Teun Remmers

Relationship between PA enjoyment and PA in children | 43 Table 3. The moderating influence of impulsivity on the association between physical activity enjoyment and PA behavior for boys and girls. % Active % Light % MVPA # PA Bouts Boys High impulsivity 3.55 (0.13 to 6.96) 2.24 (0.08 to 4.41) 1.30 (-0.81 to 3.41) 0.14 (-0.38 to 0.67) Low impulsivity -2.18 (-5.27 to 0.91) -0.34 (-2.48 to 1.81) -1.85 (-3.69 to 0.00) -0.59 (-1.12 to -0.07) Interaction 5.57 (1.34 to 9.79) 2.73 (-0.09 to 5.54) 2.84 (0.27 to 5.41) 0.70 (0.03 to 1.38) Girls High impulsivity -0.23 (-3.16 to 2.70) 0.60 (-1.60 to 2.81) -0.83 (-2.30 to 0.64) -0.19 (-0.68 to 0.29) Low impulsivity 2.57 (0.04 to 5.09) 1.38 (-0.60 to 3.35) 1.19 (-0.78 to 3.17) 0.26 (-0.32 to 0.84) Interaction -2.67 (-6.22 to 0.88) -0.67 (-3.41 to 2.07) -2.00 (-4.29 to 0.29) -0.52 (-1.22 to 0.19) Note: Linear regression analyses presenting beta’s (95% confidence intervals) adjusted for child’s age, recruitment group, maternal education, season, child’s and BMI. Bold numbers represent statistical significance at p <0.05. Discussion The present study showed that impulsivity moderates the relationship between PA enjoyment and PA behavior, but that this moderation is in opposite direction for boys and girls (see Figure 2). In boys, impulsivity strengthened the positive relationship between PA enjoyment and PA behavior while in highly impulsive girls, the relationship between PA enjoyment and PA behavior was absent. After stratification for gender and impulsivity, we found that relatively PA enjoyment was significantly associated with active behavior (i.e. all PA intensities combined) and not with light PA and MVPA. Other studies have investigated the relationship between PA enjoyment and PA behavior in children. According to the study of Dishman (2005), PA enjoyment resulted in higher PA behavior over one year, by an indirect mediated effect on self-efficacy. One study that – like our study- used the PACES and accelerometry also found no consistent association between PA enjoyment and objectively measured PA behavior, and also suggested that the association between PA enjoyment and PA behavior may be different between boys and girls (3). Wenthe et al. (2009) attempted to explain the lack of empirical support for the relationship between PA enjoyment and PA behavior by methodological imperfection regarding negatively worded questions of the PACES, the lack of within group variance of the PACES, or that PA enjoyment may have acted as a mediator in their study (3).