Eating Behaviors 14 (2013) 274–277
Contents lists available at SciVerse ScienceDirect
Eating Behaviors
Picky eating in preschool children: The predictive role of the child's temperament and mother's negative affectivity Gertrud Sofie Hafstad a, b,⁎, Dawit Shawel Abebe a, b, Leila Torgersen a, Tilmann von Soest a, c a b c
Division of Mental Health, Norwegian Institute of Public Health, P.O. Box 4404 Nydalen, 0403 Oslo, Norway NOVA, Norwegian Social Research, Postboks 3223 Elisenberg, 0208 Oslo, Norway Department of Psychology, University of Oslo, Postboks 1094 Blindern, 0317 Oslo, Norway
a r t i c l e
i n f o
Article history: Received 19 September 2012 Received in revised form 6 February 2013 Accepted 3 April 2013 Available online 10 April 2013 Keywords: Prospective study Picky eating Children Temperament Maternal mental health
a b s t r a c t Objective: The objective of this study is to describe the development and examine predictors of picky eating from 1.5 to 4.5 years of age in a community sample of children. Methods: Mothers completed a questionnaire, assessing picky eating and a range of child and maternal factors, when their children were aged 1.5 (n = 913), 2.5 (n = 777), and 4.5 (n = 727) years. Results: Picky eating increased significantly from 1.5 to 4.5 years. Lower maternal age, higher levels of child emotionality, and maternal negative affectivity at the child's age 1.5 predicted an increase in picky eating from 1.5 years to 2.5 and 4.5 years. Having siblings protected against the development of picky eating. Conclusion: Child and maternal temperament at a very early stage in the child's life increase the risk for picky eating later on.
© 2013 Elsevier Ltd. All rights reserved.
1. Introduction Picky eating, i.e. the consumption of an insufficient amount or an inadequate variety of food through rejection of food items (Galloway, Fiorito, Lee, and Birch, 2005; Lewinsohn et al., 2005), has been found to affect children's nutritional status and weight and may thus constitute a significant health concern. In particular, it has been associated with an increased risk of being underweight (Dubois, Farmer, Girard, Peterson, and Tatone-tokuda, 2007), and overweight (Faith & Hittner, 2010), as well as with a deficiency in important vitamins and fiber in children's diet (Galloway, Lee, and Birch, 2003). Early eating problems may also constitute a risk for developing more serious eating disorders in adolescence, such as anorexia nervosa (Marchi and Cohen, 1990). Accordingly, it is important to understand the factors that contribute to these behaviors. A number of cross-sectional studies have shown individual factors, such as low birth weight (Dubois et al., 2007), negative affectivity and difficult temperament (Haycraft, Farrow, Meyer, Powell, and Blissett, 2011; Jacobi, Agras, Bryson, and Hammer, 2003), to be correlated with picky eating. Environmental concomitants of picky eating, such as low maternal age, parental financial problems and maternal negative
affectivity, have as well been identified (Dubois et al., 2007; Jacobi et al., 2003). However, only one prospective study has examined factors predicting picky eating in preschool children longitudinally. This study showed that infants with a lower sucking frequency during breastfeeding were significantly more likely to be picky eaters at ages 2, 3 and 5 years (Jacobi et al., 2003). The findings also indicated older siblings may affect the presence of picky eating; boys who had older siblings were less likely to be picky eaters, whereas girls with older siblings had a higher risk of being picky eaters at the age of 5 years (Jacobi et al., 2003). Although our understanding of what characterizes children who are picky eaters has increased over the last few years, we still lack findings from longitudinal studies about child and environmental risk factors for picky eating. In particular, understanding more of the characteristics of the children and their mothers that precede the development of picky eating may help determine whether prevention is possible at an early age. The aim of this study was therefore to examine the development of picky eating, and to identify child and maternal factors that prospectively predict picky eating. 2. Methods 2.1. Participants and procedures
⁎ Corresponding author at: Norwegian Institute of Public Health, P.O. Box 4404 Nydalen, N-0403 Oslo, Norway. Tel.: +47 23070000. E-mail addresses: gertrud.sofi
