Parent Overweight Predicts Daughters’ Increase in BMI and Disinhibited Overeating from 5 to 13 Years*

Social and Behavioral

Parent Overweight Predicts Daughters’Increase in BMI and Disinhibited Overeatingfrom 5 to 13 YearsLori A. Francis,*‡ Alison K. Ventura,†‡ Michele Marini,‡ and Leann L. Birch†‡

AbstractFRANCIS, LORI A., ALISON K. VENTURA, MICHELEMARINI, AND LEANN L. BIRCH. Parent overweightpredicts daughters’ increase in BMI and disinhibitedovereating from 5 to 13 years. Obesity. 2007;15:1544–1553.Objective: To assess whether parental overweight status anddisinhibited overeating are predictive of daughters’ accel-erated weight gain and disinhibited overeating.Research Methods and Procedures: Participants were partof a longitudinal study of girls (N � 197) and their parents.Measured height and weight were used to calculate BMI[weight (kilograms)/height (meters)2]. Parents’ disinhibitedeating behavior was assessed using the Eating Inventory.Girls’ disinhibited eating was assessed using a behavioralprotocol to measure eating in the absence of hunger. Girlswere classified based on parental overweight at study entryinto four groups: neither, mother only, father only, or bothparents overweight.Results: Girls with both parents overweight had the mostrapid increases in BMI from 5 to 13 years of age; BMIincreased most slowly among the neither parent overweightgroup, with intermediate increases in BMI among motheronly and father only overweight groups. Daughters withboth parents overweight at study entry were eight timesmore likely to be overweight at age 13, controlling fordaughters’ weight at age 5. Girls with both parents over-weight had higher levels of disinhibited eating across all

ages than all other groups. Although girls in all parentalweight status groups showed increases in disinhibited eatingover time, girls with both parents overweight had largerincreases in disinhibited eating over time compared with allother groups.Discussion: Girls growing up in families differing in pa-rental overweight had divergent developmental trajectoriesfor BMI and disinhibited overeating. Findings reveal theneed to focus prevention efforts on overweight parents ofyoung children.

Key words: children, overweight, binge eating, weightgain, parents

IntroductionParents provide both genes and environments that may

promote behaviors associated with excessive weight gain inchildren (1–8). Additionally, obesity runs in families, andhaving obese parents increases obesity risk in children (9–15). Parental weight status also predicts tracking of child-hood overweight (11,16–18); an overweight child living ina family where one or more parent is overweight is likely toremain overweight throughout his or her childhood and intoadolescence and adulthood (16). Additionally, Treuth et al.(19) reported that number of overweight parents predictedfat gain among normal weight girls.

Accelerated weight gain in infancy and early childhoodpredicts later risk for overweight and obesity during child-hood, adolescence, and adulthood (20–24). Stettler et al.(23) found that accelerated weight gain in infancy had thestrength to predict obesity at age 20. Similar results werefound by Law et al. (25), providing evidence for linksbetween accelerated weight gain from 1 to 5 years of ageand higher weight status at age 22. Furthermore, acceleratedweight gain in childhood was linked to several negativehealth outcomes in adulthood, including coronary heartdisease, insulin resistance, hypertension, and type 2 diabetes(25,26).

Several studies have addressed the ways in which paren-tal overweight status may also influence the development of

Received for review August 10, 2006.Accepted in final form November 28, 2006.The costs of publication of this article were defrayed, in part, by the payment of pagecharges. This article must, therefore, be hereby marked “advertisement” in accordance with18 U.S.C. Section 1734 solely to indicate this fact.*Department of Biobehavioral Health, †Department of Human Development and FamilyStudies, and ‡Center for Childhood Obesity Research, The Pennsylvania State University,University Park, Pennsylvania.Address correspondence to Lori A. Francis, Department of Biobehavioral Health, 315 EastHealth and Human Development Building, The Pennsylvania State University, UniversityPark, PA 16802.E-mail: [email protected] © 2007 NAASO

1544 OBESITY Vol. 15 No. 6 June 2007

children’s eating behavior, dietary intake, and physical ac-tivity patterns (2,3,5,8,27,28). This study describes daugh-ters’ developing patterns of disinhibited overeating, an eat-ing style that has been consistently associated with weightgain, overweight, and obesity, among both children (29–31)and adults (32–36). Eating in the absence of hunger (EAH)1

is a behavioral measure of disinhibited overeating and ischaracterized as the tendency to consume large amounts ofpalatable foods in a short period of time in a fashion that isnot a response to hunger. Given its associations with weightgain and overweight, disinhibited eating has been proposedas a behavioral phenotype for obesity (37). Both genetic andenvironmental differences among families may shape theemergence of differences in eating styles such as disinhib-ited overeating. Additionally, overweight parents may beexhibiting distinctly different eating styles than normalweight parents and, in particular, may serve as models fortheir children’s disinhibited eating (38–40).

In this research, we address whether having one or twooverweight parents is predictive of accelerated weight gainand disinhibited eating during middle childhood and intoadolescence and whether disinhibited overeating aggregateswithin families. We hypothesized that overweight parentswould report higher levels of disinhibited overeating thanparents who are not overweight, and that, across middlechildhood, daughters growing up in families with both par-ents overweight would exhibit greater increases in BMI, bemore at risk for overweight, and have higher levels ofdisinhibited overeating than girls from families with one orneither parent overweight.

