INTRODUCTION: LBP is common already in adolescence but the association between degenerative imaging findings and low back symptoms at early age is not well known.
METHODS: Young adults (n=874) were invited to lumbar magnetic resonance imaging (MRI;1.5‐T scanner) at a mean age of 21 years. Both questionnaire data and MRI scans were available for 554 subjects. Data on low back symptoms and functional limitations at the age of 18, 19 and 21 years was used for clustering of subjects using Latent Class Analysis (LCA). The prevalences and 95% confidence intervals (CI) of MRI findings were compared between the pain clusters. Logistic regression analysis with adjustments for the degree of disc degeneration (DD) and gender was used to evaluate the association between specific imaging findings and low back symptom severity.
RESULTS: LCA produced five clusters: Subjects in ‘Always Painful’ cluster (n=65) had been painful at all time points, in ‘Recent Onset Pain’ (n=56) had increasing symptom severity, ‘Moderately Painful’ (n=73) intermediate symptoms, while in ‘Minor Pain’ (n=193) and ‘No Pain’ (n=167) clusters minor or no symptoms, respectively. The prevalence of HNP was 21 % (18 % protrusions, 3 % extrusions), SN 17 %, RT 6.1 %, HIZ 3.2 %, SLD 5.8 % and MC 0.7 %. HNP were observed more likely in the two most painful clusters compared to the two least painful clusters (p<0.001). In the logistic regression analysis, HNP were associated with the membership of the two most painful clusters (OR = 2.5; 95 % CI 1.4‐4.5). The other specific MRI findings were not associated, independently of the degree of DD, with symptom severity.
CONCLUSIONS: HNP occurred most likely in the subjects with recent onset or persistent low back symptoms. However, HNP were also detected in subjects with no symptoms. Thus, the clinical relevance of HNP must be evaluated in the context of symptoms.