Parental Perceptions of Overweight Counseling in Primary Care: The Roles of Race/ethnicity and Parent Overweight
Objective: To examine parental perceptions of primary care efforts aimed at childhood obesity prevention
Methods and Procedures: We interviewed 446 parents of children, aged 2–12 years, with an age- and sex-specific BMI ≥85th percentile; interviews occurred within 2 weeks of their child's primary care visit. We assessed parental ratings of the nutrition and physical activity advice received. Using children's clinical heights and weights and parents' self-reported heights and weights, we classified children into three categories: BMI 85th–94th percentile without an overweight parent, BMI 85th–94th percentile with an overweight parent (adult BMI ≥25 kg/m2), and BMI ≥95th percentile.
Results: In multivariate analyses, compared to parents of children with BMI ≥95th percentile, overweight parents with children whose BMI was 85th–94th percentile were more likely to report receiving too little advice on nutrition and physical activity (odds ratio (OR) 3.05; 95% confidence interval (CI) 1.49, 6.25) and to rate as poor or fair the quality of advice they received (OR 2.23; 95% CI 1.18, 4.24). Independently, African-American (OR 2.55; 95% CI 1.18, 5.51) and Hispanic/Latino (OR 2.78; 95% CI 1.27, 6.10) parents were more likely than white parents to rate as poor or fair the quality of advice they received.
Discussion: Parental overweight is associated with low subjective ratings of overweight counseling in pediatric primary care. Our findings of poorer perceived quality among racial/ethnic minority parents need further investigation.
During the past 30 years in the United States, the prevalence of overweight among youth has dramatically increased (1,2,3,4). Recent estimates indicate that ∼31% of 6- through 19-year-old children and adolescents are either overweight (BMI ≥ 95th percentile) or at risk (BMI 85th–94th percentile) in part due to their nutrition and physical activity patterns (4). Given the high rates and profound consequences of childhood overweight, its prevention has emerged as a national public health priority (5).
Primary care services could play a central role in addressing child and adolescent overweight. Continuity of care and frequent contacts with health-care providers offer opportunities to sustain individual motivation, assess progress, provide feedback, and adjust behavior change plans. Although research is needed on effective clinical approaches for the prevention and management of childhood overweight that can be implemented by primary care clinicians (6), a recent pediatric primary care based intervention to improve diet and physical activity among adolescents offers hope that such interventions can result in beneficial changes in overweight-related behaviors (7).
Children at highest risk of future obesity are likely to benefit most from prevention efforts in primary care. Overweight children have a high risk of becoming obese as adults (8,9), and parental overweight or obesity more than doubles that risk (10). For example, Whitaker et al. (10) reported that among 3- to 5-year-old children whose BMI exceeded the 85th percentile for age and sex, 21 of 55 (38%) became obese as young adults, including 24% (8/34) of children with no obese parent and 62% (13/21) with at least one obese parent. The subset of children whose BMI exceeded the 95th percentile had even higher risks of later obesity. In addition, African-American and Hispanic/Latino children are disproportionately burdened by overweight prevalence (11,12,13) and associated chronic disease (14), and substantial disparities exist in the prevalence of overweight based on socioeconomic status (15,16). No studies have examined the quality of primary care prevention efforts aimed at children at highest risk of future obesity and little is known about the role of parental overweight and parent/child socioeconomic factors in influencing quality ratings.
This study was designed to examine parents’ perceptions of primary care efforts for children at highest risk of developing obesity. Our aims were to (i) examine the nutrition and physical activity counseling parents received during their child's most recent primary care visit and (ii) evaluate associations between parents’ subjective ratings of quality of this counseling and parent/child anthropometric and socioeconomic factors, including race/ethnicity.