[email protected] (G.S. Hafstad),
[email protected] (D.S. Abebe),
[email protected] (L. Torgersen),
[email protected] (T. von Soest). 1471-0153/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eatbeh.2013.04.001
The sample consisted of Norwegian children and their mothers, who participated in the longitudinal community study ‘Tracking Problems and Opportunities’ (TOPP). Families were recruited during a scheduled visit to a child health clinic when their child was 18 months old (T1). Of
G.S. Hafstad et al. / Eating Behaviors 14 (2013) 274–277
1081 eligible families, 913 mothers (84%) completed a questionnaire at T1. Non-respondents did not differ from respondents with respect to maternal age, education, employment status, number of children, or marital status. The families were followed up through the health clinics when the children were 2.5 years old (T2) and 4.5 years old (T3). At T2, 777 mothers (85% of T1) participated, whereas 727 (80%) participated at T3. The study was approved by the Regional Committee for Medical and Health Research Ethics. 2.2. Measures 2.2.1. Child variables We used the two eating-related items from the Behaviour Checklist (Richman and Graham, 1971) designed to assess picky eating, i.e. lack of appetite and food fussiness. The items were scored 1–3, and mean scores were computed with higher scores reflecting more picky eating. The correlation between picky eating at age 1.5 and 2.5 was 0.44 (p b .001), and 0.52 (p b .001) between ages 2.5 and 4.5. Cronbach's α was 0.42 at T1, 0.48 at T2, and 0.47 at T3, with a mean corrected item-total correlation of 0.35. The low α might be expected due to the small number of items on the scale, and the average inter-item correlations are comparable to levels reported elsewhere for this type of scale. The concept, as measured by parental report, has been validated through a study including both parental reports and observational data of children in feeding situations (Jacobi et al., 2003). At T1, mothers reported child birth weight and whether they experienced any complications during delivery (yes/no). At T1 one question was included (yes/no) assessing whether the mother had experienced feeding problems in the child before the child was 3 months of age. Three items from the Behaviour Checklist (Richman and Graham, 1971), assessing difficulties falling asleep at bedtime, nightly awakenings, and unwillingness to sleep alone, were used to indicate problematic sleep-related behavior in the child. Cronbach's α was 0.64 at T1, 0.62 at T2, and 0.56 at T3. Childhood temperament was assessed at T1 with the 20-item Emotionality, Activity, Shyness and Sociability Temperament Survey (EAS) for children, which covers four dimensions (Buss and Plomin, 1984). Cronbach's α ranged from 0.71 to 0.79 for shyness, 0.62 to 0.67 for emotionality, 0.68 to 0.74 for activity, and 0.54 to 0.73 for sociability at the three assessment points. At all three time points, a sum score was computed based on parents' report of different possible chronic physical diseases of their children. Moreover, parents were asked to indicate whether their child had any physical impairment (yes/no). 2.2.2. Maternal variables Socio-demographic characteristics including age, marital status, education, and number of children in the family were assessed by selfreport at T1. The 25-item version of the Hopkins Symptom Checklist (SCL-25) measuring symptoms of anxiety and depression was used (Derogatis, Linman, Rickles, Uhlenhuth, and Covi, 1974). Two questions about suicidal ideation and loss of sexual interest or pleasure were excluded from the questionnaire after piloting. Cronbach's α was 0.90, 0.87, and 0.89 at T1, T2, and T3, respectively. Temperament was assessed by the EAS for adults, which covers five dimensions (Buss and Plomin, 1984). Cronbach's α ranged from 0.69 to 0.76 for shyness, 0.80 to 0.84 for activity, 0.56 to 0.57 for fearfulness, 0.73 to 0.75 for distress, and 0.57 to 0.61 for anger at the three time points. 3. Results Table 1 presents summary statistics for the studied variables at T1, T2, and T3. The average picky eating scores (possible range from 1 to 3)
275