Research Methods and ProceduresParticipants

Participants included 197 non-Hispanic, white familiesliving in central Pennsylvania recruited as part of a longi-tudinal study of the health and development of young girls;the sample was not recruited based on weight status. Eligi-bility criteria for girls’ participation at the time of recruit-ment included living with both biological parents, the ab-sence of severe food allergies or chronic medical problemsaffecting food intake, and the absence of dietary restrictionsinvolving animal products. Families were recruited for par-ticipation in the study using flyers and newspaper advertise-ments. In addition, families with age-eligible female childrenwithin a five-county radius received mailings and follow-upphone calls (Metromail Limited, County Durham, UK).

At study entry, participants included 197 girls who aver-aged 5.4 � 0.4 (standard deviation) years of age and theirparents, of whom 192 families were reassessed 2 years laterwhen girls averaged 7.3 � 0.3 years. A third assessment

with 183 families was conducted 2 years later when girlsaveraged 9.3 � 0.3 years, followed by a fourth assessmentwith 177 families when girls averaged 11.3 � 0.3 years anda fifth assessment with 168 families when girls averaged13.3 � 0.3 years. Attrition was primarily caused by familyrelocation outside of the study area. No significant differ-ences were found between the initial weight statuses ofparticipants lost to follow-up (n � 29) and of participantsremaining in the study through age 13 (n � 168). Thisfinding was also true for girls’, mothers’, and fathers’ BMIat daughter age 5. The University’s Institutional ReviewBoard approved all study procedures, and parents providedconsent for their family’s participation before the study began.

MeasuresParental measures used in this study were collected at

study entry, when daughters were age 5; girls’ measureswere collected when daughters were ages 5, 7, 9, 11, and 13years.

Parental MeasuresBackground Characteristics. Mothers completed a Back-

ground Questionnaire, developed in our laboratory, thatassessed family background characteristics including com-bined family income and mothers’ and fathers’ years ofeducation. Parents chose from four categories for income:less than $20,000/yr, $20,000 to $35,000/yr, $35,000 to$50,000/yr, or $50,000�/yr. For education, parents listedwhether they had the following degrees/diplomas: highschool, associates, technical/vocational school, bachelors,masters, PhD, MD, JD, or other.

Weight Status and Parental Overweight Groups. Moth-ers’ and fathers’ height and weight were measured in thelaboratory and used to calculate BMI [weight (kilograms)/height (meters)2]. Because we were specifically interestedin examining differences between overweight and normalweight parents, we divided the sample into four groupsusing the widely accepted cut-off for defining overweight inmothers (BMI � 25 kg/m2) (41). Because our purpose wasto define groups in terms of parent overweight, we used aslightly more conservative BMI cut-off of �27 kg/m2 todefine overweight in fathers to reduce the likelihood ofmisclassifying fathers whose slightly elevated BMIs reflecthigher lean body mass rather than overfatness and to pro-vide more equal group sizes. This resulted in four groups:neither parent overweight (n � 54), mother only overweight(n � 41), father only overweight (n � 40), and both parentsoverweight (n � 62).

Disinhibited Eating. Parents’ disinhibited eating behaviorwas assessed using the dietary disinhibition subscale fromthe Eating Inventory developed by Stunkard and Messick(42). The dietary disinhibition subscale consists of 16 itemsthat measure parents’ reported disinhibition of cognitivecontrol of eating or the tendency to eat in response to1 Nonstandard abbreviation: EAH, eating in the absence of hunger.

Parental Overweight Predicts Daughters’ Weight Gain, Francis et al.

OBESITY Vol. 15 No. 6 June 2007 1545

external factors such as the presence of food and emotionalstates. Respondents select among true/false and Likert-typeresponse options. Total scores for each subscale are calcu-lated by summing subscale items; higher scores reflecthigher levels of reported disinhibited eating. Internal con-sistency scores for scale items on the disinhibition scalewere 0.83 and 0.73 for mothers and fathers, respectively.

Girls’ MeasuresWeight Status. Girls’ BMI scores were generated using

procedures described above for calculating BMI in parents.Height and weight were measured by a trained staff memberfollowing procedures described by Lohman et al. (43). Age-and sex-specific BMI percentiles were used to determine theprevalence of overweight in girls in this sample. Based onstandardized reference criteria (44), a BMI score corre-sponding to the 85th percentile was used to classify girls asat-risk-for-overweight; BMI percentile scores at or abovethe 95th percentile were used to classify girls as overweight.Because our research focus is on predicting adipositychange across middle childhood, we used BMI to assessadiposity change, based on the evidence presented by Coleet al. (45) and Field et al. (46), indicating that, althoughBMI z-score is the preferred measure for assessing adiposityat a single occasion, absolute BMI is the best for assessingchange over time because within-child variability over timeis related to the child’s level of adiposity. Because allchildren in the study were girls and are in the same agecohort, standardized scores or percentile ranking were notneeded, and BMI can be used as a relative measure ofadiposity within our sample. For descriptive purposes, theBMI data for each group at each age are presented using theCenters for Disease Control and Prevention growth charts(47), which provide information on how the groups’ datarelate to the BMI percentile data.

EAH. Girls’ disinhibited eating was assessed by measur-ing their intake of palatable foods in the absence of hunger.This procedure was developed in our laboratory and hasbeen previously described elsewhere (38,48,49). Briefly, 20minutes after eating and reporting that they were full, girlswere left alone in a room and were asked to taste and rate aset of palatable snack foods. Each girl was told that she hada few minutes alone and that after rating the foods she couldeat as much or as little of the foods as she desired. Largeamounts of the foods were available so that energy intakewas not limited by availability. Other available activitiesincluded listening to music, looking at books and maga-zines, and simple table games. Energy intake was calculatedfrom gram weights obtained by pre- and post-weighinggirls’ food intake; manufacturers’ information on energycontent of foods was used to determine total energy intake,which was used as the indicator of EAH. The variable ofinterest in this study was the percentage of energy con-

sumed from the total energy available during the period inwhich the girl had ad libitum access to the snack foods.

