RESEARCH

Research and Practice Innovations

Permissive Parental Feeding Behavior IsAssociated with an Increase in Intake of Low-Nutrient-Dense Foods among American ChildrenLiving in Rural CommunitiesERIN HENNESSY, PhD, MPH; SHERYL O. HUGHES, PhD; JEANNE P. GOLDBERG, PhD, RD; RAYMOND R. HYATT, PhD;

CHRISTINA D. ECONOMOS, PhD

hmpw

efndbmo

igfaateh

ss

ABSTRACTParents play an important role in shaping children’s eat-ing habits. Few studies have evaluated the influence ofboth parenting style and parenting practices on childoutcomes such as dietary intake. During spring 2007, 99parent–child dyads from four rural US areas participatedin this cross-sectional study. Child food intake was re-ported during two interviewer-administered, parent-as-sisted 24-hour recalls. Diet quality was defined as theaverage number of low-nutrient-dense (LND) foods con-sumed. Validated questionnaires were used to assess pa-rental feeding practices and feeding style. Pearson corre-lations identified relationships among child food intake,parental feeding style typologies, and covariates. Regres-sion analyses were used to predict child diet quality.Sixty percent of children and 76% of parents were over-weight or obese. A permissive feeding style, which is

E. Hennessy is a cancer prevention fellow, NationalCancer Institute, National Institutes of Health,Bethesda, MD; at the time of the study, she was a doc-toral candidate at the Friedman School of Nutrition Sci-ence and Policy, Tufts University, Boston, MA. S. O.Hughes is an assistant professor, US Department of Ag-riculture Agricultural Research Service Children’s Nu-trition Research Center, Department of Pediatrics, Bay-lor College of Medicine, Houston, TX. J. P. Goldberg isa professor, Friedman School of Nutrition Science andPolicy, Tufts University, Boston, MA. R. R. Hyatt is anassociate professor, Department of Public Health andCommunity Medicine, School of Medicine, Tufts Univer-sity, Boston, MA. C. D. Economos is associate director,John Hancock Research Center on Physical Activity,Nutrition, and Obesity Prevention, and associate profes-sor, Friedman School of Nutrition Science and Policy,Tufts University, Boston, MA.

Address correspondence to: Erin Hennessy, PhD,MPH, National Cancer Institute, National Institutes ofHealth, 6130 Executive Blvd, EPN 4087C, Bethesda,MD 20892-7335. E-mail: hennessye@mail.nih.gov

Manuscript accepted: July 8, 2011.Copyright © 2012 by the Academy of Nutrition and

Dietetics.2212-2672/$36.00

doi: 10.1016/j.jada.2011.08.030

142 Journal of the ACADEMY OF NUTRITION AND DIETETICS

ighly responsive to a child’s requests and sets few de-ands on him or her, was the most common (n�37)

arental feeding style. This feeding style was associatedith child intake of LND foods (r�0.3; P�0.001) and

moderated the relationship between parental feedingpractices and child intake of LND foods. In the presenceof a permissive feeding style, higher levels of monitoringwere associated with child intake of LND foods (��.69;P�0.05). Parental feeding style may alter the effective-ness of parental feeding practices on children’s food in-take. More research is needed to understand the parent–child feeding relationship in the context of parentalfeeding styles and practices.J Acad Nutr Diet. 2012;112:142-148.

US children consume too many sugar-sweetened bev-erages (1) and not enough fruits and vegetables (2).It is estimated that one third of children’s daily

nergy intake is composed of low nutrient-dense (LND)oods (defined as energy-dense foods that provide modestutritional value) (3,4). These trends are central in theevelopment of overweight among children (5) and maye particularly relevant for rural children who are esti-ated to have a 25% greater risk of being overweight or

bese compared to their metropolitan counterparts (6,7).An understanding of the modifiable risk factors to food

ntake and diet quality is essential to combating therowing child obesity epidemic. Focusing on the homeood environment is warranted because eating behaviorsre established early in life (8,9) and are linked to currentnd future weight status (10,11). Parents act as the nu-rition gatekeepers (12), directly determine their child’snvironment, and indirectly influence their behaviors,abits, and attitudes (13).Two areas of parenting warrant attention: parenting

tyle and parenting practice (14). Parenting style de-cribes how parents interact with their child (15). It is

often based on two dimensions of parental behavior: re-sponsiveness/nurturance to and demandingness/controlof the child (16). Combining the two dimensions results infour parenting style typologies: authoritarian, authorita-tive, permissive, and uninvolved (16,17). This generalparenting framework has recently been applied to thefeeding domain (eg, parental feeding styles) (18,19).

