Maternal and Child Health Journal

, Volume 15, Issue 7, pp 899–909 | Cite as

Parental Feeding Practices and Concerns Related to Child Underweight, Picky Eating, and Using Food to Calm Differ According to Ethnicity/Race, Acculturation, and Income

  • Alexandra Evans
  • Jennifer Greenberg Seth
  • Shanna Smith
  • Karol Kaye Harris
  • Jennifer Loyo
  • Carol Spaulding
  • Mary Van Eck
  • Nell Gottlieb


The purpose of this study was to examine differences in parental feeding practices according to ethnicity/race, household income, parent education level, acculturation (for Hispanic participants only), and participation in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) program among parents living in a southern state in the United States. For this cross-sectional study, parents of children ages 1–5 years living throughout Texas were recruited through random digit dialing with screening questions during Fall 2006. Eligible parents who agreed to participate completed the Preschooler Feeding Questionnaire (PFQ) and a demographic questionnaire over the phone in either English or Spanish. The PFQ included five subscales: child overeating concerns, child underweight concerns, difficulty with picky eating, using food to calm, and pushing child to eat. Demographic questions assessed ethnicity/race, household income, parent education level, acculturation, and WIC participation. Structural Equation Modeling (SEM), with the demographic variables as predictors, was used to predict the five PFQ subscales. Complete data were obtained from 721 parents, 50% of whom were Hispanic. Significant differences for the PFQ subscales were noted for ethnicity/race, acculturation, and income level. Spanish-speaking Hispanic participants were significantly more worried about their child being underweight than English-speaking Hispanic participants. High-income non-WIC respondents were more likely to report that they have difficulty with picky eaters compared to WIC respondents. Spanish-speaking Hispanics and Black respondents were more likely than English-speaking Hispanics to use food to calm the child. Health practitioners need to be aware of differences in parental feeding practices and concerns among parents of diverse demographic backgrounds. Results from this study can be used to tailor health programs that promote healthy feeding practices among parents.


Parental feeding practices Low-income Acculturation Preschool-aged children 


Many children living in the United States (US) consume diets that do not meet national dietary recommendations [1, 2, 3]. Data from the most recent National Health and Nutrition Survey (NHANES) show that diets of children as young as 2 years old include too much saturated fat, sodium, and calories from nutrient-poor and calorie-dense foods [3]. Because of the high consumption of nutrient-poor foods and the associated consumption of extra calories [4], US children’s current dietary patterns may be one contributing factor to the rising prevalence of childhood obesity [5, 6, 7], a condition that now affects between 20 and 24% of preschool aged children in the US [3]. Because children’s food preferences are established around age five [8], interventions targeting the dietary behaviors of younger children are needed. However, most interventions targeting children’s dietary behaviors have focused on older children [9, 10], and more formative research to develop interventions for younger children and to determine effective avenues to deliver these types of interventions needs to be conducted.

Dietary patterns of children are greatly influenced by the foods that are made available to them by their parents [11, 12, 13]. Specific parental feeding practices, such as parental restriction, teach children to attend to external, rather than internal, satiety cues [14, 15] and have been associated with unhealthful child dietary behaviors and increased intake of calories [12]. Given the serious consequences of unhealthful dietary patterns among children, there is a need for a greater understanding of the influences on parental feeding practices.

Epidemiologic data indicate that children’s eating patterns differ by demographic factors, such as ethnicity/race [16, 17], acculturation [18, 19, 20, 21], parental education [22, 23], and household income [23, 24, 25]. It is reasonable to hypothesize that parental feeding practices differ accordingly. However, there have been very few studies that have specifically assessed parental feeding practices according to demographic factors, and more research is warranted.

The US Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) program serves low income (determined by 185% of federal poverty level) women and children up to age five by providing vouchers for nutritious foods, nutrition education, and referrals to health care. Because WIC reaches a large number of parents and children of preschool age, WIC can potentially serve as an effective avenue to implement parent-focused interventions targeting dietary patterns of young children. Recent studies have found that WIC children eat more healthfully [3, 11] and that participation in WIC is associated with a reduced risk of overweight in preschoolers [12]. However, other studies have found that children on WIC have less healthful diets [26] and more research to examine the link between WIC participation and childrens’ diets is warranted.

The primary aim of this study was to examine and compare the concerns and feeding practices of preschool parents according to WIC participation and family income. The secondary purpose was to examine the parental practices and concerns by the demographic variables of ethnicity/race, acculturation level (for the Hispanic participants only), and education level. Results from this study will be useful to guide the development of targeted interventions for preschool parents in WIC or non-WIC settings.


Design and Participants

The data presented in this manuscript were collected for a cross-sectional study that was designed to measure parent-reported dietary intake of preschool aged children and the feeding practices, parental self-efficacy, parental nutrition knowledge, attitudes and behaviors related to the purchase and preparation of foods of parents of preschool aged children. The sample included Texas households with children between the ages of 1 and 5 years. Eligible respondents for the study were identified through random-digit-dialing telephone calls by an independent market research firm. Quotas were set based on income and WIC participation to achieve three groups: (1) low-income households that were participating in WIC at the time of the survey, (2) low-income households that were WIC-eligible but did not participate in WIC at the time of the survey, and (3) higher-income households that were not eligible for WIC. Low-income was defined as 185% of the federal poverty level, based on reported monthly household income and number of people in the household. This is also the income criterion for WIC eligibility. Within each of the three groups, ethnic/racial quotas were set based on 2000 US Census data and 2006 WIC participation data to ensure a representative sample within each of these groups. Geographic quotas were set proportionally to the population density of each of 11 public health regions in Texas so that the sample represents the regional diversity of the state.

