Little research has addressed the relationships among infant temperament, eating styles and obesity risk. To address this gap, we tested whether infant temperament and eating patterns at the age of 1 year are associated with a greater increase in standardized weight status, and greater obesity risk at 6 years of age.
A secondary, prospective analysis of the Colorado Adoption Study was conducted. The main predictor variables were infant temperament (that is, emotionality, activity, attention span-persistence, reaction to food and soothability) and eating domains (that is, reactivity to food, predictable appetite and distractability at mealtime) at the age of 1 year, along with the body mass index (BMI) of biological mothers. The outcome measures were child weight and height (length) assessed at ages 1 through 6 years, from which weight-for-length and BMI were computed along with the standardized indexes (z-scores) and percentiles. Overweight/obesity status was computed at each year as well.
A primarily White sample of 262 boys and 225 girls, assessed at ages 1 through 6 years, along with their mothers.
Among boys, greater attention span-persistence was associated with reduced standardized weight status gain (β=−0.15, P<0.05) and reduced obesity risk (odds ratio (OR)=0.46, P=0.06). Among girls, greater soothability and negative reaction to food were associated with greater standardized weight status gain (β=0.19, P<0.01; and β=0.16, P<0.05, respectively) and increased obesity risk (OR=3.72, P=0.03; and OR=2.81, P=0.08, respectively). Biological mothers’ obesity status predicted obesity risk in boys (OR=3.07, P=0.01) and girls (OR=5.94, P=0.03).
Male infants with less attention span, and female infants with greater soothability or a more negative food reaction, showed greater increases in standardized weight and were more likely to be overweight/obese at the age of 6 years. The role of infant temperament in pediatric obesity onset warrants greater research.
The prevalence of childhood obesity has increased dramatically in recent decades1, 2 even among newborns.3 Yet, few early-life behavioral risk factors for childhood obesity have been identified, with most investigations having focused on parental body mass index (BMI in kg m–2) or controlling feeding styles.4 Parental obesity is arguably the strongest predictor of childhood obesity risk,5 although this partially reflects nonmodifiable genetic influences.6, 7 Restrictive feeding by mothers is associated with increased child food intake and overweight,8 although parents may be responding to their child's weight status in a bidirectional manner. Specifically, being an overweight child may elicit restrictive feeding prompts by mothers, which in turn may impede children's self-regulatory eating and promote excess BMI gain.4, 9 This phenomenon has been studied in infancy8, 10, 11 and early childhood.4 Thus, research identifying early behavioral risk factors for obesity is needed.
Studies of infant eating styles and temperament may shed light on this issue. As regard to eating patterns, rapid sucking by newborns predicted increased weight gain at 1 and 2 years.12, 13, 14 Also, consumption of a more energy-dense liquid formula compared with a less energy-dense formula in the first 112 days of life was associated with greater total energy intake and weight gain.15 This classic study by Fomon et al.15 examined 32 full-term infants who were initially measured at 6–9 days of life and then subsequently evaluated after 14 (±2), 28 (±2), 42 (±2), 56 (±2), 84 (±4) and 112 (±4) days. Another line of research has shown that bottle-fed compared with breast-fed infants are at increased risk for developing obesity,16, 17 whereas a recent study found that approximately one-fifth of working mothers fed French fries, sweetened beverages and desserts to their infants.18 Thus, the role of infant eating patterns in obesity onset is potentially relevant. As regard to temperament, research indicates that temperaments are a subclass of personality traits that appear early on in the first year of life, and are largely heritable and enduring over time.19, 20 Empirical research shows that early childhood temperament predicts emotional reactivity, social competence, coping with stress, inhibitory control, approach-avoidance behavior and self-regulation.21, 22 Regarding the association between temperament and weight status, infants who gained at least 30 percentile points in weight-for-length (Wt/L) between 6 and 12 months had more difficult temperaments than infants showing less rapid weight gain.23, 24 However, Farrow and Blissett25 reported that infant temperament did not predict 6- to 12-month weight changes. High-activity temperament was associated with reduced body fat in 8- to 12-year-old girls,26 although this did not replicate prospectively.27 Agras et al.28 found that greater anger/frustration and less soothability at the age of 5 years were associated with greater child food tantrums, which, in turn, predicted greater obesity risk at 9.5 years. Thus, the role of infant temperament in obesity onset also is potentially relevant.
