Spinal curves and health: a systematic critical review of the epidemiological literature dealing with associations between sagittal spinal curves and health.

OBJECTIVES

The purposes of this study were to (1) determine whether sagittal spinal curves are associated with health in epidemiological studies, (2) estimate the strength of such associations, and (3) consider whether these relations are likely to be causal.

METHODS

A systematic critical literature review of epidemiological (cross-sectional, case-control, cohort) studies published before 2008 including studies identified in the CINAHL, EMBASE, Mantis, and Medline databases was performed using a structured checklist and a quality assessment. Level of evidence analysis was performed as outlined by van Tulder et al (Spine. 2003;28:1290-9), and the strength of associations were determined using the procedure outlined by Hemingway and Marmot (BMJ. 1999;318:1460-7). Quality of the included articles were assessed by our own scoring system based on the STrengthening the Reporting of OBservational studies in Epidemiology checklist. Studies scoring maximum points (4/4 or 3/3) were considered to be of higher quality.

RESULTS

Fifty-four original studies were included. We found no strong evidence for any association between sagittal spinal curves and any health outcomes including spinal pain. The included studies were generally of low methodological quality. There is moderate evidence for association between sagittal spinal curves and 4 health outcomes as follows: temporomandibular disorders (no odds ratios [ORs] provided), pelvic organ prolapse (OR, 3.18; 95% confidence interval [CI], 1.46-96.93), daily function (OR range, 1.8-3.7; 95% CI range, 1.1-6.3), and death (OR, 1.40; 95% CI, 1.08-1.91). These associations are however unlikely to be causal.

CONCLUSIONS

Evidence from epidemiological studies does not support an association between sagittal spinal curves and health including spinal pain. Further research of better methodological quality may affect this conclusion, and causal effects cannot be determined in a systematic review.