Phone calls to recruit participants were made by trained interviewers between the hours of 4 and 9 pm on Mondays through Thursdays for the period of September 2006 through October 2006. Respondents were informed that the interview would take approximately 30 min to complete and were offered a $15 gift certificate to a superstore for participation. After a brief introductory statement, participants were asked to give verbal consent if they agreed to participate in the study. Four screening questions were used to determine eligibility and to categorize the respondent into one of the three groups: (1) Do you have a child between the ages of 1 and 5 years? (2) What is your total monthly household income? (3) How many people live in your household? (4) Do you participate in the WIC program? Research staff monitored phone calls throughout the process for quality assurance. Gift cards were mailed to the respondents after they completed the survey.

Three thousand nine hundred fifty-seven individuals were reached by phone. Of these, 2,105 declined participation after the purpose of the call and study was explained. Of the remaining 1,852 individuals, 1,098 were ineligible. The remaining 754 parents were eligible and agreed to participate through verbal consent. Because it was impossible to determine the eligibility of the 2,105 individuals who declined, we cannot calculate an exact response rate. Using the numbers of eligible (n = 754) and ineligible parents (n = 1,098), we determined that of the sample that was willing to participate (n = 1,852), 59% of parents were ineligible. Using this number, we calculated an approximate response rate of 47%, which is similar to response rates of other studies using only random-digit dialing as the recruitment strategy [27].

Survey responses and mailing information were stored in separate databases so that respondents could not be identified. This study was conducted in accordance with prevailing ethical principles and was approved by the University of Texas at Austin and the University of Texas Health Sciences Institutional Review Boards.


Eligible parents responded to several questionnaires including a demographic survey and the Preschooler Feeding Questionnaire (PFQ) [28].

Demographic Survey

This survey included questions about respondent’s age and gender, child’s age and gender, race/ethnicity, primary language spoken at home (used as a proxy for acculturation level [29]), employment status, educational attainment, marital status, family size, household income, and participation in the WIC program and other food assistance programs.

The Preschooler Feeding Questionnaire (PFQ)

The PFQ is a 32-item questionnaire that explores parental feeding practices and concerns hypothesized to be related to child overweight status. The first 26 items on the PFQ address how often parents engage in a variety of behaviors related to feeding their children. Parents are asked to answer using a five-point scale with the responses “Never,” “Rarely,” “Sometimes,” “Often,” or “Always” in English and “Nunca,” “Rara vez,” “A veces,” “A menudo,” or “Siempre” in Spanish. Responses were scored on a scale of 1 (Never) to 5 (Always). The last six items ask parents to report how much they agree or disagree with statements regarding whether they worry about the child’s weight status, and whether they think offering food is the best way to stop temper tantrums. Responses to the last six items were scored on a four-point scale of 1 (Disagree a lot) to 4 (Agree a lot).

The original questionnaire was developed, pilot tested, and validated by Baughcum and colleagues [28]. Baughcum et al. (2001) theorized that the items should group into six factors: (1) concern about the child being over- or underweight, (2) lack of structure during feeding, (3) concern about over- and under-eating, (4) control over the child’s food intake, (5) use of food to influence child behavior, and (6) daytime bottle use. Empirically, however, Baughcum et al. found eight factors, including (1) Difficulty in child feeding, which included six items regarding the child being a picky eater, the child having a poor appetite, and others; (2) seven items addressing the child overeating or being overweight; (3) pushing the child to eat more, which included five items such as making the child eat everything on his/her plate; (4) two items regarding using food to calm the child; (5) two items concerning the child being underweight; (6) the child’s control of feeding interactions, which included three items such as allowing the child to eat snacks whenever he/she wished; (7) structure during feeding interactions, which included three items, such as allowing the child to watch TV at meals; and (8) age-inappropriate feeding, including the two items “gave child a bottle during the day” and “mother fed child herself if child did not eat enough.” In Baughcum’s analysis, the first three factors had acceptable inter-item reliability, and the fourth and fifth had near-acceptable reliability (alphas = 0.68 and 0.69), but the last three factors had very poor reliability (alphas = 0.50, 0.37, and 0.18). To test the groupings of items into subscales for our study, a factor analysis was performed; results are described in the analysis section below.

The items are written at a low literacy level (third grade reading level as determined by the Flesch–Kincaid grade level scale), which is appropriate for individuals with lower education levels. Because 48% of the Texas population is Hispanic (, the questionnaire was translated into Spanish by a bicultural (Mexican-American), bilingual, native Spanish speaker with a nutrition background. The questions were reviewed by two native Spanish-speaking graduate students and back-translated into English by the University of Texas at Austin Survey Research Center. No discrepancies were found between the back-translated English version and the original English. We then adapted the original questionnaire’s format (i.e. self-administered) so that it could be completed using a telephone interview format. The questionnaire was field tested in English and Spanish with 235 Texas WIC clients. The field test addressed readability and interpretation of the questions and response options, and format of the questionnaire. The survey was self-administered to approximately 50% of the field test participants and phone-administered to the other 50% of the field test participants. Results from the field test indicated no difficulty in readability or comprehension by the respondents, and interpretations were consistent with the intended meaning. Therefore, no changes were made to the PFQ.


Seven hundred fifty four parents completed the phone interviews. Interviews from 33 parents were eliminated because the parents did not meet all of the eligibility criteria. The remaining 721 parents were grouped into three groups based on household income relative to family size and current participation in WIC: WIC parents (n = 244), low-income non-WIC parents (n = 274), and higher-income non-WIC parents (n = 203). The cutoff between low and higher-income was determined by the WIC eligibility guideline of 185% of poverty as determined by income level and family size.