This study tested whether infant temperament and eating styles at 1 year of age were associated with a greater increase in standardized weight status, and increased obesity risk at the age of 6 years. We tested these associations controlling for the BMI (kg m–2) of biological mothers enrolled in the Colorado Adoption Project (CAP). It was hypothesized that greater maternal BMI and less infant soothability28 would predict child obesity risk. Directional hypotheses regarding other temperaments were not specified because previous inquiries differed notably from the present in terms of temperamental conceptualization,23 temperament assessment23, 29 and participant age.26
Mother and child participants were enrolled in the CAP, a full adoption design that includes data on adopted children and their biological and adoptive parents. Data were also collected on nonadoptive families that were matched with respect to infant sex, number of siblings, and paternal age, occupational status and education.30, 31 The year 1 assessment of infant temperament and eating style measures was conducted either with the biological parents (that is, in the case of children from nonadoptive families, N=245) or with adoptive parents (that is, in the case of children from adoptive families, N=242). Biological parents were recruited through two Denver adoption agencies, which placed infants in temporary foster homes until being placed with their permanent adoptive families. Nonadoptive families were recruited at local hospitals. The mean±s.d. years of education for CAP parents were as follows: 12.1±1.8 years (biological mothers); 12.3±1.8 years (biological fathers); 14.7±2.1 years (adoptive mothers); and 15.7±2.5 years (adoptive fathers). All procedures received institutional review board approval from the University of Colorado.
This analysis used data from the years 1 through 6 annual assessments. The years 1–4 assessments were conducted at home visits by trained research staff, whereas years 5 and 6 assessments were conducted by phone interview. Children in this analysis were 262 boys and 225 girls, with mean ages of 12.22±0.51 months (year 1), 24.13±0.37 months (year 2), 36.12±0.45 months (year 3), 48.17±0.42 months (year 4), 60.15±1.19 months (year 5) and 71.97±2.27 months (year 6). In all, 52% of the children were from adoptive families and 48% were from nonadoptive families. The ethnicity of biological mothers was 93.4% white, 0.4% African American, 1.0% Asian, 2.7% Mexican American, 0.8% unknown and 1.6% other (not specified).
Five temperament variables from the Colorado Childhood Temperament Inventory (CCTI),32 as reported by mothers at the year 1 assessment, were analyzed: emotionality, activity, attention span-persistence, reaction to food and soothability. The sociability variable was not examined as this temperament has not been linked to child obesity. A sample emotionality item is ‘child gets upset easily.’ Sample items from the activity and attention span-persistence factors are, respectively, ‘child is very energetic’ and ‘child persists at a task until successful.’ A sample reaction to food item is ‘child consistently dislikes many kinds of food,’ and a sample soothability item is ‘child stopped fussing whenever someone talked to him or picked him up.’ All temperament items were rated on a 1 (‘not at all like the child’) to 5 (‘a lot like the child’) scale. Validity and reliability data for the CCTI are provided elsewhere.32, 33
Infant eating behaviors
In addition to completing the five items from the CCTI reaction to food scale, mothers completed six eating- and food-related items from the New York Longitudinal Study (NYLS).34 The rationale for combining the CCTI and NYLS items was to broaden the content coverage of the eating behavior constructs. Specifically, these 11 items were selected in advance, based on their face content, to address the following constructs of interest: food fussiness or pickiness eating,35, 36, 37 appetite and hunger,38, 39, 40 and self-regulatory eating.41, 42 These constructs or traits have been suggested as being potential contributors to childhood obesity onset, but have received limited investigation during infancy. Thus, in the present investigation, we used these 11 items. These items were jointly factor analyzed by the present authors using principal components analysis (PCA) with varimax rotation and Kaiser normalization. The PCA yielded three factors with eigenvalues of 3.56, 1.26 and 1.16, respectively. These factors accounted for 54.3% of the variance in the interitem covariance matrix (32.3, 11.5 and 10.5% for factors 1, 2 and 3, respectively). The three factors were labeled reactivity to food, predictable appetite and distractability at mealtime. One CCTI item cross-loaded on two factors and therefore was excluded. Table 1 presents the rotated component matrix and factor loadings for all 10 items. Higher scores on reactivity to food reflect a more negative reaction to or fussiness about foods. Lower scores on predictable appetite reflect greater expression of a regular or predictable appetite. Lower scores on distractability at mealtime reflect greater distractability and activity at meals.
The Cronbach α coefficients for the three factors were 0.84, 0.41 and 0.43, respectively. Factors two and three consist entirely of NYLS items, and their low α values likely reflect the small number of items per factor, the limited range of response options, and the methodology used by the NYLS (that is, qualitative interviews with parents followed by content coding of the interviews). As the NYLS eating dimensions were not psychometrically derived, it is perhaps not surprising that the resulting reliability estimates are modest. However, if one conceptualizes the three eating factors as components of a multidimensional eating behavior construct and thus computes, across all three factors, an overall stratified α coefficient,43 the resultant α estimate is 0.78.
Maternal weight and height were collected by self-report.
