OBJECTIVE. This study explored parents' perceptions about their child's appearance and health and evaluated a tool to determine parents' visual perception of their child's weight.
METHODS. Parents of children aged 2 to 17 years were surveyed concerning their child's appearance and health and opinions about childhood overweight. They also selected the sketch (from 7 choices) that most closely matched the body image of their child using 1 of 8 gender–and age-range–specific panels of sketches. Children's height and weight were measured. Respondents were grouped by child body mass index (BMI) percentile (<5th, 5–84th, 85–94th [at risk for overweight (AROW)], and ≥95th [overweight]). Those with BMI ≥5th percentile were analyzed. Logistic regression was used to examine factors influencing parental perceptions and levels of worry about their overweight or AROW child.
RESULTS. Of the 223 children, 60% were <6 years old, 42% were male, 17% were black, 35% were Hispanic, 42% were white, and 7% were other; 19% were AROW, and 20% were overweight. Few parents (36%) identified their overweight or AROW child as “overweight” or “a little overweight” using words, but more (70%) selected a middle or heavier sketch. Among parents of overweight and AROW children, 18% recalled a doctor’s concern and 26% were worried about their child’s weight. If the overweight or AROW child was age ≥6 years, parents were more likely to identify their child as “overweight” or “a little overweight” using words, select a middle or heavier sketch, and to be worried. Parents of older children were more likely to be worried if they perceived their child as less active/slower than other children or recalled a doctor’s concern.
CONCLUSIONS.Few parents of overweight and AROW children recognized their child as overweight or were worried. Recognition of physical activity limitations and physicians’ concerns may heighten the parent's level of concern. Sketches may be a useful tool to identify overweight children when measurements are not available.
obesity is prevalent among children of all ages.1 Physical, social, and emotional consequences of obesity may be evident even in very young children and may persist into adulthood.2 The achievement of behavior change, for both the family and child, is a key focus of interventions that address the multiple factors contributing to the development and persistence of childhood obesity.3 To accomplish family behavior change, parents must be well engaged in the process.
A first step in the treatment of overweight is to help families recognize their child's overweight status and achieve a willingness to make behavior changes.3,4 Parents who do not recognize that their child is overweight or do not feel that their child's overweight status represents a problem worthy of intervention may not be ready to receive counseling or interventions related to overweight.
High rates of maternal misperceptions of their child's overweight status have been reported in several studies. In a sample of mothers recruited at primary care and Special Supplemental Nutrition Program for Women, Infants, and Children sites, only 21% of the mothers of overweight, preschool children (as defined by ≥90th percentile weight for height) recognized their child as overweight.5 Similar findings were noted among a sample of parents of 3- to 5-year-old children in the United Kingdom.6 Mothers with fewer years of education were more likely to misclassify their overweight child as not overweight.5 Among mothers of overweight (body mass index [BMI] ≥95th percentile) 2- to 11-year-old children interviewed in the Third National Health and Nutrition Examination Survey, 32% considered their overweight child to be “about the right weight.”7 Furthermore, only 14% of mothers of boys and 29% of mothers of girls who were at risk for overweight (AROW) (BMI 85-94th percentile) considered their child to be “overweight.” Older children and those with higher BMI-for-age z scores were more likely to be correctly classified.7 Focus-group data from a sample of low-income mothers suggest that national growth references may not be perceived as relevant.8 Instead, these women defined overweight status by the child's social behaviors and physical limitations and would worry about their child's weight “if a child ceased to run and play with other children.”
This study was designed to assess parental perceptions and levels of concern about their child's weight and to assess their opinions about child overweight and their health behavior intentions. This will help identify factors that might influence the likelihood of families achieving the motivation to engage in obesity interventions. We were particularly interested in determining if (1) parental recall of the doctors' concerns about their child's weight, (2) their perception of their child's physical activity habits/ability compared with other children, and (3) family illness related to obesity, dietary practices, or low levels of physical activity might be associated with parental perceptions and levels of concern about their child's weight. Additionally, we sought to test newly developed, gender–and age-range–specific child sketches to determine whether a visual image could more accurately allow categorization of children's weight for height than the use of written questions when measurements are not available.
