Maternal prompts to eat, child compliance, and mother and child weight status
Article Outline
Objective
To determine if there is a relationship between maternal prompting to eat, child compliance, and mother and child weight.
Study design
Seventy-one 3- to 6-year-old children and their mothers were videotaped tasting four foods (two familiar, two novel). Maternal prompts to eat and the child’s compliance with the prompts were coded. Multiple logistic regression evaluated demographic, anthropometric, and food characteristics that predicted prompting and compliance; and demographic, behavioral, and food characteristics that predicted child body mass index Z score (BMIz) in the children of obese and non-obese mothers.
Results
Obese mothers did not prompt more than non-obese mothers, but children of obese mothers were more compliant (70.2% ± 19.4 v 59.6% ± 21.2, P = .04). Low maternal education, a novel food, and younger child age predicted prompting. Maternal obesity, a familiar food, and older child age predicted compliance. In children of obese mothers, low maternal education, more prompts to eat novel foods, fewer prompts to eat familiar foods, and fewer child bites of familiar foods predicted child BMIz (R2 = 64%). In children of non-obese mothers, none of the covariates predicted child BMIz.
Conclusions
Children of obese mothers may be more responsive to environmental cues to eat.
Abbreviations: AIC, Akaike Information Criteria , BMI, Body mass index , BMIz, Body mass index Z score , CDC, Centers for Disease Control
The prevalence of childhood obesity has increased significantly in the past 20 years,1 with growing disparities based on race and socioeconomic status.2 Although both genetics and environment contribute to obesity risk, the rapid secular increase in prevalence is driven by environmental factors that increase caloric consumption and reduce caloric expenditure. Parents are in a key position to shape the environments of children, and increasing interest has focused upon the role of parenting behaviors in contributing to obesity risk. Maternal feeding practices have received particular attention as a potential risk factor for childhood obesity.3, 4 Feeding practices may be broadly conceptualized as either restricting or promoting a child’s intake. Restrictive maternal feeding practices, such as limiting consumption of or access to particular foods, have been associated with greater adiposity and more eating in the absence of hunger in girls.5, 6 Few studies have addressed the impact of maternal feeding practices that promote intake, such as prompting children to eat, on eating behavior or adiposity. The results of these studies have been contradictory4, 7, 8, 9 with the initial study in young children showing a robust positive association between maternal prompting and child adiposity,7 but subsequent studies being unable to demonstrate an association.8, 9
Although previous studies have evaluated the relationship of maternal feeding practices with children’s eating behaviors and overweight risk,10 none have evaluated the potential relationships between maternal prompting, the child’s compliance with prompts, and mother and child weight status in a healthy, diverse, sample of 3- to 6-year-old children using direct observation of behavior. The present study therefore seeks to address three objectives: (1) to determine if maternal prompting to eat or the child’s compliance with the prompts differs by maternal obesity status; (2) to identify demographic factors and food characteristics associated with maternal prompting and child compliance; and (3) to determine if prompting and compliance are associated with the child’s weight status. We hypothesized that obese mothers and mothers with lower education would prompt their children more to eat, and that children whose mothers prompted them more to eat would have a higher mean body mass index (BMI).
Methods
Sample
Three- to six-year-old children and their mothers were recruited through postings in diverse communities. Exclusion criteria included food allergies, serious health conditions that could affect appetite or eating, or not being conversant in English. Mothers and children fasted for 2 hours before the study. Only one child per enrolled family was allowed to participate, and families were compensated $20. Written informed consent was obtained. This study was approved by the University of Michigan Medical School Institutional Review Board.
Mothers provided child age, sex, race/ethnicity (white or not white), and maternal education (<4-year college degree versus ≥4-year college degree). At the conclusion of the study protocol, mother and child heights were measured without shoes on a wall-mounted measuring tape and weights were measured without shoes or heavy clothing on a scale with accuracy to 0.1 kg (Taylor Precision). BMI was calculated and child BMIz was derived based on the Centers for Disease Control (CDC) growth charts.11 This age range is the period of adiposity rebound, during which a child’s BMI declines to a nadir and then begins to increase again.12 The significant changes in BMI, which also differ by sex, necessitate standardizing raw BMI to a value reflecting standard deviations from the mean for a particular age and sex (a Z score). Mothers were categorized as obese by the CDC definition (BMI ≥30).