Statistical AnalysesData were analyzed using the SAS version 8.2 (50).

Descriptive information was generated for all variables ofinterest. Each outcome variable was assessed for normality.ANOVA with Tukey honestly significant difference posthoc tests were used to assess differences among groups onparent weight status and disinhibited eating at study entry.Pearson’s correlation coefficients were used to assess linearassociations between parent and daughter characteristics.

Predicting Girls’ BMI Change and Disinhibited EatingStyle over Time

A mixed modeling approach (Proc MIXED; SAS Insti-tute, Inc., Cary, NC) was used to assess the effects ofparental overweight on the patterns of change in daughters’BMI and disinhibited eating from 5 to 13 years of age.Mixed modeling is a useful tool for analyzing repeatedmeasures over time, and a major advantage is its ability toretain cases with one or more missing data points (51).Determination of model fit was based on several criteria: 1)model convergence, 2) a positive definite G matrix, and 3)statistical fit comparison based on the Akaike InformationCriteria (52). For the model predicting girls’ BMI changeover time, an unstructured covariance matrix was selected,as determined by the aforementioned model fit criteria;main effects of time, parent overweight group, and a parentoverweight group by time interaction were tested in thismodel. In the model predicting girls’ EAH over time, anauto-regressive covariance structure provided the best fitmodel; main effects of time, parent weight group, and aparent weight group by time interaction were tested in thismodel. In both models (BMI and EAH) inclusion of theinteraction of parental weight status group by time provideda test of the major hypothesis, because a significant inter-action effect provides evidence for a differential pattern ofchange (or rate of growth) over time for girls in differentparental weight groups.

Examining Girls’ Overweight Prevalence over TimeLogistic regression (Proc LOGISTIC; SAS Institute, Inc.)

was used to model the outcome of girls’ overweight statusat age 13 as a function of girl’s initial BMI and parentweight status at age 5. Parent overweight was identified asa classification variable in the model, and variables werearranged so that the three class contrasts made by SASwould be neither parent overweight vs. each of the otherthree parent weight status classes. Girl’s BMI at age 5 wasincluded as a covariate in the logistic regression analysis.



ResultsParental Characteristics: Differences across ParentWeight Status Groups

Parental overweight groups did not differ in family in-come or parent education levels; all groups had the majorityof members (75% of 85%) reporting an annual familyincome over $35,000, and the majority of parents withineach group (62% to 75%) reported at least some collegeeducation. There were also no differences among groups inparent age at study entry; at the time of recruitment, parents’mean age for the total sample was 35.4 � 4.8 and 37.4 �5.4 years for mothers and fathers, respectively.

Table 1 presents data on parental weight status and dis-inhibited eating scores at study entry. Because groups werecreated based on parental weight status, maternal BMI atstudy entry was significantly higher in families in whichmothers only or both parents were overweight comparedwith families in which fathers only or neither parent wasoverweight (p � 0.001). Similarly, fathers’ BMI at studyentry was highest in families in which fathers only or bothparents were overweight (p � 0.001). Additionally, thesepatterns held for mother and father disinhibition, as is dis-cussed in more detail below.

Overweight Prevalence in GirlsAmong the total sample, at 5, 7, 9, 11 and 13 years of age,

20%, 21%, 31%, 29%, and 26% of girls, respectively, wereclassified as at-risk-for-overweight (BMI � 85th percentile)and 6%, 11%, 14%, 14%, and 11% were classified asoverweight (BMI � 95th percentile). Girls’ mean BMIincrease corresponds to a mean weight gain of �35 kg(M�5–13 � 35.4 � 11.3 kg) over the 8-year period. Increasein BMI during the period from 5 to 13 years is normative,reflecting normal growth in height and weight, and does notnecessarily reflect increasing adiposity.

Parental Overweight and Daughters’ BMI Increase from5 to 13 Years of Age

Patterns of BMI change over time differed for girlsfrom families with neither, one, or both parents over-weight, as shown in Figure 1. Results of the mixed-modelanalyses revealed a significant parental weight group bytime interaction (p � 0.01). Thus, the increase in BMIfrom age 5 to 13 was above and beyond what would beexpected for normative growth among girls from familiesin which both parents were overweight (M�5–13 � 40.8 �13.3 kg) compared with girls with neither parent over-weight [M�5–13 � 29.1 � 6.2 kg; t(181) � 4.7, p �0.001]. Girls with two overweight parents also showedgreater increases in BMI than girls from families inwhich fathers only were overweight [t(177) � 2.7, p �0.01] and mothers only were overweight [t(178) � 2.0,p � 0.05]. Girls from families in which mothers onlywere overweight showed more rapid increases in BMIover time compared with girls from families in whichneither parent was overweight [M�5–13 � 34.9 � 10.9 kg;t(182) � 2.3, p � 0.05]; there were no significant dif-ferences between the fathers only and neither parentoverweight groups. Because the mother only and fatheronly groups did not differ from each other [t(179) � 0.7,p � 0.50], these analyses were also conducted combiningthe mother only and father only groups, allowing com-parisons among the families with neither, one, or bothparents overweight. In this case, the group by time inter-action was significant (p � 0.001), further showing thatthe neither, one, and two parent overweight groups dif-fered significantly from each other in BMI increases overtime. These analyses were also conducted using a BMI�25 kg/m2 to classify overweight in fathers, and theresults were unchanged.

