This study sought to understand parental, child and neighborhood determinants and temporal patterns of BMI, %MVPA and %SED in Latino children as they approached adolescence. In this cohort, 48% of children were overweight or obese at 8-10 y and the vast majority of fathers and mothers were overweight or obese. Understanding the determinants of physical activity patterns is imperative in Latino children who are at high risk for later obesity and related cardiometabolic diseases.
First, we examined changes in BMI over the two-year period. As expected for normal growth, positive effects of child age and pubertal status at baseline on BMI were observed ,. The BMI of these children increased about 0.95 units per year which maintained them along the 75th BMI percentile, on average. Even at this young age, a very high degree of tracking of BMI was observed over the two years ,. Expected influences of paternal and maternal BMI on child BMI were demonstrated, representing genetic and environment influences ,. Baseline maternal BMI, time-varying paternal BMI, and maternal Spanish language use were positively predictive of changes in child BMI. The high burden of parental obesity and the shared home environment pose a high risk for obesity among these Latino children.
Our observation that higher maternal Spanish-language scores were associated with higher child BMI is consistent with some -, but not all reports ,. Greater US acculturation is often equated to adopting a more Westernized diet and sedentary lifestyles, increasing the risk of obesity. However, the relationship is likely more complex. A systematic review examining the relationship between acculturation and diet in Latino adults living in the US indicated that less acculturation was related to more healthful diets , but not lower energy intake, fat intake or percent energy from fat. Parental Spanish acculturation was positively associated with child obesity in two studies ,. In another study, Spanish language at home was associated with increased obesity risk in Central/South American but not Mexican origin children . The effect of parental acculturation on the child most likely will depend on the acculturation measure used, length of stay in the US, age groups, and country of origin.
The major child determinants of %MVPA were age, gender, and BMI (in boys only). %MVPA was lower in older children and was significantly higher in boys than girls, with an adjusted mean difference of 5.5%. In boys, %MVPA was negatively associated with BMI. Of the parental factors, a change in maternal BMI, maternal education and paternal age were negatively associated with %MVPA. Interestingly, increases in maternal BMI predicted decreases in child %MVPA, again reflecting family lifestyles and shared environments.
The major child determinants of %SED were age and time of assessment with %SED increasing over the two-year period. Also, %SED increased with child age and tended to be higher among girls than boys. Of the parental factors, maternal BMI changes since baseline, paternal Spanish language use and paternal age (in girls only) were associated positively with %SED. Fathers who were older and more Spanish-acculturated may have practiced more traditional Latino behaviors, impeding their children’s participation in American sports and other physical activities. In British children, ages 8-10, higher paternal age was also associated with significant increases in sedentary behaviors, but no effect was seen on MVPA .
Timing of maturation affects both physical and psychobehavioral development, however, the effects on physical activity and inactivity have been equivocal . In the Avon Longitudinal Study of Parents and Children (ALSPAC) study (n = 1351), biological maturity in boys only was inversely correlated with MVPA (r = -0.11; p = 0.01) and positively associated with sedentary activity (r = 0.10; p = 0.01) at 11 but not 13 y of age . In our study, pubertal status was not significantly associated with %MVPA or %SED.
Moderate lagged associations of %MVPA (rSpearman= 0.43, 0.39) and %SED (rSpearman= 0.38, 0.43) were seen over the two-year period, somewhat stronger than other studies . In Swedish school-aged children, low tracking in girls (r= 0.13-0.25) and low-moderate tracking in boys (r= 0.17-0.37) of MVPA and inactivity were observed over a two-year period . Similar results were seen for physical activity and inactivity in middle school girls in the Trail of Activity in Adolescent Girls (TAAG) study; intraclass correlation coefficients ranged from 0.17 to 0.22 for self-report and 0.22 to 0.29 for 6-day accelerometry .
Our study design afforded us the opportunity to examine not only the lagged effects of child BMI, %MVPA and %SED, but also their cross-lagged effects on one another. Although there were no cross-lagged effects between BMI and %MVPA, the children with higher BMI at FU1 had larger increases in %SED at FU2. The heavier children may not have been inclined or encouraged to participate in sports activities, as they approached adolescence. Although BMI was positively correlated with %SED at baseline and FU1, boys with higher levels of %SED at baseline had smaller increases in BMI at FU1, conditioned on baseline BMI.
Of the environmental factors examined, only household size was shown to influence %SED. Neighborhood disorder and victimization were not significantly associated with child BMI, %MVPA or %SED. The parents’ perceptions of disorder and victimization for the San Francisco Bay Area were relatively low; apparently, any parental concerns did not restrain their children’s level of physical activity. Our results are consistent with a systematic review of 150 studies addressing environmental correlates of physical activity . The only environmental correlates identified for children’s physical activity were father’s physical activity, child’s time spent outdoors, and school policies. Physical environment, socio-cultural environment, family structure, parental modeling, and parenting styles were unrelated to children’s physical activity. SES factors including family income, parental education and occupational status were not associated with children’s activity in the reviewed studies. Also, SES and parents’ perception of the neighborhood environment were unrelated to MVPA and sedentary time measured using accelerometry in two recent publications ,.
Most pediatric studies in the area of physical activity and health have been cross-sectional and therefore only established associations. In 897 Latino children aged 4-19 y, sedentary time was positively and strongly associated with adiposity (%fat mass) . Adjusting for age, gender and %fat mass, sedentary time also was positively associated with fasting insulin and waist circumference. In 1862 British children aged 9-10 y, MVPA and VPA were inversely associated with adiposity indexes (BMI, fat mass and waist circumference) . A few longitudinal studies have related physical activity to later health outcomes. A cross-sectional meta-analysis of 14 international studies found higher MVPA time in children and adolescents was associated with better cardiometabolic risk profiles (waist circumference, fasting insulin, triglycerides, HDL cholesterol, and systolic blood pressure) regardless of the amount of sedentary time . In the longitudinal analysis, sedentary time and MVPA at baseline were not associated with follow-up waist circumference. However, higher waist circumference at baseline was associated with increased amounts of sedentary time at follow-up. In 280 British children aged 8-10 y, higher MVPA was predictive of lower 1-y changes in BMI .
In this study, lower %MVPA was seen in children with higher BMIs, even though cross-lagged effects of %MVPA on changes in BMI were not detected, possibly attributable to low dose of MVPA. At baseline, FU1 and FU2, only 46%, 36% and 34% of children accumulated the recommended ≥60 min/d MVPA, respectively. Conversely, higher BMI at the 1-y follow-up was shown to predict higher %SED at the 2nd year follow-up, perpetuating the strong tracking of BMI in these children. The beneficial effects of physical activity on child musculoskeletal health, cardiovascular fitness, and psychosocial well-being are indisputable, but the demonstrated effect sizes tend to be low to moderate . Quantification of the long-term effects of physical activity and sedentary behaviors on health outcomes are hampered by our short-term measurement tools that capture only a snapshot of children’s activity.