Because only Hispanic participants reported speaking Spanish as their primary language, ethnicity and language were combined into a single variable with four levels: White, Black, English-speaking Hispanic, and Spanish-speaking Hispanic. For the remaining analyses, only parents who reported Hispanic, Black and White as their ethnicity were included, those with other or missing ethnicity information were dropped from further analysis (N = 57). An additional five respondents were dropped from analysis due to missing data on one or more predictors, for an N of 659 for the remaining analyses.

Analysis of the Demographic Survey

Low-income WIC parents, low-income non-WIC parents, and higher income parents were compared according to demographics (age of parent and child, gender of parent and child, ethnicity/race, income level, education level, number of people in household, and number of children in household) using the chi-square test for categorical variables and ANOVA for continuous variables (Table 1).
Table 1

Demographic characteristics of respondents and children by WIC participation and income level (n = 659)


Combined sample

WIC participants [WIC]

N = 224 (34%)

Low-income, non-WIC [LOW-NWIC]

n = 259 (39%)

Higher-income, non-WIC [HIGH]

N = 176 (27%)

Significant differences among the groups at the P < 0.05 level

Age of respondents (mean)a

33 years (±10.11)

31.4 (±11.02)

33.5 (±10.13)

33.9 (±8.60)


Female respondentb

567 (86%)

199 (89%)

228 (88%)

141 (80%)


Age of childa

3.26 (±1.35)

3.02 (±1.34)

3.41 (±1.35)

3.36 (±1.32)


Ethnicity of respondentb




218 (33%)

47 (21%)

72 (28%)

99 (56%)



83 (13%)

31 (14%)

40 (15%)

12 (7%)


 Hispanic (English speaking)

208 (32%)

66 (30%)

90 (35%)

52 (30%)


 Hispanic (Spanish speaking)

150 (23%)

80 (36%)

57 (22%)

13 (7%)


Total monthly household incomeb




121 (19%)

80 (37%)

41 (17%)




187 (29%)

102 (46%)

85 (35%)




128 (20%)

36 (17%)

90 (37%)

2 (1%)



70 (11%)


28 (11%)

42 (24%)



45 (7%)


1 (0.4%)

44 (26%)


 $5,000 or higher

84 (14%)



84 (35%)


Education levelb



 <High school

168 (26%)

86 (38%)

74 (29%)

8 (5%)


 High school graduate/GED

160 (24%)

69 (31%)

67 (26%)

24 (14%)


 >High school

331 (50%)

69 (31%)

118 (46%)

144 (82%)


 # of people in household (mean)a

4.60 (±1.38)

4.86 (±1.54)

4.69 (±1.32)

4.15 (±1.13)


 # of children <16 in household (mean)a

2.37 (±1.12)

2.60 (±1.24)

2.44 (±1.11)

1.98 (±0.80)


aANOVA for continuous variables

bchi-square test for categorical variables

Analysis of the Factor Structure of the PFQ

Univariate analyses showed that respondents used the full range of response options for all items of the PFQ. However, responses to most items were strongly skewed (for example, 86% of parents reported that they “never” think about putting their child on a diet). Consequently, the items were treated as ordinal measures for factor analyses and structural equation modeling.

First, we examined whether our Preschooler Feeding Questionnaire data followed the factor structure obtained by Baughcum et al. (2001). Responses to most items on the feeding questionnaire were strongly skewed (for example, 86% of parents reported that they “never” think about putting their child on a diet to keep from becoming overweight, 77% “never” worry that the child is eating too much, and 74% “disagree a lot” that the child is underweight). Accordingly, it seemed most appropriate to treat the items as ordinal measures. Using the Mplus structural equation modeling program (Muthen & Muthen), we ran an 8-factor confirmatory model following Baughcum’s suggested structure. The model resulted in a non-positive definite covariance matrix, an error caused by the factor of age inappropriateness; examination of the bivariate correlations revealed that the two items theoretically loading on this factor were negligibly related, suggesting that it is inappropriate to group them together on the same factor.

As were unable to replicate Baughcum’s factor structure, an exploratory factor analysis was conducted to determine the best structure for the items based on our data, treating each item as ordinal using the Mplus exploratory factor analysis procedure with the quartimin oblique rotation and the WLSMV estimator, a weighted least squares estimator appropriate for ordinal data. In the initial analysis, the eigenvalue greater than 1 criterion suggested an 8-factor solution, while the elbow criterion suggested that a 6-factor model was sufficient. For both the 6- and 8-factor solutions, however, the residual variances for several items were quite high (greater than 0.75); through an iterative process, we dropped each item with a high residual, re-ran the factor analysis, and re-examined the factor solution and residual variances. This process led us to drop seven items (all of which were thought by Baughcum et al. to load on factors 6–8), resulting in 25 items and five factors. The resulting exploratory factor structure was quite similar to Baughcum et al.’s first five factors; however, our solution included several items that cross-loaded on two or more factors. In particular, items regarding concerns about the child’s overweight status loaded strongly on both the “overweight” and “underweight” factors, and three items from the “difficulty in child eating” factor cross-loaded on either the “pushing child to eat” factor, the “using food to calm” factor, or both. Using an iterative process, we dropped each highly cross-loading item, resulting in a final set of 17 items; both the eigenvalue-greater-than-one and the elbow criterion suggested five factors were necessary. While the final model had a significant chi-square for absolute fit, χ2(44) = 95.19, P < 0.001, it had excellent relative fit (CFI (comparative fit index) = 0.98, TLI (Tucker–Lewis index) = 0.98, RMSEA (Root mean square error of approximation) = 0.04, WRMR (Weighted root mean square residual) = 0.49), and all item residuals were less than 0.75.

