Potential conflict of interest: The editor of “[The Wine Review of France]” warned (06/21//2017) A.B. is a member of the “hygienist lobby” (http://www.larvf.com/vin‐cote‐rotie‐contre‐lexomil‐audrey‐bourolleau,4540951.asp).
To The Editor:
In the study by Hajifathalian et al.1 the conclusion “(our study) suggests a protective effect of modest alcohol consumption on mortality in patients with nonalcoholic fatty liver disease (NAFLD)” deserves robust comment.1
First, while the National Health and Nutrition Examination Survey (NHANES) is a prospective cohort, Hajifathalian and colleagues performed a retrospective study (post‐hoc analysis) which can be susceptible to the phenomenon of p‐hacking: PubMed has almost 11,000 articles with hypotheses from NHANES; one surprisingly failed to find an increased mortality in NAFLD.2 Had Hajifathalian et al. preregistered (e.g., Open Science Framework) the hypothesis they tested?
Second, although liver biopsy cannot be justified routinely for diagnosis of NAFLD, the reliance on the Hepatic Steatosis Index must not be overestimated, receiver operating characteristic curves being 0.68 in obese individuals, and the proliferation of predictive models (Fatty Liver Index, the ZJU index, and Framingham Steatosis Index) highlights their limitations.3
Third, abstainers had lower physical activity and more comorbidities (diabetes, cardiovascular diseases, and cancer).1 This may explain why abstainers had a worse prognosis. Abstainer biases determine the shape of the risk relationship between alcohol consumption and mortality.4 Moreover, arbitrary categorization of alcohol use and confounding variables that have no modal distribution is not correct. Furthermore, no sensitivity analysis was provided.
Fourth, is it logical to use the term NAFLD despite the continuous use of alcohol?1
Fifth, the term “modest alcohol consumption”1 and several references entitled “moderate alcohol use” contrasted with dose‐related increases in prevalence of cancers either exponential (e.g., oral cavity, pharynx cancers) or linear (e.g., esophagus, breast) beginning at 1‐ to 2‐drink/day levels. Alcohol is a human carcinogen (class 1; International Agency for Research on Cancer). However, neither citations of robust studies avoiding selection bias (e.g., Mendelian randomisation) showing “reduction of alcohol consumption, even for light to moderate drinkers, is beneficial for cardiovascular health” nor of critical assessments showing the “French paradox” have been nurtured from studies with selection biases publicized by the alcohol lobby.4
The concept of “modest alcohol consumption”1 and its hypothetical beneficial effects from a weakly designed study is a most slippery slope: (1) Alcohol advertising has skyrocketed (http://www.huffingtonpost.com/2015/03/30/99‐ad‐buys‐ofbeer‐on‐the‐wall_n_6957198.html), and alcohol use in the United States is now a “public health crisis” (http://www.washingtontimes.com/news/2017/aug/9/alcoholabuse‐in‐us‐increasing‐as‐researchers‐warn/); (2) the one‐third of Health eHeart participants believing alcohol to be heart healthy drank substantially more and cited the lay press as the belief’s origin.5
References
1. Hajifathalian K, Torabi Sagvand B, McCullough AJ. Effect of alcohol consumption on survival in nonalcoholic fatty liver disease: a national prospective cohort study. Hepatology 2019;70:511‐521.
2. Lazo M, Hernaez R, Bonekamp S, Kamel IR, Brancati FL, Guallar E, et al. Non‐alcoholic fatty liver disease and mortality among US adults: prospective cohort study. BMJ 2011;343:d6891.
3. Shen YN, Yu MX, Gao Q, Li YY, Huang JJ, Sun CM, et al. External validation of non‐invasive prediction models for identifying ultrasonography‐diagnosed fatty liver disease in a Chinese population. Medicine (Baltimore) 2017;96:e7610.
4. Holmes MV, Dale CE, Zuccolo L, Silverwood RJ, Guo Y, Ye Z, et al. Association between alcohol and cardiovascular disease: Mendelian randomisation analysis based on individual participant data. BMJ 2014;349:g4164.
5. Whitman IR, Pletcher MJ, Vittinghoff E, Imburgia KE, Maguire C, Bettencourt L, et al. Perceptions, information sources, and behavior regarding alcohol and heart health. Am J Cardiol 2015;11:642‐646.