indicated that overall, parents reported that their children were seldom or sometimes picky with a slight, but significant, increase in the mean score from T1 to T3 (t = 96.32, df = 756, p b .001). Likewise, the proportion of parents reporting that their children were sometimes or always picky increased from 22% at T1 to 35% at T2, and to 40% at T3. We developed a random intercept model to investigate the longitudinal associations among the predictor variables at T1 and picky eating at T2 and T3 as outcomes (Table 2). As physical illness and impairments could confound the associations, we controlled for these variables in all analyses. The mean picky eating score was modeled as a combination of population characteristics (β) assumed to be shared by all individuals (fixed effects) and subject-specific characteristics (random effects). Maximum likelihood estimates were applied. In Table 2, Model 1, the effect of each predictor variable was only controlled for picky eating and physical illness at baseline (T1). By controlling for initial levels of picky eating, we were able to examine whether a variable predicted subsequent changes in picky eating. In Model 1, low birth weight, higher levels of sleep problems, shyness, emotionality, and lower levels of sociability predicted increases in picky eating over the study period. Having siblings predicted a decline in picky eating. Moreover, mothers' negative affectivity, fearfulness, and distress all predicted increases in picky eating. In Model 2, all predictors were included simultaneously and only predictors with p b .2 were kept in the final fitted model. Initial higher levels of picky eating and emotionality remained significant predictors of picky eating at 2.5 and 4.5 years. Among the maternal variables, low maternal age and high negative affectivity served as significant predictors. Moreover, having a many siblings significantly reduced the level of picky eating between the ages of 2.5 and 4.5 years. 4. Discussion The finding that the child's emotionality was a longitudinal predictor of picky eating supports the reasonably well-established association Table 1 Summary statistics for the study variables. Time 1 N = 934 Child variables M/N (SD/%) Sex (girl) 476 (50.9%) Number of siblings No sibling 453 (48.5%) One 346 (37.0%) Two 109 (11.7%) Three or more 26 (2.8%) Birth weight in grams 3550 (610) Birth complications (Yes) 173 (18.5%) Feeding problem before 3 months (Yes) 48 (5.2%) Picky eating 1.38 (0.44) Sleep problems 1.41 (0.41) Shyness 2.01 (0.62) Emotionality 3.21 (0.62) Activity 4.32 (0.55) Sociability 3.80 (0.49) Maternal variables Maternal age Education 0–10 years 10–13 years (High school) >13 years (College/university) Marital status (Married) Negative affect Shyness Activity Fearfulness Distress Anger
Time 2 N = 802
Time 3 N = 757
M/N (SD/%) M/N (SD/%) 399 (49.7%) 387 (51.1%)
1.56 (0.47) 1.45 (0.42) 2.08 (0.61) 3.20 (0.61) 4.19 (0.56) 3.87 (0.42)
1.63 1.46 2.23 3.14 4.10 3.86
(0.46) (0.41) (0.64) (0.60) (0.59) (0.44)
1.30 (0.29) 2.24 (0.60) 2.90 (0.69) 3.79 (0.60) 3.67 (0.73) 3.06 (0.69)
1.28 2.27 2.83 3.82 3.69 3.08
(0.28) (0.62) (0.70) (0.57) (0.68) (0.66)
30 (19–46) 94 475 358 639 1.35 1.24 2.98 3.76 3.64 3.02
(10.1%) (51.2%) (38.2%) (65.7%) (0.34) (0.70) (0.70) (0.64) (0.74) (0.70)
G.S. Hafstad et al. / Eating Behaviors 14 (2013) 274–277
276
Table 2 Random-intercept models for associations with early predictors on picky eating at age 2.5 to 4.5. Factors
Model 1 Unadjusted estimates
Model 2 Adjusted estimates
β (SE)
β (SE)
Child factors Initial picky eating Gender Age Number of siblings Birth weight Birth complications Feeding problems before 3 months Sleep problem Temperamental shyness Temperamental emotionality Temperamental activity Temperamental sociability
0.365 0.036 0.068 −0.066 −0.001 −0.066 0.101 0.119 0.053 0.113 0.002 0.059
Maternal factors Maternal age Education Marital Status Negative affect (SCL-23) Temperamental shyness Temperamental activity Temperamental fearfulness Temperamental distress Temperamental anger
0.001 −0.014 0.014 0.174 0.011 0.006 −0.074 −0.081 0.002
(.029)⁎⁎⁎ (.028) (.017)⁎⁎⁎ (.016)⁎⁎⁎ (b.001)⁎ (.037) (.062) (.034)⁎⁎ (.023)⁎ (.023)⁎⁎⁎
0.342 (.296)⁎⁎⁎ – 0.068 (.017)⁎⁎⁎ −0.069 (.016)⁎⁎⁎ ≤0.001 (b.001) 0.061 (.032) – – – 0.078 (.022)⁎⁎⁎
(.030) (.029)⁎
– –
(.003) (.010) (.026) (.043)⁎⁎⁎
−0.008 (.003)⁎⁎ – – 0.086 (.04)⁎ – – – – –
(.024) (.020) (.022)⁎⁎ (.019)⁎⁎⁎ (.020)
β = regression estimates (mean slopes) and SE = standard error were presented for predictors with p b .2. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.