The final factor structure included five subscales measuring: (1) Pushing child to eat, (2) Difficulty with picky eating, (3) Using food to calm, (4), Child underweight concerns, and (5) Child overweight concerns. All items had strong primary loadings (with standardized coefficients greater or equal to 0.50), with the exception of the item “How often do you offer him/her dessert after a meal to get him/her to eat foods that are good for him/her?,” which had a moderate-to-small (0.37) loading on its primary factor. Some items also had minor cross-loadings (less than 0.30) on secondary factors. Table 2 presents the loadings for each item on its primary factor. Further analyses were conducted using the latent variables within Mplus; however, to provide researchers with a revised set of scales for use in further research, we also constructed subscales based upon the primary items loading on each factor. Table 2 presents the descriptive statistics and reliability for the constructed scales. Each constructed subscale correlated quite strongly with estimated factor scores for its corresponding latent variable (picky = 0.95, overweight = 0.90; push more = 0.94; calm = 0.88; underweight = 0.96). Reliability for the constructed scales, while superior to those of Baughcum et al.’s original scales, are still not optimal, especially for using food to calm and pushing child to eat. In future studies, revising the scale to include an additional item for each of these constructs may prove helpful.
Table 2

Summary table for the PFQ subscales



Item loadings: Unstd (SE)

Item loadings: Std

Subscale mean (SD)

Inter-item reliability

Pushing child to eat

How often do you make him/her finish all his/her dinner before he/she can have a dessert?

0.88 (0.06)***


2.46 (1.03)



How often do you make him/her eat all the food on his/her plate?

0.61 (0.06)***



How often do you offer him/her dessert after a meal to get him/her to eat foods that are good for him/her?

0.37 (0.05)***



Difficulty with picky eating

How often is he/she a picky eater?

0.78 (0.02)***


2.34 (0.92)



How often is it hard for you to get him/her to eat new foods?

0.72 (0.03)***



How often do you have to make special foods for him/her because he/she is a picky eater?

0.75 (0.03)***



How often is it a struggle to get him/her to eat?

0.73 (0.03)***



How often does he/she have a poor appetite?

0.53 (0.04)***



Using food to calm

How often do you give him/her something to eat or drink if he/she is upset even if you think he/she is not hungry?

0.97 (0.05)***


1.69 (0.74)



When he/she gets fussy, how often is giving him/her something to eat or drink the first thing you do?

0.48 (0.05)***



How often do you give him/her something to eat or drink if he/she is bored even if you think he/she is not hungry?

0.65 (0.06)***



Child underweight concerns

I am worried that my child will become underweight.

0.85 (0.05)***


1.55 (0.77)



I am worried that my child is underweight right now.

0.82 (0.06)***



Child overeating concerns

How often do you worry that he/she is eating too much?

0.86 (0.03)***


1.38 (0.62)



How often do you have to stop him/her from eating too much?

0.79 (0.04)***



How often do you think about putting him/her on a diet to keep him/her from becoming overweight?

0.78 (0.05)***



How often do you get upset if he/she eats too much?

0.76 (0.04)***



*** P < 0.001

Parental Feeding Differences and Concerns Among Groups

We used Mplus to estimate the structural equation models with parental education, ethnicity/language, and WIC/income status as predictors for each of the 5 PFQ subscales, as shown in the Fig. 1. Because our predictors were categorical, we used dummy variables with reference groups selected to best address our research questions. For education, high school graduates served as the reference, for WIC and income status, WIC households served as the reference, and for ethnicity/language, English-speaking Hispanics served as the reference in order to highlight differences based on ethnicity (e.g., English-speaking Hispanics compared to Whites) and acculturation (e.g., English-speaking Hispanics compared to Spanish-speaking Hispanics). For all analyses, the P-value was set at 0.05.
Fig. 1

Structural path diagram



In our sample of 659 parents, 55% identified themselves as Hispanic, 33% as White, and 13% as Black. Twenty-three percent of the Hispanic parents reported speaking mostly Spanish at home. We found significant differences for all demographic variables among low-income WIC parents, low-income non-WIC parents, and higher income parents (Table 1). Most notably, the low-income WIC parents included 66% Hispanic parents (36% Spanish speaking) while the low-income non-WIC parents included 57% Hispanic parents (22% Spanish speaking) and the higher income parents included 37% Hispanic parents (7% Spanish speaking). In addition, the income level of the WIC parents was significantly lower than the other two groups. While 83% of the WIC participants reported an income of less than $2,000, 52% of the low-income non-WIC parents and none of the higher income parents reported an income less than $2,000. Lastly, of the WIC parents, 38% had less than a high school education, while 29% of the low-income non-WIC parents and 5% of the higher income parents had a similar level of education.

Description of Parental Feeding Practices

Overall, parents scored highest on the subscales measuring Pushing child to eat and Difficulty with picky eating, which had mean scores of 2.46, and 2.34, respectively. The subscales assessing Using food to calm and Child underweight concerns received mean subscale scores of 1.69, and 1.55, respectively. Parents scored lowest on the subscale measuring Child overweight concerns. The parents from all three groups followed this general pattern except the higher income parents who scored highest on the Difficulty with picky eating subscale (Table 3).
Table 3

Scores on subscales among 3 groups of participants


Range of responsesa

WIC participants [WIC]

N = 224 (34%)

Low-income, non-WIC [LOW-NWIC]

n = 259 (39%)

Higher-income, non-WIC [HIGH]

N = 176 (27%)

Pushing child to eat


2.51 (1.00)

2.43 (1.05)

2.44 (1.02)

Difficulty with picky eating


2.20 (0.90)

2.30 (0.91)

2.55 (0.94)

Using food to calm


1.77 (0.77)

1.66 (0.73)

1.63 (0.71)

Child underweight concerns


1.61 (0.80)

1.58 (0.77)

1.42 (0.72)

Child overeating concerns


1.49 (0.72)

1.37 (0.60)