Infant and child weight status
At years 1 through 4, weight and height were ascertained by medical records or by scale and tape measure. At years 5 and 6, weight and height were ascertained by parent report. Weight and length (or height) measures were used to compute Wt/L at year 1 and BMI at years 2–6. Wt/L z-score and percentile were computed for year 1, and BMI z-score and percentile for years 2–6, per Centers of Disease Control and Prevention growth charts.44
Handling missing data
To maximize generalizability and statistical power, multiple data imputation was performed using Schafer's NORM computer program.45 This program imputes data points by applying a Bayesian data augmentation algorithm that simulates random values of parameters and missing data from each variable's posterior distribution. First, the expectation maximization algorithm was applied to the data to generate starting values for parameters. To create a stable set of multiply imputed data, the data augmentation algorithm was repeated for 400 iterations to guarantee model convergence and this process was then repeated five times for a total of six multiple imputed data sets, consistent with previous recommendations.46 Each imputed variable was then averaged across the six data sets to give rise to a final complete data set for all 487 participants. The proportion of missing data for the year 1 variables ranged from 0 to 14.3%, with a mean (s.d.) and median value of 5.4% (4.4) and 7%, respectively. These values are well within acceptable limits.46, 47
Data analytic plan
Descriptive statistics are presented as means and s.d. Two complementary analyses tested whether child temperament and eating measures predicted changes in standardized weight status from ages 1 to 6 years and overweight/obesity status at the age of 6 years. The first analysis, hierarchical multiple regression, used as the outcome the difference between the child's standardized weight status at years 6 and 1, both assessed in z-score units. The predictor variables were maternal ethnicity, infant age, infant adoption status (that is, nonadoptive vs adoptive families) (block 1); maternal BMI (block 2); and infant emotionality, attention span, reaction to food, soothability, activity level, predictable appetite and distractability at mealtime (block 3) (reactivity to food was not modeled in inferential analyses because of its high correlation with the reaction to food temperament measure, r=0.96). This model tested the effects of temperament and eating variables after partialling out variance due to, first, demographic variables and, second, biological mother's BMI. The statistical significance and variance accounted for by each block are presented, along with the standardized β weight and P-value of each predictor. The overall significance and R2 for the full model are also presented.
The second analytic plan, logistic regression analysis, used child overweight/obesity status at 6 years of age as the outcome. Children were classified as nonoverweight/nonobese (that is, BMI <75th percentile) or overweight/obese (that is, BMI ⩾85th percentile). Children with BMIs ⩾75th and <85th percentiles were excluded from this analysis to create spread between the two outcome groups. Predictor variables were the same as those used in the previous analyses, with children scored as low or high on each respective measure based on a median split. Odds ratios (ORs) with 95% confidence intervals are presented for individual predictors. We tested the overall model fit using a χ2-test and examined the explanatory power of the predictor variables using Nagelkerke's R2 statistic. To evaluate the model's discriminatory power (that is, to correctly identify overweight/obese cases from non-cases at year 6), the percentage of cases correctly classified was examined along with the c-statistic,48 which spans from 0.5 (that is, model predictions are no better than chance) to 1.0 (that is, model always assigns higher probabilities to correct cases than to incorrect cases).
All analyses were stratified by child sex, to allow for the possibility that certain temperament or eating style influences may be present in one sex but not the other. Indeed, certain studies have found that associations between self-regulatory eating measures and overweight status or obesity risk differ by child sex.49, 50 All analyses used two-tailed significance tests with α=0.05. The Statistical Package for the Social Sciences (SPSS, Version 16, Chicago, IL, USA) was used for all analyses.
There were no significant differences on any of the infant temperament and eating measures across gender (Table 2). Similarly, there were no significant differences across gender in anthropometric measures except for Wt/L at year 1 (P<0.001) (Table 3). In all, 19 (8.4%) of the girls and 32 (12.2%) of the boys were overweight/obese at the age of 6 years.
Prediction models in boys
Hierarchical regression analysis revealed that none of the predictors in block 1 (P>0.05) or block 2 (P=0.20) were significant (Table 4a). However, in block 3, poorer attention span (P=0.02) was associated with greater changes in standardized weight status from 1 to 6 years. The overall regression model was significant (P=0.04), accounting for 8% of the variance in the outcome.
Logistic regression analysis revealed that sons of obese mothers were more likely to be overweight/obese at 6 years of age compared with sons of nonobese mothers (OR=3.07). Also, infants who are high in attention span were less likely to be overweight/obese than those low in attention span (OR=0.46) (Table 4b). Although the overall model was not significant (χ2(11)=16.37, P=0.13, Nagelkerke's R2=0.12), 87% of the cases were correctly classified and the c-statistic was 0.71 (P<0.001).
Prediction models in girls
Hierarchical regression analysis revealed that none of the predictors in block 1 (P>0.05) or block 2 (P=0.19) were significant (Table 5a). However, four variables in block 3 were significant, with the full block accounting for 9% of the variance (P=0.004). Greater negative reaction to food (P=0.02), greater soothability (P=0.01), greater predictable appetite (P=0.02) and greater distractability at mealtime (P=0.05) were associated with greater changes in standardized weight status from 1 to 6 years. The overall model was significant (P=0.01), accounting for 11% of the variance in the outcome.
Logistic regression analysis revealed that daughters of obese mothers were more likely to be overweight/obese at 6 years of age compared with daughters of nonobese mothers (OR=5.94), and highly soothable infants were more likely to be overweight/obese than less soothable infants (OR=3.72) (Table 5b). Also, girls who were high compared with low in negative reaction to food were 2.8 times more likely to be overweight/obese at year 6, although this effect did not reach significance (P=0.08). The overall model was significant (χ2(11)=22.45, P=0.02, Nagelkerke's R2=0.22), with 91% of the cases correctly classified and a c-statistic of 0.80 (P<0.001).