Parenting practices describe the behavioral strategies

that parents use to socialize their children (14,15). They

© 2012 by the Academy of Nutrition and Dietetics

iedfRtnsNpdeitpwie

om(aufarPFn5sft(ll

are embedded within the emotional climate of parentingstyle, less trait-like, and more responsive to contexts. Ithas been argued that parenting style changes the effec-tiveness of a particular parenting practice (15) and thistheory has recently been supported in the domain ofparent–child feeding (20). Although the true nature ofthe parent–child relationship is only fully understoodwhen both parenting practice and parenting style areconsidered, few studies have evaluated both constructswithin the child feeding domain (20,21).

The objectives of this study are to investigate the asso-ciation between parental feeding style and child foodintake among a diverse group of parent–child dyads liv-ing in low-income rural areas, evaluate whether parentalfeeding style or feeding practices were associated withchild diet quality, and determine whether parental feed-ing style moderated the relationship between parentalfeeding practice and child food intake. This study teststhe hypothesis that parental feeding style is positivelyassociated with the diet quality of children’s food intake.

METHODSStudy Design and PopulationThis cross-sectional study was conducted as part of alarger project and partnership between Tufts Universityand Save the Children USA. Save the Children partnerschools (grades kindergarten through five, n�83) werebroken down by region: Central Valley (California), Ap-palachia (Kentucky, Tennessee), Southeast (South Caro-lina, Georgia), and Mississippi River Delta (Mississippi,Arkansas). Rural was defined using the National Centerfor Education Statistics’ locale codes (22). One schoolfrom each region was randomly selected to participate.Participating schools had an average of 347 students, ofwhich approximately 98%�4.5% participated in the free-and reduced-price lunch program. Recruitment lettersand flyers were sent home to all 6- to 11-year-old childrenattending the four schools. A parent had to be either thebiological parent or legal guardian with whom the childlived. Children following special diets for medical reasonswere excluded.

Measurements were taken at each school on 2 separatedays (held 1 week apart) during spring 2007. All datawere collected during after school hours and took approx-imately 2 hours/day. The majority of participants in Cal-ifornia were of Spanish-speaking descent so all materialswere translated into Spanish and interpreters were onsite to provide assistance. Participants received a mone-tary gift card ($100 value) for completion of the study.The Tufts University Institutional Review Board ap-proved the study protocol and all participants providedwritten informed consent (children aged 7 years or olderalso provided assent).

MEASURESChild Dietary IntakeThe primary outcome of this study was collected throughtwo face-to-face, interviewer-administered, parent-as-sisted 24-hour dietary recalls using standard double-passprocedures (23,24). Recalls were collected approximately

1 week apart (range�5 to 7 days). All food and beverage

Jan

nformation was collected by hand using a paper datantry form (25). Interviewers followed a script that waseveloped by a trained dietetics practitioner (first author)or the study based on the Nutrient Data System foresearch (NDSR) software system (version 2006, Nutri-

ion Coordinating Center, University of Minnesota, Min-eapolis). All paper records were entered into NDSR ver-ion 2006 and final calculations were completed usingDSR version 2008. Recalls were screened to identifyhysiologically plausible dietary reports (26). This proce-ure compares reported energy intake with predicted en-rgy requirements, and examines the effects of excludingmplausible reports. Sex- and age-group–specific devia-ion cutoffs for reported energy intake in proportion toredicted energy requirements were calculated. Childrenere coded as having plausible records if reported energy

ntake was within two standard deviations of predictednergy requirements (27).Food groups were classified as belonging to one or more

f the five major food groups (ie, fruit, vegetable, dairy,eat, and grain), fats and oils, and an LND food group

4). Food variables were averaged across both recall daysnd operationalized as the number of times that partic-lar food group was mentioned during the recall. Thisorm was chosen rather than others because children areble to recall the foods they consume, but have difficultyeporting portion size (24,28,29).arent Feeding Styles and Feeding Practices. The Caregiver’seeding Styles Questionnaire (18) measures demanding-ess and responsiveness with each question scored on a-point Likert scale (never to always). Based on medianplits of these constructs, a four-level classification ofeeding style was derived representing authoritative, au-horitarian, permissive, and uninvolved feeding styles18). Description of the development, reliability, and va-idity testing of this instrument has been previously pub-ished (18,19). The Child Feeding Questionnaire (30)