Data were collected in 2003 at 7 diverse pediatric practices, which were members of the Pediatric Practice Research Group, a Chicago-area practice-based research network.9 A consecutive sample of parents/guardians with a child aged 2 through 17 years completed a survey and consented for release of anthropometric data obtained on the day of survey from their child's medical record. The study was approved by the Children's Memorial Hospital Institutional Review Board.
The survey had 6 sections and took subjects ∼3 to 5 minutes to complete. The first section gathered information on demographic characteristics (child's age, race/ethnicity, and gender; and respondent's relationship to the child and education) and medical conditions perceived to limit the child's growth or physical activity. The 5 questions in the second section evaluated the respondent's perceptions of their child's weight for height and physical activity habits/ability. The first question assessed parents' perceptions of their child's weight using the question, “I feel my child is…” (response options: underweight, a little underweight, about the right weight, a little overweight, or overweight).5 This was followed by the question, “Did your child's doctor ever tell you that your child is gaining weight too fast or is overweight?” (response options: yes, no, or I don't know). The remaining 3 questions in this section were designed to gather perceptions on the child's physical activity habits and abilities, including parental perception of days per week the child is involved in active physical exercise and the child's physical exercise amount and running speed as compared with other children. The 8 questions in the third section assessed parental levels of concern about the health effects of overweight, their child's weight status, their degree of influence on their child's food choices and physical activity habits, and the importance of parents as role models for exercise and dietary habits. Five Likert-scale response choices were provided for these questions (response options: strongly agree, agree, neutral, disagree, or strongly disagree). Section 4 included 4 questions to assess the intentions of the respondent to modify family health behaviors related to diet, physical activity, and television viewing. Section 5 asked for the history of medical illnesses (heart attack, stroke, diabetes, and high blood pressure) for the parents and grandparents. For each item, the respondent indicated whether the illness had occurred and whether they believed it was related to smoking, alcohol use, diet, low physical activity, or obesity. Section 6 presented a series of sketches of children. Respondents were requested to circle the drawing that most resembled their child. Digital images of children across age and weight spectrums were viewed by a graphic artist (S.C.M.) to create the series of sketches, which were modified to reach consensus with 2 authors experienced in child nutritional assessment (A.J.A. and H.J.B.) and who also had viewed the photographs. In each of 4 age groups, 7 sketches were provided for boys (Fig 1) and a similar panel of sketches for girls (Fig 2). (Sketches are freely available for research use on request.) The middle image in each series of sketches was developed to represent a child at the 50th BMI percentile; other sketches were not tied to particular BMI percentile categories.
Test-retest reliability of parental perception of child's weight was assessed in-person for 24 parents on 2 occasions separated by 1 to 3 days. Percentage agreement between test and retest responses for word perception of child's weight was 96% and for sketch selection was 91.7% (2 subjects chose the next heavier sketch at retest).
Respondents were excluded from analyses if they failed to select a sketch, did not answer both of the key questions (word perception of their child's weight and worry about child's weight), or had irresolvable data issues (disagreement on child gender between survey and sketch selection or invalid measurements). Data from respondents with a child having a medical condition the parent perceived as limiting growth or physical activity were also excluded from analysis to reduce variability in perceptions and concerns related to such conditions. However, children whose parent specified as limited in physical activity because of common conditions of asthma, allergy, or attention-deficit/hyperactivity disorder were not excluded.
Children's BMI percentiles were computed from the Centers for Disease Control and Prevention 2000 Reference using Epi Info 3.3 (Centers for Disease Control and Prevention, Atlanta, GA). Respondents were then categorized into 4 groups according to their child's BMI percentile (<5th, 5–84th, 85–94th [AROW], or ≥95th [overweight]). χ2 and Fisher's exact tests, as appropriate, were used to examine bivariate associations. Significance was set at P < .05.
Among parents of overweight and AROW children, the influences of various factors on 3 dependent variables (recognition that their child as overweight or a little overweight [overweight by words], selection of a middle or heavier sketch, and level of worry about their child's weight) were examined using logistic regression. Factors examined included: recall of a doctor telling the respondent that their child was overweight or gaining weight too fast (doctor concern); perception that their child was less active/slower than other children; child's gender; child's BMI group (AROW versus overweight); child's age group (2–5 vs ≥6 years); parental education (high school or less versus more than high school); ethnicity (Hispanic versus non-Hispanic); and illness in a parent or grandparent related to diet, low physical activity, or obesity. Parental recognition of the child overweight by words and sketch selection were also considered in relation to parental worry.