1.27 (0.45)

aHigher mean indicates a stronger endorsement of the items on the subscale

Comparison of PFQ Factors According to Demographic Factors

The overall structural equation model, including both the factor analysis measurement model and the demographic predictors of each of the 5 PFQ factors, had poor absolute fit, WLSMV (Weighted least squares mean and variance adjusted estimator) χ2(86) = 24.03, P < 0.01, but excellent relative fit (CFI = 0.99, TLI = 0.99, RMSEA = 0.03, WRMR = 0.57) indicating that the model performed well when compared to the independence model [29]. Standardized path coefficients and the associated significance levels for the demographic predictor on each PFQ subscale are displayed in Table 4. Although the items loading on each factor are ordinal, the latent variables estimated by Mplus are continuous; therefore, these coefficients are interpreted similarly to coefficients in an ordinary least-squares regression [30].
Table 4

Standardized path coefficients for education level, ethnicity/language, WIC participation/income on each PFQ subscale


Child overeating concerns

Child underweight concerns

Difficulty with picky eating

Using food to calm

Pushing child to eat

<HS educationa






Some college


















Spanish-Speaking Hispanic






Low-income non-WICc






High-income non-WIC






*** P < 0.001; * P < 0.05

aReference group is “high school graduate”

bReference group is “English-speaking Hispanic”

cReference group is “WIC household”

Significant demographic differences were found for Child underweight concerns, Difficulty with picky eating, and Using food to calm. Spanish-speaking Hispanic participants were significantly more worried about their child being underweight than the reference English-speaking Hispanic participants. High-income non-WIC respondents were more likely to report that they deal with picky eaters compared to the reference WIC respondents. Spanish-speaking Hispanics and Black respondents were more likely than English-speaking Hispanics to use food to calm the child. There were no significant differences in the feeding practices and concerns of Child overeating concerns and Pushing child to eat.


The purpose of this study was to examine and compare the feeding practices and concerns of parents with young children according to WIC participation, income level, ethnicity/race, acculturation level (for the Hispanic participants only), and education level.

In the full sample, the feeding practices most strongly endorsed by parents were Pushing child to eat and Difficulty with picky eating. The mean scores of these subscales indicated that most parents “sometimes” or “often” push their children to eat more or have to deal with picky eating. Parents in this sample were less likely to report that they use food to calm their children or that they are concerned about underweight. They were least concerned about their children overeating. These parental practices are incongruent with current epidemiologic data that show that approximately 25% of this sample is likely to be overweight or obese [3]. In fact, these results indicate that parents may be contributing to their children being overweight by pushing their children to eat more. However, this behavior may be very rational for parents who come from countries where under-nutrition is still a problem. Practitioners need to be culturally sensitive when explaining to parents the potential issues related to pushing a child to eat. In addition, practitioners need to be aware that parents may have difficulty assessing their own children’s weight status and need to teach appropriate feeding practices depending on the children’s weight status.

We found few significant differences among parents with different demographic backgrounds. With regard to race, the reference group of English-speaking Hispanics was similar to White and Black parents. The only significant difference found indicated that Black parents use food to calm more often than English-speaking Hispanic parents. Using food to calm or to shape behavior is a commonly reported practice among parents [31, 32]. This type of feeding practice can lead to unhealthful eating patterns (usually the food that is used to calm is nutrient-poor and calorie-dense) or child obesity if it creates an energy imbalance by interfering with the child’s ability to perceive normal hunger cues [14]. Practitioners can use this information to tailor education of feeding practices (e.g. methods to calm their child using non-food strategies) to specific ethnic/racial groups.

With regard to acculturation level, Spanish-speaking Hispanic parents were more likely than the English-speaking Hispanic parents to be concerned about their children being underweight and to use foods to calm their children. These findings are similar to an earlier study which found that Spanish-speaking Hispanic parents were more likely to report pushing their child to eat and using positive incentives to get child to eat more [33], two behaviors that have been associated with an increased dislike for the food and an increased caloric intake by the children [34, 35]. These results suggest an association between acculturation and parental feeding practices. These differences may be attributable to cultural differences between more and less acculturated parents. Compared to English-speaking Hispanics, Spanish-speaking Hispanic parents may have retained more values and concerns of their native countries. Although childhood obesity has become a health issue in Latin American countries, especially Mexico, under-nutrition and child underweight remain problems as well [36], and pushing children to eat may well be a rational behavior for natives of these countries. In addition, the use of food to calm children is a common practice in Hispanic culture, as is the belief that a parents show love for their children by feeding them well [30, 37]. An overweight child is not perceived as an obese child, but as a “healthy child,” a “cutie child” or as a child that is well fed and well-loved [38].

Although we found significant differences according to acculturation, no differences were found between English-speaking Hispanics and White parents, indicating that English-speaking Hispanic parents are more similar to white parents than to Spanish-speaking Hispanic parents in their feeding practices and concerns. These results are similar to results from a previous study [33] and have implications for intervention development. Presently, health education professionals tend to focus on ethnic and racial differences when developing culturally-specific interventions; however, it may be more worthwhile to focus on acculturation differences.

With regard to income, our results indicate one significant difference: high-income non-WIC respondents reported significantly more difficulty with picky eating than the reference WIC respondents. Picky eating has been associated with significantly lower dietary variety and diversity scores and less vegetable consumption [39, 40]. Perhaps higher income parents can afford to humor their children’s pickiness and requests for special meals. These results suggest that income level is not a significant factor that needs to be taken into consideration when teaching parents about appropriate feeding practices, except for dealing with picky eaters.

