Higher caffeine consumption, particularly from caffeinated soft drinks, was associated with lower cortical surface area among adolescents, while shorter sleep duration and differences in brain structure may help explain associations with greater mental health problems at 6- and 18-month follow-up.

Researchers analyzed data from 9,632 adolescents aged 9 to 11 years who participated in the Adolescent Brain Cognitive Development Study, a longitudinal investigation conducted across 21 US sites. Participants reported their typical weekly consumption of coffee, espresso, caffeinated tea, and soft drinks during the preceding 6 months. Structural magnetic resonance imaging was used to assess cortical thickness and surface area, while sleep duration was measured through parent reports. Linear mixed-effects models adjusted for demographic, socioeconomic, and scanner-related factors.

Primary analyses examined associations between caffeine intake and global and regional cortical thickness and surface area. Secondary analyses examined whether sleep duration mediated these associations and whether sleep duration and brain structure sequentially mediated associations with mental health problems at 6- and 18-month follow-up. Mental health outcomes were measured using the youth-reported Brief Problem Monitor, which assesses attention, internalizing, and externalizing symptoms. Baseline mental health problems were included as a covariate in the follow-up analyses.

Higher total caffeine intake was associated with lower total cortical surface area, with a standardized regression coefficient of −0.03, but not with average cortical thickness. The association with total surface area was observed at consumption levels as low as 1 cup per week. Regional analyses identified associations with lower surface area across 23 cortical regions spanning the frontal, parietal, temporal, and occipital lobes. Of these, 22 remained statistically significant in sensitivity analyses that accounted for extreme caffeine intake values.

Caffeinated soft drinks accounted for much of the observed association. Higher soft drink consumption was associated with lower total cortical surface area, with a standardized regression coefficient of −0.03, and lower surface area across 22 cortical regions. Of these, 19 remained statistically significant after sensitivity analyses. In contrast, higher coffee and espresso consumption was associated with lower average cortical thickness, with a coefficient of −0.02, particularly in frontal regions. Tea consumption was not significantly associated with global or regional cortical measurements.

Mediation analyses indicated that shorter sleep duration partially explained associations between higher total caffeine intake and lower cortical surface area. Similar findings emerged for caffeinated soft drinks. However, sleep duration did not significantly mediate associations between coffee or espresso consumption and cortical thickness, suggesting differences in the observed associations by beverage type.

The researchers also identified a sequential association among caffeine consumption, sleep duration, cortical surface area, and subsequent mental health problems. At 6-month follow-up, higher caffeine intake was associated with greater mental health problems through shorter sleep duration and lower total cortical surface area, with statistically significant indirect associations involving 11 cortical regions. At 18 months, this indirect association persisted for total cortical surface area but not for individual regions. The indirect effects were small.

Similar patterns were observed for caffeinated soft drinks. At 6 months, shorter sleep duration and lower cortical surface area sequentially mediated associations between soft drink consumption and greater mental health problems, including associations involving 10 cortical regions. At 18 months, the indirect association persisted for total cortical surface area but not regional measurements.

Sleep duration explained only part of the association between caffeine consumption and cortical structure. Associations remained after adjustment for sleep, suggesting that additional pathways may contribute. Because caffeinated soft drinks accounted for much of the observed association, sugar intake and correlated lifestyle behaviors may also have influenced the findings.

Several limitations warrant consideration. The observational design, including cross-sectional assessments of caffeine intake, sleep duration, and brain structure, prevented causal or directional conclusions. Caffeine consumption was self-reported, sleep duration was parent-reported, and beverage measurements did not account for differences in caffeine or sugar content. The highly skewed distribution of caffeine intake limited interpretation of the low-consumption threshold, while residual confounding from dietary patterns, health behaviors, stress, and adversity remained possible. Additionally, the US-based sample may limit generalizability to other populations.

Shorter sleep duration and differences in cortical structure may represent a pathway linking caffeine consumption to subsequent mental health problems during adolescence. However, further longitudinal and experimental research is needed to determine whether caffeine intake contributes directly to changes in brain development and to distinguish its effects from those of sugar and other lifestyle factors.

“Clinically, sleep duration is modifiable, and screening for caffeine use may be relevant in adolescents with sleep disturbance or emerging mental health symptoms,” wrote lead study author Jiayi Zheng, of the Department of Neuroimaging at King's College London, and colleague Divyangana Rakesh, PhD, of the same institution, in the study's conclusion.

The authors reported no biomedical financial interests or potential conflicts of interest.

Source: Biological Psychiatry: Cognitive Neuroscience and Neuroimaging