Obesity in Preschoolers: Behavioral Correlates and Directions for Treatment

Authors

  • Elizabeth S. Kuhl,

    Corresponding author
    1. Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, USA
    Search for more papers by this author
  • Lisa M. Clifford,

    1. Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, USA
    Search for more papers by this author
  • Lori J. Stark

    1. Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, USA
    Search for more papers by this author

(Elizabeth.Kuhl@cchmc.org)

Abstract

Nearly 14% of American preschoolers (ages 2–5) are obese (BMI ≥ 95th percentile for age and gender), yet this group has received little attention in the obesity intervention literature. This review examines what is known about behavioral correlates of obesity in preschoolers and the developmental context for lifestyle modification in this age group. Information was used to critically evaluate existing weight management prevention and intervention programs for preschoolers and formulate suggestions for future intervention research development. A systematic search of the medical and psychological/behavioral literatures was conducted with no date restrictions, using PubMed, PsycInfo, and MEDLINE electronic databases and bibliographies of relevant manuscripts. Evidence suggests several modifiable behaviors, such as sugar sweetened beverage intake, television use, and inadequate sleep, may differentiate obese and healthy weight preschoolers. Developmental barriers, such as food neophobia, food preferences, and tantrums challenge caregiver efforts to modify preschoolers' diet and activity and parental feeding approaches, and family routines appear related to the negative eating and activity patterns observed in obese preschoolers. Prevention programs yield modest success in slowing weight gain, but their effect on already obese preschoolers is unclear. Multi-component, family-based, behavioral interventions show initial promise in positive weight management for already obese preschoolers. Given that obesity intervention research for preschoolers is in its infancy, and the multitude of modifiable behavioral correlates for obesity in this age group, we discuss the use of an innovative and efficient research paradigm (Multiphase Optimization Strategy; MOST) to develop an optimized intervention that includes only treatment components that are found to empirically reduce obesity in preschoolers.

introduction

The prevalence of obesity (BMI ≥95th percentile for age and gender) in preschoolers has risen steadily from 5% in 1970–1971 (1) to current estimates of 13.9% (2). Obese children of ages 2–5 are more likely to remain overweight throughout the preschool and school years (3) and are at greater risk for remaining overweight as adults (4). Biological precursors to cardiac disease (5,6) are now evident in obese preschoolers. Further, obese preschoolers are subject to weight-based stigmatization from same-aged peers (7) and have more parent and teacher reported behavioral problems at entrance to kindergarten than healthy weight preschoolers (8). Interventions targeting already established obesity in preschoolers are thus clearly needed. However, little empirical attention has been given to understanding how diet, feeding, activity, sleep, and parenting behaviors can be modified to reduce obesity in this age group.

One innovative and efficient research strategy to address both gaps in the preschool obesity treatment literature is application of the Multiphase Optimization Strategy (MOST) (9). In MOST, a treatment package of only empirically selected, effective intervention components is developed prior to conducting a comparative effectiveness randomized controlled trial (RCT). The purpose of this review is to complete the first of three phases of MOST by reviewing the extant literature to identify lifestyle behaviors and theoretically based, empirically supported behavior change strategies that show promise for effective weight management in already obese preschoolers. Special emphasis will be placed on addressing the unique developmental challenges parents describe as complicating modification of lifestyle behaviors hypothesized to be behavioral correlates of obesity. The main findings of the review will then be integrated into a discussion of how the next two phases of the MOST research strategy can be applied to develop effective obesity treatments for preschoolers.

Methods and Procedures

A systematic search of the medical and psychological/behavioral literatures was conducted with no date restrictions using PubMed, PsycInfo, and MEDLINE electronic databases. Keywords included preschool, intervention studies, early intervention, prevention, overweight, obese, weight management, anthropometric, child nutrition, diet, physical activity (PA), sedentary behaviors, health, lifestyle, television, sleep, and screen time. Manuscript inclusion criteria were a preschool sample (ages 2–5 years old), peer-reviewed publication, and written in English. References from identified manuscripts were also reviewed for additional studies meeting the inclusion criteria. One-hundred and ten manuscripts met these criteria and are included within this review. Thirty-eight studies reported on relationships between weight outcomes and hypothesized behavioral correlates of obesity or examined differences in lifestyle behaviors by weight status for preschoolers are summarized in Table 1. Tables 2 and 3 summarize the eight prevention and three intervention programs, respectively, targeting obesity in preschoolers. Throughout this review, the current Center for Disease Control and Prevention's criterion (10) for overweight (BMI ≥ 85th percentile) and obese (BMI ≥95th percentile) are used, even if original articles used older terminology of at-risk for overweight and overweight. However, in instances where the international age- and gender-specific cut-offs (11) were used to describe weight status, no changes were made to the terminology.