identifies seven factors measuring parental attitudes, be-liefs, and practices about child feeding and obesity prone-ness. This study focused on the three feeding practicesubscales (restriction, monitoring, and pressure to eat),each scored on a 5-point Likert scale (higher scores indi-cating greater agreement with the construct). This in-strument has been widely used (30-33). Because theChild Feeding Questionnaire focuses on foods of low nu-trient quality (eg, sweets, snacks, and high-fat foods)regression models used only the LND food variable as theoutcome.Body Mass Index (BMI). Height and weight were obtained intriplicate for each dyad following standardized proce-dures (34). Height was measured without shoes to thenearest eighth of an inch using a portable stadiometer(Shorr Infant/Child/Adult Height/Length MeasuringBoard, Shorr, Olney, MD). Weight was measured in lightclothing to the nearest 0.1 lb on a portable digital scale(Befour PS-6600 Portable Scale, Befour, Inc, Saukville,WI). BMI was calculated from the average of the threebody weight and height measurements. BMI-for-age zscores and percentiles were calculated using the Statisti-cal Analysis Software program for the Centers for Dis-ease Control and Prevention Growth Charts (35). A childBMI �85th percentile and �95th percentile was consid-

ered overweight, whereas a child BMI �95th percentile

uary 2012 ● Journal of the ACADEMY OF NUTRITION AND DIETETICS 143

svr

was considered obese (36). Parent BMI �25 and �30 wasclassified as overweight, BMI �30 and �40 was classifiedas obese, and a BMI �40 was classified as extreme obe-sity (37).Other Demographic Data. Information on other potential co-variates was collected through a parent survey. Parentsreported their child’s date of birth, sex, and race/ethnic-ity, and their own age, sex, race/ethnicity, marital status,education level, family dynamics (number of siblings,number of adults in the home), country of birth, andnumber of years and/or months the respondent has livedin the United States.

Statistical AnalysisAnalyses were performed using SPSS (version 14.0, 2005,SPSS Inc, Chicago, IL) and SAS (version 9.1, 2004, SASInstitute Inc, Cary, NC). An � level of .05 was used for alltatistical tests. Descriptive statistics were calculated toerify normal distribution and no transformations wereequired. Independent samples t tests examined the dif-

ferences between boys and girls whereas analysis of vari-ance examined differences by race/ethnicity. The samplesize was determined to examine the association (r�0.39)between parental feeding style (authoritative) and childdietary intake (energy density). A sample size of n�47would allow adequate power (0.80) to detect this relation-ship. Two-sided Pearson correlations identified relation-ships among child dietary intake, feeding style typolo-

Table 1. Demographic characteristics of parent–(n�99) in a study to determine the influenceoutcomes such as dietary intake

Characteristic C

4SexMale 3Female 6EthnicityWhite 2African American 4Hispanic 2Education�High schoolHigh school�High schoolMarital statusMarried

4Age (y) 9�30 (n)30-39 (n)�39 (n)Body mass index 1Overweight/obese (%) 6

aParticipants lived in Mississippi (n�27), South Carolina (n�20April-June 2007.bSixteen participants completed the Spanish version of each scz score.

gies, and covariates.

144 January 2012 Volume 112 Number 1

Multiple linear regression models examined the asso-ciation between child LND food intake and each parent-ing behavior separately (feeding style and the three feed-ing practices). Indicator variables were computed for eachfeeding style typology and the feeding practice subscaleswere centered to allow for more interpretable coefficients(38). Moderated multiple regression analysis examinedthe effect of feeding style on the relationship betweenchild LND food consumption and each feeding practice.

Interaction terms were computed by multiplying eachcentered feeding practice score with each feeding styletypology indicator variable. Standard statistical testswere used to determine whether the incremental vari-ance (R2 change) explained by the interaction terms wassignificant. A post hoc analysis to probe for any signifi-cant moderation effect was conducted according to theprocedures described by Aiken and West (39) and Holm-beck (40).

RESULTSA total of 99 parent–child dyads participated in thiscross-sectional study. This full sample was used for allanalyses given that no differences were found betweenplausible (n�68) and nonplausible reporters (n�31) (datanot shown) (27,41).