Methods and Procedures
We conducted a cross-sectional survey of parents/guardians (herein referred to as “parents”) of children, aged 2–12 years, with an age- and sex-specific BMI greater than the 85th percentile. We chose to focus on this group because we were interested in informing the design of appropriate preventive interventions in the clinical setting for children at high risk. A note on usage: some divide studies to moderate excess weight gain into prevention (for BMI < 95th percentile) and treatment or management (>95th). Others use the terms primary, secondary, and tertiary prevention (<85th, 85th–95th, >95th, respectively). For the purposes of this article, we simply use the term prevention irrespective of BMI. This usage reminds us that the ultimate goal is to prevent obesity-related morbidity, and is consistent with the recommendations of the recent Institute of Medicine report, “Preventing Childhood Obesity: Health in the Balance” (17). All parents were interviewed by telephone within 2 weeks of their child's most recent primary care, well-child visit (i.e., physical exam) at any of the 14 locations of Harvard Vanguard Medical Associates, a multispecialty group practice in the Greater Boston metropolitan area. Parents were ineligible if their child had a sibling who already participated in the study, the parent could not be interviewed in English or Spanish, or if someone besides the parent took the child to their primary care visit. The institutional review board at Harvard Pilgrim Health Care approved this study.
Each parent was mailed an introductory letter <1 week after his or her child's primary care visit. The letter included a toll-free number parents could call to decline participation. Those who did not decline to be contacted were telephoned at 1 week and invited to participate. The telephone interview took an average of 12 min. Interviews were conducted in either English or Spanish and consisted of 27 closed-ended questions. The interview questions were translated and backtranslated by a certified Spanish translator and administered to Spanish-speaking parents by a research assistant fluent in Spanish. Parents who completed the interview were mailed a gift certificate worth $10 as a token of appreciation.
From the computerized data, we identified an initial sample of 744 children who were seen by Harvard Vanguard Medical Associates clinicians (included pediatricians and pediatric nurse practitioners) for a primary care visit, were between the ages of 2 and 12 years, had address and telephone information available, and had a BMI over the 85th percentile. Of these, 21 parents (3%) called the toll-free number to decline participation, 76 parents (10%) actively declined to be interviewed, and 138 parents (17%) passively declined. Of the remaining 509 with whom we completed contact, 63 were ineligible because they did not speak English or Spanish (n = 8), their child had a sibling already enrolled in the study (n = 46), someone besides the parent took the child to the visit (n = 3), or other reasons (n = 6). Thus, we completed interviews with 446 parents.
To calculate the denominator for the interview completion rate, the American Association for Public Opinion Research recommends subtracting the number of ineligibles from the initial sample (18). The number of ineligibles among the groups who opted out before contact attempts (n = 21), actively (n = 76) or passively (n = 138) declined to be interviewed was estimated at 28 by multiplying the number in these groups by the proportion ineligible among the group with completed contact (12%). The estimated number of parents eligible for interview was 653 (744 − 63 ineligible on the basis of screening questions − 28 estimated ineligible among parents with incomplete contact). The estimated completion rate was thus 68% (446 of 653).
Main outcomes. The main outcome measures were: (i) parents’ overall satisfaction with the general pediatric care they received at their most recent primary care visit, (ii) parents’ reports of receiving nutrition and physical activity counseling from their child's clinician at their most recent primary care visit, and (iii) parent's rating of the amount and quality of nutrition and physical activity advice received at the visit.
We modified questions on satisfaction and quality of care from a validated instrument developed for the Group Health Association of America (19). We asked parents (i) how satisfied they were with the general pediatric care they received; response options were “very satisfied,” “satisfied,” “unsatisfied,” and “very unsatisfied,” (ii) how they would rate the amount of advice they received about nutrition and physical activity; response options were “too little,” “about right,” and “too much,” (iii) how they would rate the quality of advice they received about nutrition and physical activity; response options were “poor,” “fair,” “good,” “very good,” and “excellent,” and (iv) did their clinician spend enough time discussing their child's nutrition (and separately, their child's physical activity); response options were “yes” or “no.”
Other measures. We used children's clinical heights and weights obtained from their electronic medical records, along with parental reports of their own heights and weights, to classify the children into three categories based on age- and sex-specific percentiles developed by the Centers for Disease Control and Prevention (20): (i) BMI 85th–94th percentile without an overweight or obese parent (BMI < 25 kg/m2), (ii) BMI 85th–94th percentile with an overweight or obese parent (BMI ≥ 25 kg/m2), and (iii) BMI ≥ 95th percentile. Because ∼90% of parents of children with a BMI ≥ 95th percentile were themselves overweight or obese, we did not further subcategorize this group of children.
