Primary outcomes: Modest (~517IU/day) vitamin D3 supplementation, approximating the daily dosage (400-600IU/day) recommended for 19-50 year-olds [3], did not enhance favorable body composition changes (i.e. increased bone and lean mass, with decreased fat mass) in basketball players following three months of strength training compared with placebo supplementation. One possible confounding variable in the present study, with regards to the lack of enhanced bone and lean mass gains, was the unexpected (pre to post) decrease in total caloric and sodium intakes seen in the vitamin D3 supplement versus the placebo group (Table 1). The decrease in calories (to meet caloric deficits from heavy training) and mineral ions (sodium ions osmotically attract water into tissues) may have attenuated bone and muscle tissue gains in the vitamin D3 supplement group, over the placebo group. Unfortunately, we were not able to control dietary intake (which was assessed using subjective recall) which may have further limited our ability to detect meaningful changes in body composition in our small randomized cohort.
With specific emphasis on bone mass, the present study mirrored the (non-significant) results obtained in a 6-month randomized-control trial performed on 32 collegiate swimmers and divers [14]. One of the stated limitations of the prior supplementation trial was that most of those participants were white (91%) and none were vitamin D deficient [14]. We thereby performed the present randomized-control supplement trial on an athletic population known to be vitamin D deficient from past experience and prior studies [1];[8]. At baseline (pre-intervention), half (n=9) of our cohort was either Vitamin D insufficient (<30ng/mL; n=7) or deficient (<20ng/mL; n=2) whereas in the swimmer study, the average 25-OH-vitamin D level across the 6-month testing period remained above 50ng/mL [14].
In addition, a large cross-sectional study performed on 604 racially diverse male athletes identified 58.6% of participants as vitamin D deficient (<20ng/mL). Similarly, no positive correlation was noted between serum 25-OH vitamin D levels versus BMD in that study [15]. After adjusting for age and ethnicity, however, calculated bioavailable vitamin D was positively associated with hip, femoral neck, and lumbar spine BMD [15]. These authors concluded that bioavailable vitamin D (which requires measurement of vitamin D binding protein) was a more accurate predictor of bone mass than serum 25-OH vitamin D levels [15]. Thus, the lack of relationship (between bone mass vs. 25-OH vitamin D) seen in our study may also reflect a limitation in measurement, as vitamin D binding protein was not quantified in the present study.
It could also be argued that a lack of concomitant calcium supplementation may have contributed to negligible differences in bone mass between our vitamin D3 versus placebo supplement groups. A prior, prospective, study performed on 11 NCAA D1 basketball players demonstrated an increase in total BMC with individualized dosing of calcium (plus a standard 400IU/day vitamin D supplement) based upon BMC and estimated dermal (sweat) calcium losses [10]. We argue, however, that the average dietary intake of calcium in our participants hovered around ~1000mg/day (Table 1), in addition to the milk they drank ad libitum with each weekly supplement. This average dietary calcium intake was well-within the range of 800-1000mg/day, as recommended by the Institute of Medicine (IOM) [16]. Thus, we do not believe that the absence of additional calcium supplementation confounded our results. We interpret our overall favorable bone mass changes (seen in both the vitamin D and placebo supplement groups) more reflective of the robust training stimulus [17] that predominated over small changes in circulating vitamin D levels.
With respect to lean and fat mass changes in our cohort, total body lean mass increased while fat mass decreased (regardless of supplement group) in response to training. Our findings contrast with a similar randomized-control trials which suggest that vitamin D supplementation decreases biomarkers for skeletal muscle breakdown (myoglobin, creatine kinase) after ultramarathons [18] and eccentric exercise [19].
Fat mass changes, in our cohort, were positively associated with serum 25-OH vitamin D levels, so that increased fat mass augmented circulating vitamin D levels both pre- and post-intervention. Previous studies performed on basketball players [20], athletes [7], and non-athletes [21] demonstrated a tendency towards a negative relationship between 25-OH vitamin D and fat mass. Accordingly, it is widely thought that adipose tissue sequesters vitamin D, with an impaired release of vitamin D in obese individuals compared with normal weight controls [22]. The present data may extend these findings to suggest that the capacity of adipose tissue to release stored vitamin D into the circulation may be enhanced in very lean and/or athletic individuals along this same continuum.
Secondary outcomes: As expected, our African American players demonstrated significantly lower baseline serum 25-OH vitamin D levels compared with Caucasian players (Table 2), as verified in other basketball player cohorts [23, 24]. Our African American players also lost significantly more body fat (%) despite a significant increase in total calories. These findings emphasize the importance of adequate caloric intake to sustain favorable anabolic effects of strength training, which are likely independent of ethnicity. Why African American players seem to adapt more favorably (i.e. higher lean mass to fat mass ratio) to a standardized strength training program remains unclear and warrants further investigation.