This study found no significant differences in parental feeding practices according to education level. Few studies have looked at the relationship between parental education level and child dietary patterns or parental feeding practices and therefore it is difficult to compare our results to other studies. One group of researchers found that obesity was not associated with maternal education in a sample of urban preschoolers [41], another study among Mexican-American mothers found that better-educated mothers were more likely to serve healthier foods [42]. Our results suggest that the messages in interventions teaching parents appropriate feeding practices do not need to be tailored according to parental education level. However, as with any educational material, the target audience’s literacy level needs to be considered and materials need to be developed at appropriate reading levels.

When comparing WIC participants to other low-income non-WIC participants, no significant differences were found. Although one study found that children who participated in WIC had less healthful diets than children not on WIC [43, 44], a recent study found that WIC-participant children were less likely to be overweight and consumed diets that were more nutrient rich compared to income-eligible nonparticipant children and were similar to the higher income children [3]. Our findings suggest that there is an opportunity to address feeding practices and concerns among participants in the Texas WIC program. Given the extensive reach of Texas WIC, this represents an opportunity to impact a large proportion of low-income families.

Strengths and Limitations

This study provided a comparison of low income WIC parents, low-income non-WIC parents, and higher-income non-WIC parents. The study design allowed examination of the influence of WIC status on feeding behaviors among low-income respondents and examination of differences related to income regardless of WIC status. This study, although a convenience sample, used quotas to achieve representation of all Texas residents, which increases the generalizability of our findings.

As in any study, this study has limitations. First, the PFQ instrument was not specifically developed for use with Hispanics or immigrant populations and may therefore not address all feeding practices relevant to these population groups. In addition, it is possible that Spanish-speakers interpreted the items or response categories differently than English speakers [45]. However, the instrument and response options were developed specifically for low-literate and low-income populations. In addition, we field-tested the instrument with a sample of Spanish-speaking and English-speaking Hispanic parents similar to the parents in the present study. The instrument was found to be understandable and relevant for our study population. Another limitation is that because we were not able to collect data from respondents who refused participation, the generalizability of the results is unknown. In addition, two of our subscales had Cronbach alpha coefficients that were barely acceptable. More instrument development needs to be done to create subscales with greater internal reliability. Also, because there was no measure of children’s weight status it is not possible to relate any of the measures to concerns about weight (and whether or not these concerns are valid). Lastly, the data are subject to the inherent limitations of using a convenience sample, using self-report data and of telephone survey research.


This study has three significant findings. First, we found that the most common feeding practices were pushing children to eat and dealing with picky eating. Parents were less likely to restrict their children from eating. This could be a concern given the rising prevalence of child overweight. Second, we found differences between less acculturated and more acculturated Hispanic parents, suggesting that it is important for health professionals to be aware of potential cultural differences and to become familiar with the cultural concerns and intergenerational transmissions of feeding practices. It is important to replace feeding practices that can potentially lead to unhealthful dietary patterns with culturally acceptable feeding practices that will lead to more healthful patterns. Third, despite the vast body of research that shows significant differences in child dietary patterns and childhood obesity according to demographic variables, we found few significant differences in parent feeding practices and concerns according to demographics. This suggests that parental feeding practices are not the cause of the disparities observed in child eating patterns and child obesity. Instead, the increased prevalence of unhealthful child eating behaviors and child obesity noted among children who are Hispanic or Black and live in low-income households may be rooted in the environment in which these children and their families live. For example, families living in low-income neighborhoods have less access to grocery stores with affordable and healthful foods compared to families living in high-income neighborhoods [46, 47]. Thus, although it is important to encourage appropriate feeding practices when trying to improve eating patterns of children, health professionals need to be aware of the significant influence of the environment on child eating behaviors as well.



The authors would like to thank the Special Supplemental Nutrition Program for Women Infants and Children (WIC) of the Texas Department of State Health Services for their ongoing support; Carol Holahan for her work in the early stages of the project; Rie Suzuki and Tara Ray for their participation in the development of the study. The research was supported by contract number 7217217217-2006 (Program Attachment # 05) from the Texas Department of State Health Services (DSHS) to the last author. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the TX DSHS. Preparation of this manuscript was made possible in part by funding by the Michael & Susan Dell Foundation.


