Nursing Research

Urine Metabolites as Indicators of Chronic Pain and Related Symptoms in Active-Duty Service Members: A Secondary Data Analysis of a Pragmatic Clinical Trial With SMART Design

Background Chronic pain is a major cause of distress and disability, and biomarkers may aid in the assessment and treatment of it. Urine metabolites may be valuable bioindicators that can provide biological insight regarding chronic pain. Objectives To investigate the relationship between a multimarker composite measure of metabolites and patient-reported outcomes scores in adults with chronic pain, using data from a pragmatic clinical trial with a sequential, multiple-assignment randomized trial design. Methods Self-reported measures and urine samples from 169 active-duty service members with chronic pain were collected. Urine was analyzed using a preestablished panel of metabolites, including four previously identified biomarkers of pain: kynurenic acid, pyroglutamic acid, ethylmalonic acid, and methylmalonate. Multivariable linear regression models—adjusted for participant characteristics such as age and sex—were used to cross-sectionally examine the relationship between 11 patient-reported outcomes (fatigue, sleep-related impairment, anxiety, depression, anger, pain catastrophizing, physical function, pain interference, satisfaction with participation with social roles, pain intensity, and pain impact score) and the four urine metabolites both individually and as a composite (urine metabolite pain indicator, or UMPI). Given the study’s small sample size and exploratory nature, a significance threshold of p ≤ .10 was used for all analyses. Results The UMPI showed statistically significant associations with five self-reported measures (fatigue, anxiety, depression, physical functioning, and pain impact score); adjusted Pearson correlations ranged from .18 to .25. Individual metabolite analyses supported these findings, with all relationships between individual metabolites and self-reported measures showing positive associations. Kynurenic acid and ethylmalonic acid showed the strongest associations, each having statistically significant relationships with four individual self-reported measures, while pyroglutamic acid had statistically significant relationships with three self-reported measures and methylmalonate with none. The UMPI demonstrated feasible reliability. Discussion Our finding of associations between the UMPI and components of the self-reported measures supports the development of the UMPI and these four urine metabolites as biomarkers for chronic pain outcomes. Further research is planned and will be essential for establishing mechanistic insight and guiding biomarker development within the context of pain management.