Food Presentation Protocol
Mother and child were seated at a table in a quiet room without distractions. Instructions were directed equally at both mother and child, who were told that they would be videotaped while sampling foods. They were told that they did not need to try the foods if they did not want to, but that we were interested in their opinions of the foods. Four foods were presented individually and sequentially, at which time each was placed on the table equidistant between mother and child and a brief verbal description, as outlined below, was provided. The order of the food presentations was randomized for all children. The researcher left the mother-child pair in the room with each food for 4 minutes, after which the food was removed and the next food presented. After each food, the researcher briefly interviewed the mother and child about their opinions of the foods, recording their comments with paper and pencil to emphasize to the mother and child the researcher’s interest in their opinions of the foods. Our intent in recording these responses was to remove some of the family’s focus from the videotape data, by suggesting that the primary study outcome of interest to the researchers was their response to the verbal inquiries about their opinions of the foods. Food was weighed on a scale (Salter) with accuracy ± 1 g before and after presentation.
The four foods differed in familiarity and taste (sweet or salty), which were chosen as the two factors on which to vary the foods because these characteristics are primary predictors of children’s food preferences.13, 14 Sweetness and saltiness are empirically palatable to humans,15 and were selected as the primary tastes on the premise that children would be most likely to eat these foods, even if unfamiliar. The novel foods were selected to closely match the familiar foods in terms of food category and sensory characteristics. Snack foods were chosen as the food stimuli for this study on the premise that these types of foods may be the focus of the greatest variability in maternal behavior during eating in obese versus non-obese mothers. The salty familiar food was a 30-g (150 kcal) bag of potato chips (Lay’s Original®). The sweet familiar food was a 45-g (150 kcal) commonly commercially available sweet, cream-filled sponge cake (Hostess Twinkie®). The salty novel food was a 30-g (140 kcal) bag of vegetable chips commercially available as a snack akin to potato chips (Terra Chips Original Flavor®). The sweet novel food was a 55-g (230 kcal) traditional, sweet Chinese moon cake. Descriptive information for the two novel foods was given during their presentation. For the vegetable chips, mother-child pairs were told, “Instead of chips made from potato, these are made from a mix of exotic vegetables: taro, sweet potato, yuca, batata, parsnip, and ruby taro.” For the moon cake, they were told “This is eaten during the Chinese Moon Festival. There are many different kinds. This one is a lotus paste moon cake. Lotus is a type of flower.” Mothers reported familiarity with each food at the conclusion of the protocol, and responses confirmed correct categorization of foods as novel or familiar to this sample (data available upon request).
Behavior of each mother-child pair was coded by two independent trained observers blind to study hypotheses. Inter-rater reliability, as indexed by intra-class correlation coefficients or Cohen’s κ as appropriate, exceeded 0.70 for all measures. Bites of each food by both mother and child were coded. Prompting was defined as the sum of physical encouragements, verbal encouragements and food offers, defined per methods in prior research.7, 8, 9, 16 Compliance with a prompt was defined per prior research methods as the child taking a bite (defined as food passing the lips) within 5 seconds of the prompt.17 Child compliance is presented as the proportion of prompts with which the child complied. Thus, a child may have taken few bites, but if all bites occurred within 5 seconds after a maternal prompt to eat, the child’s compliance would have been high. In contrast, a child may have taken many bites, but if the child rarely took a bite within 5 seconds after a maternal prompt to eat, the child’s compliance with prompts would have been low. Number of bites the child took and compliance with prompts were thus two independent scores.
Statistical Analysis
All statistical analyses were conducted using Statistical Analysis Systems, version 9.1 (SAS Institute, Cary, NC). Descriptive statistics were calculated using χ2 and t tests as appropriate to describe demographics of the sample in relation to maternal obesity. Univariate statistics were used to describe eating behavior by food type. Two-tailed P values are reported. An α level of 0.05 was used to determine statistical significance.
The data could be examined in two ways: summing behaviors across the 16 minutes of food exposures or evaluating behaviors within a single 4-minute food exposure, while accounting for the repeated measures within subjects. There are limitations and benefits to each method. Evaluating behaviors within each 4-minute food exposure allows one to statistically test the effect of food characteristics (eg, novelty) on behaviors, as well as their interactions. As will be shown, however, the variability in some of the behaviors within a single 4-minute period was limited. Therefore, summing behaviors over two foods within a category, or summing behaviors over all 16 minutes of food exposures, provided increased power to detect an effect. Each of these two methods of examining the data is used, as described below.