  1. 1.
    Munoz, K. A., Krebs-Smith, S. M., Ballard-Barbash, R., & Cleveland, L. E. (1997). Food intakes of US children and adolescents compared with recommendations. Pediatrics, 100, 323–329.PubMedCrossRefGoogle Scholar
  2. 2.
    US Department of Health and Human Services. (2000). Healthy People 2010. 2[2]. Washington: US Government Printing Office.Google Scholar
  3. 3.
    Cole, N., & Fox, M. K. (2008) Nutrient intake and diet quality of WIC participants and nonparticipants. Food and Nutrition Service.Google Scholar
  4. 4.
    Drewnowski, A., & Darmon, N. (2005). The economics of obesity: Dietary energy density and energy cost. American Journal of Clinical Nutrition, 82(1 Suppl), 265S–273S.PubMedGoogle Scholar
  5. 5.
    Ogden, C. L., Carroll, M. D., & Flegal, K. M. (2008). High body mass index for age among US children and adolescents, 2003–2006. JAMA, 299(20), 2401–2405.PubMedCrossRefGoogle Scholar
  6. 6.
    Hedley, A. A., Ogden, C. L., Johnson, C. L., Carroll, M. D., Curtin, L. R., & Flegal, K. M. (2004). Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA, 291(23), 2847–2850.PubMedCrossRefGoogle Scholar
  7. 7.
    Nicklas, T. A., Yang, S. J., Baranowski, T., Zakeri, I., & Berenson, G. (2003). Eating patterns and obesity in children—The Bogalusa Heart Study. American Journal of Preventive Medicine, 25(1), 9–16.PubMedCrossRefGoogle Scholar
  8. 8.
    Birch, L. L., & Fisher, J. O. (1998). Development of eating behaviors among children and adolescents. Pediatrics, 101, 539–549.PubMedGoogle Scholar
  9. 9.
    Flynn, M. A., McNeil, D. A., Maloff, B., Mutasingwa, D., Wu, M., Ford, C., et al. (2006). Reducing obesity and related chronic disease risk in children and youth: A synthesis of evidence with ‘best practice’ recommendations. Obesity Reviews, 7(Suppl 1), 7–66.PubMedCrossRefGoogle Scholar
  10. 10.
    American Dietetic Association. (2006). Position of the American Dietetic Association: Individual-, family-, school-, and community-based interventions for pediatric overweight. Journal of the American Dietetic Association, 106(6), 925–945.CrossRefGoogle Scholar
  11. 11.
    Haire-Joshu, D., Brownson, R. C., Nanney, M. S., Houston, C., Steger-May, K., Schechtman, K., et al. (2003). Improving dietary behavior in African Americans: The parents as teachers high 5, low fat program. Preventive Medicine, 36(6), 684–691.PubMedCrossRefGoogle Scholar
  12. 12.
    Faith, M. S., Scanlon, K. S., Birch, L. L., Francis, L. A., & Sherry, B. (2004). Parent–child feeding strategies and their relationships to child eating and weight status. Obesity Research, 12(11), 1711–1722.PubMedCrossRefGoogle Scholar
  13. 13.
    Hodges, E. A. (2003). A primer on early childhood obesity and parental influence. Pediatric Nursing, 29(1), 13–16.PubMedGoogle Scholar
  14. 14.
    Satter, E. M. (1996). Internal regulation and the evolution of normal growth as the basis for prevention of obesity in children. Journal of the American Dietetic Association, 96(9), 860–864.PubMedCrossRefGoogle Scholar
  15. 15.
    Johnson, S. L. (2000). Improving preschoolers’ self-regulation of energy intake. Pediatrics, 106(6), 1429–1435.PubMedCrossRefGoogle Scholar
  16. 16.
    Krebs-Smith, S. M., Cook, A., Subar, A. F., Cleveland, L., Friday, J., & Kahle, L. L. (1996). Fruit and vegetable intakes of children and adolescents in the United States. Archives of Pediatrics and Adolescent Medicine, 150(1), 81–86.PubMedGoogle Scholar
  17. 17.
    Trevino, R. P., Marshall, R. M., Hale, D. E., Rodriguez, R., Baker, G., & Gomez, J. (1999). Diabetes risk factors in low-income Mexican-American children. Diabetes Care, 22(2), 202–207.PubMedCrossRefGoogle Scholar
  18. 18.
    Dave, J., Evans, A., Saunders, R., Watkins, K., & Pfeiffer, K. (2009) Association of food insecurity, acculturation, and children’s fruit and vegetables intake among Hispanic children. Journal of the American Dietetic Association, 109, 697–701.PubMedCrossRefGoogle Scholar
  19. 19.
    Ayala, G. X., Baquer, B., & Klinger, S. (2008). A systematic review of the relationship between acculturation and diet among Latinos in the United States: Implications for future research. Journal of the American Dietetic Association, 108(8), 1330–1344.PubMedCrossRefGoogle Scholar
  20. 20.
    Satia-Abouta, J., Patterson, R. E., Neuhouser, M. L., & Elder, J. (2002). Dietary acculturation: Applications to nutrition research and dietetics. Journal of the American Dietetic Association, 102(8), 1105–1118.PubMedGoogle Scholar
  21. 21.
    Neuhouser, M. L., Thompson, B., & Solomon, C. C. (2004). Higher fat intake and lower fruit and vegetables intakes are associated with greater acculturation among Mexicans living in Washington state. Journal of the American Dietetic Association, 104(1), 51–57.PubMedCrossRefGoogle Scholar
  22. 22.
    Kranz, S., & Siega-Riz, A. M. (2002). Sociodemographic determinants of added sugar intake in preschoolers 2 to 5 years old. Journal of Pediatrics, 140(6), 667–672.PubMedCrossRefGoogle Scholar
  23. 23.
    Crawford, P. B., Obarzanek, E., Schreiber, G. B., Barrier, P., Goldman, S., Frederick, M. M., et al. (1995). The effects of race, household income, and parental education on nutrient intakes of 9- and 10-year-old girls. NHLBI Growth and Health Study. Annals of Epidemiology, 5(5), 360–368.PubMedCrossRefGoogle Scholar
  24. 24.
    James, W. P., Nelson, M., Ralph, A., & Leather, S. (1997). Socioeconomic determinants of health. The contribution of nutrition to inequalities in health. BMJ, 314(7093), 1545–1549.PubMedGoogle Scholar
  25. 25.
    Xie, B., Gilliand, F. D., Li, Y., & Rockett, H. R. (2003). Effects of ethnicity, family income, and education on dietary intake of adolescents. Preventive Medicine, 36, 30–40.PubMedCrossRefGoogle Scholar
  26. 26.