Table 1.  Studies examining behavioral correlates of obesity in preschoolers
inline image
Table 2.  Summary of obesity prevention studies with weight as a primary outcome
inline image
Table 3.  Summary of obesity intervention studies with weight as a primary outcome
inline image

Results

Diet, preschool eating, and parent feeding behaviors

Preschooler diet. Recent epidemiological studies consistently show preschoolers consume an average of 1,558–1,780 calories per day (12,13,14). Caloric intake at this level exceeds daily age-based recommendations of 1,200–1,400 calories (15) and is much higher than the typical intake of 1,389 calories of preschoolers in 1977 (12), when obesity affected only 5% of the preschool population. Mounting evidence suggests increases are related to changes in the nutritional quality of preschoolers' diets. In a given day, 73% of 2–3 year olds and 86–90% of 4–8 year olds (16) do not consume the recommended five servings of fruits and vegetables (FV) (17), and these percentages may be even lower if 100% fruit juice and fried vegetables (e.g., French fries) are excluded as these foods account for up to 40% of fruit and 14% of vegetable intake for preschoolers (18). Pasta, grain desserts (e.g., cookies), yeast breads, and pizza account for the largest percentage of preschoolers' food-based calories (19). Preschoolers also consume an average of 2.29 snacks per day (20), and at least one of these snacks is a high-calorie, high-fat food (e.g., chips) for an estimated 40% of low-income preschoolers (21). Beverages also account for a significant portion of preschoolers' daily caloric intake (19). Nearly 50% consume 100% fruit juice daily (22) at almost double the recommended limit of 4–6 ounces (23) and 70% are estimated to consume sugar-sweetened beverages (SSB) daily (22). Even though recommendations specify children should drink skim or 1% milk after age 2 (15), whole and 2% milk are still among the top five sources of calories for preschoolers (19).

Despite the multitude of negative dietary patterns, only SSB consumption appears to be consistently associated with increased obesity risk during the preschool years (14,21,24) and, for girls, remaining obese throughout childhood and early adolescence (26). One study (27) found snacking between meals increased obesity risk, but two studies found no relationship between snacking patterns and preschooler obesity (28,29).

Preschooler eating behaviors. A second important component of weight management for preschoolers is decreasing developmental barriers that challenge dietary modification for this age group. Young children innately prefer sweet and salty tastes and reject bitter tastes (30,31,32), which likely contributes to their preferences for and consumption of a wide variety of grains, desserts, fruits, SSBs, and few vegetables (16,18,19,21,22,33). In fact, one study found 71% of the top 24 foods disliked by preschoolers were vegetables (33). Food preferences are correlated with food consumption between ages 2 and 5 (30,31), are established by age 4 (33), and remain stable until at least age 8 (33). Expanding preschoolers' palettes to include a variety of FVs is particularly important as low FV preference was found to increase overweight/obesity risk by 5.5 in a sample of 4th graders (34).

Food familiarity is also related to preschoolers' food preference development (33,35). Preschoolers increase self-reported preference and voluntary consumption of novel foods (including FVs) with repeated exposure to these foods (36,37) and when peers and family members in their immediate eating environment are observed to enjoy eating these foods (38). However, mothers have been found not to serve foods they themselves dislike (33), which may decrease the effectiveness of modeling if healthy choices such as vegetables are not preferred. Young children are also often food neophobic and refuse new foods. Inability to manage these behaviors may account for parents discontinuing exposure to new foods after 3–5 trials (39) instead of completing the 12–15 trials needed to support the development of new taste preferences (36). Managing negative mealtime behaviors may be particularly problematic for parents of overweight preschoolers, who report food refusal makes it difficult to ensure children consume vegetables (40).

Parent feeding behaviors. A third component to weight management for preschoolers is modification of the feeding approaches thought to be associated with the development of unhealthy eating patterns and excessive weight gain. With respect to eating patterns, cross-sectional studies have found that pressuring children to eat is negatively associated with children's FV intake and diet quality (41,42) and positively associated with food neophobia (41). While preschoolers consumed less food and made more negative comments about the foods they were pressured to eat compared to when they were not pressured to eat in experimental settings (43), two observational studies found a positive association between parent pressuring and preschoolers' caloric intake (44,45). Use of pressuring may be a function of a family's socioeconomic status (SES) or ethnicity, as one study found that low-income families reported significantly higher use of pressuring feeding practices compared to high-income families (46) and a second found that non-white families were significantly more likely to use pressuring feeding practices compared to white families (41).

Maternal restriction is positively associated with intake of restricted items in the absence of hunger when available for preschool-age girls, but not boys (47). Girls whose mothers restrict food at age 5 appear to have more difficulty regulating food intake as they consume more calories after reporting satiation compared to girls whose mothers endorse less restriction at ages 5, 7, and 9 (48). Not all types of restriction may lead to negative eating patterns, however. Children have been found to eat fewer unhealthy snacks, more FV servings, and exhibit less food neophobia if their parents use covert control (control strategies are not obvious to child, e.g., not having unhealthy foods in home) and consume more healthy snacks and FV servings if their parents use overt control (control strategies are obvious to child, e.g., having unhealthy foods in home but limiting servings) (41,49).

Feeding styles, conceptualized as degree of parents' demandingness (how much parents encourage or discourage eating) and responsiveness (use of parent or child-centered strategies to shape eating behaviors) in feeding interactions (50,51), are also related to children's eating patterns. An authoritative feeding style (high demandingness and responsiveness) is associated with increased FV availability in the home, increased efforts to have children eat FVs, and children's increased consumption of vegetables (52) compared to an authoritarian feeding style (high demandingness and low responsiveness). In contrast, indulgent (low demandingness and high responsiveness) and uninvolved (low demandingness and responsiveness) feeding styles are associated with preschoolers' consumption of significantly fewer FVs and significantly more high-energy dense foods compared to an authoritarian feeding style (53).

