Description of Participants and Engagement
Of the 1435 consented youth, 7 did not provide a survey, but a parent was in the study. The resulting sample of fourth grade youth participants (n=1428, 711 males) were mostly white (n=1065, 75%) with 17% (n=238) Hispanic, 5% (n=66) reporting 2 or more races, and 2% each for American Indian, Asian, Black, and Hawaiian/Pacific Islander. Mean age was 9.6 0.4 y and mean BMI percentile was 56.6 30.1 with 24% (n=335) above the 85th percentile and 11% (n=148) above the 95th percentile.
Demographic information was not available for parents who participated only in activities other than the survey (e.g., accelerometry, family nights) because the survey was the source of demographic and behavioral information. As shown in Table 4, parents were highly educated and reported being physically active. Of the 413 parents completing the survey, mean age was 39.1 5.9y, (range 25 – 65y, n=411), 86% of the 410 reporting gender were female and 50% noted their 4th grader was a girl. Parents were mostly white (94%) with 8% Hispanic (n=32; 1 Hispanic Black/31 Hispanic White). Fewer than 5% were Asian, American Indian, Hawaiian/Pacific Islander or Black. Additional psychosocial characteristics are shown in Table 4.
Eight activity types were identified. Of the 1435 students, 777 (54%) had a parent who participated in at least 1 activity with a mean of 1.9 1.2 types of activity (range 1 – 7). These parents participated in nearly half of the activity types available to them; mean standardized PPP was 41.5% 25.4% (range 13% - 100%). The mean intensity of participation or PEI, 7.5 5.7 (range 2 – 30), was slightly more than one-fourth (27.6% 20.9%; range 4% - 100%) of the possible intensity level (i.e., standardized PEI). Overall, those who did more of the available activity types engaged in them at a higher intensity (r=.87, p<0.001, n=777). Of engaged parents, 105 (14%) did 75% or more (defined as PD) and 76 (10%) did every activity they were offered. However, only 39 (5%) participated at 75% or more of the possible intensity and 16 (2%) at their highest possible intensity. Participation in evaluation activity options was greater and more intense than in the program activities (46.0 30.7% vs 24.5 30.8% ; 30.9 24.7% vs 12.6 19.0%) but breadth and intensity of evaluation and program activities were significantly related (PPP r=0.18, p<0.001; PEI r=.27, p<0.001).
Students were evenly distributed between control (n=740; 52%) and FFF (n=695; 48%), but participating parents, compared to non-participants, had youth in the FFF treatment arm (57% vs 38%; p<0.001). Compared to controls, parents with youth in FFF engaged in more activity types (PPP 2.2 1.4 vs 1.6 0.9; p<0.001) with greater intensity (PEI 8.2 6.1 vs 6.6 4.8; p<0.001). However standardized scores showed that FFF parents did only 32.9% 19.6% of possible activity types compared to 53.0% 27.6% for controls (p<0.001). In addition the standardized intensity level was lower (p<0.001) in parents with youth in the FFF treatment (20.9% 15.1% vs. 36.7% 24.0%;p<0.001).
Parent Engagement and Activity Types
Participating parents doing < 25% of what was available to them (n=128) engaged in only four of the eight activity types and these four were the same activities that involved more than 15 participants for the 285 parents who engaged in 25 – 49.9% of the available activities (Table 5). Compared to less engaged parents, PD had a significantly greater proportion of participation in all activities except for accelerometry and take-home action packs (Table 6). Both of these activities did not require survey completion to engage and were the two most common activities notwithstanding the online parent survey. Dietary assessment for both youth and parents was popular; both were incentivized and not time intensive. PD participation in About Eating and Recipe Prep was similar even though the online program was not incentivized and had many modules, whereas recipe preparation was incentivized by providing a key ingredient, involved the child and took more than 30 minutes to do.
PD standardized PEI was higher than non-PD (66.9 20.6% vs. 21.5 12.7%; p<0.001) with a significantly greater PD PEI for the four evaluation activities and two of the program activities: Parent (2.2 2.4 vs 0.1 0.8; p<0.001) and child (2.8 2.7 vs 0.2 1.0; p<0.001) dietary assessments, accelerometry (4.4 1.6 vs. 3.8 1.9; p=.01), online survey completion (4.7 1.6 vs. 4.0 1.7; p<0.001), About Eating (0.9 1.8 vs 0.1 0.7; p <0.001), and recipe preparation (0.5 1.3 vs .02 0.8; p<0.001).
More youth in the FFF group had participating parents than control youth (64% vs 45%; p<.001). The only differences between PD and non-PD parents was that PD parents had fewer children in the FFF group than non-PD parents (27% vs 50%; p<0.001), and as shown in Table 4, their Food Acceptance subscale score of the Satter eating competence inventory was higher and child feeding styles were different. Compared to non-PD parents, PD parents were indulgent and less uninvolved. As shown in Table 6, PD more frequently than non-PD participated in parent and youth dietary assessments, recipe use and the online survey. Fewer PD than non-PD parents completed Action Packs but among the Action Pack participants, the difference in number completed (which could range from 1 to 10) was not significant between PD and non-PD parents (3.1 2.6 vs 3.5 2.6; p=.50). No other differences were noted in demographics, stress level, eating competence, activity level, or other eating behaviors (Table 4).
Relationship Between BMI and Parent Engagement
Obese/overweight parent (n=198) PPP, PEI, and standardized values did not differ from parents with normal/underweight (n=228) BMI. BMI percentile of youth with engaged parents (n=737) was significantly lower than youth without participating parents (n=636; 55.5 8.0 vs. 58.1 7.0; p<0.001). Youth BMI percentile was lower for those whose parents participated in all the available activities (n=74) compared to less those with 1 – 99 % engagement (49.2 10.0 vs. 56.2 8.3; p<0.001). Of the youth with engaged parents, 174 (23%) were overweight or obese at baseline; their parent PEI and PPP were lower than parents of normal/underweight youth (6.5 5.0 vs 7.8 5.9, p=0.005; 1.7 1.1 vs 2.0 1.3, p=0.019).
Of all youth, 354 (26%) were overweight or obese at baseline. Pattern of change in adjusted BMI percentile over 12 months did not differ between youth of PD and non-PD parents. However, BMI percentile for youth of PD (n=85) was lower (p<0.001) at baseline (49.0 0.8) and 12 months later (48.7 0.8) than for non-PD parents (n=1015; 57.1 0.2, 56.9 0.2 respectively). Among participating parents, standardized intensity and breadth of activity were inversely related to the youth BMI percentile, (n=739; PEI r=-.0.39, p<0.001; PPP r=-0.34, p<0.001).