Burnout and work-engagement profiles, career resources, and psychological distress among early-career doctoral healthcare professionals: a cross-sectional latent profile and network analysis

Abstract

Early-career doctoral healthcare professionals combine clinical responsibilities with academic-performance expectations, yet heterogeneity in their burnout and work-engagement patterns and their associations with career resources and psychological distress is poorly characterized. We conducted a cross-sectional survey of 386 doctoral healthcare professionals within 3 years of recruitment in Sichuan, China. Burnout dimensions, work engagement, and four career-resource domains were assessed. Anxiety and depressive symptoms were measured using the two-item Generalized Anxiety Disorder scale (GAD-2) and the two-item Patient Health Questionnaire (PHQ-2), respectively. We conducted latent profile analysis using a Gaussian mixture model with a diagonal covariance structure to identify distinct burnout–engagement profiles. Multinomial regression examined profile correlates, robust linear regression compared symptom scores, and a regularized partial-correlation network described conditional associations. Positive anxiety and depression screens occurred in 79 (20.47%) and 51 (13.21%) participants. A three-profile solution was retained: resilient-engaged ( n  = 65), moderate ( n  = 156), and high-strain ( n  = 165; entropy = 0.92; mean posterior probability = 0.97). Although information criteria decreased through six classes, higher-class solutions accumulated boundary variances or produced classes below 5% under constrained variance. Compared with the resilient-engaged profile, each standard-deviation increase in career-development planning was associated with lower odds of moderate (odds ratio [OR] = 0.26, 95% confidence interval [CI] 0.13–0.52) and high-strain membership (OR = 0.16, 95% CI 0.07–0.36). After adjustment, moderate and high-strain profiles had higher GAD-2 scores by 0.81 and 1.65 points and higher PHQ-2 scores by 0.52 and 1.35 points, respectively. Four-class, posterior-probability, and low-loading-item sensitivity analyses preserved the career-planning association and symptom gradient. Emotional exhaustion was the highest-strength node in 95.90% of bootstrap samples, but strength and bridge-expected-influence stability were moderate (CS = 0.40), and the exact bridge ranking remained scoring-sensitive. Burnout–engagement profiles followed a severity gradient, with the high-strain profile comprising 42.75% of the sample and showing the greatest psychological distress. Career planning was the most consistent inverse profile correlate, but longitudinal studies are needed before causal or intervention claims can be made.