Eating behaviors shaped from parent feeding practices and styles appear to impact child weight status. Pressuring feeding practices (43,48,54,55) and an authoritarian feeding style (51,53) have been associated with lower child weight whereas restrictive feeding practices (47,48,54,56,57,58) and an indulgent feeding style (51,53,59) have been associated with higher child weight. While not feeding specific, one study found an authoritative general parenting style was associated with higher BMI in girls (54) and a second, longitudinal study found an authoritative general parenting style at ages 4 and 5 decreased the odds of preschoolers becoming overweight or obese 2 years later by two times compared to a permissive or neglectful parenting style and by five times compared to an authoritarian parenting style (60).

Focus group data from mothers of overweight and obese preschoolers (40,61) provide additional support for the relationship between an indulgent feeding style and children's excess weight gain (51), but contrast findings regarding the relationship between restriction and excess weight gain (47,56,58). Specifically, mothers describe several barriers to setting food limits including anticipation of tantrums, not wanting to deny children food when they report being hungry, not wanting to deprive them of sweets and snack foods (generally or as rewards), and not receiving support from other family members when trying to limit preschoolers' consumption of unhealthy foods (40,61). While comparison studies are lacking, discrepancies may reflect differences in feeding approaches that are moderated or mediated by ethnicity and SES as studies examining restriction have included primarily middle-class, white samples (41,47,48,49,54,58) and studies examining feeding styles have included primarily low-income, minority samples (50,51,52,53,59). Evidence also suggests that maternal use of restriction in white families is moderated by perceptions that children are overweight (49,62). While this relationship has not been examined for minority families, minority families with preschoolers are less likely to accurately perceive their children as overweight (28,63). Future research is clearly needed to examine the differential effects of cultural and demographic variables on feeding behaviors and their relation to preschoolers' weight gain. In addition, because most studies examining feeding and weight have been cross-sectional in design, longitudinal studies, particularly with minority populations, are needed to help clarify the directionality and possible causality in the relationship between parent feeding approaches and child weight (64).

Treatment implications. The American Academy of Pediatrics (AAP) recommends five FV servings daily, limiting or eliminating SSB intake, decreasing consumption of high-calorie/low-nutrient foods, and limiting the frequency of eating out for weight management in children (17). Parents of obese preschoolers will likely have difficulty adhering to these dietary recommendations if interventions do not teach effective strategies for increasing preschoolers' acceptance of new, healthy foods (especially FVs), managing negative mealtime and food refusal behaviors, setting limits on children's intake of high-calorie/low-nutrient foods, and improving feeding consistency between family members. Directly exposing preschoolers to new foods through taste tests may provide a more structured strategy that facilitates parents' completion of the 12–15 exposure trials needed for preschoolers to develop preferences for new foods. For example, one randomized trial demonstrated that parent-led daily vegetable taste tests over 14 days resulted in increased preferences for and voluntary consumption of vegetables for healthy weight preschoolers compared to nutritional information or no intervention (37).

Teaching parents how to manage negative child behaviors is also important because they have specifically requested this help (40) and because persistent tantruming for food at age 3 increases overweight risk at age 9.5 (65). Child behavior management strategies such as differential attention (balancing praise of desired behaviors with ignoring undesired behaviors) and time out may be beneficial to include as components of weight management programs for obese preschoolers given their effectiveness at decreasing tantrums and food refusal behaviors and improving eating and dietary adherence in young children with chronic health conditions (66). Additional strategies may include meal planning to decrease eating out, setting and enforcing rules (e.g., only FVs for snacks), and generating a list of nonfood reinforcers for good behavior and compliance. Stimulus control strategies (e.g., covert control) (41,49) are also important as the absence of unhealthy foods (stimuli) decreases visual cues for these items, which may in turn decrease tantrums and parents conceding to requests for these foods. Children may also be more likely to make healthier snack choices if these are the only foods available. While not empirically supported at this time, elements of the Satter Competence Model (67) may be helpful in shaping elements of authoritarian and authoritative feeding styles that are associated with positive eating habits and healthy child weight (51,52,53). This model suggests that parents should provide healthy food choices for children at meal and snack times and children should decide how much of the provided food to eat. Use of child behavior modification and stimulus control strategies will likely be necessary in order to implement this practice. Finally, preschoolers may be more likely to accept dietary changes if they are made for the family as a whole (33,42), which requires consistency in feeding approaches between parents and other caregivers (e.g., grandparents).

Child lifestyle

Physical activity. The AAP recommends 60 min of moderate activity daily as a recommendation for child weight management (17). Objective measures of activity suggest preschoolers are exceeding this goal by two to four times when assessed in both daycare/preschool and home settings (68,69,70,71,72). These rates may even be underestimated as three studies sampled activity at 60-s epochs instead of the recommended 15-s epochs to best capture preschoolers' sporadic, short bursts of activity (73). While preschoolers are often less active on weekends compared to week days (68,69), one study found that activity on the weekend was still consistent with AAP recommendations for 70% of preschoolers in their sample (69). Finally, higher rates of PA have been found among preschool boys, preschoolers who spend more time outdoors, and preschoolers whose parents engage them in more active play (74). Despite empirical evidence that preschoolers are generally meeting AAP activity recommendations, parents report many barriers to increasing their preschoolers' activity including high costs and low availability of structured PA programs, supervision difficulties, parents not liking PA themselves, and inclement weather (75,76).