Children were on average aged 9.0�1.5 years, mostlyfemale (n�60) with 60% classified as overweight or obese(Table 1). There was no difference in BMI z score by sex

dyads living in rural areas of the United Statesa

renting style and parenting practices on child

Parent

™™™™™™™™™™™™™™™™™ n ™™™™™™™™™™™™™™™™™3

495 (87% mothers)

294822b

253044

46™™™™™™™ mean�standard deviation ™™™™™™™3.5

214929

.9c 32.3�8.676

ucky (n�27), and California (n�25). Data were collected from

rt measure.

childof pa

hild

™™™

90

982

™™™.0�1

.2�00

), Kent

elf-repo

(P�0.17) or race/ethnicity (P�0.42). Most parents were

bbwi

wL(aP(aaifPi

(casc

sarttcpstcftm

iw

mothers (n�83) and the mean parental BMI was32.3�8.6 corresponding to 24% overweight, 33% obese,and 19% extremely obese. Table 2 shows the mean num-er of food groups as well as average energy intake acrossoth recall days. For parents, a permissive feeding styleas the most common (n�37) and an authoritative feed-

ng style was the least common (n�15).A permissive feeding style was positively associatedith energy intake (r�0.25; P�0.01), consumption ofND foods (r�0.29; P�0.001), sweetened beverages

r�0.21; P�0.03), fats and oils (r�0.28; P�0.01), meatnd beans (r�0.26; P�0.01), and milk intake (r�0.25;�0.01), but not with intake of whole grain foods

r��0.24; P�0.02). An uninvolved feeding style was neg-tively associated with energy intake (r��0.26; P�0.01)nd sweetened beverages (r��0.26; P�0.01). An author-tative feeding style was also inversely related with LNDoods (r��0.21; P�0.04) and fats and oils (r��0.23;�0.02). No relationships were found between an author-

tarian feeding style and any child diet variable.A permissive feeding style was positively associated

��.3) with a higher child intake of LND foods whileontrolling for covariates. This model (Table 3, Model 1)ccounted for 34% of the variation in child intake. Re-triction, monitoring, and pressure to eat were not asso-

Table 2. Child dietary intake and parental feedinrural areas of the United States (n�99)

Food item or practice

Total energyLow-nutrient-densea foodsFruitNo juiceVegetableDark green onlyGrainNo snack foodsWhole grains onlyMeat and beansMilkLow-fat milkFats and oilsSweetened beveragesFeeding practice (30)Restrictionb

Pressure to eatb

Monitoringc

Feeding style (18)AuthoritativeAuthoritarianPermissiveUninvolved

aDefined following the criteria set forth by Kant and colleagusweeteners: sugar, syrup, candy, carbonated and non-carbonadairy desserts: cookies, cakes, pies, pastries, ice cream, pudmiscellaneous: coffee, tea, condiments.bScored on a 5-point Likert scale (disagree�1 to agree�5).cScored on a 5-point Likert scale (never�1 to always�5).

iated with child intake. Table 3, Models 5 through 7,

Jan

how the results of the moderated multiple regressionnalyses. There was significant interaction between pa-ental feeding style and the feeding practices of restric-ion and monitoring (Table 3, Models 5 and 6). The rela-ionship between restriction and child LND foodonsumption was significant when in the presence of aermissive feeding style. The addition of this term re-ulted in an R2change of 0.06, F(3, 86)�2.83; P�0.04, andhis model explained an additional 38% of the variance inhild diet quality. Both a permissive and an uninvolvedeeding style moderated the relationship between paren-al monitoring and child LND food consumption. Thisodel resulted in an R2change of 0.16, F(3, 86)�8.41;

P�0.001 and explained an additional 40% of the variancein child diet quality.

The post hoc probing suggests that for the permissivefeeding style group there was little difference betweenlow and high restrictive feeding practices and child LNDfood intake (��.46; P�0.22). For the unpermissive group(ie, not permissive), higher restrictive feeding practiceswere associated with lower child LND food intake(���.86; P�0.05). In the presence of a permissive feed-ng style, higher levels of monitoring were associatedith a higher child intake of LND foods (��.69; P�0.05)

whereas the opposite relationship was true for the unper-

les and practices of parent–child dyads living in

hild Parent

™™™™™™™™ mean�standard deviation ™™™™™™3,712�3525.5�2.31.6�0.90.8�0.72.7�1.80.2�0.34.2�1.23.3�1.10.5�0.62.9�1.23.0�1.00.5�0.72.1�1.21.1�0.8

3.5�0.83.0�1.03.9�0.9

™™™™™™™™™™™™™™™™™™ n ™™™™™™™™™™™™™™315263721

visible fat: butter, oil, dressings, cream cheese and gravies;verages (including fruit drinks and sweetened tea); baked andand cheese cakes; salted snacks: potato, corn, tortilla chips;

g sty

C

41

4

es (4):ted bedings,

missive group (���.88; P�0.05). The post hoc probing of

uary 2012 ● Journal of the ACADEMY OF NUTRITION AND DIETETICS 145

c

cc(rgaTlsctc

the significant uninvolved�monitoring interaction termshowed no significant trends.

