Effects of Parent and Child Behaviours on Overweight and Obesity in Infants and Young Children From Disadvantaged Backgrounds

Systematic Review With Narrative Synthesis

Catherine Georgina Russell; Sarah Taki; Rachel Laws; Leva Azadi; Karen J. Campbell; Rosalind Elliott; John Lynch; Kylie Ball; Rachael Taylor; Elizabeth Denney-Wilson

Disclosures

BMC Public Health. 2016;16(151) 

In This Article

Results

The initial search process resulted in 4062 documents. Removal of duplicates resulted in 3114 articles, and the screening process reduced this further to 80. After analysing full document texts from the 80 potentially eligible articles, 22 studies remained. An additional 10 studies were included, six after searching reference lists and citation searching and four studies when the search was repeated. A total of 32 articles reporting on 31 studies were included in the review (Fig. 2).

Figure 2.

Flowchart depicting the search strategy and selection of articles for inclusion in the review on the effects of parents and children on overweight or obesity in disadvantaged children

Study Characteristics

The studies were heterogeneous in their design, context, focus and quality although the majority (16 of 32) of the studies used a cross-sectional self-reported survey to collect data from parents about themselves and their children. Longitudinal observational designs were also used (n = 6). Only one Randomised Controlled Trial (RCT) was included. The majority of the studies (n = 26) were conducted in the USA, with the remainder from various countries in Europe (n = 4) and Mexico (n = 1). A large proportion of the studies from the USA (n = 10) were of mixed or minority ethnic groups of low income. Two studies were of Indigenous populations of Native Americans and Native Alaskans from the USA and Canada, respectively. Other ethnicities were German (n = 1), Mexican (n = 3), Flemish (n = 1) and Dutch (n = 2). Three studies did not report ethnicity. Participants were typically recruited from primary care settings (e.g. Special Supplemental Nutrition Program for Women, Infants and Children, WIC, n = 19), preschools or kindergartens (n = 5) and community centres (n = 4). Just four studies reported using a theoretical model or conceptual framework to inform their study design. Articles were published between 2001 and 2014.

Quality Assessment

The quality rating for the studies ranged from 25 % to 100 %, with four articles rated at 25 %. Nine studies received a rating of 100 % with 13 rated at 75 %. Six further studies received a 50 % rating.

Synthesis of Findings

Key findings reported here reflect the pathways proposed in Fig. 1. A summary of each study is presented in Additional file 1: Table S1 https://static-content.springer.com/esm/art%3A10.1186%2Fs12889-016-2801-y/MediaObjects/12889_2016_2801_MOESM1_ESM.docx.

Evidence for the Association Between Parenting and Child Eating (Pathway A)

Twelve of the 32 studies examined associations between parenting and child eating (Pathway A, Fig. 1, Additional file 1: Table S1 https://static-content.springer.com/esm/art%3A10.1186%2Fs12889-016-2801-y/MediaObjects/12889_2016_2801_MOESM1_ESM.docx). The examined variables included parental feeding styles, parental capacity for resilience and an array of parental feeding practices such as control, pressure to eat and rewarding for eating. Seven studies[29,44,48–52] focused on various elements of dietary intake and one study[29] focused on dietary quality. Four studies[48–50,53] examined associations between parenting practices and soft drink consumption; three[29,49,51] focused on fruit and vegetable consumption; and one on formula feeding.[54]

Parental Feeding Practices. Several studies showed associations between maternal feeding practices and dietary outcomes in infants and children. Wijtzes et al.,[50] in a high quality study, reported that parental modelling, 'pressure to eat', 'restriction' and 'monitoring' were predictive of Dutch children's dietary intakes. Two other studies of moderate quality[29,52] also reported cross-sectional associations between various parental feeding practices and children's diets. However, null relationships between other parental feeding practices and infants' or children's dietary intakes were also reported in these two studies. Another high quality study[53] also noted that amongst Flemish mothers with lower levels of education parental modelling (avoiding negative modelling) was not associated with children's soft drink consumption.

