TY - JOUR
T1 - Personality disorders in patients with fibromyalgia
AU - Magen, Eli
AU - Geishin, Akim
AU - Weizman, Abraham
AU - Magen, Israel
AU - Merzon, Eugene
AU - Ashkenazi, Shai
AU - Vinker, Shlomo
AU - Israel, Ariel
N1 - Publisher Copyright:
Copyright © 2024. Published by Elsevier Inc.
PY - 2026/3
Y1 - 2026/3
N2 - Background Fibromyalgia (FM) frequently co-occurs with mental disorders, but population-scale estimates for personality disorders (PDs) and cluster-specific patterns over extended horizons are limited. We quantified cross-sectional prevalence and 20-year cumulative incidence of recorded PD diagnoses in FM versus matched controls, and assessed the impact of diagnostic opportunity. Methods We performed a retrospective matched-cohort study within a nationwide Israeli health system. Adults with FM ( N = 15,869) were matched 1:5 to controls ( N = 79,345) on sex, age, and enrollment year. Outcomes were ICD-10 PD codes (F60–F61; ancillary F63–F69). Cross-sectional prevalence at baseline and end-of-follow-up was compared using conditional logistic regression (odds ratios [ORs], 95% CIs) with false-discovery-rate. Among participants PD-free at baseline, 20-year cumulative incidence (“risk”) was estimated with risk differences (RD, percentage points) and risk ratios (RR) using Wilson and log-Wald intervals. A surveillance-calibrated sensitivity analysis scaled RRs by the FM:control psychiatry-visit ratio to approximate adjusted RRs (aRRs). Results At baseline, any PD was more prevalent in FM than controls (1.84% vs 0.49%; OR 3.79, 95% CI 3.26–4.42; q < 0.001), with pronounced elevations in Cluster B—especially borderline PD (F60.31)—and smaller signals in Cluster C (avoidant, F60.6). At the end of follow-up, contrasts persisted or strengthened (any PD 2.85% vs 0.57%; OR 5.11, 4.48–5.82; q < 0.001). Among those PD-free at baseline, 20-year risk remained higher in FM (any PD 1.03% vs 0.08%; RD 0.95 pp., 0.78–1.14; RR 12.87, 9.62–17.22), driven by Cluster B (borderline 0.20% vs 0.02%; RD 0.19 pp.; RR 12.34, 6.48–23.51). Surveillance calibration attenuated but did not eliminate Cluster-B excess (e.g., borderline aRR ≈ 3.07). Conclusions FM is associated with a selective, durable enrichment of Cluster-B PDs—most notably borderline—observable cross-sectionally and as long-term cumulative risk, only partly explained by diagnostic opportunity. Findings support routine PD screening within FM care and motivate mechanistic work on immuno-neuroendocrine and frontolimbic pathways.
AB - Background Fibromyalgia (FM) frequently co-occurs with mental disorders, but population-scale estimates for personality disorders (PDs) and cluster-specific patterns over extended horizons are limited. We quantified cross-sectional prevalence and 20-year cumulative incidence of recorded PD diagnoses in FM versus matched controls, and assessed the impact of diagnostic opportunity. Methods We performed a retrospective matched-cohort study within a nationwide Israeli health system. Adults with FM ( N = 15,869) were matched 1:5 to controls ( N = 79,345) on sex, age, and enrollment year. Outcomes were ICD-10 PD codes (F60–F61; ancillary F63–F69). Cross-sectional prevalence at baseline and end-of-follow-up was compared using conditional logistic regression (odds ratios [ORs], 95% CIs) with false-discovery-rate. Among participants PD-free at baseline, 20-year cumulative incidence (“risk”) was estimated with risk differences (RD, percentage points) and risk ratios (RR) using Wilson and log-Wald intervals. A surveillance-calibrated sensitivity analysis scaled RRs by the FM:control psychiatry-visit ratio to approximate adjusted RRs (aRRs). Results At baseline, any PD was more prevalent in FM than controls (1.84% vs 0.49%; OR 3.79, 95% CI 3.26–4.42; q < 0.001), with pronounced elevations in Cluster B—especially borderline PD (F60.31)—and smaller signals in Cluster C (avoidant, F60.6). At the end of follow-up, contrasts persisted or strengthened (any PD 2.85% vs 0.57%; OR 5.11, 4.48–5.82; q < 0.001). Among those PD-free at baseline, 20-year risk remained higher in FM (any PD 1.03% vs 0.08%; RD 0.95 pp., 0.78–1.14; RR 12.87, 9.62–17.22), driven by Cluster B (borderline 0.20% vs 0.02%; RD 0.19 pp.; RR 12.34, 6.48–23.51). Surveillance calibration attenuated but did not eliminate Cluster-B excess (e.g., borderline aRR ≈ 3.07). Conclusions FM is associated with a selective, durable enrichment of Cluster-B PDs—most notably borderline—observable cross-sectionally and as long-term cumulative risk, only partly explained by diagnostic opportunity. Findings support routine PD screening within FM care and motivate mechanistic work on immuno-neuroendocrine and frontolimbic pathways.
KW - Borderline personality disorder
KW - Cumulative incidence
KW - Fibromyalgia
KW - Matched cohort
KW - Personality disorders
KW - Psychoneuroimmunology
KW - Surveillance bias
UR - https://www.scopus.com/pages/publications/105028808979
U2 - 10.1016/j.jpsychores.2026.112536
DO - 10.1016/j.jpsychores.2026.112536
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
C2 - 41534256
AN - SCOPUS:105028808979
SN - 0022-3999
VL - 202
JO - Journal of Psychosomatic Research
JF - Journal of Psychosomatic Research
M1 - 112536
ER -