Daily loneliness and alcohol use in Alcohol Use Disorder: an 8-week EMA study of within-person associations
Abstract
Background and Aims: Loneliness has been suggested as a risk factor for alcohol consumption (AC), potentially contributing to a cycle whereby loneliness increases drinking, which in turn exacerbates loneliness through alcohol-related consequences. However, longitudinal evidence is inconsistent, and little is known about short-term, within-person associations in individuals with alcohol use disorder (AUD). We examined concurrent and lagged bidirectional associations between loneliness and AC in daily life and whether these associations varied according to social contact, mood, age, and gender. Design: Ecological momentary assessment (EMA) study with daily assessments over 8 weeks. Within-person associations were examined using fixed-effects panel regression models; multilevel models examined moderation by age and gender. Setting: A real-life context with data collection between December 2025 and June 2026 in Germany. Participants: The final analytic sample comprised 58 adults meeting at least two DSM-5 criteria for AUD (74% men; mean age = 48.8 years, SD = 8.3). Intervention: Observational study without interventions. Measurements: Daily alcohol use (yes/no) and quantity (grams), loneliness, social contact, perceived social support, and mood valence and calmness were assessed using smartphone-based EMA. Concurrent and prospective associations across consecutive days were examined, accounting for autoregressive stability. Findings: EMA compliance was high (96.7%). Greater-than-usual loneliness was associated with lower probability of alcohol use during the same retrospective assessment interval (b = −0.092, 95% CI [−0.151, −0.033], pFDR = .006) and lower alcohol quantity (b = −15.89 g, 95% CI [−24.16, −7.61], pFDR = .003). Prospective associations in either direction were small, with confidence intervals constraining their plausible magnitude. The concurrent association between loneliness and alcohol quantity was stronger on days with social contact, higher positive mood valence and calmness, and among men. Mood also moderated the prospective association between loneliness and subsequent alcohol use, although simple slopes were not significant. Conclusions: In adults with AUD, loneliness was associated with lower AC concurrently whereas prospective associations in either direction were small. The concurrent association varied by social and mood context and gender, suggesting that loneliness alone may provide limited information about imminent drinking risk.
Related
- Loneliness and alcohol use in alcohol use disorder: A longitudinal within-person analysis
- From Daily Attitudes to Alcohol‐Related Harm: A Within‐Person Serial Mediation Test Informed by the Theory of Planned Behavior
- Validation of within-person associations with coverage and momentary EMA self-reports using an objective standard (physical activity)
- Associations of digital self-efficacy and social network connectedness with loneliness among urban older adults in kerman, iran: a cross-sectional study
- Core schema instability in youth at clinical high risk for psychosis: within-person variability as an index of self-belief precision
BibTeX
@misc{kraplin2026daily,
title = {{Daily loneliness and alcohol use in Alcohol Use Disorder: an 8-week EMA study of within-person associations}},
author = {Anja Kräplin and Fabian Arntz and Gianna Spitta and Jan Luca Wiedenhöft and Noemi Keng and Yoonha Park and Katharina Hoffmann and Shuyan Liu and Louisa Diallo and Michael N. Smolka},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/46ty8_v1},
url = {https://osf.io/preprints/psyarxiv/46ty8_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Daily loneliness and alcohol use in Alcohol Use Disorder: an 8-week EMA study of within-person associations
AU - Anja Kräplin
AU - Fabian Arntz
AU - Gianna Spitta
AU - Jan Luca Wiedenhöft
AU - Noemi Keng
AU - Yoonha Park
AU - Katharina Hoffmann
AU - Shuyan Liu
AU - Louisa Diallo
AU - Michael N. Smolka
PY - 2026
DA - 2026/09/29
PB - PsyArXiv
DO - 10.31234/osf.io/46ty8_v1
UR - https://osf.io/preprints/psyarxiv/46ty8_v1/
AB - Background and Aims: Loneliness has been suggested as a risk factor for alcohol consumption (AC), potentially contributing to a cycle whereby loneliness increases drinking, which in turn exacerbates loneliness through alcohol-related consequences. However, longitudinal evidence is inconsistent, and little is known about short-term, within-person associations in individuals with alcohol use disorder (AUD). We examined concurrent and lagged bidirectional associations between loneliness and AC in daily life and whether these associations varied according to social contact, mood, age, and gender. Design: Ecological momentary assessment (EMA) study with daily assessments over 8 weeks. Within-person associations were examined using fixed-effects panel regression models; multilevel models examined moderation by age and gender. Setting: A real-life context with data collection between December 2025 and June 2026 in Germany. Participants: The final analytic sample comprised 58 adults meeting at least two DSM-5 criteria for AUD (74% men; mean age = 48.8 years, SD = 8.3). Intervention: Observational study without interventions. Measurements: Daily alcohol use (yes/no) and quantity (grams), loneliness, social contact, perceived social support, and mood valence and calmness were assessed using smartphone-based EMA. Concurrent and prospective associations across consecutive days were examined, accounting for autoregressive stability. Findings: EMA compliance was high (96.7%). Greater-than-usual loneliness was associated with lower probability of alcohol use during the same retrospective assessment interval (b = −0.092, 95% CI [−0.151, −0.033], pFDR = .006) and lower alcohol quantity (b = −15.89 g, 95% CI [−24.16, −7.61], pFDR = .003). Prospective associations in either direction were small, with confidence intervals constraining their plausible magnitude. The concurrent association between loneliness and alcohol quantity was stronger on days with social contact, higher positive mood valence and calmness, and among men. Mood also moderated the prospective association between loneliness and subsequent alcohol use, although simple slopes were not significant. Conclusions: In adults with AUD, loneliness was associated with lower AC concurrently whereas prospective associations in either direction were small. The concurrent association varied by social and mood context and gender, suggesting that loneliness alone may provide limited information about imminent drinking risk.
ER -