Drivers of international variation in prevalence of disabling low back pain: findings from the Cultural and Psychosocial Influences on Disability study

Key points

1. 01Pain propensity (number of painful body sites outside the low back) was the single strongest predictor of future disabling LBP, with a population attributable fraction of 39.8%.
2. 02Disabling LBP prevalence varied sevenfold across 45 occupational groups (6% in Japanese sales workers to 46% in Nicaraguan nurses), far exceeding within-country nurse-vs-office-worker differences.
3. 03Mean pain propensity index correlated with group-level disabling LBP prevalence (Spearman r = 0.58) across occupational groups.
4. 04Cultural beliefs about back pain and most group-level social/economic factors were not independently associated with disabling LBP at follow-up.
5. 05Combined individual pain propensity and somatizing tendency accounted for over 50% of the population attributable fraction (PAF 54.9%, 95% CI 47.5-62.3%).

How it was conducted

Design

Prospective longitudinal cohort with baseline and follow-up questionnaires at mean 14 months

Participants

11,710 workers aged 20-59 from 45 occupational groups (nurses, office workers, manual workers) in 18 countries; 9,055 (77.3%) completed follow-up

Countries

18 countries across Europe, Asia, Latin America, Oceania, and the Middle East (2006-2011)

Primary exposure

Pain propensity index: number of anatomical sites outside the low back painful in the 12 months before baseline (0-9 sites)

Primary outcome

Disabling LBP in the past month at follow-up (pain making it difficult/impossible to dress, do housework, or cut toenails)

Analysis

Random intercept Poisson regression with prevalence rate ratios (PRRs) and population attributable fractions (PAFs)

What they found

• Pain propensity index PRR rose progressively from 1.4 (95% CI 1.2-1.5) for index=1, to 2.6 (95% CI 2.2-3.1) for index >=6, versus index=0.
• PAF for individual pain propensity >0: 39.8% (95% CI 34.0-45.7%).
• PAF for female sex: 20.3%; older age (30-59 vs 20-29 years): 16.3%; somatizing tendency (>=1 vs 0 distressing symptoms): 15.1%.
• Combined PAF for pain propensity and/or somatizing tendency: 54.9% (95% CI 47.5-62.3%).
• At follow-up, 22% of participants (2,003/9,055) reported disabling LBP; 83% of these had also reported LBP at baseline.
• Disabling LBP ranged from 6% (Japanese sales workers) to 46% (Nicaraguan nurses) across groups.
• Spearman correlation between group mean pain propensity index and group disabling LBP prevalence: r = 0.58.
• Without risk-factor adjustment, geometric SD of observed/expected prevalence ratios was 1.68 (95th centile from simulation: 1.23); after adjustment for pain propensity alone: 1.58 (95th centile: 1.37); after full model: 1.49 (between 75th and 95th centile).
• Women had PRR 1.4 (95% CI 1.2-1.5) vs men; age 50-59 vs 20-29 years PRR 1.4 (95% CI 1.2-1.6); somatizing tendency >=2 vs 0 symptoms PRR 1.4 (95% CI 1.3-1.6).
• Only one group-level factor was significant: lack of social security for long-term unemployment (PRR 1.3, 95% CI 1.0-1.6).

Limitations

• Occupational groups were not nationally representative, limiting generalizability to general populations.
• Risk factors and outcomes both relied on self-report questionnaires; pain propensity could reflect reporting tendency rather than true pain experience.
• Ergonomic exposures were not directly measured; heavy manual tasks may have been incompletely captured within-group.
• Pain propensity index was measured at baseline, so reverse causation (LBP promoting pain elsewhere) cannot be fully excluded, though a longitudinal design was used to reduce this risk.

Why it matters

For patients

People who experience pain at multiple body sites are at substantially higher risk of future disabling low back pain, independent of their job type or country.

For clinicians

Screening for widespread pain propensity and somatizing tendency may identify patients at highest risk of disabling LBP, and interventions targeting general pain sensitivity could complement spine-specific ergonomic approaches.

For readers

International differences in disabling back pain appear rooted in general pain predisposition rather than spine-specific or workplace factors, pointing to a need for research into what keeps musculoskeletal pain propensity low in countries like Japan, Pakistan, and Sri Lanka.