Availability and Accessibility of Food to Children. Just one (high quality) study considered home food availability and accessibility reporting these constructs mediated associations between lower maternal education and Flemish children's increased soft drink consumption.[53]

Parental Feeding Style. Two high quality studies[53,55] reported relationships between parental feeding style and children's food or beverage intakes: De Coen[53] noted that Flemish parents' permissive parental feeding style was associated with children's increased soft drink consumption.[53] While Chaidez and Kaiser[55] reported that overweight or obese Latino mothers' use of indulgent feeding was positively associated with their toddler's intake of fat, saturated fat and sweetened beverages but not total energy six months later[55] and that maternal authoritative feeding style was predictive of toddlers' lower consumption of sweetened beverages. Two other lower quality studies also reported positive associations between parental feeding styles and children's food or beverage intakes, with parental feeding pressure, indulgent coaxing, restrictive feeding, authoritarian and uninvolved feeding styles being important.[44,51]

General Parenting. Lim and colleagues[49] (quality rating 75 %) reported that low-income African American caregivers' (n = 317) greater capacity for resilience was associated with children's higher fruit and vegetables and lower soda intakes at baseline, and soft-drink consumption four years later.

Evidence for the Association Between Parent Behaviour and Child Physical Activity/Sedentary Behaviour (Pathway B)

Two studies examined associations between parenting and children's physical activity or sedentary behaviour (Pathway B, Fig. 1) and both reported positive findings. In the higher quality study Wijtzes and colleagues[56] describe that parental (n = 268) modelling (maternal TV viewing) mediated almost a quarter of the effect of low maternal educational level as a predictor children's TV viewing time. Dawson-McClureet al's[57] lower quality (score 50 %) pre-post intervention with a low-income US sample of mixed ethnicity (76 % Afro-Caribbean) also reported associations between parenting behaviours and child physical activity and sedentary behaviour, but not weight.

Evidence for the Association Between Child Eating and Child Weight (Pathway C)

Children's Dietary Intake. Several studies examined relationships between dietary intake and body weight in a variety of disadvantaged groups.[24,58–63] Layte et al's high quality study[63] reported negative associations between diet quality and children's weight at three years in a large (n = 11134) Irish study. Six papers tested associations between sweetened beverage consumption and weight[24,58,59,61,62,64] with the only high (100 %) quality[58] study reporting associations between sugar-sweetened beverage (SSB) consumption and weight. Other studies reported that vegetable intake[64] and diet quality[59] were not associated with weight whilst consumption of high fat containing snacks was associated with greater obesity in Mexican children.[61]

Children's Eating Behaviours/Appetitive Traits. Just one, good quality (100 %), study examined associations between parental feeding and children's eating behaviours or appetitive traits. Powers et al.[65] reported no relationship between the African American WIC mothers' reports of the preschool children's 'food responsiveness' or 'desire to drink' and child BMI z-score.

Evidence for the Association Between Child Physical Activity/Sedentary Behaviour and Weight (Pathway D)

Two studies examined associations between physical activity or sedentary behaviour and children's weight (Pathway D, Fig. 1) and both of these studies reported associations between the time spent watching TV and children's weight. Layte and colleagues'[63] longitudinal prospective cohort study (n = 790) was rated as high quality, whilst Dennison et al's[33] cross-sectional study of WIC participants (n = 1761) was rated as good (75 %) quality and controlled for other variables such as the child's age, sex and ethnicity.

Evidence for the Association Between Parental Feeding practices or Parenting Behaviours and Child Weight (Pathway E)

There were 21 studies included in this review that examined the correlation between parenting behaviours and child weight. Variables considered were breastfeeding, formula feeding and an array of parental feeding behaviours such as pressure to eat and restriction.

Breastfeeding or Formula Feeding. Fifteen studies, of variable quality, investigated relationships between breastfeeding, formula feeding, bottle-feeding and weight. Thirteen of the studies were completed in the USA, ten of which collected data from WIC, including both studies examining milk bottle-feeding. Many of these studies reported a low prevalence of any breastfeeding or breastfeeding extending beyond 6 months of age.[58,63,66,67] However, four good or high quality analyses demonstrated that breastfeeding appeared protective against future weight gain in non-Hispanic White children attending WIC clinics,[66,67] a mixed ethnicity sample[35] and in an Irish sample.[63] By contrast, breastfeeding did not appear to be related to later weight in non-Hispanic Blacks or Hispanic children,[59,64] and in another study of good quality no associations were reported between feeding mode and weight[68] in a predominantly White sample.

