The review pooled longitudinal, prospective, population- and community-based studies examining whether baseline anxiety and depression predict later pain in young people. Investigators identified 42 studies drawn from 31 cohorts and conducted searches up to 3 July 2026. Short statements: anxiety, depression, and combined anxiety/depression were each positively associated with subsequent pain in unadjusted models. In multi-variable models that accounted for confounders, however, only depression remained significantly associated with later pain. The authors reported this pattern across diverse cohorts and ages spanning 0–24 years.
Longer explanation: the team used a three-level meta-analysis to synthesise effect estimates and also provided narrative synthesis where meta-analysis was not possible. Two independent reviewers undertook screening and extraction; risk of bias was appraised with ROBINS-E, and the certainty of pooled outcomes was graded with GRADE. The methods aimed to reduce study-selection bias and to standardise how temporality was evaluated, but the source did not provide exhaustive study-level dose or measurement details for every included cohort in the abstract; readers should consult the supplementary material for study-specific operationalisations.
Baseline pain showed a consistent positive association with later anxiety, depression, and combined anxiety/depression. Investigators reported that these associations persisted after adjustment for confounders. In short: pain predicted later mental health symptoms both before and after statistical adjustment. This pattern appeared in multiple cohorts and across varying follow-up durations.
The study team emphasised bidirectionality: while depression retained an adjusted association with later pain, pain predicted later anxiety and depression even after controlling for candidate confounders. The degree to which pain predicted each mental health outcome varied across studies, and the meta-analytic pooling revealed heterogeneity in effect sizes.
Data indicated moderate to considerable heterogeneity across pooled outcomes. The authors explored potential moderators and found that heterogeneity was partially explained by sex, baseline age, follow-up duration, pain type (for example, general versus specific pain sites), the informant reporting pain (self-report versus parent report), how pain and anxiety/depression were operationalised, and the risk-of-bias rating for each study. Some cohorts used clinical scales; others used screening instruments or single-item measures — the source did not summarise every instrument in the abstract.
Practical takeaway: pooled estimates should be interpreted in light of this variability. The review team reported moderator analyses but also acknowledged that residual heterogeneity persisted after accounting for measured study characteristics.
The authors applied ROBINS-E for bias assessment and GRADE for certainty appraisal. Two of 14 outcomes were judged high certainty, while the remainder ranged from very low to moderate certainty. Short fact: most outcome comparisons did not reach high certainty, reflecting study limitations and variability.
Expanded point: risk-of-bias concerns, inconsistencies between studies, and indirectness in some outcome operationalisations contributed to lower GRADE ratings for many comparisons. The abstract did not list which two outcomes reached high certainty; readers should consult the full text and supplementary tables for the study-level GRADE profiles and ROBINS-E assessments.
Investigators concluded that temporal associations between pain and anxiety/depression are present in young people and that these relationships are, in places, bidirectional. Authors noted that such findings should inform screening and management strategies, and be considered by clinicians and public-health planners working with children, adolescents, and young adults. Short directive: screening for co-occurring mental health symptoms in youth with persistent pain appears warranted based on these longitudinal associations.
Research implications: the evidence base would benefit from harmonised measurement of pain and mental health, longer follow-up in population samples, and clearer reporting of confounder adjustment strategies. The source did not supply precise lists of recommended covariates across all cohorts; those interested should consult the full paper for details.
The review has several concrete limitations reported by the authors. Heterogeneity across studies was moderate-to-considerable, certainty of evidence varied (two of 14 outcomes high; remaining very low–moderate), and operational differences in exposures and outcomes limited comparability. The article is a preprint and has not yet been peer reviewed; the authors and medRxiv explicitly advise that findings should not be used to guide clinical practice until formal peer review is complete. The study team stated that all data produced are available upon reasonable request.
Specific unanswered questions noted in the source include which mechanisms drive the adjusted association of depression with later pain, how sex and developmental stage modify bidirectional effects, and which pain measurement approaches yield the most reproducible longitudinal signals. These questions remain open pending further harmonised longitudinal research and peer-reviewed synthesis.