Smoking trajectories and brain health in middle-aged adults: a 15-year longitudinal study
Background. Smoking is a prevalent and preventable risk factor for poor brain health. These habits often establish in early adulthood and have lasting consequences, yet most research focuses on later-life smoking, overlooking important prevention windows. We examined smoking trajectories from early adulthood in relation to midlife brain health. Methods. In 1,900 adults (65.6% female, mean age 32.5 years) from the Dutch population-based Generation R Study (ORACLE subset), smoking was assessed five times over 15 years, from before pregnancy to midlife. K-means clustering identified smoking trajectories, alongside measures of current smoking, pack-years, and abstinence years. At mean age 47.7 years, MRI assessed gray and white matter volumes, white-matter hyperintensity volume, cerebrovascular markers, and vertex-wise cortical thickness and surface area. Findings. Five trajectories were identified: non-smoking (63%), quitting before pregnancy (12%), quitting during/after pregnancy (11%), relapsing (9%), and persistent smoking (4%). All compared with non-smoking, quitting before pregnancy was associated with 6.57 cm3 larger gray matter (95%CI[0.27,12.87],p=.041), 7.71 cm3 larger white matter (95%CI[0.77,14.65],p=.030), and larger cortical surface area in bilateral middle/inferior temporal regions. Quitting during/after pregnancy was associated with more lacunes (OR=2.39; 95%CI[1.23,4.68],p=.011), but similar brain volumes. Relapsing was associated with 0.15 standard deviations more white matter hyperintensity volume (95%CI[0.00,0.30],p=.045) and thinner regions in the right cortex. Persistent smoking had double white matter hyperintensities (0.31 standard deviations; 95%CI[0.10,0.50], p=.003) and widespread bilateral thinner cortices. Similarly, single-timepoint measures like current smoking (vs never smoking) were linked to 9.72 cm3 smaller gray matter volume (95%CI[-16.53,-2.19],p=.005), 0.17 standard deviations more hyperintensity volume (95%CI[0.03,0.32],p=.015), and widespread thinner cortices. Each additional pack-year was associated with 0.45 cm3 smaller gray matter volume (95%CI[-0.85,-0.06],p=.025); each additional abstinence year was associated with 0.01 standard deviations less hyperintensity volume (95%CI[-0.02,0.00],p=.022). Interpretation. Smoking trajectories from early adulthood and midlife are associated with midlife brain health, supporting early and sustained smoking cessation.
Authors
- Ingmar H. A. Franken (ORCID: https://orcid.org/0000-0002-7853-2694)
- Janna Cousijn (ORCID: https://orcid.org/0000-0002-7699-2582)
- Hanan El Marroun (ORCID: https://orcid.org/0000-0002-9763-5015)
- Frank J. Wolters
- Julia Neitzel (ORCID: https://orcid.org/0000-0001-5739-466X)
- Isabel K. Schuurmans (ORCID: https://orcid.org/0000-0002-2312-4087)
- Ryan L. Muetzel (ORCID: https://orcid.org/0000-0003-3215-1287)
- Fleur de Ruijter
Institutions
- Erasmus MC (NL)
- Erasmus University Rotterdam (NL)
Publication Details
- Journal
- medRxiv
- Published
- 2026-11-15
- DOI
- https://doi.org/10.64898/2026.11.13.26368016
- Primary Topic
- Smoking Behavior and Cessation
- Type
- preprint