Table 1. Mothers’ and fathers’ BMI and disinhibited eating scores at study entry, by parental overweight group

Parental overweight group

Neither parentoverweight(n � 54)

Mother onlyoverweight(n � 41)

Father onlyoverweight(n � 40)

Both parentsoverweight(n � 62)

Mothers’ BMI 21.7b* (1.4) 30.1a (4.6) 22.0b (1.7) 30.8a (6.3)Fathers’ BMI 24.5b (1.2) 24.6b (1.9) 31.1a (3.8) 31.1a (3.6)Mothers’ disinhibited eating score† 5.1b (3.3) 8.8a (4.0) 4.9b (3.3) 8.5a (3.3)Fathers’ disinhibited eating score† 4.3b (2.3) 3.6b (2.0) 5.7a (3.1) 5.5a (3.2)

* Means within a group followed by the same letter are not statistically different by Tukey’s honestly significant difference, p � 0.05.† Mothers’ and fathers’ disinhibited eating was measured using the disinhibition subscale from the Three Factor Eating Questionnaire (42),with scores ranging from 0 to 16.



Parental Overweight Predicting Girls’ OverweightPrevalence at Age 13

Results of the logistic regression analysis indicated that,compared with families in which neither parent was over-weight, girls with both parents overweight at study entrywere 8.1 times more likely to be overweight at age 13, whengirls’ initial BMI at age 5 was included in the model as acovariate (p � 0.05). The contrasts of mother only over-weight or father only overweight vs. neither parent over-weight were not statistically significant. Girls’ initial BMI atage 5 was a significant covariate (p � 0.001).

Associations between Parental Overweight andDisinhibited Eating

Group means for parents’ disinhibited eating at studyentry are shown in Table 1. As shown, mothers’ disinhibitedeating was significantly higher in the two groups of familieswith overweight mothers (both parents and mothers onlygroups) in comparison with the two groups in which moth-ers were not overweight (neither parents nor fathers onlygroups). A similar pattern was noted for fathers; fathers’disinhibited eating was significantly higher in the two fam-ily groups with overweight fathers (both parents and fathers

only) compared with families in which fathers were notoverweight (neither parent and mother only groups). Fa-thers’ disinhibition scores were consistently lower thanmothers’ disinhibition scores. Table 2 presents correlationsbetween parents’ disinhibited eating and girls’ EAH. Al-though mothers’ disinhibited eating was not significantlyrelated to daughters’ disinhibited eating style (EAH) whendaughters were age 5 or 7, mothers’ disinhibited eating wassignificantly and positively related to girls’ EAH at ages 9,11, and 13. Fathers’ disinhibited eating was not related togirls’ EAH at any point in time.

Parental Overweight and Daughters’ Disinhibited Eatingfrom 5 to 13 Years of Age

The overall pattern of change in girls’ disinhibited eatingfrom 5 to 13 years for the parent overweight groups isshown in Figure 2, which shows the percentage of totalavailable energy consumed in the EAH procedure for eachgroup across time. There was a significant interaction be-tween parental weight status group and time (p � 0.001) inthe mixed-model analysis, which indicates that there isevidence that the parental weight status groups differedsignificantly in daughters’ change in disinhibited eating

Figure 1: Girls’ BMI trajectories from 5 to 13 years of age by parental overweight group, plotted on the Centers for Disease Control andPrevention Growth Chart for Girls (47). Age 5, n � 197; age 7, n � 192; age 9, n � 183; age 11, n � 177; age 13, n � 168.



over time. However, analysis of the data omitting age 5reveals that the interaction term is no longer significant (p �0.3470), indicating that the interaction in the complete anal-ysis is caused solely by the relative shifting of means fromage 5 to age 7. Examination of the relationships in Figure 2,although the lines do cross, most clearly shows a relativelysimilar pattern over time for the different parent overweightgroups. Therefore, we examined the main effects, and therewas a significant main effect for parental weight statusgroup (p � 0.01), indicating that there were significantdifferences among groups in girls’ level of disinhibitedeating. Post hoc analyses revealed that girls with both par-ents overweight consumed a higher percentage of calories

during the EAH procedure across all ages than all othergroups. The effect of time was significant (p � 0.001),showing that daughters’ disinhibited eating increased overtime for all groups. Additionally, post hoc analyses revealedthat girls with both parents overweight had larger increasesin disinhibited eating over time compared with all othergroups (p � 0.05). As was the case with the model predict-ing girls’ BMI over time, results were unchanged when weused BMI �25 kg/m2 to classify overweight in fathers.

DiscussionResults from this study revealed that during childhood

and into early adolescence, girls developing in families

Table 2. Correlations between parents’ disinhibited overeating style at study entry and girls’ disinhibitedovereating across time*†

Daughters’ EAH

Age 5 Age 7 Age 9 Age 11 Age 13

Mothers’ disinhibited eating �0.12 0.03 0.24‡ 0.23‡ 0.19§Fathers’ disinhibited eating 0.07 0.11 0.12 �0.01 �0.02

EAH, Eating in the Absence of Hunger.* Mothers’ and fathers’ disinhibited overeating style was measured using the disinhibition subscale from the Three Factor EatingQuestionnaire (42).† Daughters’ disinhibited overeating was measured using the EAH protocol.‡ p � 0.01.§ p � 0.05.

