The association between diabetes mellitus and low back pain: a systematic review and meta-analysis

Systematic review and meta-analysis.

Both diabetes mellitus (DM) and low back pain (LBP) are leading global causes of disability, and emerging evidence suggests a link between them. Understanding the DM-LBP association is clinically important as it may facilitate more integrated treatment strategies; however, the exact nature of this relationship remains poorly evaluated. This systematic review and meta-analysis aimed to compare the comorbid prevalence of LBP and DM and to evaluate their longitudinal relationship in adult populations.

A systematic search was conducted across PubMed, EMBASE, CINAHL and Google scholar from inception to October 2024. This review included English-language studies involving adults (≥ 18 years) with DM (type 1, type 2, or prediabetes) and LBP. Eligible study designs included cohort studies, randomized controlled trials, case-control, twin, and cross-sectional studies. Exclusion criteria applied to review articles, single case studies and conditions including gestational diabetes, fractures, malignancies, infections, and post-traumatic- or rheumatoid arthritis. Two independent reviewers performed study screening, data extraction, and methodological quality assessment using Newcastle-Ottawa Scale. Random-effects meta-analysis and regressions were performed using R software. Primary outcomes centered on the prevalence and longitudinal incidence of LBP and DM across diverse global populations.

Twenty-six studies (28 datasets) involving 1,359,721 participants from 17 countries were included, comprising 20 cross-sectional, 4 cohort and 2 case-control designs, with a predominant representation from Asia (62%). Methodological quality assessment (via Newcastle-Ottawa Scale) classified 28.6% of datasets as low risk of bias (Good), 46.4% as moderate risk (Fair), and the remainder as high risk of bias (Poor). The pooled LBP prevalence difference was 8.7 percentage points higher in the diabetic population compared to the non-diabetic population (95% CI: 4.4-13). In contrast, the pooled DM prevalence difference was 7.4 percentage points higher in the LBP population compared to the non-LBP population (95% CI: 3.8-11). The pooled odds ratio (OR) for LBP in patients with DM, compared with those without DM, was 1.66 (95% CI: 1.34-2.06). Univariate meta-regression model identified mean age as a significant moderator (p = 0.033). Utilizing adjusted risk estimates, the longitudinal analysis of the overall risk of LBP in patients with DM was non-significant (RR 1.22; 95% CI: 0.91-1.62). A sex difference was however observed (p = 0.037) where men had a reduced risk (RR 0.83; 95% CI: 0.71-0.96) and women possibly an increased risk (RR 1.42; 95% CI: 0.88-2.32). Conversely, LBP was modestly associated with an increased risk of developing DM (RR 1.15; 95% CI: 1.02-1.29), though this association was slightly stronger in women. (RR 1.31; 95% CI: 1.11-1.56).

The certainty of evidence for all outcomes was very low. Evidence was limited by large statistical heterogeneity (I2 > 95%), potential publication bias, and clinical indirectness arising from non-standardized definitions of LBP. Thus, while this review highlights a potential bidirectional association between diabetes and low back pain, the findings must be interpreted with caution. Limited longitudinal data also suggest that LBP was modestly associated (15%) with a higher risk of developing DM compared to non-LBP peers - potentially pointing to a stronger association in women (31%). These findings may be considered in the clinical setting for both patients with DM and for those with LBP. The low certainty of evidence highlights the critical need for more standardized, high-quality longitudinal research.

The review as registered in PROSPERO (ID: CRD42025643242) and the primary funding was received from non-profit grant givers (ALF, FoUU and Skåne University hospital, Lars Holmqvist, and Sparbanksstiftelsen Varberg funds).