Assessing Quitter Effects: Disaggregating Current Nondrinkers
Bibliographic record
Abstract
To the Editor: The recent article by Lee et al.1 provides an interesting analysis using more than 12,000 respondents to the U.S. Health and Retirement Study (HRS) aged 55 and older interviewed in 2002. The study focuses on alcohol consumption and all-cause mortality while taking into account a number of risk factors that may confound this relationship. However, we have substantial reservations about their management of a key variable, namely, long-term drinking status. They state: “Subjects who reported no alcohol intake in 2002 but reported alcohol use in 2000 or 1998 were excluded (n=2,323) to minimize the chances of classifying subjects who stopped drinking because of worsening health as nondrinkers.” In our view, this is far from adequate in that it still leaves people in the nominal “nondrinker” category who drank some years and not others or who drank before 1998 and not more recently. The authors' exclusion of “recent quitters,” as noted in their description of the sensitivity analysis, does not adequately address this issue. Persons who were heavy drinkers for a number of years, quit some years before, and were abstainers in 2002 may still be at higher risk of long-term chronic complications from alcohol and of early mortality. Including them with nondrinkers confounds the interpretation. For example, in heavy drinkers who stop drinking, the risk of alcohol-related cancer can take up to 20 years to decrease to that of an abstainer.2 Overall, the mortality risk of ex-drinkers based on a comprehensive systematic meta-analysis of data sets mainly stemming from the United States was estimated to be RR1.44 for women (95% confidence interval (CI)=1.28–1.61) and RR1.21 for men (95% CI=1.10–1.32).3 A more-refined approach would have been to include all persons in the analysis with a disaggregation of the “current nondrinker” category into three groups, such as lifetime abstainer, current abstainer—former heavy drinker, and current abstainer—former moderate drinker, and to include these three groups in the analysis in Tables 1 and 2. These tables show that the nondrinker category had 5,672 subjects, the largest of the six categories under “number of alcoholic drinks consumed”; therefore, a breakdown into three groups would not complicate the analysis. The main problem with analyses of older populations that do not adequately control for sick-quitter effects is that they may lead to overall false conclusions and recommendations. Searching for the best control group in alcohol epidemiology is cumbersome,4,5 but without an adequate control, such recommendations are not valid. Most population-level studies fail to find beneficial health effects of moderate consumption,6 and this also needs to be taken into consideration when making recommendations. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this letter. Financial Disclosure: Funded by the Centre for Addiction and Mental Health, Toronto, Ontario, Canada. Author Contributions: Both authors contributed substantially to this letter. J. Rehm noted some of the challenges in the paper by Lee et al. that this letter focuses on. N. Giesbrecht developed a draft. J. Rehm revised the letter. Sponsor's Role: None.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.036 | 0.184 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.003 | 0.001 |
| Research integrity | 0.006 | 0.004 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".