Factors were first examined individually using Stata (Intercooled Stata 8.0, StataCorp, College Station, TX). However, significant interactions between the child's age group and perception that their child was less active/slower than other children required that factors be examined separately by child age group. Given the small number of parents of children 2 to 5 years old who identified their child as overweight by words or who were worried about their child's weight, additional modeling of these dependent variables for this age group was not attempted. Similarly, because few parents of older children selected 1 of the lighter sketches, additional modeling of this dependent variable for this age group was not attempted. Only univariate models are reported, because the sample size did not support multivariate modeling or the inclusion of practice as a random effect, which would have accounted for the cluster design.
Among the 499 parents or guardians who accompanied an age-eligible child to the office, 396 (79%) completed a survey, and for 261 (52%) of these, the child's height and weight were measured at the visit. Many subjects were not present for a well-child check and, therefore, did not have both height and weight recorded for that day. Among the 261 subjects with a survey and child growth measurements, 38 were excluded from analysis (13 excluded because of medical conditions, 12 with incomplete survey responses, 3 because of data inconsistencies, and 10 whose child was <5th BMI percentile). Thus, analyses were limited to surveys regarding the 223 children at ≥5th BMI percentile, which included 20 children whose parent reported them as having a limitation in growth or physical activity because of asthma, allergy, or attention-deficit/hyperactivity disorder.
Characteristics of the respondents and their children are shown in Table 1. Eighty-seven percent of respondents were the mother (11% father and 2% other relationship). Children of more highly educated parents were less likely to be overweight or AROW, and boys were more likely to be overweight than girls. Only 22% of the parents of overweight children and 14% of the parents of AROW children recalled their child's doctor ever mentioning that their child was gaining weight too fast or was overweight.
Parents of overweight children were the most likely to report that their child was a little overweight or overweight (Fig 3; Table 1). Similarly, they more often selected sketches of a heavier child (Fig 4; Table 1). The sensitivity of parents' choice of overweight or a little overweight to identify overweight or AROW in their child was 36% (95% confidence interval [CI]: 27–47%), whereas the sensitivity of selecting a middle or heavier sketch for their overweight or AROW child was 70% (95% CI: 60–79%). Specificity was 84% (95% CI: 77–90%) and 79% (95% CI, 71–86%) for identification of overweight and AROW children by words and by sketch. Among the parents of overweight and AROW children, all who identified their child as overweight by words selected a middle or heavier sketch for their child, and all of the parents who selected a lighter sketch did not identify their child as overweight by words.
Parents of overweight and AROW children did not perceive their child as more physically limited than nonoverweight children in terms of days per week the child engaged in physical activity or in comparison to physical activity habits/abilities of other children (Table 1). However, among parents of overweight and AROW children, the perception that their child was either less physically active or ran slower than other children was strongly related to child age group (<6 years, 4% [2 of 48] perceived as less active/slower versus ≥6 years, 38% [15 of 40] perceived as less active/slower; Fisher's exact test, P < .001).
A majority of parents agreed that they could influence their child's food choices and amount of physical activity (Table 2). However, parents of overweight children were less likely to perceive themselves as having the ability to influence their child's physical activity. Among parents of overweight and AROW children, 26% were worried about their child's weight. Parents in all of the groups held similar opinions about the effects of overweight on children and parental influences on children's physical activity and eating habits.
Overall, 44% of subjects reported a parent or grandparent with a history of an illness related to diet, low physical activity, or obesity (heart attack: 15%, stroke: 6%, diabetes: 21%, and hypertension: 28%). The frequency of overall history of an illness or of individual types of illness related to diet, low physical activity, or obesity did not significantly vary by child BMI percentile group (all P > .05, data not shown). An additional 26% of subjects reported 1 of these illnesses in a parent or grandparent but did not relate their occurrence to diet, low physical activity, or obesity.
A majority of parents reported that they already practiced healthy behaviors related to limiting purchasing of junk food and limiting their child's intake of sweetened beverages or juice (Table 3). Approximately one half of the parents reported limiting their child's television viewing. Fewer parents reporting exercising for 30 minutes ≥5 days per week (39%), but many (48%) had good intentions. Responses to these questions did not vary by child BMI percentile group (all P > .05, data not shown). Furthermore, the occurrence of an illness related to diet, low physical activity, or obesity in a parent or grandparent was not significantly associated with these health behavior intentions (all P > .05, data not shown).
