Background: Evidence on the association between objective measures of tobacco exposure and depression remains limited. This study examined the relationship between serum cotinine and depression in a nationally representative U.S. population. Methods: Data were obtained from the National Health and Nutrition Examination Survey (NHANES) 2005-2018. Serum cotinine was analyzed as a biomarker of nicotine exposure. Depression was defined as a Patient Health Questionnaire-9 (PHQ-9) score ≥10. Weighted logistic regression and restricted cubic spline models were used to assess linear and nonlinear associations. Subgroup and sensitivity analyses were also performed. Results: A total of 29,514 participants were included. Serum cotinine concentrations were positively associated with depression. Each one-unit increase in log-transformed cotinine was associated with higher odds of depression (OR = 1.07, 95% CI: 1.06-1.09). Compared with the lowest tertile, participants in the highest tertile had increased odds of depression (OR = 1.72, 95% CI: 1.45-2.04). Restricted cubic spline analysis indicated a linear association between log-transformed serum cotinine concentrations and depression after excluding participants with serum cotinine concentrations at the lower limit of detection. The association remained robust after additional adjustment for smoking status (OR = 1.05, 95% CI: 1.03-1.07) and subgroup analyses. Conclusions: Serum cotinine concentrations were positively associated with depression in a nationally representative U.S. population. These findings suggest a potential link between nicotine exposure and depression and underscore the importance of further investigating the relationship between tobacco exposure and mental health.

Background: Ecological Momentary Assessments (EMAs), a newer method assessing alcohol use, present advantages over Timeline Follow Back (TLFB). EMAs have progressed from telephone calls to smartphone applications, providing opportunities for growth in a naturalistic assessment. This narrative review provides an overview of smartphone-based EMAs in alcohol clinical trials showing strengths/weaknesses, implications of designs, assessment durations, validity, and compliance. Methods: A literature review was conducted from March 2020 to April 2025, with no limitations on article time frame as EMAs are newer. PubMed, PsycINFO, EMBASE, and Web of Science were searched for English peer-reviewed articles. Results: 100 final articles were included. EMAs can collect real-time data, resulting in nuanced alcohol use data, often revealing that TLFB underreports alcohol consumption due to recall bias. EMA validity is influenced by design, intoxication while reporting, and increased awareness of drinking patterns. EMA compliance increases if the study design avoids participant burden, provides incentives, and accounts for participants who react to EMAs by changing their alcohol use patterns. However, this narrative review is limited to focusing on breadth due to research on a high number of different EMAs with few being investigated repeatedly. Conclusions: Additional clinical trials with the same EMAs are needed to continue to establish their validity and refine existing protocols for researchers and clinical trial specialists. EMAs offer advantages including fine-grained/real-time data on alcohol use, avoiding recall bias, and a higher external validity. EMA recommendations include providing drink options, training participants on platform usage, and designing studies based on specific aims.

ABSTACTAlcohol consumption remains a major public health challenge and a leading preventable cause of morbidity and mortality worldwide. Although the Alcohol Use Disorders Identification Test (AUDIT) is widely used to identify hazardous alcohol use, reliance on total scores may obscure clinically meaningful heterogeneity in drinking behaviors. This study aimed to identify latent alcohol consumption typologies among Colombian adults using AUDIT response patterns and to examine their sociodemographic and psychosocial correlates. Latent class analysis (LCA) was conducted using ten dichotomized AUDIT items from 14,131 adults who reported alcohol consumption during the previous 30 days in the 2019 Colombian National Survey of Psychoactive Substance Use. Competing models were evaluated according to statistical fit, parsimony, classification accuracy, and interpretability. A three-class solution was selected, comprising Low Risk (50.7%), Hazardous Binge Drinking (38.5%), and Harmful/Dependence (10.8%) profiles. Although the two higher-risk classes showed similarly high probabilities of heavy episodic drinking, the Harmful/Dependence profile was characterized by substantially greater probabilities of dependence symptoms and alcohol-related consequences. Early alcohol initiation, peer heavy drinking, and depressive symptoms were associated with more severe profiles, whereas female sex and higher socioeconomic status were protective. These findings highlight the value of person-centered approaches for identifying heterogeneous alcohol use patterns and informing targeted prevention, screening, and intervention strategies.