Decreased pain sensitivity among people with schizophrenia: a meta-analysis of experimental pain induction studies

Abstract

Patients with schizophrenia report reduced pain sensitivity in clinical studies, but experimental studies are required to establish pain sensitivity as a potential endophenotype. We conducted a systematic review of electronic databases from database inception until April 15, 2015, including experimental studies investigating pain among patients with schizophrenia spectrum disorder vs healthy controls. A random-effect meta-analysis yielding Hedges' g ±95% confidence intervals (CIs) as the effect size (ES) measure was conducted. Primary outcome was a pooled composite of pain threshold and pain tolerance; secondary outcomes included these parameters individually, plus sensory threshold, physiological pain response, and pain intensity or unpleasantness. Across 17 studies, patients with schizophrenia spectrum disorder (n = 387; age, 30.7 ± 6.9 years; females, 31.9%; illness duration, 7.0 ± 5.7 years) were compared with controls (n = 483; age, 29.5 ± 7.4 years; females, 31.0%). Patients had elevated pain threshold/pain tolerance vs controls (ES = 0.583; 95% CI, 0.212-0.954; P = 0.002; studies = 15). Results were similar in antipsychotic-free individuals (ES = 0.599; 95% CI, 0.291-0.907; P < 0.0001; studies = 8), with trend-level significance in antipsychotic-treated individuals (ES = 0.566; 95% CI, -0.007 to 1.125; P = 0.047; studies = 9). Likewise, patients with schizophrenia had increased pain tolerance (ES = 0.566; 95% CI, 0.235-0.897; P = 0.0001; studies = 6), sensory threshold (ES = 1.16; 95% CI, 0.505-1.727; P < 0.0001; studies = 5), and pain threshold (ES = 0.696; 95% CI, 0.407-0.986; P < 0.001; studies = 9), as well as reduced physiological response to noxious stimuli (ES = 0.456; 95% CI, 0.131-0.783; P = 0.006) and pain intensity/unpleasantness ratings (ES = 0.547; 95% CI, 0.146-0.949; P = 0.008). Findings were similarly significant in antipsychotic-free patients with schizophrenia (analysable parameters = 4) and antipsychotic-treated individuals (analysable parameters = 2). Finally, greater psychiatric symptoms moderated increased pain threshold, and younger patient age moderated increased pain tolerance. Decreased pain sensitivity seems to be an endophenotype of schizophrenia spectrum disorders. How this alteration links to other dimensions of schizophrenia and physical comorbidity-related help-seeking behaviour/morbidity/mortality requires further study.