Development and validation of the Sports Supplements Beliefs Scale

Abstract

It has been proposed that the use of sports supplements by athletes might lead to the use of banned substances. This has been termed the gateway hypothesis. Given this hypothesis, if we accept that athletes use non-banned sports supplements because they believe that they will be effective, a measure of athletes’ beliefs about supplements might allow practitioners to identify athletes at risk of doping. We report the five-stage development of the Sports Supplements Beliefs Scale (SSBS). In study 1 we evaluated athletes’ beliefs about sports supplements by conducting semi-structured interviews on 16 athletes. Inductive and deductive analyses resulted in a pool of 26 items. In study 2 we recruited a panel of experts and athletes to evaluate the content validity of the 26 items. 15 items were eliminated at this stage. In study 3 we subjected the responses of 171 athletes to exploratory factor analysis to determine the factor structure of the scale. A two-factor model emerged, with one strong six-item factor, a less coherent four-item factor, and one item that cross loaded. In study 4, responses of a sample of 412 team sports athletes were subjected to confirmatory factor analysis. Of three competing models tested, a six-item single-factor model demonstrated best model fit (χ2/df=2.894, RMSEA=0.068; 90% CI=0.038 to 0.099, p>0.146, SRMR=0.0246, CFI=0.987, TLI=0.978, AIC=50.045, EVCI=0.122). Factor loadings ranged from 0.4 and 0.9. All t-values were statistically significant (p<0.001) and ranged from 10.3 to 13.3. In study 5 we examined relationships between scores on the six-item scale and supplement use. Linear regression indicated that higher scores were significantly associated with the use of a greater number of supplements (β=0.534, p<0.001, r2=0.285) and higher frequency of supplement use (β=−0.517, p<0.001, r2=0.267). Scores of users and non-users of supplements differed significantly (mean differences=6.37±0.56, U=8, p<0.001), with discriminant function analysis indicating that scores correctly predicted 76% of sport supplement users and 66% of non-users (Wilks Lambda=0.760 χ2=110.988, p<0.001). Whilst future research will be required to demonstrate its predictive validity, the SSBS has utility in the assessment of athletes’ beliefs about sports supplements. In the context of the gateway hypothesis, SSBS scores might play a meaningful role in identifying at risk athletes and in evaluating interventions.