We asked parents to report their age, education, household income, and their racial/ethnic background. Parents reported their child's mean hours of television (TV)/video viewing per week based on questions from the National Longitudinal Study of Youth (21), the number of TV in the household, and whether the child had a TV in his or her bedroom. Parents also reported their child's weekly servings of fast food using a question adapted from a longitudinal study of adults (22), and their child's weekly servings of sugar-sweetened beverages (soda, juice, or other sweetened drinks). Finally, we also asked parents to rate the usefulness of several potential services their clinical practice could offer on childhood overweight prevention, including group classes, telephone advice, and one-on-one visits with nutrition specialists.
We first examined the bivariate relationships of socioeconomic characteristics with overweight risk categories using χ2 analyses for categorical variables and the Wilcoxon rank-sum test for ordinal or continuous variables. We also examined the bivariate relationship between overweight risk category and socioeconomic characteristics with parents’ report of receiving counseling on overweight-related behaviors during their child's most recent primary care visit. We then used multiple logistic regression models to assess the independent associations of overweight risk category and sociodemographic characteristics with parents’ rating of their satisfaction with, amount of, and quality of nutrition and physical activity advice received at the visit. We report odds ratios (ORs) and 95% confidence intervals (CIs) for each predictor. Because some parents in the study may have attended the same clinical practice site, we used logistic regression with generalized estimating equations (23) to adjust for the possible effect of clustering by clinical practice site. We conducted all analyses using SAS, version 8.0 (SAS Institute, Cary, NC).
Characteristics of study participants are summarized in Table 1. Of the 446 parents/guardians interviewed, 89% were the child's mother, 8% were the child's father, and 3% were the child's grandparent. Parents’ mean (s.d.) BMI was 26.8 (5.5) kg/m2; 57% of parents were overweight. Mean BMI (s.d.) percentile of children in the study was 94.0 (4.2).
Table 1. Parent and child characteristics by overweight risk group
Table 1 compares parent and child characteristics across overweight risk groups. Compared to children with a BMI ≥95th percentile, those with a BMI between the 85th and 95th percentiles without an overweight parent were more likely to have older parents (P = 0.005), have parents with higher educational attainment (P = 0.003), and more likely to be white (P = 0.01). In addition, they were less likely to have a TV in their bedrooms (P ≤ 0.0001), had fewer hours of screen time exposure (P = 0.02), and consumed less fast food servings per week (P ≤ 0.0001) than children with a BMI ≥ 95th percentile. In all characteristics and behaviors, we found no significant differences between children with a BMI between the 85th and 95th percentiles with an overweight parent and those with a BMI ≥ 95th percentile (Table 1).
Table 2 shows topics discussed by parents and clinicians at the most recent primary care visit. Parents were least likely to report receiving counseling on removing the TV from the room where their child sleeps, and on reducing TV viewing time.
Table 2. Nutrition and physical activity topics discussed during the most recent primary care visit
Parents of children with a BMI ≥95th percentile were more likely to report receiving counseling on nutrition and physical activity than parents of children with a BMI between the 85th and 95th percentiles (Table 3). We observed few differences by socioeconomic status in reported receipt of counseling on overweight-related behaviors (Table 3). In some cases (i.e., fast food intake and sugar-sweetened beverages), racial/ethnic minorities and low socioeconomic status parents reported receiving more counseling than white and high socioeconomic status parents (Table 3). We also found that parents of children 2–6 years of age were much less likely than parents of 7- to 13-year-olds to report receiving counseling on any of the topics.
Table 3. Bivariate predictors of receiving counseling on overweight-related behaviors during a primary care visit
Overall, 98% of parents reported they were “satisfied” or “very satisfied” with the general pediatric care received at their child's visit and there were no material differences among overweight risk groups or parent/child sociodemographic characteristics. However, 13% of parents reported they received “too little” advice about nutrition and physical activity during their visit, 17% rated as “poor” or “fair” the nutrition and physical activity advice received, 23% said their clinician did not spend enough time discussing nutrition, and 22% said their clinician did not spend enough time discussing physical activity.