Figure 2: Girls’ disinhibited eating from 5 to 13 years of age in the EAH protocol by parental overweight status groups. Means representthe percentage of total energy consumed from the total energy available during the ad libitum intake period. Plotted values are mean �standard error. Age 5, n � 197; age 7, n � 192; age 9, n � 183; age 11, n � 177; age 13, n � 168.



differing in parental overweight had divergent trajectoriesof weight gain and disinhibited overeating. The data pro-vided support for the primary hypothesis that, relative togirls without overweight parents, girls living in familieswith overweight parents had greater BMI increases acrossthis period. Additionally, daughters’ BMI increase from 5 to13 years, as well as their risk for becoming overweight, wasrelated to the number of overweight parents. Relative togirls with neither parent overweight, girls who had twooverweight parents were 8.1 times more likely to be over-weight at age 13. Overweight parents also reported higherlevels of disinhibited eating than normal weight parents, andpatterns of increase in girls’ disinhibited overeating duringthis same period of development also differed across paren-tal weight status groups. At age 5, there were no systematicdifferences in BMI or disinhibited eating among girls, butdiverging trajectories for BMI and disinhibited eating pro-duced differences in both outcomes over time. Effects ofparent overweight seem to be additive; girls with two over-weight parents had the greatest gains in BMI and increasesin disinhibited eating from 5 to 13 years of age.

The results of this study corroborate and extend previousstudies reporting that having two overweight parents ele-vates children’s risk for rapid weight gain and becomingoverweight during childhood and that having one parentoverweight also elevates risk, relative to having neitherparent overweight (9,11,16,21). In Figure 1, the BMI datafor the four groups are imposed on the Centers for DiseaseControl and Prevention BMI percentiles growth charts (47)to show the differing patterns of change across groupsagainst the backdrop of the reference data. The mean BMIof girls with neither parent overweight consistently tracksjust above the 50th percentile from 5 to 13 years of age. Incontrast, the BMI trajectory among girls with both parentsoverweight crosses percentiles, from 5 to 9 years of age,beginning at the 75th percentile, approaching the 90th per-centile by age 9, and remaining at that level at 11 and 13years of age. The trajectory of upward crossing of BMIpercentiles, as shown by the both parent overweight groupin this study, has been identified by the American Academyof Pediatrics as indicative of growth patterns placing chil-dren at higher risk for childhood overweight and obesity(53). These data, revealing that girls in this group were morethan eight times more likely to become overweight, provideadditional confirmation on this point.

Previous research has shown that disinhibited eating isconsistently related to overweight among adults (29–33)and children (34–36), and these findings provide additionalevidence for associations among disinhibited eating andelevated weight status among young girls. In addition, thesefindings revealed that, over time, disinhibited overeatingincreased among all groups, suggesting that such increasesmay be normative. These findings revealed group differ-ences in patterns of change in BMI and differences in

disinhibited eating; daughters of two overweight parentshad significantly greater increases in BMI, significantlygreater increases in disinhibited eating, and greater levels ofdisinhibited eating than girls with one or neither parentoverweight. However, this pattern of association is notsufficient to causally link disinhibited eating to greaterincreases in BMI, given the lack of an experimental design.We have proposed elsewhere that restrictive parenting prac-tices, which tend to be used by white, middle class parentswith daughters who are more overweight, can promotedisinhibited overeating. However, influence in parent-childdyads is bi-directional, suggesting that children’s weightstatus can be both a cause and a consequence of parentalfeeding practices (54); therefore, these findings raise thepossibility of causal links, which must be addressed infuture studies.

Mothers’ but not fathers’ disinhibited eating was associ-ated with daughters’ disinhibited eating, but these linksbetween mother’s and daughters’ disinhibited eating did notemerge until later childhood (ages 9, 11, and 13), perhapsbecause daughters’ disinhibited eating was lowest at age 5and increased among all groups from age 5 to 13. Althoughassociations between daughters’ and mothers’ disinhibitedeating have been previously reported (55,56), relationshipsbetween fathers’ and daughters’ have not been evaluated.Both Jacobi et al. (56) and Cutting et al. (55) found astronger association between mothers’ and daughters’ eat-ing behavior than fathers’ and daughters’ eating behavior.These findings reveal that mothers and fathers are bothinfluential in shaping daughters’ patterns of BMI changeand that the development of disinhibited eating among girlswas distinctly different only among families in which bothparents were overweight. These findings are consistent withan additive model of parental influence, with both mothers’and fathers’ genetic and environmental input contributing todaughters’ patterns of weight gain and the development ofdisinhibited eating, although perhaps in different ways. Ourfindings are consistent with relatively extensive findingsshowing maternal influence on daughters’ eating and weightstatus (31,38,55,57–62); there are very few data regardingfathers’ influence.

Research is needed to determine the mechanisms of in-tergenerational transmission of eating style from parent tochild. This sample was selected to include only families inwhich daughters were living with both biological families.In such cases, parents both provide genes and shape theenvironments in which these genetic predispositions may beexpressed. For example, there is experimental evidence thatovereating can be fostered in children by experience withrestrictive feeding environments (49) and that maternalmodeling may be important (55). As indicated above, thereis some evidence that disinhibited overeating has a geneticbasis and can serve as a behavioral phenotype for obesity(37,63). These findings are consistent with this perspective



and suggest how gene environment correlations shape thedifferential emergence of disinhibited eating among chil-dren growing up in families with and without overweightparents. For example, if overweight parents are geneticallypredisposed to disinhibited eating, and children share thosegenetic predispositions, parental displays of disinhibitedeating can foster the acquisition of disinhibited eating bychildren.