To determine if maternal prompting, child compliance, or consumption differed based on maternal obesity, we summed behaviors across all four food exposures and performed bivariate unadjusted analyses. For those variables that differed significantly based on maternal obesity in unadjusted analyses, we performed regression analyses testing for potential confounding by child age, sex, race, and maternal education.
Predictors of prompting and compliance were evaluated for each 4-minute food exposure individually, which allowed statistical testing of the effect of specific food characteristics on outcomes. Potential predictors of prompting were child sex, race, age, and BMIz; maternal obesity; maternal education; novelty and sweetness of the food; number of child bites; and number of mother bites. The number of bites the child took was tested as a predictor of prompting because we hypothesized that mothers would prompt more when children were eating less. The number of bites the mother took was tested as a predictor of prompting because we hypothesized that mothers who ate more of the foods would also prompt their children to eat more. Potential predictors of compliance were all of these except child bites because compliance is defined by the child’s bites. The number of bites taken by the mother was included in the model predicting compliance because we hypothesized that mothers who modeled eating while prompting would have children who complied more with the prompts. We used mixed models accounting for repeated measures within subjects to evaluate the unadjusted relationship of these potential predictors with prompting and compliance. To identify the most robust predictors of prompting and compliance for inclusion in adjusted models, Akaike Information Criteria (AIC) was used to identify the most parsimonious model with the best fit for predictors of each outcome. Once the base model was identified, repeated measures were accounted for in the model and interactions were tested.
We evaluated the unadjusted relationship between covariates and child BMIz in children of obese and non-obese mothers. In these models, predictors included behaviors summed across food exposures. We stratified these analyses on maternal obesity because maternal obesity has very strong predictive value for future overweight in children,18 and we therefore hypothesized that the predictors of child BMIz in children of obese mothers may differ from the predictors of child BMIz in the children of non-obese mothers.
We next created two linear regression models predicting child BMIz; one for children of obese mothers and one for children of non-obese mothers. AIC statistics were again used to identify the most parsimonious model with the best fit for the proposed predictors in each of the two models. Potential predictors included child age, race, sex, and maternal education, prompts to eat novel foods and familiar foods, compliance with prompts to eat novel and familiar foods, and the number of bites the child took of novel and familiar foods. Once the base model was identified, repeated measures were accounted for in the model and interactions were tested.
Results
Seventy-one mother-child pairs participated. Sample demographics are presented in Table I. The sample of children was about half male, slightly more than half of the children were white, and about half of mothers had an education that was less than a 4-year college degree. Twenty-six mothers (36.6%) and nine children (12.7%) were obese. There were no differences in child sex, race, age, or BMIz by maternal obesity status. There was a trend toward obese mothers having a lower educational level (P = .053). Patterns of consumption and behaviors by food type are described in Table II. Mother-child pairs ate about half of the food that was presented. Mothers prompted their children to eat 17.5 ± 12.1 times (range 1 to 35) over the entire session and children complied with 63.5 ± 21.1% of prompts (range 0 to 100%).
Table I. Demographic and anthropometric characteristics of children of obese and non-obese mothers
| Total n = 71 | Mother obese n = 26 | Mother not obese n = 45 | P value | |
|---|---|---|---|---|
| Child gender (n, %) | .10 | |||
 Male | 39(54.9) | 11(42.3) | 28(62.2) | |
| Female | 32(45.1) | 15(57.7) | 17(37.8) | |
| Child race (n, %) | .85 | |||
| White | 42(59.1) | 15(57.7) | 27(60.0) | |
| Not white | 29(40.9) | 11(42.3) | 18(40.0) | |
| Child age in years (Mean, SD) | 5.0(1.1) | 4.7(0.9) | 5.1(1.1) | .17 |
| Child BMIz (Mean, SD) | .50(1.12) | .76(1.07) | .34(1.14) | .13 |
| Child overweight (BMI ≥95th %ile)(n, %) | .60 | |||
| Yes | 9(12.7) | 4(15.4) | 5(11.1) | |
| No | 62(87.3) | 22(84.6) | 40(88.9) | |
| Maternal educational level (n, %) | .053 | |||
| ≥4-years college | 38(53.5) | 10(38.5) | 28(62.2) | |
| <4 years college | 33(46.5) | 16(61.5) | 17(37.8) |
Table II. Consumption data and eating behavior by food type
| Mean (SD) | |||||
|---|---|---|---|---|---|
| Familiar salty food | Familiar sweet food | Novel salty food | Novel sweet food | Total | |
