohol consumption less than twice a month (n = 28). Cardiovascular disease might make people change their alcohol intake, and hence we also repeated the analyses excluding those with diagnosed myocardial infarction or stroke at baseline (n = 715) and those who used antihypertensive medication at baseline (n = 1646).” (p. 283)

15. Ruitenberg et al. described their series of drinking questions as follows: “First, we asked participants whether they ever drank alcohol. If the answer was affirmative, we asked about the frequency of drinking. People who reported that they drank alcohol at least twice a month were further asked about the average amounts of specific beverages (wine, beer, liquor, and fortified wine [eg, sherry, port]) that they drank. Participants were furthermore asked if they had changed their pattern of alcohol consumption during the preceding 5 years (less than they used to drink, more than they used to drink) and if they had consumed more than six alcoholic beverages on one day during the last year (binge drinking)” (p. 282). The “less than they used to drink” and “more than they used to drink” responses hint that this follow-up question was asked of current drinkers only. Hence the “changed in their drinking in the past five years” may not have been asked of nondrinkers.

16. Ruitenberg et al. explained: “The results essentially did not change with iterative exclusion of participants with changed drinking patterns, those with a history of alcoholism (0.1%), those who drank less than twice a month (0.5%), or those who were binge drinkers (6.2%). The hazard ratio for light-to-moderate drinking after exclusion of participants with changed drinking patterns was 0.57 (95% CI 0.36–0.89), and after exclusion of binge drinkers was 0.56 (0.36–0.88). The hazard ratio did not change much after exclusion of people with prevalent stroke or myocardial infarction at baseline (hazard ratio light-to-moderate drinking 0.55 [0.33–0.92]), but slightly decreased after exclusion of participants who used antihypertensive medication (0.48 [0.27–0.86]). Use of medications that contraindicated alcohol consumption (anxiolytics, hypnotics, and antidepressants) was reported by 691 (13%) participants. Again, exclusion of these individuals did not alter the results appreciably (0.60 [0.37–0.98]).” (pp. 284–285)

17. Both Neafsey and Collins’ and Peters et al.'s reviews cited Järvenpää et al. as addressing the problem.

18. Cited by Anstey et al. as testing for the impact of former drinkers on overall results.

19. According to Shin et al.: “Past heavy drinkers were defined on the following basis (Rosen et al., Citation1993): (i) a minimum of three months before the assessment during which no more than two drinks per week was consumed; (ii) at least one decade of drinking a minimum of 42 drinks per week; and (iii) a lifetime alcohol consumption of 30,000 drinks or over.” (p. 1076)

20. Cited by Anstey at al. as finding no difference in cognitive measures between abstainers and former drinkers at baseline.

21. Neafsey and Collins wrote: “Similarly, when Ganguli et al compared the minimal drinking group with life-long nondrinkers (excluding quitters), the lower odds of decline on the MMSE of 0.05 (95% CI: 0.01–0.26) remained significant; this also was seen for the ratio for the moderate drinking group for decline on the MMSE of 0.27 (95% CI: 0.09–0.84).” (p. 472)

22. Zuccala et al. wrote: “In this study, alcohol consumption was associated with a lower probability of cognitive impairment also after adjusting for comorbidity; thus, that the presence of ‘sick quitters’ among abstainers is an unlikely explanation for this association. Noticeably, the protective effects of alcohol on the risk of cardiovascular and cerebrovascular disease have been confirmed even in studies that excluded subjects who abstained from alcohol because of poor health (Dufouil, Ducimetiere, & Alperovitch, Citation1997; Marmot and Brunner, Citation1991).” (p. 1746)

23. The coding employed was clarified via a personal communication (email from Kirsten Mehlig, November 10, 2011).

24. Clinton Wright clarified that an initial question to respondents asked if they “had every taken alcohol in their life and the amount imbibed” (personal communication, email from Wright to Roizen, January 3, 2012); a negative response to this question, we infer, supplied this analysis with its “never” drinkers.