Puberty Appears to Shield Boys From Fatty Liver Disease, But Not Girls, Study Finds

One of the most common liver conditions in children may follow a surprising developmental timetable. Metabolic dysfunction-associated steatotic liver disease, or MASLD, the modern name for the accumulation of fat in liver cells driven by metabolic disturbances rather than alcohol, has been rising steadily among young people worldwide. Now a longitudinal study from Beijing suggests that the hormonal upheaval of puberty itself may reshape that risk, and that the effect is strikingly different in boys and girls. According to the research, boys who progress into mid and late puberty show a substantially lower likelihood of having MASLD than boys still in the early stages, while girls show no comparable protective pattern at all.

The study, published in Pediatric Research by a team led by Huiming He and Yifei Hu of Capital Medical University together with colleagues including Sten H. Vermund of the University of South Florida, drew on 1,544 children aged 10 to 13 years enrolled in the PROC study, a pediatric cohort based in the Shunyi District of Beijing. The researchers assessed each child at two follow-up visits, classifying pubertal development using Tanner staging, the standard clinical system that grades physical maturation on a five-point scale originally described by James Tanner and colleagues in the late 1960s. Because Tanner staging captures the visible progression of secondary sexual characteristics, it offers a practical, repeatable measure of how far along puberty a child has traveled at any given moment.

The central question was deceptively simple: does advancing through puberty change a child’s odds of having a fatty liver? To answer it, the team turned to Bayesian statistical methods rather than the classical frequentist approaches that dominate most medical literature. Bayesian logistic regression was used to estimate the cross-sectional association between pubertal stage and MASLD at each visit, while Bayesian mixed-effects models were deployed to verify the association longitudinally, tracking the same children over time and accounting for the fact that repeated measurements from one child are not independent. The Bayesian framework produces a full probability distribution for each effect, summarized here as adjusted odds ratios with 95 percent highest density intervals, or HDIs, a credible range that, unlike a confidence interval, can be interpreted directly as containing the true effect with high probability.

The headline result concerns boys. Compared with boys in early puberty, those in mid-puberty had roughly a third of the odds of MASLD, with an adjusted odds ratio of 0.34 and a 95 percent HDI of 0.11 to 0.68. By late puberty the association strengthened further: the odds ratio fell to 0.22, with an HDI of 0.03 to 0.59. Because the entire interval lies below 1.0 in both cases, the authors could conclude that the protective association is credible rather than a statistical fluke. Across the cohort as a whole, the overall prevalence of MASLD declined as pubertal stage advanced, but this aggregate trend concealed a sharp asymmetry between the sexes.

For girls, the story was different in a way that matters clinically. No significant association between pubertal progression and MASLD risk emerged in the female participants, at any stage. The researchers did not stop at their own cohort, however. To guard against the possibility that the Beijing findings were an artifact of a single population, they performed an external validation using data from the National Health and Nutrition Examination Survey, or NHANES, a nationally representative American survey. Because NHANES does not record Tanner stages in the same way, the team defined puberty biochemically, using total testosterone concentrations in boys and estradiol concentrations in girls as hormonal markers of pubertal maturation. This is a meaningful methodological choice: previous work has shown that gonadal steroid levels correspond reasonably well to clinician-assessed sexual maturity ratings, making circulating hormones a defensible proxy when physical staging is unavailable.

The NHANES analysis replicated the core finding. Boys at late puberty again showed markedly lower odds of MASLD, with an adjusted odds ratio of 0.25 and a 95 percent HDI of 0.01 to 0.95, while girls once more showed no significant association. The convergence of two independent datasets, one Asian and one American, one based on physical staging and one on hormone measurements, lends considerable weight to the conclusion that the sex-specific pattern is real rather than a peculiarity of the PROC cohort. It also demonstrates the value of Bayesian replication: the American estimates, though based on different puberty definitions, landed in the same protective range as the Chinese ones.