Six studies have examined the relationship between PA and excess weight gain for preschoolers. In the first study (77), healthy weight preschool boys, but not girls, were found to be significantly more active at preschool than overweight boys. A second study (78) found moderate PA and vigorous PA did not differentiate healthy from obese preschoolers, but that engaging in low levels of vigorous PA increased the odds of being obese. Similarly, a third study found greater rates of vigorous PA and very vigorous activity (71) decreased preschoolers' risk for being overweight. Two additional studies examined the association between duration of PA and weight, with one reporting reduced overweight risk for preschoolers engaging in ≥30 min of PA or exercise each day (21) and a second reporting decreased overweight risk for preschoolers engaging in 60–120 min of active play compared to <30 min of active play daily (29). Finally, the longitudinal benefits of being physically active during the preschool years to weight management are evident as one study found higher rates of PA at ages 3–5 predicted lower BMI at ages 5–7 (79) and a second found higher rates of PA at ages 4 predicted smaller gains in BMI, triceps, and skinfolds across 8 years (80).

Treatment implications. The paucity of studies directly comparing PA for obese and healthy weight preschoolers and inconsistencies in how activity is measured and defined within these studies challenges our understanding of the contribution of PA to the onset and maintenance of obesity during the preschool years. While vigorous activity appears to decrease obesity risk for preschoolers (21,29,71,77) exactly how much daily activity at this level is needed to achieve this protection is not clear. Positive associations between increased activity during the preschool years and decreased BMI as children age (79,80) indicates the importance of activity to long-term weight management. However, parents of already obese preschoolers may not be motivated to modify their child's activity patterns as they self-report children who are already active enough and that increasing activity has not been effective in their own attempts to manage weight (61). These barriers are only strengthened given published data suggesting that some obese preschoolers may already be meeting AAP activity recommendations (81,82) and that preschoolers participating in a weight management intervention experienced BMI z-score (BMIz) decreases without further increasing their activity level (82). Studies are clearly needed to determine what level and duration of activity is sufficient for weight management in this age group. Additionally, while moderate activity (burning 3–7 kcal/min) has been defined for adolescents (83), a more developmentally appropriate and objective definition is needed for preschoolers to help parents discern whether children are meeting daily activity recommendations and to increase consistency between studies assessing PA in this age group. Finally, it is important to discern whether parents of obese preschoolers perceive similar, or additional, barriers to those reported by parents of healthy weight preschoolers (75,76) as presence of these barriers may further compromise intervention efforts to increase PA.

Obesity interventions for preschoolers should include education on the unique developmental factors of PA for this age group including that moderate activity should be encouraged through play and games, not organized sports (83). Parents will also likely benefit from suggestions for creative, short, high-movement activities that can be completed in- or outdoors, at home, and require minimal to no supervision. Indeed, enhanced education (e.g., importance of modeling, how to set PA goals) resulted in parent-reported increases in parent-child active play compared to standard education for 2–5-year-old women, infants, and children participants (84). Similarly, parents of preschoolers who received education on increasing PA, identifying opportunities for PA within their community, and the importance of PA to health and overall development reported significantly greater increases in their children's outdoor play and rates of vigorous play over 3 years compared to parents of preschoolers who received no intervention (85). While encouraging, the impact of these interventions on weight management was not measured and no objective measure of activity (e.g., actigraphs) was collected. Finally, future research should examine the importance of making smaller behavioral changes, such as taking the stairs instead of the elevator or walking to the library instead of driving, to achieving an overall healthy lifestyle.

Sedentary activity. Preschoolers spend ∼65–84% of their time at home (68,86) and 50–89% of their time at daycare/preschool (87,88,89) engaged in sedentary activity (SA). While many developmentally appropriate activities for preschoolers are sedentary in nature (e.g., coloring), empirical investigation of SA in preschoolers has primarily focused on television viewing. Population-based studies suggest that 30–68% of preschoolers (90,91,92) exceed the AAP's recommendation of ≤2 h screen-use per day (17), a practice that nearly triples their risk for becoming obese (92). Upwards of 15% of preschoolers are estimated to watch as much as 5 h of television daily (90). In addition to the home setting, one study found 36% of center-based and 70% of home-based daycares permit television viewing at rates of 1.2 and 3.4 h per day, respectively (93). Television viewing is related to negative dietary habits. One study found each additional hour of television viewing was associated with increases in fast food, SSB, fruit juice, snack food, and daily caloric intake and decreases in FV intake for 3 year olds (94). Shaping healthy screen-use habits during the preschool years is important as children who exceed AAP recommendations as preschoolers are more likely to do so at age 6 (90) and because exceeding this recommendation during the preschool years increases obesity risk at age 7 (29).

