Pain Phenotypes in Endometriosis: A Population‐Based Study Using Latent Class Analysis

ABSTRACT

Objective

To identify pain phenotypes in patients with endometriosis and investigate their associations with demographics, clinical characteristics, comorbidities and pain‐related quality of life (QoL).

Design

Cross‐sectional, single‐centre, population‐based study.

Setting

Referral university centre in Quebec City, Canada.

Population

Patients diagnosed with endometriosis were enrolled consecutively between January 2020 and April 2024.

Methods

Latent class analysis was used to identify pain phenotypes. A three‐step approach of latent class analysis, involving logistic regression models, was applied to assess the associations between pain phenotypes and demographics, clinical characteristics, comorbidities and pain‐related QoL.

Main Outcome Measures

Pain phenotypes; demographic, clinical and comorbidity predictors of phenotype membership; association between QoL and pain phenotypes.

Results

A total of 352 patients were included. Two pain phenotypes were identified with distinct clinical presentations: one (54% of the participants) with more severe and frequent pain symptoms and poorer QoL and the other (46% of the participants) with mild and less frequent pain symptoms. The high pain phenotype was associated with previous treatment failure, painkiller use, familial history of endometriosis, low annual family income and comorbidities, including painful bladder, fibromyalgia, migraines, lower back pain, irritable bowel syndrome, anxiety and depression or mood disorders. The presence of endometrioma was associated with the low pain phenotype. Phenotype membership was associated with distinct QoL profiles ( p < 0.001). The mean QoL score was higher in the high pain phenotype (59; 95% CI, 56–62) than in the low pain phenotype (33; 95% CI, 29–37).

Conclusion

Patients with endometriosis can be categorised into two distinct phenotypes that correlate with QoL and patient characteristics. Validation in other populations is necessary and could aid the development of specialised or personalised interventions.