Affective response to physical activity as a deep phenotype in a non-randomized pilot study

Large-scale genomic studies are beginning to identify genetic predictors of physical activity (PA). For those genetically predisposed to engage in low PA, a behavioral intervention may target a malleable factor that mediates genetic predisposition to low PA (i.e., intermediate phenotype) to mitigate the genetic influences. In a non-randomized exercise promotion pilot study, we test the feasibility of examining affective response to PA (how one feels during PA) as an intermediate phenotype between genetic variation and PA adherence. We hypothesized that three single nucleotide polymorphisms (SNPs; rs8044769 and rs3751812 in FTO; rs6265 in BDNF), identified from a prior systematic review, would be predictive of affective response to PA, and that affective response to PA would mediate the SNP-PA link. Forty five healthy, low-active adults received a 12-week print-based PA promotion program. Baseline affective response to PA was assessed using the Feeling Scale, a single-item measure of affective valence. Moderate to vigorous PA (MVPA) was assessed using accelerometers pre- and post-intervention. We examined the three SNPs in a weighted genetic score. Age, sex, body mass index, race, and neighborhood walkability were potential covariates. Affective response to PA and MVPA at follow-up (minutes/day over 4–7 days) were regressed on variation in SNPs, controlling for covariates. One unit increase in genetic score was associated with a 0.14 higher mean Feeling Scale, though was not statistically significant (p = 0.13). Among individual SNPs, having an additional FTO rs8044769 C allele was associated with a mean Feeling Scale score of 0.53 units higher (p = 0.015), which was statistically significant after applying the corrected p-value of 0.016. The genetic score or individual SNPs were not predictive of MVPA 12 weeks later, thereby mediation analyses were not performed. The preliminary findings demonstrate the promise of the intermediate phenotype approach.