between a ‘difficult’ temperament and problematic eating behavior in young children (Haycraft et al., 2011; Jacobi et al., 2003). One possible explanation for this is that emotional temperament may lead to difficult interactions between child and parents, which could influence the atmosphere during mealtimes and thus the child's eating behavior. Previous research has shown that parental pressure on the child to eat during mealtimes leads to negative affect and more pickiness among children (Galloway et al., 2005), and it is possible that such interactions may evolve into contraproductive cycles. The present study showed that mother's negative affectivity at an early point in the child's life has implications for the later development of the child's eating patterns, which is in accordance with results from cross-sectional studies showing such a link (Blissett, Meyer, and Haycraft, 2007). This further strengthens the notion that pickiness is influenced by environmental factors, and, possibly, by interactional patterns with caregivers. The results are also in accordance with research showing that maternal anxiety influences the mother's sensitivity to the child in feeding situations, and may therefore lead to more controlling feeding practices in infancy (Micali, Simonoff, and Treasure, 2010). These may, in the long run, again lead to picky eating behavior in the child (Galloway et al., 2005). Having siblings predicted a decrease in picky eating over time, suggesting that siblings may somehow buffer against picky eating problems. Only one other study has so far examined the influence of having siblings on picky eating, finding that cross-sectionally there was an interaction between gender and older siblings (Jacobi et al., 2003). The present study adds that having siblings at an early age serve as a protective factor against the development of picky eating later on. One likely explanation for the finding is that older siblings serve as role models for eating behaviors. It is also plausible that parental attention toward each child during mealtime situations decreases with increasing number of children. This again could be beneficial to children's eating behaviors, as much attention or the feeling of being pressured to eat may increase pickiness (Blissett et al., 2007; Galloway et al., 2005).
This is one of very few studies to prospectively examine predictors of picky eating in preschool children. A main strength of this study was its longitudinal design, examining the course of picky eating over time, and identifying risk factors. The study also has some shortcomings. First, information about both children and mothers was assessed by mother's report only. Thus, it is difficult to disentangle the information provided about the child from the mother's own psychological status and response style. However, we were able to control for socioeconomic status and negative affect in the analyses, thereby providing information about relationships corrected for maternal factors in the multiple regression analyses. Second, our study assessed picky eating by survey measures only and provides no measures of clinical level of impairment. This may affect the applicability of the findings to high-risk groups. Nevertheless, the intention of this study was to examine picky eating in the normal population, and for this purpose the method used is considered adequate. Moreover, we used only two items to assess picky eating. Although the use of single-item has been shown to provide an adequate measure of picky eating in previous publications, the wording of the items used in this study was not identical to the studies by Jacobi et al. (2003). This, taken together with the lack of weight assessment, does not warrant any firm conclusions about whether we have captured a transient or a chronic eating problem in these children. Overall, the findings indicate that both child and maternal temperament at a very early stage in the child's life increase the risk for picky eating later on. Such knowledge may help to identify children in the normal population that are at risk for developing abnormal eating patterns throughout the preschool years. Role of funding sources Funding for this study was provided by the Norwegian Research Council. The Norwegian Research Council had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.
