Racial-Ethnic Differences in Osteoarthritis Pain and Disability: A Meta-Analysis

Abstract

Osteoarthritis (OA), a leading cause of disability and pain, affects 32.5 million Americans, producing tremendous economic burden. Although some findings suggest that racial/ethnic minorities experience increased OA pain severity, other studies have shown conflicting results. This meta-analysis examined differences in clinical pain severity between African Americans (AAs) and non-Hispanic whites with OA. Articles were initially identified between October 1 and 5, 2016, and updated May 30, 2018, using PubMed, Web of Science, PsycINFO, and the Cochrane Library Database. Eligibility included English-language peer-reviewed articles comparing clinical pain severity in adult black/AA and non-Hispanic white/Caucasian patients with OA. Nonduplicate article abstracts (N = 1,194) were screened by 4 reviewers, 224 articles underwent full-text review, and 61 articles reported effect sizes of pain severity stratified by race. Forest plots of the standard mean difference showed higher pain severity in AAs for studies using the Western Ontario and McMasters Universities Osteoarthritis Index (0.57; 95% confidence interval [CI], 0.54-0.61) and non-Western Ontario and McMasters Universities Osteoarthritis Index studies (0.35, 95% CI, 0.23-0.47). AAs also showed higher self-reported disability (0.38, 95% CI, 0.22-0.54) and poorer performance testing (-0.58, 95% CI, -0.72 to -0.44). Clinical pain severity and disability in OA is higher among AAs and future studies should explore the reasons for these differences to improve pain management. PERSPECTIVE: This meta-analysis shows that differences exist in clinical pain severity, functional limitations, and poor performance between AAs and non-Hispanic whites with OA. This research may lead to a better understanding of racial/ethnic differences in OA-related pain.

Keywords: Disability; Meta-analysis; Osteoarthritis; Pain; Race/ethnicity.

Figure A1:

Funnel Plot of WOMAC & Non-WOMAC Pain Measures

Figure A2:

Funnel Plot of Disability Measures

Figure 1:

PRISMA Diagram

Figure 2:

Forest Plot of WOMAC Pain Measures Higher scores=worst outcome/most severe pain

Figure 3:

Forest Plot of Non-WOMAC Pain Measures Studies that reported using more than one pain scale are listed multiple times. Higher scores=worst outcome/most severe pain. Scales used by Lavernia 2004, (SF-36v2), Neuberger 2012 (OHS, OKS), and Yang 2012 (KOOS) are reverse coded, therefore mean scores were recalculated by subtracting from the maximum possible value for that scale.

Figure 4:

Forest Plot of Self-Reported Disability Measures Studies using WOMAC: Allen 2009; Ang 2003; Cruz-Almeida 2014; Gandhi 2008; Golightly 2005; Lavernia 2004; Lavernia 2010; MacFarlane 2018; Vina 2018. Studies using AIMS2: McIlvane 2008, Parmelee 2012; Smith 2016. Studies using SF-12 or SF-36v2: Groeneveld 2008; Kwoh 2015; Yang 2012. Other scales used include SIP: Cano 2006; GCPS: Cruz-Almeida 2014; and KOOS: Yang 2012. Studies the reported using more than one pain scale are listed multiple times. Higher scores=worst outcome/less mobility; The SF-12, SF-36v2 and KOOS are reverse coded, therefore mean scores were recalculated by subtracting from the maximum possible value for that scale.

Figure 5:

Forest Plot of Performance Testing Disability Measures Performance measures include Gait velocity: Blanco 2012; Walking test: Burns 2007; and Short Physical Performance Battery (SPPB): Cruz-Almeida 2014. Lower scores= worst outcome/less mobility.