DISCUSSIONIn this population of rural children living in low-incomecommunities, higher consumption of LND foods, a corre-late of poor diet quality (4,42), was associated with apermissive parental feeding style.

This adds to the growing empirical evidence that beinga permissive parent in food-specific situations is associ-ated with poor diet quality (43) and higher weight among

Table 3. Regression models predicting the average number of low-nStates

Model��standarderror

Multiple linear regressionab

Model 1: Feeding styleAuthoritarian 1.2�.7Uninvolved .9�.7Permissive 1.3�.6Model 2: Restriction �.3�.3Model 3: Monitoring .2�.3Model 4: Pressure to eat .1�.3Moderated multiple linear regressionac

Model 5: Feeding style and restrictioninteraction

Restriction �1.3�.6Authoritarian 1.0�.7Permissive 1.2�.6Uninvolved .4�.7Authoritarian�restriction .3�.8Permissive�restriction 1.7�.7Uninvolved�restriction 1.1�1.0Model 6: Feeding style and monitoring

interactionMonitoring �.9�.6Authoritarian .6�.6Permissive 1.1�.6Uninvolved .6�.6Authoritarian�monitoring �.5�.7Permissive�monitoring 1.6�.7Uninvolved�monitoring 1.9�.9Model 7: Feeding style and pressure

to eat interactionPressure to eat �.0�.8Authoritarian 1.0�.7Permissive 1.2�.7Uninvolved .7�.7Authoritarian�pressure to eat �.3�.9Permissive�pressure to eat �.0�.8Uninvolved�pressure to eat �.2�.9

aReference group for feeding style (authoritative).bModels 1-4 adjusted for child characteristics (sex, age, race/ethnicity, body mass indecSignificant F change presented for Models 5-7 adjusting for child ethnicity, child body*P�0.05.**P�0.01.

hildren (44). A permissive feeding style may lead to c

146 January 2012 Volume 112 Number 1

hild-initiated snacking and higher energy intake be-ause the parent lacks control of the feeding situation44). This parental feeding style, characterized as highlyesponsive and less demanding, may be undesirableiven the current obesogenic environment that providesn abundance of nutrient-poor, high-energy food (43,45).his may be especially problematic for families living in

ow-income communities. For instance, parents in thistudy anecdotally expressed that they had limited finan-ial resources. Because of this constraint they often hado say no to their child’s requests for new items (such aslothing and toys), but food was the one area where they

t-dense foods reported by children living in rural areas of the United

andard � t P value Adjusted R2

0.34.2 1.8 0.07.2 1.2 0.23.3 2.0 0.05*.1 �1.0 0.31 0.31.1 0.8 0.57 0.22.0 0.2 0.60 0.22

0.38*

.5 �2.2 0.03*

.2 1.5 0.14

.3 1.9 0.06

.1 0.7 0.52

.1 0.4 0.73

.4 2.5 0.01**

.1 1.1 0.260.47**

.4 �1.6 0.12

.1 1.0 0.31

.2 1.9 0.07

.1 1.0 0.32

.1 �0.8 0.44

.4 2.5 0.01**

.3 2.2 0.03*0.22

1.0 �0.0 0.99.2 1.6 0.12.3 1.8 0.07.1 1.1 0.29.1 �0.3 0.77.0 0.0 0.98.0 �0.2 0.83

re) and parent characteristics (age, education, marital status).index, z score, and parent age.

utrien

St

�

�

�

�

�

���

x z scomass

ould more often say yes to requests (characteristic of a

ssf

permissive feeding style). Despite having good intentions,parents may be unaware that this behavior has unin-tended consequences with regard to their child’s foodintake.

The findings of this study and others (46-48) also lendsupport to Darling and Steinberg’s (15) model that par-enting style—and in this context parental feeding style—may change the effectiveness of particular parentingpractices on child outcomes. Much of the parent–childfeeding literature has shown that restrictive feedingpractices can increase preference for and intake of palat-able foods (14,49). These findings are consistent withattribution theory, which predicts that child-feedingstrategies that restrict children’s access to snack foodsactually make the restricted foods more attractive (50).However, this study suggests that restriction may oper-ate differently depending upon the feeding style of theparent. When a permissive parent either restricts or mon-itors his or her child’s intake, the actions appear to back-fire and lead to the child eating more of the restrictedfood. Conversely, restricting or monitoring a child’s in-take of certain foods may be successful if a parent doesnot express a permissive feeding style. This suggests thatinterventions solely focused on addressing child feedingpractices may be less effective or perhaps ineffective un-less parental feeding style is addressed. Dietetics practi-tioners should be aware that the way in which parentscommunicate with their children during feeding situa-tions may influence their child’s eating behavior. Know-ing the parent–child dynamic around food can help prac-titioners provide relevant advice and guidance to parentsthat may be more effective in controlling the diet qualityof their child’s food intake. A simple screening tool wouldalso allow practitioners to assess parental feeding styleand practices quickly and be able to counsel the parentaccordingly. No tool currently exists and work is neededto develop a valid and reliable measure.

