A 2-year prospective cohort study of overuse running injuries. The Runners and Injury Longitudinal Study (TRAILS)

Key points

1. 0166% of runners sustained at least one overuse injury over 2 years; women were injured more often than men (73% vs 62%)
2. 02Knee stiffness was the sole significant predictor in multivariable analysis: each 1 N-m/deg increase raised injury odds by 18% (OR = 1.18, 95% CI 1.021-1.374, P = .03)
3. 03The effect of knee stiffness was strongest in runners weighing more than 80 kg
4. 04Flexibility, Q-angle, arch height, rearfoot motion, lower extremity strength, weekly mileage, footwear type, and prior injury were NOT significant risk factors
5. 05The multivariable model explained only 12.3% of variance, indicating most injury predictors remain unidentified

How it was conducted

Design

Prospective 2-year observational cohort study (TRAILS)

Participants

300 recreational runners aged 18-60 who were injury-free for at least 6 months and running a minimum of 5 miles per week

Baseline measures

Training history, demographics, anthropometrics, flexibility, strength (isokinetic dynamometry), 3D gait biomechanics, and psychosocial questionnaires

Injury monitoring

Biweekly email queries throughout 24 months; positive responses led to physician diagnosis using Marti et al grading (grades 1-3)

Primary analysis

Bivariate screening followed by multivariable logistic regression; training pace and body weight included as covariates

What they found

• 199 of 300 runners (66%) sustained at least 1 overuse injury; 111 of 199 injured runners (56%) were injured more than once
• 73% of female runners and 62% of male runners sustained at least 1 injury (P = .046)
• Knee was the most common first-injury site (28%), followed by foot (21%), hip (13%), ankle (12%), and leg (12%)
• 83% of first injuries occurred during year 1 (n = 166)
• Knee stiffness OR = 1.18 (95% CI 1.021-1.374, P = .03) per 1 N-m/deg increase in multivariable analysis; mean knee stiffness was 6.89 N-m/deg in injured group vs 6.72 N-m/deg in uninjured group
• SF-12 mental component score: injured mean 47.8 vs uninjured 49.5 (P = .01) in bivariate analysis; not significant in multivariable model
• PANAS negative affect score: injured mean 14.7 vs uninjured 13.6 (P = .02) in bivariate analysis; not significant in multivariable model
• No significant differences between groups in weekly mileage (20.4 vs 19.9 miles/week, P = .75), Q-angle, arch index, any flexibility measure, or any strength measure after sex adjustment
• Rearfoot motion variables (touchdown angle, eversion range of motion, maximum eversion velocity) were all non-significant predictors
• Ground-reaction forces were non-significant after adjusting for training pace and body weight

Limitations

• Analysis pooled all overuse injury types, potentially masking risk factors specific to knee, foot, or other individual diagnoses
• Extreme-value subgroups (high rearfoot motion, very high mileage) were not analysed separately, so threshold effects may have been missed
• All predictors were measured at baseline only; changes in biomechanics or training load during the 2-year period were not captured
• The final multivariable model explained only 12.3% of variance, indicating the major injury predictors in recreational runners remain unknown

Why it matters

For patients

Recreational runners, especially those weighing more than 80 kg, may reduce overuse injury risk by reducing knee stiffness through shorter stride length or softer running surfaces, while common beliefs about flexibility, arch type, and footwear choice appear less important.

For clinicians

Knee joint stiffness during early stance is the only biomechanical variable independently associated with overuse injury risk after accounting for body weight and pace; standard assessments of flexibility, Q-angle, arch height, rearfoot motion, and strength add limited predictive value for injury screening.

For readers

This is one of the largest and most comprehensive prospective running injury studies to date, and it challenges several long-held clinical assumptions about what causes overuse running injuries.