| Amount consumed by mother-child pair (g) | 16.3(8.7) | 34.4(15.4) | 15.1(8.4) | 24.9(18.4) | 90.8(35.8) |
| Amount consumed by mother-child pair (kcals) | 81.5(43.5) | 114.6(51.8) | 70.5(39.2) | 104.1(76.9) | 371.2(149.2) |
| Maternal bites | 8.0(5.4) | 2.5(1.5) | 10.8(5.7) | 3.7(3.1) | 25.2(11.5) |
| Child bites | 18.4(10.0) | 7.0(4.9) | 11.8(7.9) | 3.8(3.8) | 41.2(19.0) |
| Maternal prompts to eat | 2.4(3.0) | 3.2(3.4) | 7.2(5.7) | 4.6(4.7) | 17.5(12.1) |
| Proportion of prompts with which child complied | 0.82(0.37) | 0.79(0.30) | 0.65(0.28) | 0.53(0.32) | 0.63(0.21) |
Obese mothers and their children ate significantly more of the familiar sweet food than did non-obese mothers and their children (40.7 ± 11.6 g v 31.0 ± 16.2 g, P = .01), but consumption of the other food types did not differ by maternal obesity. There were no differences in the number of bites taken by either mother or child of each of the 4 foods by maternal obesity status (data available on request). Prompting did not differ between obese and non-obese mothers (18.7 ± 13.8 v 16.9 ± 11.1, P = .55). However, the children of obese mothers were significantly more likely to be compliant with maternal prompts to eat (70.2% ± 19.4 vs 59.6% ± 21.2, P = .04). The difference in compliance by maternal obesity status was primarily driven by differences in compliance with prompts to eat novel foods. Children of obese mothers complied with 67.4 ± 20.4% of prompts to eat novel foods, compared with 51.6 ± 22.7% for the children of non-obese mothers (P = .005). There was no difference in child compliance with maternal prompts to eat familiar foods by maternal obesity status (83.0 ± 23.4% for children of obese mothers vs 80.0 ± 23.3% for children of non-obese mothers, P = .61). The associations between greater consumption of the sweet familiar food and maternal obesity, as well as greater child compliance (with overall prompts and for novel foods) and maternal obesity, were not eliminated in regression models controlling for child age, sex, race, and maternal education (data available on request).
The unadjusted relationships of the covariates with maternal prompting and child compliance within each 4-minute food exposure are shown in Table III; available at www.jpeds.com. Predictors of greater maternal prompting included younger child age, a novel food, and more bites of the food taken by the mother. Predictors of greater compliance by the child with the maternal prompts to eat included older child age and a familiar food. Adjusted models including only the combinations of covariates that accounted for the greatest proportion of the variance of each outcome, developed per our stated methods, are presented in Table IV. Predictors of prompting were low maternal education, a novel food, and younger child age (R2 = .19). The role of maternal education in predicting prompting differed significantly for novel versus familiar foods. Lower maternal education was only associated with more prompts for the novel foods (β = 2.1 ± .9, P = .04), but not for the familiar foods (β = .31 ± .51, P =.62). Predictors of compliance were maternal obesity, a familiar food, and older age of the child (R2 = .14). We tested all interactions among these covariates and found none to be significant.
Table III. Unadjusted relationships of covariates with prompting and compliance
| Maternal prompting | Compliance | |
|---|---|---|
| Slope (β) (SE) | Slope (β) (SE) | |
| Child gender female | 0.10(0.56) | −.03(.04) |
| Child race not white | 0.11(0.57) | 0.03(0.04) |
| Child age | −1.27(0.25)⁎ | .06(.02)⁎ |
| Child BMIz | 0.22(0.25) | −0.02(0.2) |
| Mother obese | 0.45(0.58) | .06(.04) |
| Low maternal education | 1.18(0.55) | .05(.04) |
| Novel food | 3.04(0.53)⁎ | −0.21(0.04)⁎ |
| Sweet food | −0.92(0.55) | −0.06(0.04) |
| Child bites | −0.06(.03) | - |
| Mother bites | 0.16(.05)⁎ |
⁎ P < .01. |
Table IV. Adjusted models for predictors of maternal prompting to eat and child compliance
| Characteristic | Maternal prompting⁎ | Compliance† | ||
|---|---|---|---|---|
| Slope (β) (SE) | P value | Slope (β) (SE) | P value | |
| Mother obese | - | - | .08(.04) | .06 |
| Low maternal education | 1.20±1.26 | .06 | - | - |
| Novel food | 3.04±.50 | <.0001 | −.21(.04) | <.0001 |
| Child age | −1.27±.24 | <.0001 | .07(.02) | .0005 |
| R2 | 0.19 | 0.14 | ||
⁎ Potential predictors of prompting were child sex, race, age, and BMIz, maternal obesity, maternal education, novelty and sweetness of the food, number of child bites, and number of mother bites. |
† Potential predictors of compliance included all potential predictors of prompting except child bites. |
The unadjusted relationships between covariates and child BMIz among children of obese and non-obese mothers are shown in Table V; available at www.jpeds.com. In children of obese mothers, younger child age, lower maternal education, and more prompts to eat novel foods predicted a higher BMIz in the child. None of the covariates in unadjusted analyses were significant predictors of BMIz in children of mothers who were not obese.
