Secondary Logo

Journal Logo

Feeding Behavior Problems in Children With Cystic Fibrosis in the UK: Prevalence and Comparison With Healthy Controls

Duff, Alistair James Aitken*†; Wolfe, Susan Pamela*; Dickson, Corrine; Conway, Steven Phillip*; Brownlee, Keith Graham*

Journal of Pediatric Gastroenterology and Nutrition: April 2003 - Volume 36 - Issue 4 - p 443-447
Original Articles: Hepatology and Nutrition

Background Feeding behavior problems contribute to inadequate dietary intake for many patients with cystic fibrosis (CF). However, to establish effective intervention programs, more needs to be known about the occurrence and distribution of these difficulties. The aims of this study were to establish the prevalence and range of disruptive child behaviors (DCB) in patients with CF and the inappropriate parental responses (IPR) during mealtimes and to compare the results with those of healthy children.

Methods In study A, parents of 108 patients (aged 1–7 years) completed a Behavioural Paediatric Feeding Assessment Scale comprising two domains: DCBs and IPRs during mealtimes. Parents rated the frequency of the behaviors and responses and identified those they considered problematic. In study B, data from the CF group (n = 69, aged 1–12 years) were compared with 69 age- and sex-matched control subjects.

Results Parents of children with CF aged 5 to 8 years recorded significantly more DCBs than those in all other age ranges. These parents also reported significantly more IPRs than did parents of children aged 9 to 12 years and 13 to 17 years. Parents of children with CF reported significantly more DCBs and IPRs than did those of the control subjects. There were significantly more problematic DCBs and IPRs in the CF group than in the control group for children aged 5 to 8 years and 9 to 12 years but not for those aged 1 to 4 years.

Conclusions Parents of children with CF report more feeding behavior problems than do those of healthy control subjects. The high prevalence of feeding behavior problems in older children suggests that preventative and reactive interventions must continue throughout childhood and vary according to the child's developmental abilities.

*Regional Paediatric Cystic Fibrosis Unit (RPCFU), and †Department of Clinical and Health Psychology, St. James University Hospital (SJUH), Leeds, United Kingdom

Received December 20, 2000; accepted July 24, 2002.

Address correspondence and reprint requests to Alistair James Aitken Duff, Regional Paediatric CF Unit, Children's Day Hospital, St James's University Hospital, Leeds LS9 7TF, UK (e-mail:

Cystic fibrosis is the most common inherited genetic condition affecting white people. Recent advances in therapy have resulted in improved nutritional status of people with CF, which is one of the major factors contributing to improved longevity (1–3). Despite this, many children remain malnourished (4). Energy needs vary widely and can be in excess of 150% of the normal estimated requirement for age (5). Exploration of patients' difficulties in meeting these requirements has traditionally focused on underlying physiologic and disease-related abnormalities (6–9). For young children, in whom chronic infection is less common, attention focuses on the contribution of feeding behavior problems during mealtimes to inadequate energy intake. Such children are vulnerable to a range of behavior problems (10), and because nutrition is an essential part of CF care, it is understandable that parents become anxious about their child's food intake. Some studies have suggested these children are uniquely “at risk” of developing behavior problems during mealtimes (11,12), which are the most frequently cited area of difficulty for parents of young children with CF (13–17). These children have also been shown to have lower energy intakes (16), which can be improved by as much as 40% through joint behavioral and dietetic intervention (17–21). Meta-analysis has concluded that behavior therapy is as effective as oral supplementation and enteral and parenteral feeding in improving weight gain of young people with CF (22).

In 1992, the US CF Consensus Conference on Nutrition made behavior management recommendations (23). Such guidelines can be effective in helping parents manage problematic behaviors during mealtimes and may possibly prevent malnutrition (18,19). In the United Kingdom, dietitians routinely provide this advice, particularly during infancy. However, structured joint psychological and dietetic intervention is seldom available because of limited resources.

Most studies have focused on the impact of feeding behavior problems in small groups of preschool children with CF. Little attention has been paid to the mealtime interactions of older and adolescent children and their parents. Consequently, this study had two main aims:

  • To establish the prevalence and nature of disruptive mealtime behaviors and inappropriate parental responses in a large sample of children with CF, aged 1 to 18 years (study A); and
  • To compare the results from the patient group with those from a group of healthy control subjects (study B).
Back to Top | Article Outline


Study A

Parents of Children with CF

Parents (n = 112) of children (aged 1–17 years) who attended the Regional Paediatric CF Unit (RPCFU) between September 1997 and February 1998 were asked to participate in the study during routine outpatient visits (agreed, n = 108). Parents completed two questionnaires during routine outpatient appointments: a General Information Questionnaire (GIQ) designed to elicit basic demographic information, and a Behavioural Paediatric Feeding Assessment Scale (BPFAS).

