Back

Causation Between Smoking Quantity and Depressive Symptoms in Young Adults: Evidence from Novel Cross-Lagged Twin Models

Singh, M.; Hunter, M. D.; Assary, E.; Verhulst, B.; Peterson, R. E.; Maes, H. H. M.; Dolan, C. V.; Eley, T. C.; Neale, M. C.

2025-11-19 psychiatry and clinical psychology
10.1101/2025.11.18.25340516 medRxiv
Show abstract

BackgroundSmoking and depression often co-occur in young adults, potentially due to reciprocal causal effects and/or shared underlying etiological factors. Here, we tested causal hypotheses between smoking quantity (cigarettes per day; CigDay) and depressive symptoms (DepSx) using novel Biometrical Cross-Lagged Panel Models (CLPMs), which integrate twin developmental and direction-of-causation analyses for more robust causal inference in longitudinal twin studies. MethodsStudy sample included 10,034 participants (61.6% females; 4,112 twin pairs) from the Twins Early Development Study, with up to six repeated assessments spanning ages 21-29. The Biometrical CLPM provided three key innovations over standard twin CLPM: distinct autoregressive processes of latent genetic and environmental factors; cross-lagged genetic and environmental liabilities, contrasting confounding and causation; and cross-sectional causal effects, distinguishing between proximal (short-term) and distal (lagged) causation. ResultsIn the two largest waves assessed approximately four years apart, genetic liabilities for both variables remained highly stable, while environmental influences showed considerable temporal variation. In the standard CLPM, CigDay predicted DepSx four years later. However, in the Biometrical CLPM, the cross-lagged phenotypic associations were attenuated and non-significant when accounting for the differential genetic and environmental developmental processes. Adding the cross-sectional causal paths revealed significant bidirectional proximal effects, with a stronger CigDay[->]DepSx effect. Across all six waves with intervals up to two years, analyses consistently showed significant bidirectional cross-lagged associations. ConclusionsThe findings support a reinforcing feedback loop between CigDay and DepSx during young adulthood, involving bidirectional effects that persist for up to two years but dissipate over longer intervals.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.