In multivariate analyses adjusted for child's age, sex, and race/ethnicity, and parent's age, education, and household income, compared with parents of children with a BMI ≥95th percentile, overweight parents with children whose BMI was 85th–94th percentiles were more likely to report (i) receiving too little advice on nutrition and physical activity (OR 3.05; 95% CI 1.49, 6.25), (ii) their clinician did not spend enough time discussing nutrition (OR 2.84; 95% CI 1.62, 5.15) or physical activity (OR 2.63; 95% CI 1.46, 4.71), and (iii) to rate as poor or fair the quality of nutrition and physical activity advice received (OR 2.23; 95% CI 1.18, 4.24) (Table 4).
Table 4. Multivariate adjusted predictors of parental ratings of the quality of overweight counseling in primary care
Independent of overweight risk group, African-American (OR 2.55; 95% CI 1.18, 5.51), Hispanic/Latino (OR 2.78; 95% CI 1.27, 6.10), and Asian (OR 5.43; 95% CI 1.83, 16.13) parents were more likely than white parents to rate as poor or fair the quality of nutrition and physical activity advice they received. In addition, African-American and Hispanic/Latino parents were more likely than white parents to report their clinician did not spend enough time discussing physical activity (Table 4). Lower parental educational attainment was an independent predictor of a poor or fair rating of the quality of nutrition and physical activity advice received.
Individualized counseling and support from a nutritionist or health-care provider was chosen by 68% of respondents as a “very useful” service they could have been offered by their primary care practice in support of childhood overweight prevention. A separate clinic for overweight children within their primary care practice (56%) and family-based assistance in making healthy lifestyle changes (55%) were also rated as “very useful” services. Other services, including group classes on nutrition and physical activity (45%) and telephone advice (38%) were reported to be less frequently endorsed as “very useful.”
In this study of overweight and at risk of overweight children, we found that parents were, on average, very satisfied with the general pediatric care they received at their child's visit. However, overweight parents with children whose BMI was 85th–94th percentile reported less counseling and lower subjective ratings of the nutrition and physical activity advice they received, even though their children's dietary patterns and sedentary behaviors were nearly the same as those of children with a BMI ≥95th percentile. We also found that African-American, Hispanic/Latino, and Asian parents reported lower quality ratings of the nutrition and physical activity counseling they received. These differences persisted even after controlling for other socioeconomic factors related to subjective measures of quality of care including household income and education.
In a recent report, the Institute of Medicine recommended that child health professionals routinely monitor and track BMI and offer relevant evidence-based counseling and guidance to improve nutrition, increase physical activity, and decrease sedentary behaviors (17). Although many aspects of diet and physical activity are important in preventing overweight, both epidemiologic and experimental evidence from the past decade support targeting TV viewing, physical activity, intake of sugar-sweetened beverages, and fast food consumption. Several studies have examined clinician counseling on nutrition and physical activity during primary care visits (24,25), but no studies have reported on parent's perceptions of the nutrition and physical activity advice they received at their child's primary care visit and little is known about primary care prevention efforts aimed at children at highest risk of developing obesity.
In our study, overweight parents with children whose BMI was 85th–94th percentile reported poorer ratings for the nutrition and physical activity advice they received at their child's most recent primary care visit. In addition, despite reporting that their child had dietary and sedentary practices that were almost identical to overweight children, overweight parents with children whose BMI was 85th–94th percentile were less likely to receive counseling on TV viewing, fast food, sugar-sweetened beverages, and physical activity. Several factors may explain why this group of parents reported less counseling and had lower subjective ratings of counseling. First, these children might not be recognized by clinicians as being at high risk of later obesity. Several studies have shown parental obesity to be one of the strongest and most reliable predictors of later obesity in children (10,26) and some investigators have found that targeting parental obesity exclusively is effective in treating childhood overweight (27). Greater recognition of children at highest risk of overweight can be achieved through the use of BMI charts (28) and gathering a history of parental obesity. Second, the high prevalence of overweight in a pediatric practice might put a strain on available resources for managing children with a BMI <95th percentile, and many insurance carriers may deny claims submitted for children with a BMI between the 85th and 94th percentile. Third, clinicians may lack the confidence, proficiency, and training to address childhood overweight (29,30) and this might be especially true for addressing parental obesity in pediatric primary care.