Table V. Unadjusted relationship between covariates and child BMIz among children of obese and non-obese mothers
| Characteristic | Children of obese mothers n = 26 | Children of non-obese mothers n = 45 |
|---|---|---|
| Slope (β) (SE) | Slope (β) (SE) | |
| Child gender female | 0.58(0.42) | 0.21(0.35) |
| Child race not white | 0.60(0.41) | 0.02(0.35) |
| Child age | −0.53(0.20)⁎ | −0.05(0.16) |
| Low maternal education | 1.24(0.36)⁎⁎ | 0.10(0.35) |
| Prompts to eat novel foods | 0.05(0.02)⁎⁎ | −0.03(0.02) |
| Prompts to eat familiar foods | 0.01(0.06) | −0.01(0.02) |
| Compliance with prompts to eat novel foods | −1.85(1.01) | .009(.77) |
| Compliance with prompts to eat familiar foods | −0.05(0.95) | 0.42(0.76) |
| Child bites of novel foods | −0.02(0.02) | .000(0.02) |
| Child bites of familiar foods | −0.03(0.01) | −.002(0.01) |
⁎ P < .05. |
⁎⁎ P < .01. |
The ability to predict BMIz in children of obese versus non-obese mothers in adjusted models differed significantly. The combination of a limited number of the potential covariates resulted in an adjusted model able to account for a significant proportion of the variance in BMIz of children of obese mothers. In children of obese mothers, predictors of child BMIz were low maternal education, more maternal prompts to eat novel foods, fewer maternal prompts to eat familiar foods, and fewer child bites of familiar foods (R2 = .64) (Table VI). In contrast, no combination of the potential covariates resulted in a model that could predict the BMIz of children of non-obese mothers with statistical significance.
Table VI. Adjusted model for predictors of child BMIz among children of obese mothers⁎
| Slope (β) ± SE | P value | |
|---|---|---|
| Low maternal education (versus high) | .95(.30) | .005 |
| Maternal prompts to eat novel foods | .05(.01) | .001 |
| Maternal prompts to eat familiar foods | −.14(.05) | .008 |
| Child bites of familiar foods | −.03(.01) | .047 |
| R2 | 0.64 | |
⁎ Potential predictors included child age, race, sex, and maternal education, prompts to eat novel foods and familiar foods, compliance with prompts to eat novel and familiar foods, and the number of bites the child took of novel and familiar foods. |
Discussion
Our findings parallel prior work in several important ways. First, obese mothers in our study did not prompt their children to eat more than non-obese mothers, which is consistent with prior findings in large samples of self-report data.3, 4 In addition, mothers self-report applying less pressure to eat when their daughters are perceived as more overweight.5 We observed that when the child’s adiposity is greater, obese mothers prompt their children less to eat what might be perceived as familiar “junk food,” but prompt their children more to eat what might be perceived as healthier, novel food. Maternal pressure to eat was modified in our sample by maternal obesity status and by the type of food in question. Our study also adds observational data to corroborate work indicating that mothers with low educational levels report prompting their children to eat more.3
There are few prior observational studies of maternal prompting to eat in relation to mother or child weight status, and they have had contradictory results. All studies were limited by small, primarily white samples, and few overweight children. More parental prompting to eat has been very significantly associated with greater child relative weight in some studies,7, 16 moderately associated with lower BMI in others,19 and not associated with child weight status in others.8, 9 None of these studies evaluated the child’s response to the parental prompting to eat or the type of food involved with the prompts, which our findings suggest may be a significant limitation in their methodologies.