Although there is evidence for the heritability of disin-hibited eating behavior (63,64), de Castro et al. (64) foundthat 40% of the variance in disinhibited eating behavior wasexplained by a shared environmental component. Proven-cher et al. (63) found that only a small amount of thefamilial resemblance in eating behavior was caused bygenetics; family environment was particularly important fordietary disinhibition. A recent study by Faith et al. (37)tested the hypothesis that our measure of disinhibited eat-ing, EAH, was a genetic marker for childhood obesity. Theyfound some support for this hypothesis among preschool-aged children, but only in boys and not girls. They con-cluded that eating in the absence of hunger might be morehighly influenced by environmental as opposed to geneticfactors. In this study, consistent with Faith et al., we failedto note initial associations between daughters’ disinhibitedeating and weight status when girls were age 5, when levelsof disinhibited eating were lowest among girls. However,our results reveal that parallels in patterns of change forgirls’ BMI and disinhibited eating emerged as developmentproceeded, providing some evidence for disinhibited eatingas a behavioral phenotype.

Study limitations include a sample of primarily middle-income, well-educated, and non-Hispanic white familieswith daughters, and only families in which daughters wereliving with both biological parents. This precludes general-izing the findings to boys, to single parent families, or toother ethnic and income groups, including those with over-weight prevalence rates that are higher than this sample. Toclassify parents, we used parental BMI, based on measuredheights and weights. We also used BMI as our primarymeasure of daughter overweight. BMI is not a direct mea-sure of adiposity and may have led to misclassification ofsome parents; however, such misclassification would tendto reduce the likelihood of obtaining relationships amongparental weight status and daughters’ weight gain. Dataobtained from DXA on girls’ fat mass corroborated the BMIfindings among daughters (data not shown). The use oflongitudinal data is a positive feature, as was the inclusionof fathers in the research, because it allowed us to providesome initial information regarding paternal influence ondaughters’ developing weight status and eating style duringchildhood and early adolescence.

In conclusion, although no epidemiologic data are avail-able on the prevalence of overweight specifically amongparents in the United States, among 20- to 54-year-old

adults, an age range that would include most parents, abouttwo thirds are overweight. These estimates suggest that themajority of children in the United States today are living infamilies with one or two overweight parents; only a smallminority of children are growing up in families with neitherparent overweight. These findings suggest that children infamilies with overweight parents are at elevated risk forobesity and that this demographic pattern may further ac-celerate the obesity epidemic. These findings indicate that,at least for girls, the effects of parental overweight onweight gain and disinhibited eating seem to be additive;girls growing up in families with at least one overweightparent showed accelerated patterns of weight gain from age5 to 13 years relative to girls with neither parent overweightand that girls with two overweight parents show even moresubstantial weight gain over the same period.

The findings of this study corroborate previous research(10) indicating the elevated risk for childhood obesity con-ferred by having overweight parents and underscores theimportance of developing prevention approaches that targetoverweight parents with young children. To inform suchprevention efforts, research is needed to pinpoint the envi-ronmental and behavioral mediators of family resemblancesin adiposity. Such efforts would include exploring the fea-sibility of approaches to prevent the transmission of disin-hibited eating from parents to children. One approach couldfocus on reducing disinhibited overeating behaviors amongparents to reduce children’s exposure to parental models ofdisinhibited overeating. Alternative approaches include pro-viding anticipatory guidance for parents regarding ap-proaches to parenting and child feeding that can promotechildren’s continued responsiveness to hunger and satietycues in controlling energy intake and providing alternativesto parents’ use of feeding practices that can foster disinhib-ited overeating among children (38).

AcknowledgmentsThe services provided by the General Clinical Research

Center of the Pennsylvania State University were appreci-ated. This research was supported by NIH Grants HD32973and M01 RR10732.

References1. Patrick H, Nicklas TA. A review of family and social deter-

minants of children’s eating patterns and diet quality. J AmColl Nutr. 2005;24:83–92.

2. Cullen KW, Lara KM, de Moor C. Familial concordance ofdietary fat practices and intake. Family Commun Health.2002;25:65–75.

3. Wardle J, Guthrie C, Sanderson S, Birch L, Plomin R.Food and activity preferences in children of lean and obeseparents. Int J Obes Relat Metab Disord. 2001;25:971–7.

4. Trost SG, Kerr LM, Ward DS, Pate RR. Physical activityand determinants of physical activity in obese and non-obesechildren. Int J Obes Relat Metab Disord. 2001;25:822–9.



5. Davison KK, Birch LL. Child and parent characteristics aspredictors of change in girls’ body mass index. Int J ObesRelat Metab Disord. 2001;25:1834–42.

6. Maes HH, Neale MC, Eaves LJ. Genetic and environmentalfactors in relative body weight and human adiposity. BehavGenet. 1997;27:325–51.

7. Mamun AA, Lawlor DA, O’Callaghan MJ, Williams GM,Najman JM. Family and early life factors associated withchanges in overweight status between ages 5 and 14 years:findings from the Mater University Study of Pregnancy and itsoutcomes. Int J Obes Relat Metab Disord. 2005;29:475–82.