Why would puberty protect boys’ livers but not girls’? The study itself does not identify a mechanism, and the authors are careful not to overclaim, but the broader literature offers several plausible threads. Puberty triggers a surge of sex steroids, and testosterone in particular has been linked in adult studies to favorable liver fat profiles, whereas estrogen’s effects on hepatic fat handling appear more complex. Body composition also diverges dramatically at puberty: boys gain proportionally more lean skeletal muscle mass, and muscle is a major site of glucose and fatty acid disposal, while girls accumulate relatively more adipose tissue, particularly subcutaneous fat. Research on obese adolescents has documented altered lipid fluxes and fat-cell dynamics in the subcutaneous adipose tissue of girls that may predispose them to less favorable fat distribution patterns. Differences in insulin sensitivity trajectories across puberty, which dip mid-puberty before recovering, may also play out differently in the two sexes.

The stakes are considerable. MASLD in childhood is not a benign curiosity. Longitudinal natural-history studies of pediatric fatty liver disease have documented its persistence over a decade, and cohort analyses have linked childhood and adolescent steatosis to elevated long-term mortality in young adults, as well as increased risk of youth-onset type 2 diabetes and adverse cardiovascular structure in children. Estimates of the global burden in adolescents have been rising, with projections suggesting continued growth through 2035. Prior work has also hinted at sexual dimorphism in pediatric fatty liver disease, with systematic reviews of Asian children noting sex differences in prevalence, and studies of obese youth reporting gender differences in disease presence. The new PROC analysis adds a developmental dimension to this picture: the risk landscape is not static across childhood but shifts as puberty advances, and it shifts differently for each sex.

Methodologically, the study has notable strengths. The longitudinal design, with two follow-ups of the same children, allows the temporal relationship between maturation and liver status to be examined rather than merely inferred from a snapshot. The Bayesian mixed-effects framework appropriately handles within-child correlation across visits. The use of Tanner staging, assessed at both visits, captures pubertal development directly rather than relying on age, which is a poor proxy given the wide normal variation in pubertal timing. And the external validation in NHANES addresses one of the most common weaknesses of single-cohort studies: generalizability. The authors also note prior PROC findings, including evidence that reversing steatotic liver disease promotes healthier cardiovascular structures in children and that suboptimal hydration increases the risk of incident MASLD, situating the new results within an ongoing research program on pediatric metabolic liver health in Beijing.

The practical implication, the authors argue, is that prevention resources and health messages for MASLD during puberty could be tailored by sex. If mid-to-late puberty genuinely lowers hepatic fat risk in boys, then boys in early puberty, particularly those with overweight or obesity, may represent a window of heightened vulnerability in which intervention is most valuable. Girls, who show no such developmental reprieve, may need sustained attention throughout adolescence. The researchers caution that their observational design cannot prove that puberty itself causes the decline in boys, and residual confounding by body composition, diet, or physical activity cannot be excluded. Still, the consistency of the sex-specific pattern across a Chinese cohort staged by Tanner criteria and an American cohort staged by hormones makes a compelling case that the biology of adolescence, not merely the passage of time, shapes which children’s livers accumulate fat. As pediatric MASLD continues its global climb, understanding that developmental asymmetry may prove essential to catching the disease where and when it strikes.

Subject of Research: Sex-specific associations between pubertal development and metabolic dysfunction-associated steatotic liver disease in children

Article Title: Sex-specific associations between pubertal development and MASLD in children: evidence from the PROC study

Article References: He, H., Guan, M., Abudumijiti, T., Li, M., Zou, Y., Vermund, S. H., Huang, D., & Hu, Y. (2026). Sex-specific associations between pubertal development and MASLD in children: evidence from the PROC study. Pediatric Research. https://doi.org/10.1038/s41390-026-05398-2

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05398-2

Keywords: MASLD, puberty, Tanner staging, pediatric liver disease, PROC study, NHANES, Bayesian analysis, sex differences, longitudinal cohort, hepatic steatosis, adolescent health, testosterone