Clinical Scorecard: Cumulative training load may signal musculoskeletal injury

At a Glance

Key Highlights

• Across 11 studies, higher sensor-derived training loads were associated with musculoskeletal injury; the exploratory pooled effect estimate was 2.74.
• Cumulative load magnitude had a pooled effect estimate of 2.86 across five studies, but the evidence was rated very low certainty.
• The ACWR estimate was 2.64 across five studies, with substantial heterogeneity and no statistically significant association under the conservative primary inference method.
• Different effect measures, populations, and study designs limit comparisons; the review did not establish that cumulative load magnitude performs better than ACWR.
• The evidence did not establish injury prediction, prevention through training-load modification, or numerical thresholds for clinical or training decisions.

Guideline-Based Recommendations

Diagnosis

• The review evaluated associations between inertial-sensor-derived training load and musculoskeletal injury; it did not provide diagnostic criteria.

Management

• The review found insufficient evidence to recommend specific numerical training-load thresholds for clinical or training decisions.
• The evidence did not establish that modifying training load prevents musculoskeletal injury.

Monitoring & Follow-up

• Wearable inertial sensors were used to quantify external training load through cumulative load magnitude and ACWR.
• The review did not establish that these measures can predict which athletes will be injured.

Risks

• All outcomes were rated very low certainty; none of the included studies had low overall risk of bias.
• Confounding was generally inadequately controlled, and sensor, exposure, injury-definition, and data-processing differences limited comparability.
• Some studies may have included overlapping cohorts, and publication bias could not be excluded.
• Several studies reported injury proportions above 10%, which could make odds ratios diverge from relative risks.

Patient & Prescribing Data

The evidence comprised 508 athletes in six sports cohorts and one military cohort; the median study sample size was 45 athletes.

No treatment regimen or validated load threshold was established. The review did not show that changing training load prevents injury.

Clinical Best Practices

• Interpret sensor-derived training-load associations cautiously; the overall pooled estimate combined different effect measures and was described as exploratory.
• Do not treat the absence of detected heterogeneity for cumulative load magnitude as evidence that study populations or methods were homogeneous.
• Do not infer that cumulative load magnitude is superior to ACWR; the measures were not compared directly within the same cohorts.
• Do not use the review as evidence that training-load metrics identify which athletes will be injured or that load modification prevents injury.

Related Resources & Content

• Systematic review and meta-analysis, BMC Musculoskeletal Disorders (full citation and URL not provided in the source)