8. Davison KK, Birch LL. Obesigenic families: parents’ phys-ical activity and dietary intake patterns predict girls’ risk ofoverweight. Int J Obes Relat Metab Disord. 2002;26:1186–93.

9. Safer DL, Agras WS, Bryson S, Hammer LD. Early bodymass index and other anthropometric relationships betweenparents and children. Int J Obes Relat Metab Disord. 2001;25:1532–6.

10. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH.Predicting obesity in young adulthood from childhood andparental obesity. N Engl J Med. 1997;337:869–73.

11. Lake JK, Power C, Cole TJ. Child to adult body mass indexin the 1958 British birth cohort: associations with parentalobesity. Arch Dis Childhood. 1997;77:376–81.

12. Klesges RC, Klesges LM, Eck LH, Shelton ML. A longi-tudinal analysis of accelerated weight-gain in preschool-chil-dren. Pediatrics. 1995;95:126–30.

13. Danielzik S, Langnase K, Mast M, Spethmann C, MullerMJ. Impact of parental BMI on the manifestation of over-weight 5–7 year old children. Eur J Nutr. 2002;41:132–8.

14. Maffeis C, Talamini G, Tato L. Influence of diet, physicalactivity and parents’ obesity on children’s adiposity: a four-year longitudinal study. Int J Obes Relat Metab Disord. 1998;22:758–64.

15. Agras WS, Hammer LD, McNicholas F, Kraemer HC.Risk factors for childhood overweight: a prospective studyfrom birth to 9.5 years. J Pediatr. 2004;145:20–5.

16. Magarey AM, Daniels LM, Boulton TJ, Cockington RA.Predicting obesity in early adulthood from childhood andparental obesity. Int J Obes Relat Metab Disord. 2003;27:505–13.

17. Pietilainen KH, Kaprio J, Rasanen M, Winter T, RissanenA, Rose RJ. Tracking of body size from birth to late adoles-cence: contributions of birth length, birth weight, duration ofgestation, parents’ body size, and twinship. Am J Epidemiol.2001;154:21–9.

18. Williams S. Overweight at age 21: the association withbody mass index in childhood and adolescence and parents’body mass index. A cohort study of New Zealanders born in1972–1973. Int J Obes Relat Metab Disord. 2001;25:158 –63.

19. Treuth MS, Butte NF, Sorkin JD. Predictors of body fat gainin nonobese girls with a familial predisposition to obesity.Am J Clin Nutr. 2003;78:1212–8.

20. Toschke AM, Beyerlein A, von Kries R. Children at highrisk for overweight: a classification and regression trees anal-ysis approach. Obes Res. 2005;13:1270–4.

21. Reilly JJ, Armstrong J, Dorosty AR, et al. Early life riskfactors for obesity in childhood: cohort study. Br Med J.2005;330:1357–9.

22. Kinra S, Baumer JH, Smith GD. Early growth and child-hood obesity: a historical cohort study. Arch Dis Childhood.2005;90:1122–7.

23. Stettler N, Kumanyika SK, Katz SH, Zemel BS, StallingsVA. Rapid weight gain during infancy and obesity in youngadulthood in a cohort of African Americans. Am J Clin Nutr.2003;77:1374–8.

24. Tanaka T, Matsuzaki A, Kuromaru R, et al. Associationbetween birthweight and body mass index at 3 years of age.Pediatr Int. 2001;43:641–6.

25. Law CM, Shiell AW, Newsome CA, et al. Fetal, infant, andchildhood growth and adult blood pressure—a longitudinalstudy from birth to 22 years of age. Circulation. 2002;105:1088–92.

26. Barker DJP, Eriksson JG, Forsen T, Osmond C. Fetalorigins of adult disease: strength of effects and biologicalbasis. Int J Epidemiol. 2002;31:1235–9.

27. Burke V, Beilin LJ, Dunbar D. Family lifestyle and parentalbody mass index as predictors of body mass index in Austra-lian children: a longitudinal study. Int J Obes Relat MetabDisord. 2001;25:147–57.

28. Birch LL, Davison KK. Family environmental factors influ-encing the developing behavioral controls of food intake andchildhood overweight. Pediatr Clin North Am. 2001;48:893–907.

29. Shunk JA, Birch LL. Girls at risk for overweight at age 5 areat risk for dietary restraint, disinhibited overeating, weightconcerns, and greater weight gain from 5 to 9 years. J AmDietetic Assoc. 2004;104:1120–6.

30. Fisher JO, Birch LL. Eating in the absence of hunger andoverweight in girls from 5 to 7 y of age. Am J Clin Nutr.2002;76:226–31.

31. Johnson SL, Birch LL. Parents’ and children’s adiposity andeating style. Pediatrics. 1994;94:653–61.

32. Bellisle F, Clement K, Le Barzic M, Le Gall A, Guy-GrandB, Basdevant A. The Eating Inventory and body adiposityfrom leanness to massive obesity: a study of 2509 adults. ObesRes. 2004;12:2023–30.

33. Carmody TP, Brunner RL, Stjeor ST. Dietary helplessnessand disinhibition in weight cyclers and maintainers. Int JEating Disord. 1995;18:247–56.

34. Hays NP, Bathalon GP, McCrory MA, Roubenoff R, Lip-man R, Roberts SB. Eating behavior correlates of adultweight gain and obesity in healthy women aged 55–65 y. Am JClin Nutr. 2002;75:476–83.

35. McCrory MA, Suen VMM, Roberts SB. Biobehavioral in-fluences on energy intake and adult weight gain. J Nutr.2002;132:3830S–4S.