    Ziegler, P., Hanson, C., Ponza, M., Novak, T., & Hendricks, K. (2006). Feeding Infants and Toddlers Study: Meal and snack intakes of Hispanic and non-Hispanic infants and toddlers. Journal of the American Dietetic Association, 106, S107–S123.PubMedCrossRefGoogle Scholar
  27. 27.
    Life Sciences Research Office, Federation of American Societies for Experimental Biology. (1995). Third report on nutrition monitoring in the United States. Washington, DC: US Government Printing Office.Google Scholar
  28. 28.
    Baughcum, A. E., Powers, S. W., Johnson, S. B., Chamberlin, L. A., Deeks, C. M., Jain, A., et al. (2001). Maternal feeding practices and concerns and their relationships to overweight in early childhood. Journal of Developmental and Behavioral Pediatrics, 22(6), 391–408.PubMedCrossRefGoogle Scholar
  29. 29.
    Lara, M., Gamboa, C., Kahramanian, M. I., Morales, L. S., & Bautista, D. E. (2005). Acculturation and Latino health in the United States: A review of the literature and its sociopolitical context. Annual Review of Public Health, 26, 367–397.PubMedCrossRefGoogle Scholar
  30. 30.
    Muthen, L. K., & Muthen, B. O. (2007). Mplus user’s guide (5th ed.). Los Angeles, CA: Muthen & Muthen.Google Scholar
  31. 31.
    Baughcum, A., Burklow, K., Deeks, C., Powers, S., & Whitaker, R. (1998). Maternal feeding practices and childhood obesity. Archives of Pediatrics and Adolescent Medicine, 152, 1010–1014.PubMedGoogle Scholar
  32. 32.
    Patrick, H., & Nicklas, T. (2005). A review of family and social determinants of children’s eating patterns and diet quality. Journal of American College of Nutrition, 24, 83–92.Google Scholar
  33. 33.
    Seth, J. G., Evans, A. E., Harris, K. K., Loyo, J. J., Ray, T. C., Spaulding, C., et al. (2007). Preschooler feeding practices and concerns: Differences among Spanish- and English-speaking WIC clients. Family & Community Health, 30(3), 257–270.Google Scholar
  34. 34.
    Fischer, L. L., Birch, D., marlin, D., & Kramer, L. (1982). Effects of instrumental eating on children’s food preferences. Appetite, 3, 125–134.Google Scholar
  35. 35.
    Herzler, A. A. (1983). Children’s food patterns—A review. II. Family and group behavior. Journal of the American Dietetic Association, 83, 555–560.Google Scholar
  36. 36.
    Rivera, J., Barquera, S., Gonzalez-Cossio, T., Olaiz, G., & Sepulveda, J. (2008). Nutrition transition in Mexico and in other Latin American countries. Nutrition Reviews, 62, S149–S157.CrossRefGoogle Scholar
  37. 37.
    Johnson, S., Clark, L., Goree, K., O’Connor, M., & Marquez-Zimmer, L. (2008). Healthcare providers’ perceptions of the factors contributing to infant obesity in a low-income Mexican-American Community. JSPN, 13, 180–190.PubMedGoogle Scholar
  38. 38.
    González, C., & Alcañiz L. (2006) Gordito doesn’t mean healthy: What every latina mother needs to know to raise happy, healthy kids. New York: Penguin Publishing Group.Google Scholar
  39. 39.
    Carruth, B. R., Skinner, J., Houck, K., Mran, J., I. I. I., Coletta, F., & Ott, D. (1998). The phenomenon of “Picky Eater”: A behavioral marker of eating patterns of toddlers. Journal of the American College of Nutrition, 17, 180–186.PubMedGoogle Scholar
  40. 40.
    Jacobi, C., Agras, W. S., Bryson, S., & Hammer, L. (2003). Behavioral validation, precursors, and concomitants of picky eating in childhood. Child & Adolescent Psychiatry, 42, 76–84.CrossRefGoogle Scholar
  41. 41.
    Whitaker, R., Pepe, M., Wright, J., Seidel, K., & Dietz, W. (1998). Early adiposity rebound and the risk of adult obesity. Pediatrics, 1001, e5.CrossRefGoogle Scholar
  42. 42.
    Olvera-Ezzell, N., Power, T., & Cousins, J. H. (1990). Maternal socialization of children’s eating habits: Strategies used by obese Mexican-American mothers. Child Development, 61, 395–400.PubMedCrossRefGoogle Scholar
  43. 43.
    Dave, J. (2007). Assessing the home, parental, intrapersonal, and demographic factors influencing fruit and vegetables intake among Hispanic children in elementary schools. Unpublished Dissertation.Google Scholar
  44. 44.
    Devaney, B., Kalb, L., Briefel, R., Zavitsky, T., Clusen, N., & Ziegler, P. (2004). Feeding infants and toddlers study: Overview of the study design. Journal of the American Dietetic Association, 104, S8–S13.PubMedCrossRefGoogle Scholar
  45. 45.
    Rosal, M. C., Carbone, E. T., & Goins, K. V. (2003). Use of cognitive interviewing to adapt measurement instruments for low-literate Hispanics. Diabetes Education, 29(6), 1006–1017.CrossRefGoogle Scholar
  46. 46.
    Andreyeva, T., Blumenthal, D., Schwartz, M., Long, M., & Brownell, K. (2008). Availability and prices of foods across stores and neighborhoods: The case of New Haven, Connecticut. Health Affairs, 27, 1381–1388.PubMedCrossRefGoogle Scholar
  47. 47.
    Morland, K., Wing, S., Diez Roux, A., & Poole, C. (2002). Neighborhood characteristics associated with the location of food stores and food service places. American Journal of Preventive Medicine, 22, 23–29.PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  • Alexandra Evans
    • 1
  • Jennifer Greenberg Seth
    • 2
  • Shanna Smith
    • 3
  • Karol Kaye Harris
    • 2
  • Jennifer Loyo
    • 2
  • Carol Spaulding
    • 2
  • Mary Van Eck
    • 4
  • Nell Gottlieb
    • 2
  1. 1.Michael & Susan Dell Center for Advancement of Healthy Living, Health Promotion and Behavioral SciencesUniversity of Texas School of Public HealthAustinUSA
  2. 2.Department of Kinesiology and Health Education, College of EducationThe University of Texas at AustinAustinUSA
  3. 3.Division of Statistics & Scientific Computation, College of Natural SciencesThe University of Texas at AustinAustinUSA
  4. 4.Texas Department of State Health ServicesNutrition Services SectionAustinUSA

Personalised recommendations