Despite these identified risks, parents of healthy and overweight preschoolers report difficulty enforcing television limits due to reliance on television as a babysitter, a source of entertainment during inclement weather, an educational tool, and because some do not perceive the quantity of time their preschooler spends in front of screens as problematic (40,95,96). Furthermore, even though it is associated with increased risk of becoming overweight (97), parents report that removing televisions from preschoolers' bedrooms is one of the most difficult recommendations from the AAP to follow (96) due to use of the television as a sleep aid for children (40,95). This is a significant problem given that 37–43% of preschoolers reportedly have a television in their bedroom (91,97) and because having a television in the bedroom increases television viewing by an average of 8 h per week for children between the ages of 4 and 7 (98).

Treatment implications. Educating parents on the definition of screen time (includes all screen activities, even if educational), screen-use recommendations, and the negative health risks of excessive screen use are important first steps to decreasing preschoolers' SA. While relation to BMI was not assessed, adherence to AAP screen recommendations increased from 64.2% to 70.5% for families receiving women, infants, and children services who received similar educational messages (99). Education should be supplemented with strategies for modifying the home environment to facilitate compliance with screen limitations. Stimulus control (e.g., covering or removing screens) is one strategy, as children may be more likely to engage in nonscreen activity if toys and PA items (e.g., balls) are more visible and readily accessible. Child behavior management strategies such as differential attention and time out will likely be necessary for parents to successfully enforce screen restrictions. Parents should also be provided with recommendations for alternate activities to replace screen use. For example, talking during meals rather than watching television or reading instead of watching a DVD. One study found that preschoolers whose parents implemented similar activity modifications reported watching significantly less television per week and were significantly more adherent to screen activity recommendations compared to those that did not receive the intervention (100). Finally, parents should be given strategies for improving preschoolers' sleep hygiene to decrease their reliance on television as a sleep aid.

Sleep. Emerging data suggests that sleep should be a component of weight management interventions for preschoolers. Receiving inadequate sleep significantly increases obesity risk in children and adolescents (101,102) even after controlling for known risk factors of pediatric obesity such as parental obesity and screen time (27,29,103). Sleep deprivation has the greatest impact on concurrent and future weight status of younger compared to older children (29,90,104,105). Sleep deprived obese youth are also at increased risk for other diseases compared to nonobese youth such as hyperglycemia (106) and higher fasting insulin, peak insulin, and insulin resistance (106,107). Although investigations of pathways linking short sleep duration to obesity have not established causation, experimental studies with adults have found sleep deprivation to be associated with alterations of metabolic and hormone processes associated with the regulation of hunger and appetite (108). Dysregulation of these processes are posited to increase hunger and appetite (109) especially for calorie dense foods (108), and to increase fatigue, which in turn, decreases PA (27,103).

Treatment implications. Existing data consistently show increased odds of overweight and obesity and subsequent disease risk among children who receive inadequate sleep. Given that this research is in its infancy, studies involving the manipulation of sleep as a means of promoting changes in weight status have yet to be conducted. Evidence linking sleep deprivation to changes in the regulation of hunger and appetite in adults (108) indicate that modification of sleep behaviors may be an important target for pediatric obesity intervention and prevention programs for preschoolers. Treatment models for sleep (110,111,112) and pediatric obesity (81,113) are typically behaviorally based, include parent training in behavior management, and promote decreases in sedentary behaviors; these overlaps increase the ease of adding sleep hygiene to obesity interventions. Developing a consistent sleep-wake schedule, encouraging bedtime routines, and eliminating televisions from the sleep environment will be important for ensuring preschoolers meet age-based sleep recommendations (12–14 h sleep for 2–3 year olds; 11–13 h sleep for 3–5 year olds) (114). Teaching parents to generalize child behavioral management strategies (e.g., positive reinforcement to improve limit setting and using differential attention to decrease bedtime resistance) will be important to helping them establish the behavioral and environmental changes to support positive sleeping habits. Incorporating sleep management into interventions for pediatric obesity could thus present a more comprehensive treatment approach that has the potential to improve treatment outcomes by addressing underlying factors that may challenge parental efforts to change unhealthy eating habits and behaviors.

Prevention and intervention efforts

Prevention. Prevention programs targeting “high-risk” populations (e.g., minority and low SES groups) have been the primary approach to addressing the rising obesity prevalence in preschoolers. Of the 13 published preschool obesity prevention models, only eight (100,115,116,117,118,119,120,121,122,123,124) have examined program impact on child weight outcomes. We review these eight studies below by delivery setting and detailed outcomes are reported in Table 2. Relevant findings from the other five prevention programs (13,37,84,85,125) were integrated into previous sections of this review.

Daycare/preschool setting. Seven RCTs (100,116,117,119,120,121,122,123,124) have examined the efficacy of single (100,120,122,123,124) and multi-component (116,117,119,121) obesity prevention programs delivered in daycare/preschool settings on child BMI. For the three single-component interventions, one 2-year program (122,123) targeted improving diet and decreasing cardiovascular risk factors in socially disadvantaged preschoolers enrolled in Head Start by reducing total and saturated fat in school meals and snacks. At the 1- and 2-year assessment, intervention children consumed significantly fewer calories from fat and saturated fat and fewer total calories at year 1, but not year 2, compared to nonintervention children (122). For intervention children, caloric intake from total fat increased significantly for boys only while intake of saturated fat decreased significantly for girls only at the 4-year follow-up (126). Total serum cholesterol was significantly lower for intervention children, but not controls at year 1 (124) and was associated with change in BMI at year 4 (126). However, changes in BMI percentile and weight-to-height ratio were not significant for either group (124), with the exception of greater weight-to-height ratio increases among white intervention children compared to white control children at year 1 (124).