More work is also needed to identify and examine otherchild feeding practices that represent some of the positivepractices that parents may employ (eg, teaching theirchildren about healthy eating) (51) and those that teaseapart the complexity by which parents control theirchild’s eating behavior (52). Combining research efforts tomeasure parental feeding style together with a more com-prehensive view of child-feeding practices will give muchneeded insight into the complex nature of the parent–child feeding relationship.

This study has some limitations. It is important torecognize that the parent–child feeding relationship islikely bidirectional and the findings presented here do notinfer causality. Studies are needed to understand thepathway linking parenting behaviors and child food in-take and diet quality that include measures of specificchild behaviors (eg, temperament). The small sample sizedid not allow for certain subgroup analyses to be con-ducted (eg, separating boys vs girls or mothers vs fa-thers), and logistical constraints limited the ability toassess usual intake to 2 weekday dietary recalls. In ad-dition, the sample represents a diverse group of parent–child dyads living in rural areas. Generalizability may belimited given the lack of representation from other groupsand the randomly selected, but convenient sample from

which the participants were recruited. Despite these

Jan

hortcomings, this study provides new direction for re-earch to identify the modifiable risk factors within theamily setting that may influence children’s food intake.

CONCLUSIONSParental feeding style was found to be a potential modi-fiable risk factor for children’s consumption of LND foods.A permissive feeding style was positively associated withand moderated the relationship between parental feedingpractices and child LND food intake. This offers a poten-tial area for dietetics practitioners to intervene. Futureresearch should expand this study to parent–child dyadsfrom other cultural backgrounds and also incorporate amore comprehensive set of parental feeding practices.Results from these studies can provide a better under-standing of the parent–child feeding relationship andguidance for developing behavior change interventions.

STATEMENT OF POTENTIAL CONFLICT OF INTEREST:No potential conflict of interest was reported by the au-thors.

FUNDING/SUPPORT: This research was supported byawards from the Centers for Disease Control and Preven-tion (Public Health Dissertation Research Award, grantno. 1R36DP001325-01) and New Balance Foundation,with additional funding provided by Save the Children,US Programs, and the Robert Wood Johnson Foundation(grant no. 59458).

ACKNOWLEDGEMENTS: The authors thank the fol-lowing Save the Children USA partners: Vivica Kraak,nutrition and physical activity advisor; the nutrition andphysical activity specialists; and community partners, allof whom assisted with the study recruitment and datacollection. The authors are especially grateful for thechildren and parents who participated in this study.

References1. Reedy J, Krebs-Smith SM. Dietary sources of energy, solid fats, and

added sugars among children and adolescents in the United States.J Am Diet Assoc. 2010;110(10):1477-1484.

2. Guenther PM, Dodd KW, Reed J, Krebs-Smith SM. Most Americanseat much less than recommended amounts of fruits and vegetables.J Am Diet Assoc. 2006;106(9):1371-1379.

3. Ballew C, Kuester S, Gillespie C. Beverage choices affect adequacy ofchildren’s nutrient intakes. Arch Pediatr Adolesc Med. 2000;154(11):1148-1152.

4. Kant AK. Reported consumption of low-nutrient-density foods byAmerican children and adolescents: Nutritional and health correlates,NHANES III, 1988 to 1994. Arch Pediatr Adolesc Med. 2003;157(8):789-796.

5. French SA, Story M, Jeffery RW. Environmental influences on eatingand physical activity. Annu Rev Public Health. 2001;22:309-335.

6. Tai-Seale T, Chandler C. Nutrition and overweight concerns in ruralAmerica: A literature review. In: Rural Healthy People 2010: A Com-panion Document to Healthy People 2010. College Station, TX: TexasA&M University System Health Science Center, School of Rural Pub-lic Health, Southwest Rural Health Research Center; 2003.

