PTSD Symptom Trajectories in the Wake of Traumatic Events: a Systematic Review and Meta-Analysis of 20 Years of Research

Introduction: Psychological trauma and subsequent Posttraumatic Stress Disorder (PTSD) and other stress- and trauma-related mental health problems are globally ubiquitous. Efficient trauma-related mental health care depends on reliable information about expected amounts of affected people in need of these services and timely targeting and tailoring of provided services based on accurate distinctions between adaptive psychological trauma reactions and emerging mental health problems. Over the past 20 years, statistical growth mixture modeling (GMM) methods have been shown to be a robust and clinically relevant tool to investigate prototypical PTSD symptom trajectories and its etiological mechanisms, prognostic risk indicators and treatment outcomes. We conducted a comprehensive systematic review of the currently available literature on GMM studies of PTSD symptom trajectories following potentially traumatic events. Subsequently, we applied meta-analysis to estimate the relative prevalences of the most observed PTSD symptom trajectories and identify socio-demographic, trauma-related and methodological moderators of prevalence variability.Methods: Article identification combined traditional systematic search methods with machine learning and large language model–assisted approaches. Eligible studies applied statistical Growth Mixture Methods to longitudinal assessments of PTSD symptoms using validated continuous measures at three or more time points across minimally 12 weeks in any language. To ensure consistency across studies and diminish the potential impact of classification decisions at the individual study-level, we developed a classification framework to guide the standardized interpretation of PTSD symptom trajectories. Pooled relative prevalence estimates for each trajectory were calculated using generalized linear mixed-effects models with a binomial distribution and logit link. Moderation analyses were conducted using meta-regression models.Results: We included 90 studies representing 104 independent samples and a total of 209,161 participants. Five core trajectories of PTSD symptoms were identified. The pooled relative prevalences were 74.0% (95% CI: 69.1%, 78.3%) for the low symptom trajectory, 13.6% (95% CI: 11.3%,16.4%) for the recovery trajectory, 8.5% (95% CI: 6.6%, 11.0%) for the worsening trajectory, 8.4% (95% CI: 6.6%, 10.7%) for the high symptom trajectory, and 19.8% (95% CI: 14.2%, 26.9%) for the partially symptomatic trajectory. Several sample-level, methodological, and trauma-related moderators were significantly associated with variation in relative prevalence, including education level, sample size, number of assessments, entropy, trauma type, and military versus civilian composition.