Greenwood · Molecular and cellular biochemistry 2003 · prospective non-randomized cohort study · n=?

Creatine supplementation during college football training does not increase the incidence of cramping or injury.

Cited 85 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective non-randomized cohort study with self-selected supplement groups.

PubMed 12701814 · record verified 2026-08-26

What was done

NCAA Division IA college football players were observed across three seasons (1998–2000) during training and competition in variable weather conditions (8–40 °C, 19–98% relative humidity). In an open-label design, athletes elected whether to take creatine (loading dose of 15.75 g/day for 5 days, followed by ~5 g/day maintenance) or non-creatine supplements. Athletic trainers logged all treated injuries, heat illnesses, and missed practices across multiple categories.

What was found

Creatine users accounted for 34% to 56% of players over the study period. The proportion of injuries occurring in creatine users was generally proportional to or lower than the overall usage rate: - Muscle cramping: 37/96 (39%) - Heat/dehydration: 8/28 (36%) - Muscle tightness: 18/42 (43%) - Muscle pulls/strains: 25/51 (49%) - Non-contact joint injuries: 44/132 (33%) - Contact injuries: 39/104 (44%) - Illness: 12/27 (44%) - Missed practices due to injury: 19/41 (46%) - Players lost for the season: 3/8 (38%) - Total injuries and missed practices combined: 205/529 (39%)

Why it matters

These multi-season field data challenge concerns that creatine supplementation in warm or humid training conditions elevates the risk of muscle cramps, heat-related illness, or soft tissue injuries.

Limits

Supplementation was self-selected and open-label rather than randomized and blinded, creating potential confounding by position, starter status, training volume, and diet. The abstract does not report the total participant count (n), athlete-exposure hours, or formal statistical hypothesis testing (such as p-values or confidence intervals).

Cited by