A second, 30-week prevention program based in Thailand aimed to reduce BMI and weight-for-height in preschoolers by supplementing school-based physical education (control group) with an additional 35-min of PA (intervention). Post-treatment, the intervention group demonstrated significantly greater decreases in obesity prevalence based on tricep skinfold thickness ≥95th percentile and intervention girls, but not boys, had a 68% decrease in the likelihood of an increasing BMI slope compared to the control group (120). However, changes in BMI and weight-for-height were not significant for either group. Clear conclusions about the impact of the intervention on reducing the risk for obesity cannot be made, however, as the absolute value of BMI does not take into account child age and gender (127).

The third study (100) examined the efficacy of a seven-session school-based educational program (six-weekly sessions, followed by one session 1 month later) targeting reduction of screen time on changes in BMIz for preschoolers from predominately white rural communities. At 6 months, television/video use decreased for children in the intervention group and increased for controls. Adherence to AAP screen recommendations increased for children in the intervention group but decreased among controls. Treatment groups did not differ significantly in BMIz changes.

With respect to multi-component prevention programs, one study in Scotland examined the efficacy of a 24-week program on reducing BMIz, increasing PA, and decreasing SA in preschoolers through an enhanced PA program. At 6-month post-treatment, intervention children demonstrated significant improvements in global motor skills and significantly less moderate-vigorous PA compared to children in preschools with unmodified PA programs (control group) but groups did not differ significantly on SA. Groups did not differ on BMIz at the 6- or 12-month time points (121).

Two separate RCTs (116,117) were conducted to examine the impact of a 14-week prevention program targeting diet, PA, and SA compared to a general health control intervention (e.g., dental hygiene, fire safety) for minority preschoolers attending Head Start. No significant between groups differences were found on any outcomes when this program was examined with Latino preschoolers (117). However, African American (116) preschoolers receiving this program had significantly smaller gains in BMIz at 1- and 2-year follow-up assessments and significantly lower increases in calories from saturated fat at 1-year follow-up compared to those who received a general health control intervention. Treatment groups did not differ significantly on PA or SA.

Finally, a longitudinal study conducted in France (119) examined the impact of a school-based prevention program (Epideamiologie et preavention de l'obeasitea Infantile; EPIPOI) in reducing overweight (BMI >90th percentile, based on French reference curves) and improving diet, PA, and SA for children ages 3–4 who completed treatment compared to school-archived anthropomorphic data for same-age preschoolers who received no treatment. Those receiving treatment were randomized by classroom to either basic strategy (EPIPOI-1) or basic strategy and education (EPIPOI-2). The EPIPOI-1 intervention consisted of educating parents, family physicians, and teachers about the health consequences of overweight. Parents of children whose BMI ≥75th percentile were encouraged to follow-up with their pediatrician, who were offered pediatric obesity treatment training. Children randomized to EPIPOI-2 received the EPIPOI-1 intervention and ten, 20-min education sessions on diet, activity, and sedentary behaviors. At the conclusion of 2 years, prevalence of overweight increased for all groups, but was less for children receiving intervention compared to the control group. Participant's community SES also differentiated outcomes. For participants from high SES communities, EPIPOI-2 children demonstrated significantly smaller increases in median BMIz compared to EPIPOI-1 and control children. Among children from lower SES communities, children in both intervention groups had lower median BMIz and smaller gains in BMIz, compared to the control group. Outcome data for change in targeted lifestyle behaviors were not reported.

Home setting. Only one home-based prevention program (118) targeting weight-for-height z as an outcome has been reported. This 16-week program taught overweight/obese Native American mothers how to apply parenting skills to modify diet and PA habits of their 9-month to 3-year-old children. At post-treatment, mothers in the intervention group significantly decreased restrictive parent feeding practices; however, there was only a trend toward reduced weight-for-height z and daily caloric intake for intervention families compared to the control group who received only general parenting skills.

Community setting. Only one community-based prevention program (115) has been reported, which evaluated the impact of a multi-setting (daycares, preschools, home-based services, and community health care centers), multi-component (diet, PA, and SA) intervention on changes in BMIz for cohorts of 2- and 3.5-year-old Australian children. Within intervention communities, health and childcare providers were educated on strategies to promote child health behaviors, marketing and health-related resource materials were disseminated, and policies were implemented to improve diet and activity in early childhood educational and daycare centers. Obesity prevalence decreased for children in both cohorts. Decreases in weight and BMIz were significantly greater for the 3.5-year-old cohort only compared to children in nonintervention communities. Across both intervention groups, parents reported significant increases in child intake of FVs, water, and milk and significant decreases in juice from pre to postintervention. Parent-reported screen use was also significantly lower for intervention compared to control group children at post-treatment. Changes in PA were not significant for any group.