Contributors Tilmann von Soest and Gertrud Sofie Hafstad designed the study and wrote the protocol. Leila Torgersen and Gertrud Sofie Hafstad conducted literature searches and provided summaries of previous research studies. Dawit Shawel Abebe, Gertrud Sofie Hafstad and Tilmann von Soest conducted the statistical analysis. Gertrud Sofie Hafstad wrote the first draft of the manuscript and all authors contributed to and have approved the final manuscript. Conflict of interest All authors declare that they have no conflicts of interest.
References Blissett, J., Meyer, C., & Haycraft, E. (2007). Maternal mental health and child feeding problems in a non-clinical group. Eating Behaviors, 8(3), 311–318. http://dx.doi.org/ 10.1016/j.eatbeh.2006.11.007. Buss, A. H. H., & Plomin, R. (1984). Temperament: Early developing personality traits. Hillsdale, NJ: Erlbaum. Derogatis, L. R., Linman, R. S., Rickles, K., Uhlenhuth, E. H., & Covi, L. (1974). The Hopkins Symptom Checklist (HSCL): A self-report symptom inventory. Behavioral Science, 19, 1–15. Dubois, L., Farmer, A., Girard, M., Peterson, K., & Tatone-tokuda, F. (2007). Problem eating behaviors related to social factors and body weight in preschool children: A longitudinal study. International Journal of Behavioral Nutrition and Physical Activity, 4(9), 1–10. http://dx.doi.org/10.1186/1479. Faith, M. S., & Hittner, J. B. (2010). Infant temperament and eating style predict change in standardized weight status and obseity risk at 6 years. International Journal of Obesity, 34, 1515–1523. Galloway, A. T., Fiorito, L., Lee, Y., & Birch, L. L. (2005). Parental pressure, dietary patterns, and weight status among girls who are “picky eaters”. Journal of the American Dietetic Association, 105(4), 541–548. http://dx.doi.org/10.1016/j.jada.2005.01.029. Galloway, A. T., Lee, Y., & Birch, L. L. (2003). Predictors and consequences of food neophobia and pickiness in young girls. Journal of the American Dietetic Association, 103(6), 692–698. http://dx.doi.org/10.1053/jada.2003.50134. Haycraft, E., Farrow, C., Meyer, C., Powell, F., & Blissett, J. (2011). Relationships between temperament and eating behaviors in young children. Appetite (Retrieved from http://www.sciencedirect.com/science/article/pii/S0195666311000560)
G.S. Hafstad et al. / Eating Behaviors 14 (2013) 274–277 Jacobi, C., Agras, W. S., Bryson, S., & Hammer, L. D. (2003). Behavioral validation, precursors, and concomitants of picky eating in childhood. Journal of the American Academy of Child and Adolescent Psychiatry, 42(1), 76–84. http://dx.doi.org/ 10.1097/00004583-200301000-00013. Lewinsohn, P. M., Holm-Denoma, J. M., Gau, J. M., Joiner, T. E., Striegel-Moore, R., Bear, P., et al. (2005). Problematic eating and feeding behaviors of 36-month-old children. The International Journal of Eating Disorders, 38(3), 208–219. http://dx.doi.org/10.1002/ eat.20175. Marchi, M., & Cohen, P. (1990). Early childhood eating behaviors and adolescent eating disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 29, 112–117.
277
Micali, N., Simonoff, E., & Treasure, J. (2010). Pregnancy and post-partum depression and anxiety in a longitudinal general population cohort: The effect of eating disorders and past depression. Journal of Affective Disorders, 131(1–3), 150–157. http://dx.doi.org/10.1016/j.jad.2010.09.034 (Elsevier B.V.). Richman, N., & Graham, P. J. (1971). A behavioural screening questionnaire for use with three-year-old children. Preliminary findings. Journal of Child Psychology and Psychiatry, 12(1), 5–33. http://dx.doi.org/10.1111/j.1469-7610.1971.tb01047.