There are several limitations to our study. It is possible that our sample was self-selected to contain mothers who were particularly attuned to and interested in their child’s eating behavior. Thus, we may have included mothers who prompt children more to eat, or children who are particularly picky eaters. Although our sample was diverse and controlling for race/ethnicity and maternal education did not eliminate the observed associations, our sample was relatively small. Maternal obesity is so highly associated with lower education and minority ethnicity that it is difficult to disentangle whether observed differences based on maternal obesity are because of the independent effect of maternal obesity, or because of the confounders of race and socioeconomic status.3 A larger sample with greater power would allow further confirmation. The number and types of foods used as stimuli was also small, and therefore it is unknown if the observed behaviors would extrapolate to other types of foods. It would be of particular interest to determine if the findings extrapolate to behaviors with vegetables. We also did not have consumption data in grams for mother and child individually, which would be of interest in future studies. Finally, our methodology of direct observation in the laboratory setting allowed tighter experimental control, but the laboratory setting may have altered behavior in a way that the natural environment would not. As with all studies of this nature, there are strengths and weaknesses to both direct observation in a laboratory setting with tight experimental control, as well as observation in natural environments. Disentangling the relationship between eating behavior and weight status will ultimately require contributions from both methodologies.
Our study contributes observational data to the growing body of questionnaire-based research on maternal feeding behaviors and child weight status. It is as yet unknown if the children of obese mothers are more responsive to all environmental prompts to eat (ie, peer prompts in the lunchroom or television commercials), or if the prompts of obese mothers differ qualitatively from the prompts of non-obese mothers. A growing body of evidence suggests that maternal feeding behaviors are related to child obesity risk. Just as restrictive and controlling maternal feeding practices may contribute to obesity risk, so may prompting, particularly in children of obese mothers. Further work is needed to determine the developmental underpinnings of this phenomenon and the limits of its effect.
We would like to thank Niko Kaciroti, PhD, for his biostatistical support, as well as Howard Bauchner, MD, Susana Patton, PhD, and Kyung Rhee, MD, MPH, for their helpful suggestions on earlier versions of the manuscript.
References
- . Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002 . JAMA . 2004;291:2847–2850
- . Epidemic increase in childhood overweight . JAMA . 2001;286:2845–2848
- Maternal feeding practices and beliefs and their relationships to overweight in early childhood . J Dev Behav Pediatr . 2001;22:391–408
- . Parental feeding style and the inter-generational transmission of obesity risk . Obes Res . 2002;10:453–462
- . Learning to overeat (maternal use of restrictive feeding practices promotes girls’ eating in the absence of hunger) . Am J Clin Nutr . 2003;78:215–220
- . Eating in the absence of hunger and overweight in girls from 5 to 7 years of age . Am J Clin Nutr . 2002;76:226–231
- . The effects of parental influences on children’s food intake, physical activity, and relative weight . Int J Eat Disord . 1986;5:335–346
- . Can mothers influence their child’s eating behavior? . J Dev Behav Pediatr . 1999;20:88–92
- . Relations between parental mealtime practices and children’s food intake . Appetite . 1994;22:245–257
- . Parent-child feeding strategies and their relationships to child eating and weight status . Obes Res . 2004;12:1711–1722
- . CDC Growth Charts . Advance Data . 2000;314:1–28
- . Early adiposity rebound and the risk of adult obesity . Pediatrics . 1998;101:E5
- . Dimensions of preschool children’s food preferences . J Nutri Educ . 1979;11:77–80
- . Preschool children’s food preferences and consumption patterns . J Nutr Educ . 1979;11:189–192
- . Preferences for sweet and salty in 9- to 15-year-old adult humans . Science . 1975;190:686–687
- . Parental influences on children’s eating behavior and relative weight . J App Behav Anal . 1983;16:371–378
- Parent and child mealtime behavior in families of children with cystic fibrosis . J Pediatr . 2000;136:195–200
- . Predicting obesity in young adulthood from childhood and parental obesity . N Engl J Med . 1997;337:869–873
- . BEACHES (an observational system for assessing children’s eating and physical activity behaviors and associated events) . J App Behav Anal . 1991;24:141–151
The American Heart Association had no role in (1) the study design; (2) the collection, analysis, and interpretation of data; (3) the writing of the report; or (4) the decision to submit the paper for publication. The first draft of the manuscript was written by Dr Lumeng, and no form of payment was given to anyone to produce this manuscript.Supported by the American Heart Association Fellow-to-Faculty Transition Award 0275040N to Dr Lumeng and in part by K24 HD 042489 to Dr Bauchner.
PII: S0022-3476(06)00308-8
doi:10.1016/j.jpeds.2006.04.009
© 2006 Mosby, Inc. All rights reserved.