Clinical and research implications. Obesity prevention programs have yielded modest improvements in lifestyle behaviors thought to be associated with weight gain in preschoolers, but the impact of these changes on weight and BMI has typically been small. Prevention programs yielded significant changes in BMI measures compared to no treatment in four of eight studies (115,116,119,120), but only two programs (115,119) reported significant reductions in the prevalence of overweight/obesity in preschoolers. Furthermore, the only prevention program that differentiated preschoolers by weight outcome found that those ≥85th BMI percentile did not benefit from the prevention intervention (116). Limited long-term follow-up data challenge efforts to draw conclusions about the extent to which the reviewed programs meet the primary aim of preventing future obesity for high-risk preschool children. Additionally, most prevention programs have occurred in daycare/preschool settings, targeting children and not parents as the agents of change, which may limit generalizability of program goals to the home setting. The limited impact of prevention programs on child BMI may, in part, be a reflection of the multifaceted and complex nature of obesity, and indicate the need for more intensive and comprehensive programs of intervention.

Interventions. Only four studies (81,82,113,128) have examined the effectiveness of weight management intervention programs for already obese preschoolers (see Table 3) and only one was an RCT. The first study (128) examined the efficacy of a clinic-based program involving sessions every 1–3 months that targeted dietary and PA modifications through parent education and goal setting with a sample of Singaporean preschoolers. At 12 months, 20.2% of the obese children achieved a healthy weight status; however, no data was provided on the type or degree of diet or PA changes and there was no control group.

Three studies have examined the efficacy of two family-based behavioral weight management interventions for preschoolers (81,82,113). In the first study (113), obese preschoolers completing a 12-month (10-weekly sessions, followed by monthly sessions), family-based weight management program targeting diet, PA, and child management demonstrated a decrease in percent overweight by the conclusion of treatment that was maintained at 1-year follow-up. Preschoolers also experienced a significant decrease in fat intake and daily calorie intake from baseline to 10 weeks. Data on PA changes was not reported. While these data are encouraging, lack of a control group precludes our understanding of whether the decrease in overweight was due to treatment or time.

Two studies (81,82) examined the efficacy of a 24-week, clinic and home-based, multi-component (diet, PA, and child management) intervention for obese preschoolers. This program, called Learning about Activity and Understanding Nutrition for Child Health (LAUNCH), placed particular emphasis on addressing developmental barriers to modification of lifestyle behaviors in preschoolers. In the first study, BMIz and percent over ideal body weight decreased for treatment completers but increased or remained the same for noncompleters at post-treatment (6 months). At the 6-month follow-up, decreases in BMIz and percent ideal body weight were greater for treatment completers compared to noncompleters. Treatment completers maintained an activity level consistent with AAP recommendations across treatment (81).

In a pilot RCT of the same intervention (82) decreases in BMIz, BMI percentile, and weight gain were significantly greater for LAUNCH compared to children who received an enhanced usual care visit with a pediatrician (Pediatrician Counseling) at post-treatment and 6-month follow-up. LAUNCH also demonstrated significantly greater reductions in daily caloric intake from baseline to 6-month follow-up compared to Pediatrician Counseling. Parents participating in LAUNCH lost significantly more weight at post-treatment and 6-month follow-up and reported significantly greater reductions in aversion to mealtime compared to parents participating in Pediatrician Counseling at post-treatment. Compared to Pediatrician Counseling, LAUNCH families also demonstrated significantly greater increases in the number of fresh FVs and significantly greater decreases in the number of high-calorie beverages in their homes from baseline to post-treatment and significantly greater decreases in the number of high-calorie foods in the home from baseline to 6-month follow-up. PA did not change significantly for either group and levels were consistent with AAP recommendations at all time points for all participants.

Clinical and research implications. Initial findings from intervention studies are promising and demonstrate that intensive, multi-component intervention programs may be more effective at weight management than prevention programs for already obese preschoolers. However, limited data is provided on intervention effects on modifications to targeted lifestyle behaviors. Family-based behavioral interventions (81,82,113) appear to be effective at decreasing preschoolers' daily caloric intake, and one focusing specifically on environmental changes yielded positive outcomes with respect to decreasing the obesogenic nature of homes for intervention participants (81,82). Thus, while intervention studies provide some insight on the utility of specific intervention components in reducing obesity in preschoolers, this clearly remains an understudied area of research. Further, methodological limitations including small sample sizes, inconsistent inclusion of control groups, and failure to include preschoolers from ethnic/racial and SES backgrounds that are disproportionately affected by obesity, challenges our efforts to draw conclusions about the overall effectiveness of these interventions and particularly their generalizability. Future research should thus examine the extent to which specific lifestyle modifications optimize weight outcomes for obese preschoolers.

Additionally, obese preschoolers may be less likely to benefit from these programs if their weight is not recognized as problematic. Across development, pediatricians are less likely to identify obesity and recommend treatment for preschoolers (129,130). Only 10.5–29% of parents accurately identify their preschoolers as obese (63,131,132,133), with the lowest rates of accuracy occurring within minority (134) and low-income (63) families. Parental perceptions of preschoolers' weight are influenced by normalization of excess weight gain, overriding concerns of undernutrition and growth deficits, the belief that growth alone will moderate weight (63,90,131,132,133,134), absence of a comorbid health condition (61,135), and biased attributions of genetic factors in determining weight. Ensuring pediatricians have the necessary tools and knowledge to effectively screen, discuss, and make treatment recommendations for obese preschoolers has the potential to improve rates of obesity detection. However, future research is needed to determine the best methods for allaying parent concerns about underweight and making salient to parents how failure to modify lifestyle behaviors in support of weight management for already obese preschoolers may result in greater negative health outcomes for their children (61,135).