This is the first prospective study to show that infant temperament and eating patterns at 12 months of age are associated with change in standardized weight and obesity status at the age of 6 years. These associations were independent of biological mothers’ BMI and differed by child sex. Among boys, poorer attention span predicted greater weight gain. This finding is consistent with previous research in that (a) attention problems, aggression and oppositional defiant behaviors often co-occur;51, 52 and (b) behavior problems and oppositional defiant disorder in childhood predict future overweight/obesity status.53, 54, 55, 56 Interestingly, Carey et al.57 found that children who scored higher on an index of difficulty showed greater weight-for-height percentile gains at ages 4–5 and 8–9 years. Other studies have found that infants with greater distress to limitations gained more weight in the first 2 months of life.58 Behavioral mechanisms linking early attention span to obesity risk need to be identified. Reduced attention span may impede the development of self-regulatory eating and satiety recognition,41 promote eating in response to external cues such as portion size,59, 60 encourage mindless eating,61 or facilitate other obesity-related behaviors. Thus, our findings regarding reduced attention span and standardized weight gain in male infants builds upon current knowledge regarding the potential role of an emotional/difficult temperament in obesity risk. It is also possible that our findings for attention span reflect poor effortful control. Indeed, recent studies have established a prospective link between poorer inhibitory control and childhood obesity onset.62, 63, 64 Future research needs to resolve the unique contributions of attention span, self-control and negative emotionality to excess weight gain during childhood.
Among girls, greater soothability predicted greater change in standardized weight status and obesity onset at the age of 6 years. This effect was in the opposite direction to that which we predicted, as it seems that parents would be more inclined to use food to placate the less easily soothed infant compared with the more easily soothed infant. Thus, behavioral mechanisms underlying this association need to be identified. Parents may use food to comfort the highly soothable infant, which may over time enhance the reinforcing properties of foods.65, 66 Future research should explore such reciprocal behavioral mechanisms. Increased food reward has been linked to obesity status in children and adults.65, 67 Another possibility is that highly soothable girls are comforted more easily by watching television, which parents might then encourage. Increased television viewing has been linked to high-calorie snack food consumption68, 69, 70 and increased obesity risk.71 Future research needs to examine the extent to which television is used as a soothing strategy by parents, whether this differs by child temperament, and how this may lead to overconsumption. Interview-based research with parents of 6- to 13-year-old children indicates that many caregivers use television as a safe and easy soothing strategy.72 Interestingly, there is experimental evidence that television viewing can have an analgesic effect in 7- to 12-year-old children undergoing venipuncture.73
Given the novelty and potential importance of the soothability finding, we conducted a post hoc analysis (data not presented) to further probe the soothability effect in girls. Results indicated that standardized weight increased in both low- and high-soothability girls in a monotonic and parallel fashion until the age of 4 years, but then diverged between 4 and 6 years. This period overlaps with adiposity rebound, a possible critical period for obesity onset,74, 75 and may be a developmentally important juncture when temperament-based differences in diet and physical activity become apparent.76 Differences in peer relations, group socialization and emotional factors during this period might also have contributed to these findings. Interestingly, Agras et al.28 found that less soothability at the age of 5 years predicted increased obesity risk at 9.5 years. Differences in both temperament assessment and sampling frame (1–6 years vs 5–9.5 years) might account for the discrepant results. Although this study assessed soothability using the CCTI,32 Agras et al. used the Child Behavior Questionnaire to assess this trait.77 Wells et al.29 also found that less infant soothability was associated with increased skinfold thickness at ages 2.5–3 years. Future research should further explore the association between soothability and obesity risk, with an eye toward potential developmental changes over time.
Among girls, a more negative reaction to food predicted increased weight gain. This finding may reflect early control struggles in the infant–caregiver feeding relationship, resulting in food rejection by the infant. To compensate, parents might engage in assertive or coercive feeding strategies, which have been linked to overconsumption and obesity in children.4, 78 However, few empirical studies have examined these issues in infants.8, 79 Negative food reaction may also reflect picky or fussy eating,37 although few studies have examined these traits in relation to child weight status.35, 80 In response to a child's food dislikes and fussiness at mealtime, parents might overfeed their infant in an effort to placate her by offering calorically dense foods. Future studies should explore these issues.