Two studies of moderate quality (quality score 75 %), with a predominantly (75 %) Hispanic WIC population[69] and the other with Mexican mothers[61] reported that formula feeding increased the risk of overweight. In the high quality study of Gibbs and Forste[35] which included a large (n = 1527) sample of mixed ethnicity reported a positive association between predominant formula feeding versus predominant breastfeeding and weight between the ages of 9 and 24 months, when controlling for several mother and child confounders. Bogen et al.[66] and Layte et al.[63] also demonstrated that the protective effect of breastfeeding was reduced with some formula use.

Bottle Feeding. Two high quality studies[35,58] reported associations between bottle-feeding practices and weight. May et al.[58] noted that bottle use at 18 months (which often contained SSBs) was predictive of child overweight in their small (n = 134) ethnically diverse sample, whilst Gibbs and Forste[35] reported that putting a child to bed with a bottle was associated with higher weight at 24 months.

Bonuck and colleagues published three studies[4,70,71] on bottle-feeding and weight in predominantly Hispanic samples recruited from WIC centres. Results were mixed, with bottle use being predictive of obesity but not overweight,[4,71] being predictive in toddlers but not pres-choolers[4] and not predicting weight or energy intake in 12–13 month olds.[70] These studies were rated as low-poor quality.

Age of Introduction to Solid Foods. Four studies examined relationships between the age at which solid foods were introduced to infants and weight gain or dietary intakes. The two high quality studies reported associations between earlier introduction of solid foods and greater weight gain.[35,63] In contrast, two other studies reported no associations between the age of introduction of solids and Mexican infants' (n = 810)[61] or Black and Hispanic infants' (n = 96)[69] weight.

Parental Feeding Practices. Eight of the included studies investigated relationships between parental feeding practices and weight in infants or children. Two higher quality studies noted an association between parental use of 'pressure' and infant weight. Thompson et al.[44] and Powers[65] reported that parental 'pressuring' was cross-sectionally and prospectively associated with lower infant weight in low-income African American samples (n = 217 and n = 296 respectively). This relationship was not observed by Worobey and colleagues[72] in their small, cross-sectional sample (n = 49) of low-income Hispanic mothers of four-year-old children (quality score 50 %). 'Feeding restriction' was also associated with higher BMI in some[44,73] but not all[72] studies.

Starling Washington et al.[32] noted (quality 75 %) variable relationships between the measured feeding practices and chid weight whereas Hurley et al.[73] (n = 297) and Faith et al.[24] (n = 1797), reported no differences in feeding practices and child weight (quality 25 % and 50 % respectively).

These relationships may differ according to parental weight status: In Powers' high quality study[65] in which several child- and parent-related confounders were controlled for, greater maternal 'restriction' and 'control' were both associated with higher child BMI z-score in obese mothers only. In non-obese mothers maternal 'restriction' was associated with lower child BMI z-score whereas 'control' was not related to child BMI z-score in non-obese mothers. Murashima et al.[52] reported cross-sectional associations between parents' use of 'control' and 'contingency' and children's higher BMI scores (quality 75 %). Likewise, another study[60] found that when controlling for the child's age, only one of the thirteen measured parental feeding behaviours addressing maternal control of child eating ('child takes food from refrigerator/pantry between meals') was associated with obesity and none were associated with overweight.

Parental Feeding Styles. Associations between parental feeding style and infant or child weight were examined in two studies. Chaidez and Kaiser's high quality study[55] found that indulgent and authoritative parental feeding styles did not cross-sectionally associate with toddler weight in a Latino sample (n = 94). Hughes et al.[74] reported that indulgent and authoritarian, but not authoritative or uninvolved parental feeding were associated with BMI in preschool children (quality 25 %).

General Parenting. Two good quality studies examined the effects of general parenting measures on weight-related outcomes in children. Lim and colleagues[49] reported that higher caregiver resilience was associated with a lower relative risk of children remaining overweight or obese versus normal weight over the four years of the study, but did not predict whether a child transitioned from normal weight to overweight or obese. Starling Washington et al.[32] found that scores on 'response to distress' (higher in normal weight group) and 'cognitive growth fostering' (higher in obese group) were significantly different between the two groups at enrolment, but only 'cognitive growth fostering' was different six months later. None of the scores on the measure of the proximal and distal home environments were significantly different between the two groups.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....