36. Provencher V, Drapeau V, Tremblay A, Despres JP, Le-mieux S. Eating behaviors and indexes of body compositionin men and women from the Quebec family study. Obes Res.2003;11:783–92.

37. Faith MS, Berkowitz RI, Stallings VA, Kerns J, Storey M,Stunkard AJ. Eating in the absence of hunger: a geneticmarker for childhood obesity in prepubertal boys? Obes Res.2006;14:131–8.



38. Birch LL, Fisher JO, Davison KK. Learning to overeat:maternal use of restrictive feeding practices promotes girls’eating in the absence of hunger. Am J Clin Nutr. 2003;78:215–20.

39. Fisher JO, Mitchell DC, Smiciklas-Wright H, Birch LL.Parental influences on young girls’ fruit and vegetable, mi-cronutrient, and fat intakes. J Am Diet Assoc. 2002;102:58–64.

40. Hanson NI, Neumark-Sztainer D, Eisenberg ME, Story M,Wall M. Associations between parental report of the homefood environment and adolescent intakes of fruits, vegetablesand dairy foods. Public Health Nutr. 2005;8:77–85.

41. NIH, National Institute of Heart, Lung, and Blood Insti-tute. Clinical Guidelines on the Identification, Evaluation,and Treatment of Overweight and Obesity in Adults. Wash-ington, DC: National Institutes of Health, 1998.

42. Stunkard AJ, Messick S. The three-factor eating question-naire to measure dietary restraint, disinhibition and hunger.J Psychosomatic Res. 1985;29:71–83.

43. Lohman TG, Roche AF, Martorell R, eds. AnthropometricStandardization Reference Manual. Champaign, IL: HumanKinetics Books; 1988.

44. Kuczmarski RJ, Flegal KM. Criteria for definition of over-weight in transition: background and recommendations for theUnited States. Am J Clin Nutr. 2000;72:1074–81.

45. Cole TJ, Faith MS, Pietrobelli A, Heo M. What is the bestmeasure of adiposity change in growing children: BMI, BMI%, BMI z-score or BMI centile? Eur J Clin Nutr. 2005;59:419–25.

46. Field AE, Laird N, Steinberg E, Fallon E, Semega-JannehM, Yanovski JA. Which metric of relative weight best cap-tures body fatness in children? Obes Res. 2003;11:1345–52.

47. Centers for Disease Control and Prevention. NCHS—United States Clinical Growth Charts. Hyattsville, MD: U.S.Department of Health and Human Services; 2005.

48. Fisher JO, Birch LL. Restricting access to foods and chil-dren’s eating. Appetite. 1999;32:405–19.

49. Fisher JO, Birch LL. Restricting access to palatable foodsaffects children’s behavioral response, food selection, andintake. Am J Clin Nutr. 1999;69:1264–72.

50. SAS Institute, Inc. Statistical Analysis Software. Cary, NC:SAS Institute; 2001.

51. Singer JD. Using SAS PROC MIXED to fit multilevel mod-els, hierarchical models, and individual growth models. JEduc Behav Stat. 1998;23:323–55.

52. Littell RC, Milliken GA, Stroup WW, Wolfinger RD,Schabenberger O. SAS for Mixed Models. 2nd ed. Cary, NC:SAS Institute; 2006.

53. Krebs NF, Jacobson MS. Prevention of pediatric overweightand obesity. Pediatrics. 2003;112:424–30.

54. Holden GW, Miller PC. Enduring and different: a meta-analysis of the similarity in parents’ child rearing. PsycholBull. 1999;125:223–54.

55. Cutting TM, Fisher JO, Grimm-Thomas K, Birch LL. Likemother, like daughter: familial patterns of overweight aremediated by mothers’ dietary disinhibition. Am J Clin Nutr.1999;69:608–13.

56. Jacobi C, Agras WS, Hammer L. Predicting children’sreported eating disturbances at 8 years of age. J Am AcadChild Adolesc Psychiatry. 2001;40:364–72.

57. Agras S, Hammer L, McNicholas F. A prospective study ofthe influence of eating-disordered mothers on their children.Int J Eating Disord. 1999;25:253–62.

58. Birch LL, Fisher JO. Mothers’ child-feeding practices influ-ence daughters’ eating and weight. Am J Clin Nutr. 2000;71:1054–61.

59. Fisher JO, Mitchell DC, Smiciklas-Wright H, Birch LL.Maternal milk consumption predicts the trade-off betweenmilk and soft drinks in young girls’ diets. J Nutr. 2001;131:246–50.

60. Hill AJ, Franklin JA. Mothers, daughters and dieting: inves-tigating the transmission of weight control. Br J Clin Psychol.1998;37:3–13.

61. Johnson RK, Panley CV, Wang MQ. Associations betweenthe milk mohters drink and the milk consumed by their school-aged children. Family Econ Nutr Rev. 2001;13:27.

62. Pike KM, Rodin J. Mother, daughters, and disordered eating.J Abnormal Psychol. 1991;100:198–204.

63. Provencher V, Perusse L, Bouchard L, et al. Familial re-semblance in eating behaviors in men and women from theQuebec Family Study. Obes Res. 2005;13:1624–9.

64. de Castro JA, Lilenfeld LRR. Influence of heredity ondietary restraint, disinhibition, and perceived hunger in hu-mans. Nutrition. 2005;21:446–55.



Parent Overweight Predicts Daughters’ Increase in BMI and Disinhibited Overeating from 5 to 13 Years*

Documents