María Flores-López, Raquel Reviriego, Nuria García-Marchena, Nerea Requena-Ocaña, Juan Jesús Ruiz-Ruiz, Jesús Herrera-Imbroda, Antonio Bordallo, Juan Suarez, Patricia Moreno-Peral, Juan Ángel Bellón, Fernando Rodríguez de Fonseca, Francisco Javier Pavón-Morón, Antonia Serrano
Background/Objectives: Age at alcohol initiation is consistently associated with later alcohol use disorder (AUD), but it remains uncertain whether initiation across distinct developmental stages identifies clinically different adult phenotypes. We examined whether alcohol initiation during early, middle, or late adolescence was associated with graded differences in addictive, psychiatric, and treatment-related complexity among adults with AUD. Methods: This cross-sectional study included 436 treatment-seeking adults with lifetime AUD recruited from addiction programs in Spain. Primary analyses included 409 participants who reported alcohol initiation during early (10-13 years; n = 85), middle (14-16 years; n = 182), or late adolescence (17-21 years; n = 142). Participants underwent structured assessment with the Psychiatric Research Interview for Substance and Mental Disorders (PRISM). Between-group comparisons, ordinal logistic regression, and exploratory latent class analysis (LCA) were performed. Results: Earlier initiation was associated with graded increases in comorbid non-alcohol substance use disorders (67.1%, 64.8%, and 47.2% across early, middle, and late initiation; FDR-adjusted p = 0.003), childhood conduct disorder (21.2%, 9.3%, and 4.2%; FDR-adjusted p < 0.001), and higher rates of lifetime therapeutic community treatment and mutual-support participation. Median age at AUD diagnosis was 19.0, 23.5, and 30.5 years, respectively (p < 0.001). In the model including age at AUD diagnosis, childhood conduct disorder (OR = 3.42, 95%CI 1.50-7.76) and mood disorder (OR = 2.13, 95%CI 1.41-3.21) were associated with earlier initiation. LCA identified a high externalizing/polysubstance complexity class (15.2%) enriched for earlier initiation (p < 0.001). Conclusions: Developmental timing of alcohol initiation was associated with a multidomain adult AUD phenotype. These findings suggest that initiation age could serve as a simple historical prompt for broader assessment; however, its incremental clinical utility was not tested, and it should not be interpreted as evidence of causality or a validated prediction tool.