Recognition of and Concern About the Child's Weight Status Among Parents of Overweight and AROW Children
We next limited the data to responses from parents of overweight and AROW children to examine factors influencing their perception of their child weight status and their level of concern about their child's current weight. Child age influenced parental recognition of overweight and their level of concern about the child's weight. Parents with a child ≥6 years of age were more likely to recognize their child as overweight by words than were those with a younger child (odds ratio [OR]: 7.0; 95% CI: 2.3–21.4); parents of 56% (23 of 40) of the older children versus only 18% (9 of 48) of parents of the younger children recognized this. Selection of a middle or heavier sketch was more frequent for those with older children (OR: 3.1; 95% CI: 1.1–8.4); parents of 82% (33 of 40) of the older children versus 60% (19 of 48) of parents of the younger children selected a middle or heavier sketch. Additionally, parents of children ≥6 years old were more likely than parents of younger children to be worried about their child's weight (OR: 4.0; 95% CI: 1.3–12.4); parents of 40% (16 of 40) of the older children versus only 15% (7 of 48) of parents of the younger children were worried.
Interactions between the significant factors, child age group and parental perceptions of their child's physical activity habits/ability, led us to further examine the influences on parental perceptions and concern separately by age group. Because of the low numbers of parents of younger children who identified their child as overweight by words or were worried about their child's weight and the low numbers of parents of older children who did not select a middle or heavier sketch, an examination of factors influencing overweight by words and worry about their child's weight for younger children and sketch selection for older children were not conducted. Thus, for parents of younger children, we only examined influences on sketch selection and found none of the other 7 factors (doctor concern, perception that their child was less active/slower than other children, child's gender, child's BMI group, parental education, ethnicity, and illness in a parent or grandparent) to be significant.
For parents of older children, we examined influences of the remaining 7 factors on the dependent variables, recognition of overweight by words and worry about their child's weight. Parents of older children who perceived their child as less active/slower than other children were more likely to recognize their child as overweight by words (OR: 9.8; 95% CI: 1.8–52.8). Recall of a doctor's concern also increased parental recognition of overweight by words, but this only approached significance (OR: 8.5; 95% CI: 0.95–76.6). Parents of older children were more likely to be worried about their child's weight if they perceived their child as less active/slower than other children (OR: 6.3; 95% CI: 1.5–26.0), recalled a doctor's concern (OR: 8.5; 95% CI: 1.5–49.4), or recognized their child as overweight using words (OR: 16.1; 95% CI: 2.1–121.1). Multivariate models were not attempted because of the low sample size. Other factors considered did not approach significance for either dependent variable.
A needed step before entering a treatment program that applies strategies to accomplish behavior change is to recognize the problem that is the focus of the treatment. Recognition of the problem will facilitate the development of the degree of concern that will motivate behavior change. In this sample, few parents of overweight and AROW children felt that their child was overweight or even a little overweight, and few were worried about their child's weight. This foretells difficulty with recruiting families into treatment programs for child overweight and achieving behavior changes in response to counseling on child overweight conditions. One ray of hope is the association of recognition and concern with recall that the child's doctor has told the parent that their child is gaining weight too fast or is overweight. This type of advice is likely to be given more routinely as determination of children's BMI in compliance with the guideline for the prevention of pediatric overweight and obesity10 becomes more common. In our sample, physician advice on the child's overweight status was infrequently recalled, even by the parents of the most overweight children. With heightened national concerns about obesity,11 parents will also receive similar messages through media and other sources. Given the strong relationship between parental recognition and concern about their child's overweight, ways to foster an accurate perception of their child's weight status should be developed and tested.
Among parents of older overweight or AROW children, the perception that their child's activity habits/ability was more limited in comparison with other children was associated with recognition of their child as overweight. This supports the qualitative research findings8 and suggests that use of measures of fitness that are user-friendly in primary care, for example, a step test,12 may be one way to increase parental recognition of a child's overweight or AROW status.
