Early obesity : family-based risk factors and treatment interventions

Abstract

Background: The high prevalence of preschool obesity is a global concern. In order to support families through obesity interventions, we need a better understanding of underlying family-based risk factors. Specifically, there is a gap in the knowledge related to challenges that families face in everyday life. Thus, valid and reliable instruments to assess child and parental behaviors are required. Further, obesity treatment in early childhood seems to be more effective than treatment of adolescents but the support offered to parents needs to be optimized.

Aims: The overall aim of this thesis was to examine the family’s role in early childhood obesity. Study I: To examine associations between infant growth and known early risk factors. Study II and III: To validate two questionnaires on child and parental behaviors related to obesity and to examine associations between potential confounders.

Study IV: To describe the conceptual frame and design of a novel parent-only treatment program for early childhood obesity, the More and Less (ML) study. Study V: To examine the effects of a parent-only program as compared to standard care as a treatment for preschool obesity (primary outcome body mass index standard deviation score; BMI SDS). To assess the acceptance and feasibility of the parent-only program.

Materials: Three samples of parents and preschoolers from Stockholm County were examined: Study I. 197 one-year-old children (52% girls, mean BMI SDS -0.4) and their parents (mean age 35 years, mean body mass index (BMI) 29, 54% had a university degree and 13% born in a non-Nordic country) participating in the Early Stockholm Obesity Prevention Project (Early STOPP) recruited from child health care centers in Stockholm County. Study II and III. A school sample of 431 parents of preschoolers recruited via 25 preschools/schools and a clinical sample of 47 parents from the ML study recruited through child health care centers. In this sample, 80% of the children were of normal weight and 20% had overweight or obesity (mean age 5.5 years, mean BMI SDS 0.2). The parents were 39 years old on average with mean BMI of 24; 70% had a university degree and 13% were born in a non-Nordic country. Study IV-V. 177 children aged 4-6 years with obesity (56% girls, mean age 5.2 years, mean BMI SDS 3.2) and their parents (mean age 38 years, mean BMI 29, 57% of foreign background, 40% had a university degree) were randomized to either parent-group treatment (n=89) or to standard treatment (n=88).

Methods: Study I. Infant BMI SDS at 3, 6 and 12 months and rapid weight gain during the first year of life was compared between children at high and low risk of developing obesity based on parental BMI (n=144 high risk and n=53 low risk) and education level (n=57 high risk and n=139 low risk), adjusting for early life risk factors. Study II. We translated and validated the Lifestyle Behavior Checklist (LBC), a questionnaire measuring obesity-related child behaviors (Problem scale) and parents’ confidence (Confidence scale) in handling these behaviors. Parents’ understanding of the translated questions was assessed with cognitive interviews. Confirmatory factor analysis (CFA) was used to assess psychometric properties. We also examined associations between the LBC and the Child Feeding Questionnaire (CFQ), which measures parental feeding practices, and sociodemographic factors. Study III. We validated the Child Eating Behavior Questionnaire (CEBQ) with CFA. We also examined associations between child eating behaviors and CFQ parental feeding practices with structural equation modelling (SEM), adjusting for sociodemographic factors. Parents’ concern for their child being overweight was used as a mediator in the model. Study IV-V. We compared a parent-only program (10 sessions at 1.5 h/week) based on skills training in evidence-based positive parenting practices to standard treatment focused on lifestyle changes. BMI SDS (primary outcome) was measured at 3 and 6 months follow-up, adjusting for sociodemographic factors. Acceptance of the parent-only program by parents was assessed by mean scores on evaluation forms and by reviewing interviews with participants. The interviews were evaluated with thematic analysis.

Results: Study I. Child BMI SDS during the first year of life was associated to parental education level but not to parental BMI. The associations could not be explained by previously known risk factors. No associations were found for rapid weight gain. Study II. A five factor structure of the LBC proved best fit to the data, introducing a new factor, Screen time. The validity of the LBC was proven by: correlations to the CFQ, associations to child BMI SDS and different scorings of parents of normal weight and overweight/obese children. The LBC Confidence scale proved to be unidimensional and was not associated to any child or parental characteristics. Study III. An eight factor structure of the CEBQ proved best fit to the data. Child’s small appetite was associated to higher levels of parental pressuring feeding practices. A large appetite in the child was not directly associated to restrictive feeding practices but indirectly via parental concern for the child being overweight. Study IV-V. Children in the parent-only group reduced their BMI SDS after 3 (0.21) and 6 months (0.42) compared to an increase of 0.01 at 3 months and 0.02 at 6 months in the standard treatment group (p < 0.001). The parent-only group children were four times more likely to reach a clinically significant reduction of 0.5 in BMI SDS. Children of Swedish parents with a university degree succeeded better in treatment. The program was highly accepted by parents.

Conclusions: Parental education level is important for infant weight development as early as the first year of life, independent of parental BMI and other known early risk factors for childhood obesity. To be able to help families in treatment, we need to know what challenges the family faces in everyday life. Thus, valid and reliable instruments to assess child and parental behaviors are required. The LBC and the CEBQ are two such instruments. The associations found between child eating behavior and parental feeding practices suggest an important role for child health care practitioners to support appropriate feeding practices. Further, a parent -only program including skills training in positive parenting practices outperformed standard treatment of preschool obesity regarding child weight status. The program was well accepted by parents. This thesis strengthens the evidence for early initiated obesity interventions and elucidates considerations for reaching families of different socioeconomic backgrounds.