7. Lutfiyya MN, Lipsky MS, Wisdom-Behounek J, Inpanbutr-MartinkusM. Is rural residency a risk factor for overweight and obesity for U.S.children? Obesity. 2007;15(9):2348-2356.

8. Birch LL, Fisher JO. Development of eating behaviors among childrenand adolescents. Pediatrics. 1998;101(3 pt 2):539-549.

9. Davison KK, Birch LL. Childhood overweight: A contextual model andrecommendations for future research. Obes Rev. 2001;2(3):159-171.

10. Skinner JD, Bounds W, Carruth BR, Morris M, Ziegler P. Predictorsof children’s body mass index: A longitudinal study of diet and growth

in children aged 2-8 y. Int J Obes Relat Metab Disord. 2004;28(4):476-482.

uary 2012 ● Journal of the ACADEMY OF NUTRITION AND DIETETICS 147

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

11. Strauss RS, Knight J. Influence of the home environment on thedevelopment of obesity in children. Pediatrics. 1999;103(6):e85.

12. Wansink B. Nutritional gatekeepers and the 72% solution. J Am DietAssoc. 2006;106(9):1324-1327.

13. Ritchie LD, Welk G, Styne D, Gerstein DE, Crawford PB. Familyenvironment and pediatric overweight: What is a parent to do? J AmDiet Assoc. 2005;105(5)(suppl 1):70-79.

14. Ventura AK, Birch LL. Does parenting affect children’s eating andweight status? Int J Behav Nutr Phys Act. 2008;(5):15.

15. Darling N, Steinberg L. Parenting style as context: An integrativemodel. Psychol Bull. 1993;113(3):487-496.

16. Baumrind D. Current patterns of parental authority. Dev PsycholMonograph. 1971;Part 2(4):1-103.

17. Maccoby E, Martin J. Socialization in the context of family: Parent-child interactions. In: Hetherington EM, ed. Handbook of child psy-chology: Socialization, Personality, and Social Development. Vol 4.New York, NY:Wiley;1993:1-101.

18. Hughes SO, Power TG, Fisher JO, Mueller S, Nicklas TA. Revisitinga neglected construct: Parenting styles in a child-feeding context.Appetite. 2005;44(1):83-92.

19. Hughes SO, Anderson CB, Power TG, Micheli N, Jaramillo S, NicklasTA. Measuring feeding in low-income African-American and Hispanicparents. Appetite. 2006;46(2):215-223.

20. van der Horst K, Kremers S, Ferreira I, Singh A, Oenema A, Brug J.Perceived parenting style and practices and the consumption of sug-ar-sweetened beverages by adolescents. Health Educ Res. 2007;22(2):295-304.

21. Vereecken C, Rovner A, Maes L. Associations of parenting styles,parental feeding practices and child characteristics with young chil-dren’s fruit and vegetable consumption. Appetite. 2010; 55(3):574-581.

22. National Center for Education Statistics. Urban\rural classificationsystems. US Department of Education Web site. http://nces.ed.gov.Accessed February 1, 2006.

23. Lytle LA, Murray DM, Perry CL, Eldridge AL. Validating fourth-grade students’ self-report of dietary intake: Results from the 5 A DayPower Plus program. J Am Diet Assoc. 1998;98(5):570-572.

24. Livingstone MB, Robson PJ. Measurement of dietary intake in chil-dren. Proc Nutr Soc. 2000;59(2):279-293.

25. Marcotte L, Hennessy E, Dwyer J, et al. Validity and reliability of acalcium checklist in early elementary-school children. Public HealthNutr. 2008;11(1):57-64.

26. Huang TT, Roberts SB, Howarth NC, McCrory MA. Effect of screen-ing out implausible energy intake reports on relationships betweendiet and BMI. Obes Res. 2005;13(7):1205-1217.

7. Kranz S, Mitchell DC, Smiciklas-Wright H, Huang SH, KumanyikaSK, Stettler N. Consumption of recommended food groups amongchildren from medically underserved communities. J Am Diet Assoc.2009;109(4):702-707.

8. Lytle LA, Nichaman MZ, Obarzanek E, et al. Validation of 24-hourrecalls assisted by food records in third-grade children. The CATCHCollaborative Group. J Am Diet Assoc. 1993;93(12):1431-1436.

9. Matheson DM, Hanson KA, McDonald TE, Robinson TN. Validity ofchildren’s food portion estimates: A comparison of 2 measurementaids. Arch Pediatr Adolesc Med. 2002;156(9):867-871.