Discussion

The high prevalence of obesity in preschoolers and mounting evidence that obese preschoolers are experiencing negative physical, behavioral, and emotional comorbidities of this chronic health condition makes evident the urgent need for effective weight management programs for this age group. The preschool years are an optimal time for obesity intervention because eating and activity patterns are being established during this stage of development and because parents have higher control over preschoolers' daily environment compared to during the school-age and adolescent years. Given the multitude of parent-reported and developmental barriers to modifying lifestyle behaviors for preschoolers, efficient and efficacious interventions comprised of only the most potent treatment components and sequence of components are needed to lessen parent burden and decrease barriers to treatment success. Current pediatric obesity treatment interventions (including those for preschoolers) are generally comprised of multiple components that target diet (decrease high-calorie/low nutrient—dense foods and beverages, increase low-calorie/nutrient-dense foods and beverages) and activity (decrease SA and increase PA). While inclusion of each of these components makes intuitive sense, the differential effects of each component on weight management during the preschool years are not known. Studies are needed to identify the most potent components of interventions as well as the most optimal combination and sequence of components for improving weight outcomes among preschoolers.

MOST (9) is an efficient and innovative research approach that could achieve these goals. This review completed the first phase of applying MOST to designing interventions for preschool obesity by empirically identifying the following components that show promise in effective weight management for preschoolers: decreasing preschoolers' screen use (29,92,97) consumption of high-calorie/high-fat foods and beverages (21,24,26,82,113), vigorous PA (71,72,77,78,80), increasing sleep (27,29,103,104,105,106), modifying parent feeding approaches to be consistent with elements of an authoritarian and authoritative feeding styles (52,53,59), and modifying general parenting to be consistent with an authoritative style (54,60), all within the context of family-based, behavioral intervention (81,82,113).

In the second phase of MOST, a series of randomized experiments would be conducted to discern which of the aforementioned components, or combination of components, is most causally related to weight management for preschoolers. While different experimental designs may be used, a factorial experimental design permits estimation of interactions between components and help to identify synergistic or dampening effects when components are combined. For example, decreasing caloric intake is associated with weight loss in preschoolers (82,113). However, studies have not been designed to test which dietary behavior (s) (e.g., decreasing high-calorie/high-fat foods and beverages, increasing FVs, or both) is most important to successful weight management. Each of these dietary recommendations may be effective independently but effects may be dampened when combined.

The third phase of MOST includes refinement of only effective intervention components identified during the second phase. Randomized experimentation is again implemented to discern optimal levels of these components (e.g., specific amount of a behavior included in a recommendation or the duration of the entire treatment) and the extent to which optimal levels vary depending upon individual or group characteristics (e.g., ethnicity, SES, child stage of development) (9). For example, if completing 60 min of moderate activity was deemed an effective component of weight management for preschoolers in phase 2, the refining phase might consist of testing whether the same effect could be achieved by adhering to this goal 5 days per week instead of 7. Furthermore, this phase may also include examination of whether the same activity recommendation is appropriate for 2–3 year olds compared to 4–5 year olds.

The outcome of the first three phases of MOST is optimization of an intervention that includes empirically selected components that are most causally related to weight management in preschoolers. The final phase of MOST would consist of conducting a standard RCT to compare the effectiveness of the optimized intervention to either standard care or an alternative established intervention. If the new, optimized intervention is more effective, then researchers can begin the next phase of translational research by testing the effectiveness of the intervention in real-world settings where obese preschoolers are most likely to be seen for treatment including primary and tertiary care settings.

In summary, research examining behavioral risk factors specific to obesity in preschoolers and effective weight management programs for this age group is limited but growing. Being overweight and obese as a preschooler is associated with greater SSB intake, watching more television, having a television in a child's bedroom, getting inadequate sleep, and low rates of PA (particular higher intensity activities). Multiple barriers unique to the preschool years challenge caregivers' efforts to change dietary and activity patterns in this age group. Parental motivation to address these barriers and make lifestyle changes to promote weight management may be compromised by family and ethnic norms associated with weight and weight-related behaviors, the extent to which parents perceive their child's weight to be problematic, and parental willingness to set limits and follow-through on changes to diet and PA. Prevention programs have yielded modest success in slowing weight gain in preschoolers across all weight categories, but may not be sufficient to address established obesity in preschoolers. Direct intervention to reduce established obesity, on the other hand, has more consistent promise. Application of innovative treatment research strategies such as MOST have the potential to advance the obesity treatment literature by identifying the behavioral correlates of obesity in preschoolers as well as the most effective intervention strategies for weight management in this age group.

ACKNOWLEDGEMENT

This review was supported by grants K24 DK 059492 (L.J.S.) and T32 DK063929 (Scott Powers, PhD, ABPP) from the National Institutes of Health.

DISCLOSURE

The authors declared no conflict of interest.

Ancillary