Lack of recognition of the child's overweight status and lower levels of concern about the child's weight were common among the parents of the younger children. Although lower levels of concern may be appropriate, because risk of long-term overweight status for overweight children begins at 3 years old but increases with child age,13–15 the lack of recognition of child overweight status may foretell difficulty recognizing and addressing health behaviors that have led to the overweight status in young children. Misperception of a young child's overweight status has been reported among samples of Hispanic and black parents.5,8,16 Our sample includes black, Hispanic and white parents, but because so few parents of younger children perceived their child as overweight by words, we were unable to examine differences in perception by racial/ethnic group. For older children, ethnicity did not significantly influence perception of the child as overweight by words, which would support the findings of Maynard et al.7 Unlike other studies, we did not find child gender to influence parent's correct recognition of a child's overweight status.5,7
Many families reported illnesses among the child's parents or grandparents that they perceived to be related to diet, physical inactivity, or obesity. Additional evaluations of the impact of family history should be examined in larger samples. The health-belief model includes framing counseling in the context of the perceived recognition of the consequences of the health behavior.4 The effects of counseling the parent on the consequences for the child of unhealthy dietary and physical activity habits are unknown. Focus-group findings indicate that parents would welcome a greater emphasis on the consequence of unhealthy behaviors to motivate them to change family health practices.17
Previous research has shown that the parents' activity and dietary patterns can be used to predict the children's risk of obesity,18 and most parents agreed that their exercise and eating habits influence the habits of their children. However, the ability of the pediatrician to influence parental habits, as recommended by the American Academy of Pediatrics guideline,10 has not been tested. The vast majority of parents felt that they could influence their child's food choices and amount of physical activity. Parents of overweight children were less confident in their ability to influence their child's physical activity. We did not examine how child age may influence these perceptions, because parents of older children may recognize how the child's independence places limitations on parental control over such factors. Additionally, other parental characteristics, including health conditions, obesity status, dietary and physical activity habits, and mental health conditions may impact parent's perceived ability to influence their child's health habits. These merit additional examination.
Most parents perceived themselves as already practicing healthy behaviors related to junk foods and sweetened beverages. Parents were less likely to report success with physical exercise for themselves and television viewing limitations for their children. However, many intended to improve their habits. In-office counseling specific to the parental stage of change in the behavioral process may be an important influence on the parent's ability to make change. As little as 3 minutes of clinician counseling directed to smokers to encourage quitting has been shown to be effective,19 and brief strategies of counseling on habits related to overweight need to be similarly evaluated.
Our child sketches more sensitively assessed the child's BMI group than did parental report by words. These sketches might prove useful to categorize siblings or other child household members when weight and height information is not available. Adult silhouettes have been successfully used in the past to categorize parental weight status20 and would be an interesting addition to studies on parental perceptions. One additional consideration may be to examine the influence of the order of the sketches on sketch choice.21,22
The major limitation of this study is its relatively small sample size. The influences of factors other than child age on perceptions and parental levels of concern about child overweight should be viewed as exploratory because of the limited sample size. However, the finding that so few parents of overweight or AROW children perceive their child as overweight and that most lack concern about their child's weight status is an important finding. Ways clinicians can heighten parental awareness and concern need to be examined. The linkage between perception and level of concern for the parent is also a new finding and merits additional investigation in a larger sample. Second, we had no way of verifying what messages health providers had conveyed previously to parents about their AROW and overweight children's health. Third, the children in the sample were relatively young. Evaluations among parents of older children would be beneficial. Fourth, measurements of the children were those obtained by the practice staff members. Inaccuracies may have led to BMI group misclassification. However, we feel that our conclusions would not have changed by misclassification of a few individuals.
In this sample of parents, we found low levels of recognition of children's overweight status and relatively few parents of overweight or AROW children to be concerned about their child's weight. Few parents recalled that their child's doctor had ever counseled that their child was overweight or gaining weight too fast, but parents of older overweight and AROW children who recalled this advice were more likely to recognize their child's overweight. Those who recognized their child as overweight were more likely to be worried. Identification of counseling strategies for pediatric practice that heighten parental awareness and concern about their child's overweight condition may be a needed first step in motivating families to participate in intervention programs that address child overweight.
Research support was provided by the Society for Pediatric Research in the form of a Medical Student Research Program Grant (to Dr Eckstein) and the Children’s Memorial Hospital Resident Research Program.