Multiple studies have shown that African-American and Hispanic/Latino children are disproportionately burdened by overweight prevalence (11,12,13). Behaviors associated with overweight are also more prevalent among minority youth, including higher levels of TV viewing and more TV in bedrooms (31), higher consumption of sugar-sweetened beverages (32), and lower levels of physical activity among African-American and Hispanic/Latino youth compared to white youth (33,34). In this study, we found that African-American, Hispanic/Latino, and Asian parents were more likely than white parents to rate as poor or fair the quality of nutrition and physical activity advice they received. However, African-American, Hispanic/Latino, Asian, and white parents reported similar rates of counseling by their clinicians. Our results suggest that the poor ratings reported by African-American, Hispanic/Latino, and Asian parents might have been due to the content and duration of the counseling rather than not receiving counseling. Although the reasons for such disparities are likely multifactorial, disparities may arise from communication gaps between parents and their child's providers (35). Studies suggest that both patients and providers enter medical encounters with cultural explanatory models of health and illness. These models influence the interactions of patients and providers, the effectiveness of communication and decision-making processes, and the subsequent health behaviors of patients (36,37). Efforts to enhance parent-provider communication may help reduce disparities on overweight prevention counseling among racial/ethnic minorities.
In our analyses, we found that parents infrequently reported receiving counseling on TV viewing and removing TV from the room where their child sleeps. Our findings are in contrast to those of Kolagotla and Adams (25) who found that 73% of clinicians reported asking their overweight patients and families about TV viewing and video games during ambulatory visits. However, the study by Kolagotla and Adams. obtained clinicians’ reports about counseling while our study relied on parental report. Although our results may indicate low clinician counseling rates about TV, they may also reflect the topics that parents were more open to discussing. In a recent study of parent perspectives on messages delivered in pediatric primary care, Ariza et al. (38) found that parents were hesitant about clinician recommendations to limit TV viewing and thought the suggestion to limit TV viewing to 1–2 h per day was too restrictive. Further studies should examine counseling strategies for parents to reduce their child's TV viewing. Our findings of less counseling reported by parents of preschool age children suggest that clinicians may be missing an opportunity for preventive counseling at a young age.
Several limitations should be considered when interpreting our study. First, although mothers in the study had diverse racial/ethnic backgrounds, their educational and income levels were relatively high. Our results may not be generalizable to more socioeconomically disadvantaged populations. On the other hand, the narrow socioeconomic status of the study population may help reduce confounding by these factors. Second, this study relied on subjective parent-reported data of clinical practice vs. objective, observed data. Thus, parents may have chosen responses that seemed “correct” or more socially desirable, which can result in an overestimate of positive opinions about their general pediatric care. The tendency for respondents to give socially desirable answers is a limitation of all survey-based research. It is important to note that parents’ reports of counseling received may not represent the “truth.” However, parents’ beliefs and actual practices have direct effects on overweight prevention, and the perceived recommendations of their clinicians are important for many parents. Third, parents in the study recalled events that took place 1–2 weeks before the interview. The observed association between overweight risk category and lower perceived quality of overweight counseling could have stemmed from recall bias in that parents of overweight children may have been more likely than parents of at risk of overweight children to remember being counseled.
In conclusion, parental overweight was associated with low subjective ratings of overweight counseling in primary care. Family-based approaches to overweight counseling may be helpful in pediatric primary care given the high rates of parental obesity among overweight children. Our findings of racial/ethnic differences in subjective ratings of quality underscore the need for enhanced communication between clinicians and their racial/ethnic minority patients.
This work was supported in part by a grant from the Deborah Munroe Noonan Memorial Foundation. E.M.T. is supported in part by the Harold Amos Medical Faculty Development Program of the Robert Wood Johnson Foundation. S.L.G. was supported by a grant from the Centers for Disease Control and Prevention, Prevention Research Centers Grants U48DP000064-1. E.M.T. had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. We thank Elizabeth Pratt, Kate Saltus, Krista Garrod, and Catherine Gauthier for their research assistance. This work is solely the responsibility of the authors and does not represent official views of the Centers for Disease Control and Prevention.
The authors declared no conflict of interest.