A key methodological finding was the derivation of three empirically derived eating factors (that is, reactivity to food, predictable appetite and distractability at mealtime). This demonstrates not only that psychometrically developed eating factors can be derived from parental report, but that such measures can evidence predictive validity. A more predictable appetite and greater distractability at mealtime predicted greater weight gain for girls between years 1 and 6 in the hierarchical regression model. Controlling for maternal BMI, the temperament and eating variables accounted for 9% variance in weight change from years 1–6. Future research should replicate and extend these results. One recommendation is to develop additional items for eating factors two and three in an effort to boost the reliability of these eating dimensions, given that their suboptimal reliabilities are a limitation of this study. A second recommendation is greater research to establish the overall construct validity of these measures and their broader nomological network.81
Maternal BMI was a strong predictor of infants’ weight status in the logistic regression analyses, with infants of biologically obese mothers at an approximate threefold (boys) and sixfold (girls) increased risk of becoming overweight/obese compared with infants of nonbiologically obese mothers. These findings underscore the importance of maternal weight status in the development of childhood obesity. Moreover, as biological mothers did not reside with their adopted children, this implicates the strong role of genetic factors in the development of childhood obesity.82, 83
Our findings suggest potential practical implications for pediatricians and caregivers, pending greater research. Greater awareness should be paid to the role of infant temperament and its potential influence on parenting practices that may increase obesity risk. Caregivers of high soothable female infants might be particularly cautious about using food or television to calm their child, whereas caregivers of infant girls with negative food reaction might require additional patience and food exposures to promote fruit and vegetable consumption.84, 85, 86 For caregivers of infant boys with low attention span, it may be especially important during development to teach recognition of satiety cues or to restructure the home environment to discourage eating in response to external cues.87, 88, 89, 90
The strengths of this study include use of a long-term prospective design, the large sample size, a comprehensive data-analytic approach, and infant temperament and eating measures that have been rarely examined. Also, use of a full adoption study design allowed us to test for any influence of child temperament and eating style on weight status above and beyond that of biological mothers’ BMI, which solely reflects genetic influences that are known to be a strong determinant of child obesity status.5, 6 Limitations include the fact that directly measured weights and heights were not available for use in years 5 and 6 (contrary to the previous years), the CAP's use of primarily Caucasian families, the lack of intermediary diet and physical activity measures, and potential unreliability of weight and height measures from physicians’ offices. Although parental reported heights and weights for children are less optimal than direct measurements, there are data to suggest that the former are sufficiently valid for community-based samples.91 Also, participating children evidenced, on average, lower BMI z-scores and percentile ranks relative to normative data.44 This might be due to the geographical location from which the sample was recruited, as Colorado residents show higher levels of physical exercise and lower overweight/obesity rates.92
In summary, infants with less attention span (boys), as well as greater soothability or more negative food reaction (girls), showed greater increases in standardized weight and were more likely to be overweight/obese at the age of 6 years. Future research needs to replicate these findings and identify underlying behavioral mechanisms. Infant temperament and eating patterns may contribute to childhood obesity in a sex-specific manner that potentially operates through parenting styles.
Ogden CL, Carroll MD, Flegal KM . High body mass index for age among US children and adolescents, 2003–2006. JAMA 2008; 299: 2401–2405.
Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM . Prevalence of overweight and obesity in the United States, 1999–2004. JAMA 2006; 295: 1549–1555.
Kim J, Peterson KE, Scanlon KS, Fitzmaurice GM, Must A, Oken E et al. Trends in overweight from 1980 through 2001 among preschool-aged children enrolled in a health maintenance organization. Obesity (Silver Spring) 2006; 14: 1107–1112.
Faith MS, Scanlon KS, Birch LL, Francis LA, Sherry B . Parent-child feeding strategies and their relationships to child eating and weight status. Obes Res 2004; 12: 1711–1722.
Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH . Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med 1997; 337: 869–873.
Faith MS, Pietrobelli A, Nunez C, Heo M, Heymsfield SB, Allison DB . Evidence for independent genetic influences on fat mass and body mass index in a pediatric twin sample. Pediatrics 1999; 104: 61–67.
Maes HH, Neale MC, Eaves LJ . Genetic and environmental factors in relative body weight and human adiposity. Behav Genet 1997; 27: 325–351.
Fisher JO, Birch LL, Smiciklas-Wright H, Picciano MF . Breast-feeding through the first year predicts maternal control in feeding and subsequent toddler energy intakes. J Am Diet Assoc 2000; 100: 641–646.
Francis LA, Hofer SM, Birch LL . Predictors of maternal child-feeding style: maternal and child characteristics. Appetite 2001; 37: 231–243.
Taveras EM, Scanlon KS, Birch L, Rifas-Shiman SL, Rich-Edwards JW, Gillman MW . Association of breastfeeding with maternal control of infant feeding at age 1 year. Pediatrics 2004; 114: e577–e583.
Birch LL . Development of food acceptance patterns in the first years of life. Proc Nutr Soc 1998; 57: 617–624.
Stunkard AJ, Berkowitz RI, Schoeller D, Maislin G, Stallings VA . Predictors of body size in the first 2 y of life: a high-risk study of human obesity. Int J Obes Relat Metab Disord 2004; 28: 503–513.
Stunkard AJ, Berkowitz RI, Stallings VA, Schoeller DA . Energy intake, not energy output, is a determinant of body size in infants. Am J Clin Nutr 1999; 69: 524–530.