0. Birch LL, Fisher JO, Grimm-Thomas K, Markey CN, Sawyer R,Johnson SL. Confirmatory factor analysis of the Child Feeding Ques-tionnaire: A measure of parental attitudes, beliefs and practices aboutchild feeding and obesity proneness. Appetite. 2001;36(3):201-210.

1. Anderson CB, Hughes SO, Fisher JO, Nicklas TA. Cross-culturalequivalence of feeding beliefs and practices: The psychometric prop-

erties of the child feeding questionnaire among Blacks and Hispanics.Prev Med. 2005;41(2):521-531.

148 January 2012 Volume 112 Number 1

2. Corsini N, Danthiir V, Kettler L, Wilson C. Factor structure andpsychometric properties of the Child Feeding Questionnaire in Aus-tralian preschool children. Appetite. 2008;51(3):474-481.

3. Kaur H, Li C, Nazir N, Resnicow K, Birch LL, Ahluwalia JS. Confir-matory factor analysis of the child-feeding questionnaire among par-ents of adolescents. Appetite. 2006;47(1):36-45.

4. Lohman TG. Advances in Body Composition Assessment. Champaign,IL: Human Kinetics; 1992. Current Issues in Exercise Science SeriesMonograph No. 3.

5. 2000 CDC Growth Charts. Centers for Disease Control and Preven-tion Web site. http://www.cdc.gov/growthcharts. Accessed October 2,2007.

6. Barlow SE. Expert committee recommendations regarding the pre-vention, assessment, and treatment of child and adolescent over-weight and obesity: Summary report. Pediatrics. 2007;120(suppl 4):164-192.

7. National Heart Lung and Blood Institute (NHLBI) Obesity EducationInitiative Expert Panel. Clinical Guidelines on the Identification,Evaluation, and Treatment of Overweight and Obesity in Adults: TheEvidence Report. Bethesda, MD: National Institutes of Health; 1998.

8. Baron RM, Kenny DA. The moderator-mediator variable distinctionin social psychological research: Conceptual, strategic, and statisticalconsiderations. J Pers Soc Psychol. 1986;51(6):1173-1182.

9. Aiken LS, West SG. Multiple Regressions: Testing and InterpretingInteractions. Thousand Oaks, CA: Sage; 1991.

0. Holmbeck GN. Post-hoc probing of significant moderational and me-diational effects in studies of pediatric populations. J Pediatr Psychol.2002;27(1):87-96.

1. Nielsen SJ, Adair L. An alternative to dietary data exclusions. J AmDiet Assoc. 2007;107(5):792-799.

2. Kant AK, Graubard BI. Predictors of reported consumption of low-nutrient-density foods in a 24-h recall by 8-16 year old US childrenand adolescents. Appetite. 2003;41(2):175-180.

3. Hoerr SL, Hughes SO, Fisher JO, Nicklas TA, Liu Y, Shewchuk RM.Associations among parental feeding styles and children’s food intakein families with limited incomes. Int J Behav Nutr Phys Act. 2009;6:55.

4. Hughes SO, Shewchuk RM, Baskin ML, Nicklas TA, Qu H. Indulgentfeeding style and children’s weight status in preschool. J Dev BehavPediatr. 2008;29(5):403-410.

5. Saxton J, Carnell S, van Jaarsveld CH, Wardle J. Maternal educationis associated with feeding style. J Am Diet Assoc. 2009;109(5):894-898.

6. van Strien T, Bazelier FG. Perceived parental control of food intake isrelated to external, restrained and emotional eating in 7-12-year-oldboys and girls. Appetite. 2007;49(3):618-625.

7. Van Strien T, van Niekerk R, Ouwens MA. Perceived parental foodcontrolling practices are related to obesogenic or leptogenic child lifestyle behaviors. Appetite. 2009;53(1):151-154.

8. Wake M, Nicholson JM, Hardy P, Smith K. Preschooler obesity andparenting styles of mothers and fathers: Australian national popula-tion study. Pediatrics. 2007;120(6):e1520-e1527.

9. Faith MS, Kerns J. Infant and child feeding practices and childhoodoverweight: The role of restriction. Matern Child Nutr. 2005;1(3):164-168.

0. Birch LL. Development of food preferences. Annu Rev Nutr. 1999;19:41-62.

1. Musher-Eizenman D, Holub S. Comprehensive Feeding PracticesQuestionnaire: Validation of a new measure of parental feeding prac-tices. J Pediatr Psychol. 2007;32(8):960-972.

2. Ogden J, Reynolds R, Smith A. Expanding the concept of parental

control: A role for overt and covert control in children’s snackingbehaviour? Appetite. 2006;47(1):100-106.