We thank the research associates who assisted with study processes and the following practices, their staff members, and the named physicians who guided data collection efforts at their offices: Barbara Bayldon, MD (Children's Memorial Pediatrics-Uptown, Chicago, IL); Sandra Sanguino, MD, MPH (Children's Memorial Pediatrics-Lincoln Park, Chicago, IL); Bennett Kaye, MD (Children's Health Care Associates, Chicago, IL); Lori Walsh, MD (Glenbrook Pediatrics, Glenview, IL); Lisa McKenna, MD (Near North Health Service Corporation, Winfield-Moody Health Center, Chicago, IL); Tim Wall, MD (Pediatric Health Associates, Naperville, IL); and Dennis Vickers, MD (Infant Welfare Society, Chicago, IL).
- Accepted July 5, 2005.
- Address correspondence to Helen J. Binns, MD, MPH, Children's Memorial Hospital, 2300 Children's Plaza, 157, Chicago, IL 60614. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
Dr Eckstein's current address is Harvard Longwood Psychiatry Residency Training Program, Boston, MA 02215.
Dr Mikhail's current address is Child and Adolescent Health Associates, Chicago, IL 60610.
Mr Millard's current address is DynoMed, a ChartLogic Company, Indianapolis, IN 46202.
- ↵Barlow SE, Dietz WH. Obesity evaluation and treatment: expert committee recommendations. Pediatrics.1998;102(3). Available at: www.pediatrics.org/cgi/content/full/102/3/e29
- ↵Baughcum AE, Chamberlin LA, Deeks CM, Powers SW, Whitaker RC. Maternal perceptions of overweight preschool children. Pediatrics.2000;106:1380–1385
- ↵Jain A, Sherman S, Chamberlin LA, Carter Y, Powers SW, Whitaker RC. Why don't low-income mothers worry about their preschoolers being overweight? Pediatrics.2001;107:1138–1146
- ↵LeBailly S, Ariza A, Bayldon B, Binns HJ, for the Pediatric Practice Research Group. The origin and evolution of a regional pediatric practice-based research network: practical and methodological lessons from the Pediatric Practice Research Group. Curr Probl Pediatr Adolesc Health Care.2003;33:124–134
- ↵American Academy of Pediatrics, Committee on Nutrition. Prevention of pediatric overweight and obesity. Pediatrics.2003;112:424–430
- ↵US Department of Health and Human Services. The Surgeon General's Call to Action to Prevent and Decrease Overweight and Obesity. Rockville, MD: Office of Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, National Institutes of Health, and Washington, DC: US Department of Health and Human Services, Public Health Service, Office of the Surgeon General; 2001. Available at: www.surgeongeneral.gov/topics/obesity/calltoaction/CalltoAction.pdf. Accessed April 7, 2005
- ↵Trevino RP, Marshall RM Jr, Hale DE, Rodriguez R, Baker G, Gomez J. Diabetes risk factors in low-income Mexican-American children. Diabetes Care.1999;22:202–220
- Guo SS, Chumlea WC, Roche AF. Predicting overweight and obesity in adulthood from body mass index values in childhood and adolescence. Am J Clin Nutr.2002;76:653–658
- ↵Ariza AJ, Greenberg RS, LeBailly SA, Binns HJ; Pediatric Practice Research Group. Parent perspectives on messages to be delivered after nutritional assessment in pediatric primary care practice. Ann Fam Med.2005;3(suppl 2):S37–S39
- ↵Fiore MC, Bailey WC, Cohen SJ, et al. Treating Tobacco Use and Dependence. Clinical Practice Guideline. Rockville, MD: US Department of Health and Human Services, Public Health Service; 2000:56–59
- ↵Stunkard AJ, Sorensen T, Schulsinger F. Use of the Danish adoption register for the study of obesity the thinness. In: Kety SS, Rowland LP, Sidman RL, Matthysse SW, eds. Genetics of Neurological and Psychiatric Disorders. New York, NY: Raven Press; 1983
- ↵Tourangeau R, Couper MP, Conrad F. Spacing, position, and order: interpretive heuristics for visual features of survey questions. Public Opin Q.2004;68:368–393
- ↵Krosnick JA, Alwin DF. An evaluation of a cognitive theory of response-order effects in survey measurement. Public Opin Q.1987;51:201–219
- Copyright © 2006 by the American Academy of Pediatrics