Agras WS, Kraemer HC, Berkowitz RI, Korner AF, Hammer LD . Does a vigorous feeding style influence early development of adiposity? J Pediatr 1987; 110: 799–804.
Fomon SJ, Filer Jr LJ, Thomas LN, Rogers RR, Proksch AM . Relationship between formula concentration and rate of growth of normal infants. J Nutr 1969; 98: 241–254.
Gillman MW, Rifas-Shiman SL, Kleinman K, Oken E, Rich-Edwards JW, Taveras EM . Developmental origins of childhood overweight: potential public health impact. Obesity (Silver Spring) 2008; 16: 1651–1656.
Gillman MW, Rifas-Shiman SL, Camargo Jr CA, Berkey CS, Frazier AL, Rockett HR et al. Risk of overweight among adolescents who were breastfed as infants. JAMA 2001; 285: 2461–2467.
Grzywacz JG, Tucker J, Clinch CR, Arcury TA . Individual and job-related variation in infant feeding practices among working mothers. Am J Health Behav 2010; 34: 186–196.
Buss A, Plomin R . A Temperament Theory of Personality Development. Wiley: NY, 1975.
Buss A . Personality: Evolutionary Heritage and Human Distinctiveness. Lawrence Erlbaum Associates: Hillsdale, NJ, 1988.
Rothbart MK, Sheese BE, Conradt L . Child temperament. In: Corr PJ, Matthews G (eds). Cambridge Handbook of Personality Psychology Cambridge. Cambridge University Press: UK, 2009, pp 177–190.
Shiner R, Caspi A . Personality differences in childhood and adolescence: measurement, development, and consequences. J Child Psychol Psychiatry 2003; 44: 2–32.
Carey WB . Temperament and increased weight gain in infants. J Dev Behav Pediatr 1985; 6: 128–131.
Bates JE, Freeland CA, Lounsbury ML . Measurement of infant difficultness. Child Dev 1979; 50: 794–803.
Farrow C, Blissett J . Does maternal control during feeding moderate early infant weight gain? Pediatrics 2006; 118: e293–e298.
Anderson SE, Bandini LG, Dietz WH, Must A . Relationship between temperament, nonresting energy expenditure, body composition, and physical activity in girls. Int J Obes Relat Metab Disord 2004; 28: 300–306.
Anderson SE, Bandini LG, Must A . Child temperament does not predict adolescent body composition in girls. Int J Obes (Lond) 2005; 29: 47–53.
Agras WS, Hammer LD, McNicholas F, Kraemer HC . Risk factors for childhood overweight: a prospective study from birth to 9.5 years. J Pediatr 2004; 145: 20–25.
Wells JC, Stanley M, Laidlaw AS, Day JM, Stafford M, Davies PS . Investigation of the relationship between infant temperament and later body composition. Int J Obes Relat Metab Disord 1997; 21: 400–406.
Eley TC, Deater-Deckard K, Fombonne E, Fulker DW, Plomin R . An adoption study of depressive symptoms in middle childhood. J Child Psychol Psychiatry 1998; 39: 337–345.
Petrill SA, Lipton PA, Hewitt JK, Plomin R, Cherny SS, Corley R et al. Genetic and environmental contributions to general cognitive ability through the first 16 years of life. Dev Psychol 2004; 40: 805–812.
Rowe DC, Plomin R . Temperament in early childhood. J Pers Assess 1977; 41: 150–156.
Emde R, Hewitt J, Kagan J . Infancy to Early Development: Genetic and Environmental Influences on Developmental Change. Oxford University Press: New York, 2001.
Thomas A, Chess S, Birch H . Temperament and Behavior Disorders in Children. New York University Press: NY, 1968.
Dubois L, Farmer AP, Girard M, Peterson K . Preschool children's eating behaviours are related to dietary adequacy and body weight. Eur J Clin Nutr 2007; 61: 846–855.
Galloway AT, Lee Y, Birch LL . Predictors and consequences of food neophobia and pickiness in young girls. J Am Diet Assoc 2003; 103: 692–698.
Dovey TM, Staples PA, Gibson EL, Halford JC . Food neophobia and ‘picky/fussy’ eating in children: a review. Appetite 2008; 50: 181–193.
Wardle J, Guthrie CA, Sanderson S, Rapoport L . Development of the children's eating behaviour questionnaire. J Child Psychol Psychiatry 2001; 42: 963–970.
Llewellyn CH, van Jaarsveld CH, Boniface D, Carnell S, Wardle J . Eating rate is a heritable phenotype related to weight in children. Am J Clin Nutr 2008; 88: 1560–1566.
Berkowitz RI, Moore RH, Faith MS, Stallings VA, Kral TV, Stunkard AJ . Identification of an obese eating style in 4-year-old children born at high and low risk for obesity. Obesity (Silver Spring) 2010; 18: 505–512.
Birch LL, Fisher JO . Development of eating behaviors among children and adolescents. Pediatrics 1998; 101: 539–549.
Johnson SL, Birch LL . Parents’ and children's adiposity and eating style. Pediatrics 1994; 94: 653–661.
