Prevalence of Abnormal Vitamin D Levels Among Division I NCAA Athletes

Researchers measured 25-hydroxyvitamin D in the blood of 223 Division I college athletes over one summer. About one in three had levels below the sufficient range, with a mean of 40.1 ng/mL across the group. Athletes with dark skin tone were far more likely to have low levels, and that pattern held once sex, race, sport and season were accounted for.

Why this study was done

Vitamin D is linked to muscle power and bone mass, and low levels have been tied to muscle injuries and stress fractures. Muscle tissue from people with low vitamin D shows shrinkage in fast twitch type II fibers, used for sprinting and jumping. The team noticed a high rate of low vitamin D in athletes presenting with stress fractures, so they measured how common abnormal levels were across a college squad and which athletes were most at risk.

How it was run

A single-institution cohort study at the University of Southern California, using a retrospective chart review. One resting blood sample was taken from each participant between June and August 2012, and serum 25-hydroxyvitamin D measured by mass spectrometry. Levels were classed sufficient at 32 ng/mL or above, insufficient 20 to under 32, deficient below 20. Levels were obtained for 223 of the 559 athletes competing that year: 121 men and 102 women, 114 white, 78 black, 12 Hispanic, 7 Asian, 12 other. Logistic regression tested sex, race, skin tone, body mass index, sport type, indoor or outdoor, and season.

What they found

Mean 25-hydroxyvitamin D was 40.1 plus or minus 14.9 ng/mL, median 37, range 13 to 92. Of the 223 athletes, 148 (66.4 percent) were sufficient and 75 (33.6 percent) abnormal: 68 (30.5 percent) insufficient and 7 (3.1 percent) deficient.

On single-factor analysis:

• Dark skin tone: odds ratio 15.18, 95 percent confidence interval 7.56 to 30.52, P less than 0.0001. Abnormal levels affected 64.9 percent of dark-skinned athletes against 10.8 percent of light-skinned ones.
• Black race: odds ratio 19.12, interval 8.88 to 41.20, P less than 0.0001.
• Hispanic: odds ratio 6.07, interval 1.66 to 22.15, P = 0.0063.
• Male sex: odds ratio 2.83, interval 1.57 to 5.13, P = 0.0006.

Several factors were not significant. Body mass index showed nothing: odds ratio 0.63 for overweight (P = 0.40), 0.71 for obese (P = 0.55). Indoor sport reached only P = 0.052 and winter sport P = 0.063. After adjusting for everything at once, only dark skin tone survived, adjusted odds ratio 15.2, confidence interval 7.5 to 30.5. Every other variable was dropped.

Limitations

This is a prevalence snapshot at one university in a sunny city, taken in a single summer window, so it says nothing about winter levels or other regions. Only 40 percent of eligible athletes were tested, inviting selection bias, and 108 of 223 were football players. The authors note their per-sport analysis lacked power except for football, age was not tested because the range was too narrow, and body mass index was left out of the adjusted model because it was missing for nearly half the cohort. No diet or supplement intake was recorded. The paper labels itself Level 1 evidence, a strong grade for a retrospective single-site cohort. One author declares royalties from Arthrex, a surgical device company. No supplement or ingredient company funding is declared. The article carries a standard publisher copyright line with no Creative Commons statement.

What it means in practice

The useful finding is narrow: skin tone was the one factor that held up once everything else was accounted for. That fits known biology, since more melanin means less vitamin D made from sunlight. Read the odds ratios with their wide confidence intervals in mind. The study measures blood levels only. It does not test supplementation, does not track injuries or performance, and cannot show that correcting a low level changes any outcome. It describes elite college athletes at one institution, so the prevalence figures should not be read as applying generally.