【Introduction】
The relationship between Modic changes (MCs) and low back pain (LBP) is still a matter of debate. Few long-term studies have explored this clinical issue. This study aims to investigate the association between the emergence of new MCs and LBP over an 11-year duration.
【Method】
The subjects were 163 community residents (48 males and 115 females, mean age 60 years). They all underwent lumbar spine magnetic resonance imaging (MRI) in both 2004 and 2015. Exclusion criteria encompassed individuals with a history of vertebral fractures and lumbar spine surgeries. MCs were assessed within the L1-L2 to L5-S1 range and categorized into Types 1, Type 2, Type 3, or normal. We defined cases in which the baseline MRI was normal, and any types of MCs occurred in any lumbar spine level at follow-up as the MCs occurrence. The presence of LBP and the Roland Morris Disability Questionnaire (RDQ) was evaluated at both baseline and follow-up. RDQ scores were standardized as norm-based scores (nRDQ score). Additionally, a history of LBP treatment during the study period was considered. As a confounding factors, intervertebral disc degeneration at each spinal level at baseline was evaluated using the Schneiderman classification (ranging from grade 1 to grade 4), and the cumulative sum of these grades (DDD score) was employed as an indicator of disc degeneration. Demographic data were compared between the MCs occurrence group and the non-occurrence group using the Mann-Whitney U test and chi-square test. Furthermore, a multivariate logistic regression analysis was used to investigate the associations between MCs occurrence and LBP at follow-up, decrease in nRDQ score from baseline, and the history of LBP treatment. This analysis was adjusted for age, sex, body mass index, DDD scores, presence of LBP and MCs at baseline. The significance level for analysis was set at 5%.
【Results】
There were 75 (46%) subjects in MCs occurrence group and 88 (54%) in the non-occurrence group. MCs occurrence group had significantly higher age (66.0 vs. 62.0, p=0.014), DDD score (10.0 vs. 9.0, p=0.017), and the prevalence of MCs at baseline (58.7% vs. 42.0%, p=0.034). Multivariate logistic regression analysis revealed a statistically significant association between MCs occurrence and LBP at follow-up (Odds Ratio [OR]: 2.78; 95% Confidence Interval [CI]: 1.08-7.17), as well as history of LBP treatment (OR: 2.31; 95% CI: 1.13-4.69). Conversely, no significant association was observed between MCs occurrence and decrease in nRDQ score.
【Discussion】
Our longitudinal study supported previous cross-sectional studies reporting that MCs were associated with LBP. However, it was not clear which MC types were responsible for LBP. Further research including other MRI findings and sensitivity analysis should be needed to validate these results.