Cronbach LJ, Schoonemann P, McKie D . Alpha coefficients for stratified-parallel tests. Educ Psychol Meas 1965; 25: 291–312.
Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat 11; 2002: 1–190.
Schafer J . Analysis of Incomplete Multivariate Data. Chapman and Hall/CRC Press: London, 1997.
Rubin DB . Multiple Imputation for Nonresponse in Surveys. John Wiley: NY, 1987.
Peng CY, Harwell M, Liou SM, Ehman LH . Advances in missing data methods and implications for educational research. In: Sawilowsky S (ed). Real Data Analysis. Information Age Publishing: Chrlotte, NC, 2007, pp 31–78.
Hanley JA, McNeil BJ . The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143: 29–36.
Hill C, Llewellyn CH, Saxton J, Webber L, Semmler C, Carnell S et al. Adiposity and ‘eating in the absence of hunger’ in children. Int J Obes (Lond) 2008; 32: 1499–1505.
Faith MS, Berkowitz RI, Stallings VA, Kerns J, Storey M, Stunkard AJ . Eating in the absence of hunger: a genetic marker for childhood obesity in prepubertal boys? Obes Res 2006; 14: 131–138.
Faraone SV, Biederman J, Weber W, Russell RL . Psychiatric, neuropsychological, and psychosocial features of DSM-IV subtypes of attention-deficit/hyperactivity disorder: results from a clinically referred sample. J Am Acad Child Adolesc Psychiatry 1998; 37: 185–193.
Hinshaw SP . On the distinction between attentional deficits/hyperactivity and conduct problems/aggression in child psychopathology. Psychol Bull 1987; 101: 443–463.
Lumeng JC, Gannon K, Cabral HJ, Frank DA, Zuckerman B . Association between clinically meaningful behavior problems and overweight in children. Pediatrics 2003; 112: 1138–1145.
Mustillo S, Worthman C, Erkanli A, Keeler G, Angold A, Costello EJ . Obesity and psychiatric disorder: developmental trajectories. Pediatrics 2003; 111: 851–859.
Pine DS, Cohen P, Brook J, Coplan JD . Psychiatric symptoms in adolescence as predictors of obesity in early adulthood: a longitudinal study. Am J Public Health 1997; 87: 1303–1310.
Mamun AA, O’Callaghan MJ, Cramb SM, Najman JM, Williams GM, Bor W . childhood behavioral problems predict young adults’ bmi and obesity: evidence from a birth cohort study. Obesity (Silver Spring) 2009; 17: 761–766.
Carey WB, Hegvik RL, McDevitt SC . Temperamental factors associated with rapid weight gain and obesity in middle childhood. J Dev Behav Pediatr 1988; 9: 194–198.
Darlington AS, Wright CM . The influence of temperament on weight gain in early infancy. J Dev Behav Pediatr 2006; 27: 329–335.
Fisher JO, Arreola A, Birch LL, Rolls BJ . Portion size effects on daily energy intake in low-income Hispanic and African American children and their mothers. Am J Clin Nutr 2007; 86: 1709–1716.
Orlet Fisher J, Rolls BJ, Birch LL . Children's bite size and intake of an entree are greater with large portions than with age-appropriate or self-selected portions. Am J Clin Nutr 2003; 77: 1164–1170.
van Ittersum K, Wansink B . Do children really prefer large portions? Visual illusions bias their estimates and intake. J Am Diet Assoc 2007; 107: 1107–1110.
Francis LA, Susman EJ . Self-regulation and rapid weight gain in children from age 3 to 12 years. Arch Pediatr Adolesc Med 2009; 163: 297–302.
Anzman SL, Birch LL . Low inhibitory control and restrictive feeding practices predict weight outcomes. J Pediatr 2009; 155: 651–656.
Tsukayama E, Faith MS, Toomey SL, Duckworth AL . Self-control protects against overweight status in the transition from childhood to adolescence. Arch Pediatr Adolesc Med 2010; 164: 631–635.
Temple JL, Legierski CM, Giacomelli AM, Salvy SJ, Epstein LH . Overweight children find food more reinforcing and consume more energy than do nonoverweight children. Am J Clin Nutr 2008; 87: 1121–1127.
Epstein LH, Leddy JJ, Temple JL, Faith MS . Food reinforcement and eating: a multilevel analysis. Psychol Bull 2007; 133: 884–906.
Saelens BE, Epstein LH . Reinforcing value of food in obese and non-obese women. Appetite 1996; 27: 41–50.
Wiecha JL, Peterson KE, Ludwig DS, Kim J, Sobol A, Gortmaker SL . When children eat what they watch: impact of television viewing on dietary intake in youth. Arch Pediatr Adolesc Med 2006; 160: 436–442.
Matheson DM, Killen JD, Wang Y, Varady A, Robinson TN . Children's food consumption during television viewing. Am J Clin Nutr 2004; 79: 1088–1094.
