Heterogeneous Dose–Response Analyses of Alcohol Abuse and Dependence

• Published in: Alcoholism, clinical and experimental research clinical and experimental research, 2019-02, Vol.43 (2), p.299-308
• Main Authors: Gruenewald, Paul J., Mair, Christina
• Format: Article
• Language: English
• Subjects: Abuse; Adolescent; Adult; Alcohol abuse; Alcohol Drinking - epidemiology; Alcohol use; Alcohol Use Disorder; Alcohol withdrawal; Alcoholism - epidemiology; Alcohols; Binge drinking; Dependence; Dose-Response Relationship, Drug; Drinking; Drinking behavior; Drinking Problems; Drug abuse; Drug tolerance; DSM‐IV; Ethanol - adverse effects; Female; Heterogeneity; Humans; Male; Middle Aged; Models, Psychological; Physiological responses; Social conditions; Young Adult
• ISSN: 0145-6008; 1530-0277
• DOI: 10.1111/acer.13940

Background Survey‐based estimates of the prevalence of alcohol abuse, dependence, and disorders in the United States rely upon self‐reports of drinking patterns (e.g., binge drinking), social problems (e.g., trouble at work), physiological responses to use (e.g., tolerance), and desistance from use (e.g., withdrawal). Diagnostic criteria derived from these reports enable prevalence estimates of abuse and dependence, but moderating structural relationships among symptom groups may lead some light and moderate drinkers to appear to exhibit an alcohol use disorder (AUD). Methods A dynamic model of drinking and problems predicts that symptoms of dependence will moderate relationships between drinking measures and symptoms of abuse. Wave 1 National Epidemiologic Survey on Alcohol and Related Conditions data on DSM‐IV diagnoses of abuse and dependence were used to test predictions from this model and assess whether moderating effects were observed among lighter and heavier drinkers (those who drink 1 to 3 vs. 4 or more drinks on average). A dose–response model that accounts for other known sources of risk heterogeneity related to drinking and problems enabled us to test these predictions. Results As expected from previous work, symptoms of abuse and dependence and dependence criteria were nonlinearly related to drinking patterns; more symptom reports appeared and criteria were met among less frequent drinkers who drank more on each occasion and this pattern of dose–response was substantially moderated among heavier drinkers. Controlling for these effects, relationships between drinking and symptoms of abuse were moderated among respondents who met more dependence criteria. These effects were observed among both lighter and heavier drinkers. Conclusions Moderating relationships observed between measures of drinking, abuse, and dependence criteria among lighter and heavier drinkers suggest that the same etiologic forces are at play among all drinking groups. Greater symptoms of dependence among lighter drinkers may lead to greater reports of symptoms of abuse and an AUD diagnosis. We provide evidence of reduced negative feedback control over drinking across groups of drinkers who meet more dependence criteria using a dose‐response model that accounts for other sources of risk heterogeneity. More symptoms of abuse and dependence appeared among less frequent drinkers who drank more on each occasion, with this pattern substantially moderated among heavier drinkers. T