An unexpected racial difference, however, was our finding that male African American players spent significantly less time in the sun than male Caucasian players. Leisure time in the sun during the summer [14] and use of tanning beds in the spring [7] have been previously (and positively) associated with serum 25-OH vitamin D levels, although another study conducted on basketball players demonstrated no correlation between sun exposure and serum 25-OH vitamin D levels [24]. The biological explanation suggests that African Americans have less Vitamin D binding protein (VDBP) and equal biologically active vitamin D levels as white Americans [9];[15]. However, our data supports the possibility of a novel behavioral component to sustained low serum 25-OH vitamin D levels in African American athletes living in the Midwest, worthy of further examination from a cultural perspective (i.e. anecdotally, our African American players strongly preferred remaining indoors, even during summer).
Tertiary outcomes: Bone mass was our main (body composition) tissue of interest and like other randomized control trials, we did not see an augmentation of bone mass with Vitamin D3 supplementation. We thereby performed preliminary analyses on mineral ions to assess other possible influencers of bone mass. As such, previous research has suggested that under-replaced sweat calcium losses may contribute to the paradoxical osteopenia seen in runners [25];[26], cyclists [27], and basketball players [1] over a competitive season. Sweat calcium losses, however, have yet to be associated with changes in BMD or bone resorption markers as expected [28]. Alternatively, interest in sweat sodium (rather than calcium) losses as a contributor to athletic osteopenia has been similarly hypothesized [29]. With ~30% of total body sodium located within the skeleton [30] and osteopenia associated with chronic hyponatremia in humans [31];[32], it is tempting to speculate that consistent sweat [Na+] losses over time may contribute to decreased bone mass in athletes. We thereby tested this hypothesis by measuring the sodium content in pharmacologically stimulated sweat during pre-intervention testing. Paradoxically, resting sweat sodium concentrations ([Na+]) were positively associated with both total body BMC and BMD (Figure 5a). We hypothesized that the positive relationship between sweat [Na+] and total bone mineral content represented an additional (extrarenal) regulatory route for sodium secretion/excretion, in response to increased sodium ingestion when the capacity for skeletal sodium storage has been maximized. Accordingly, the estimated dietary sodium intake of our collective cohort averaged >3000mg/day (Tables 1 and 2), which was well above the IOM recommendations of 1500-2300mg/day [33]. There were tendencies for total body BMC to be related to self-reported dietary intakes of sodium and calcium (Figure 5b and 5c), and between sweat [Na+] vs. sodium intake (Figure 5d)(as demonstrated elsewhere [34]) which would trend towards supporting this hypothesis in larger cohorts. These trends encourage further exploration of the complex interplay between sodium, calcium, bone metabolism, and vitamin D which have been proposed but under-investigated [35];[36].
Of final note, two male athletes, allocated into the vitamin D supplement group during summer training, sustained fifth metatarsal fractures after our supplement trial ended. A cross-sectional study of professional basketball players support a significant relationship between history of stress fractures and higher vitamin D levels [23]. However, outside of basketball players, the literature remains mixed with regards to vitamin D and fracture risks in young, active, populations, as one report suggested that stress fractures were associated with lower (<40ng/mL) serum 25-OH vitamin D levels in soldiers [37] while another demonstrated no relationship between injury and serum 25-OH vitamin D levels in swimmers and divers [14]. Any potential disadvantage of vitamin D supplementation on bone health and fracture risk in basketball players thereby warrants further exploration.
Limitations: We acknowledge the limitations of extrapolating our largely negative results to a broader population, due to our small sample size (N=18), modest weekly dosage (4000IU), short intervention period (12-weeks), highly specialized population (collegiate basketball players) and mix of ethnicities and genders within group analyses. However, we hope these data can launch future investigations involving larger cohorts of competitive athletes which are historically difficult to study. To justify our small cohort, it is important to note that the number of basketball players per team remain limited (10-15/team), with small cohort numbers representative of the sport in general [1, 20, 24]. Additionally, we acknowledge that higher dosages of vitamin D (>4000IU/day) have been suggested to promote ergogenic benefits,[1] especially in athletes who are deficient. However, excessive supplementation (>70,000 IU/week) has also been shown to be detrimental [38]. We chose this conservative dosage because it reflects the current guidelines for healthy individuals [3];[16] and the maximum dosage we could reliably dispense within our Institutions’ regulatory constraints for individuals with normal (mean) baseline levels (>30ng/mL) of serum 25-OH vitamin D. Lastly, the dietary intakes of vitamin D, in our cohort of basketball players, were below recommended thresholds (~200 IU/day), but similar to a previous cross-sectional study performed on 41 collegiate athletes [7]. Increased dietary intake of vitamin D was positively associated with increased serum 25-OH vitamin D at baseline (pre-intervention) in our athletes, demonstrating a potential confounding effect of not controlling vitamin D in the athletes’ diet during the intervention period.