Matheson DM, Wang Y, Klesges LM, Beech BM, Kraemer HC, Robinson TN . African-American girls’ dietary intake while watching television. Obes Res 2004; 12 (Suppl): 32S–37S.
Jordan A . Television viewing and weight status in children and adolescents. J Child Media 2007; 1: 45–54.
Jordan AB, Hersey JC, McDivitt JA, Heitzler CD . Reducing children's television-viewing time: a qualitative study of parents and their children. Pediatrics 2006; 118: e1303–e1310.
Bellieni CV, Cordelli DM, Raffaelli M, Ricci B, Morgese G, Buonocore G . Analgesic effect of watching TV during venipuncture. Arch Dis Child 2006; 91: 1015–1017.
Whitaker RC, Pepe MS, Wright JA, Seidel KD, Dietz WH . Early adiposity rebound and the risk of adult obesity. Pediatrics 1998; 101: E5.
Dietz WH . Critical periods in childhood for the development of obesity. Am J Clin Nutr 1994; 59: 955–959.
Janz KF, Levy SM, Burns TL, Torner JC, Willing MC, Warren JJ . Fatness, physical activity, and television viewing in children during the adiposity rebound period: the Iowa Bone Development Study. Prev Med 2002; 35: 563–571.
Rothbart MK . Measurement of temperament in infancy. Child Dev 1981; 52: 69–78.
Birch LL, Marlin DW . I don’t like it; I never tried it: effects of exposure on two-year-old children's food preferences. Appetite 1982; 3: 353–360.
Taveras EM, Rifas-Shiman SL, Scanlon KS, Grummer-Strawn LM, Sherry B, Gillman MW . To what extent is the protective effect of breastfeeding on future overweight explained by decreased maternal feeding restriction? Pediatrics 2006; 118: 2341–2348.
Galloway AT, Fiorito L, Lee Y, Birch LL . Parental pressure, dietary patterns, and weight status among girls who are ‘picky eaters’. J Am Diet Assoc 2005; 105: 541–548.
Cronbach LJ, Meehl PE . Construct validity in psychological tests. Psychol Bull 1955; 52: 281–302.
Cardon LR . Genetic influences on body mass index in early childhood. In: Turner JR, Cardon LR, Hewitt JK (eds). Behavior Genetic Approaches in Behavioral Medicine. Plenum Press: New York, London, 1995, pp 133–143.
Faith MS, Rha SS, Neale MC, Allison DB . Evidence for genetic influences on human energy intake: results from a twin study using measured observations. Behav Genet 1999; 29: 145–154.
Birch LL . Effects of experience on the modification of food acceptance patterns. Ann NY Acad Sci 1989; 561: 209–216.
Birch LL . Preschool children's food preferences and consumption patterns. J Nutr Educ 1979; 11: 189–192.
Mennella JA, Nicklaus S, Jagolino AL, Yourshaw LM . Variety is the spice of life: strategies for promoting fruit and vegetable acceptance during infancy. Physiol Behav 2008; 94: 29–38.
Wansink B, Cheney MM . Super bowls: serving bowl size and food consumption. JAMA 2005; 293: 1727–1728.
Wansink B, Kim J . Bad popcorn in big buckets: portion size can influence intake as much as taste. J Nutr Educ Behav 2005; 37: 242–245.
Fisher JO, Birch LL . Eating in the absence of hunger and overweight in girls from 5 to 7 y of age. Am J Clin Nutr 2002; 76: 226–231.
Davison KK, Birch LL . Childhood overweight: a contextual model and recommendations for future research. Obes Rev 2001; 2: 159–171.
Garcia-Marcos L, Valverde-Molina J, Sanchez-Solis M, Soriano-Pérez MJ, Baeza-Alcaraz A, Martinez-Torres A et al. Validity of parent-reported height and weight for defining obesity among asthmatic and nonasthmatic schoolchildren. Int Arch Allergy Immunol 2006; 139: 139–145.
Brock DW, Thomas O, Cowan CD, Allison DB, Gaesser GA, Hunter GR . Association between insufficiently physically active and the prevalence of obesity in the United States. J Phys Act Health 2009; 6: 1–5.
We thank Dr Tanja Kral for her assistance deriving CDC z-scores and percentiles for anthropometric measurements.
The authors declare no conflict of interest.
About this article
Cite this article
Faith, M., Hittner, J. Infant temperament and eating style predict change in standardized weight status and obesity risk at 6 years of age. Int J Obes 34, 1515–1523 (2010). https://doi.org/10.1038/ijo.2010.156
- infant feeding
- familial transmission
International Journal of Eating Disorders (2020)
Journal of the American Heart Association (2020)
Childhood Obesity: The Relationship Between Negative Emotionality, Emotion Regulation, and Parenting Styles
Journal of Child and Family Studies (2019)
A biopsychosocial approach to processes and pathways in the development of overweight and obesity in childhood: Insights from developmental theory and research
Obesity Reviews (2019)
Association of Infant Temperament With Subsequent Obesity in Young Children of Mothers With Gestational Diabetes Mellitus
JAMA Pediatrics (2019)