  • 84% of the parents was aged 25 to 44 years;
  • 80% were either married or cohabiting; and
  • 46.3% had paid employment.
Back to Top | Article Outline

Children with CF

Fifty-one of the 108 children were girls. Twenty-three were aged 1 to 4 years, 33 were aged 5 to 8 years, 27 were aged 9 to 12 years, and 25 were aged 13 to 17 years (mean, 8.4 years). These age categories were based on developmental groupings of cognitive (24) and social (25) capabilities and concur with the groupings adopted by UK schools (1–4 years, preschool; 5–8 years, infants; 9–12 years, junior; 13–17 years, comprehensive). Twenty-eight children had been diagnosed with CF during the first week of life, 34 by 2 weeks, 54 by 1 month, 74 by 3 months, and 96 by 6 months (median diagnostic age, 2.6 months; SD, 2.75). Nineteen children were diagnosed through neonatal screening. Eighteen had meconium ileus, 23 had chest symptoms, 40 had gut symptoms, and 8 had a previously affected sibling. At the time of sampling, the mean weight, height, and body mass index (BMI) standard deviation scores (SDS) were −0.39 (SD, 1.12), −0.53 (SD, 1.05), and −0.09 (SD, 1.14), respectively. Fifty-eight parents perceived their child's condition to be “moderate” or “severe;” the remainder believed it to be “mild.”

Back to Top | Article Outline


The BPFAS (15) was developed in Canada specifically to explore behaviors associated with poor nutritional intake. It comprises two domains: disruptive child behaviors (DCB, 25 items) and inappropriate parental responses (IPR, 10 items). Each statement contains a descriptive phrase, and participants rate the “frequency of occurrence” on a 5-point Likert scale (“never” to “always,” scoring 1–5). The sum of the statement ratings gives a frequency score (FS) for each domain. Participants also indicate whether the statement is considered problematic by circling “yes” or “no.”

Little normative or validity data exists for the BPFAS at present. Items were taken from related instruments and supplemented with original statements based on clinical presentations. Good internal consistency has been demonstrated (26) with Cronbach α values (27) of 0.88, 0.84, and 0.74 for the full-scale and child and parent domains, respectively. Equally, satisfactory test–retest reliability has been established for the entire scale and its domains (15).

For the purposes of this study, the measure was modified for use with a UK population (BPFAS-UK), “anglicizing” words and spellings. On the DCB domain, two statements were added. These related to the location of mealtimes in the home and dietary variety. One question relating to enteral tube feeding was removed because this was considered not to be a “behavioral” issue (total statements = 26; minimum FS scoring range 26–130). On the IPR domain, two statements were added. These pertained to using bribes to encourage children to eat and ratings of their satisfaction with their child's weight and growth (total statements = 12; FS scoring range, 12–60). Cronbach α values (27) were recalculated for the entire scale and for the child and parent subscales. These were 0.90, 0.86, and 0.81, respectively, and represent even better internal consistency than the original measure.

Back to Top | Article Outline

Study B

This study compared data obtained from parents of CF children with those from children with no chronic illness, matched for age and sex. It was thought that children aged 13 to 17 years (n = 25) would be more independent of their parents and that parents' views alone would not accurately reflect feeding behavior patterns in this age group. Consequently, they were excluded from this part of the analysis.

Back to Top | Article Outline

Control Children

Parents of healthy children aged 1 to 12 years were recruited via local school and preschool nursery groups. These were selected from those attended by the children in the CF group to reflect socioeconomic status. Three nurseries declined to assist in recruiting parents. Consequently, three nurseries where no CF patients attended assisted in the recruitment of control group participants.

In nurseries and infant schools (age groups 1–4 years and 5–7 years, respectively), the class teacher gave parents the questionnaires, whereas children in junior schools (aged 7 years and older) were sent home with them in sealed envelopes. The GIQ was modified by removing the section related to CF and by adding a further question about the child's current health status to establish an effective screen for chronic ill health. In both cases, parents returned the questionnaires by mail.

Both groups of parents were comparable in age and family size. Significantly more parents in the control group were in paid employment than in the CF group (control, 61%; CF, 38%;P2) < 0.05). There were also significantly fewer single parents in the control group compared with the CF group (control, 6%; CF, 23%;P2) < 0.01).

Back to Top | Article Outline


For both studies, differences between group means were calculated by t test (two-tailed), and frequency data were determined by χ2. Within-group variance was calculated by a one-way analysis of variance (ANOVA) with post hoc Tukey HSD tests being performed for DCB and IPR FS.

Back to Top | Article Outline


Study A

The mean DCB FS was 52.8 (SD, 12.8; range, 33–94). A mean of 3.4 statements (SD, 5.5) were cited as being “problematic” to parents. The mean IPR FS was 21.9 (SD, 7.3; range, 12–46). A mean of 1.7 statements (SD, 2.8) were cited as being “problematic” to parents.

Overall, 76 parents (70%) indicated that CF made them more anxious about food intake, and 30 (28%) were unhappy with their child's growth. However, only 7 parents (6.5%) specifically requested intervention from either a dietitian or a child psychologist.

Parents of boys reported significantly more DCBs and IPRs than those of girls (4.3 and 2.4, respectively;P < 0.05AQ).

There were no other significant differences between any other demographic variable (e.g., employment status) and the mode of presentation on the DCB and IPR variables.

The hypothesis that parents of younger children would report more feeding behavior problems than those of older children was only partially supported. Table 1 shows that mothers of children aged 5 to 8 years recorded a significantly higher DCB FS than those of children in the other age groups and a significantly higher IPR FS than those of children aged 9 to 12 years and 13 to 17 years.



Back to Top | Article Outline

Study B

One hundred sixty-four questionnaires were dispatched to schools, and 143 were returned (87%). Of these, 26 parents indicated that their child had a chronic health condition (e.g., asthma, diabetes), and 48 were grossly incomplete largely because of respondents failing to answer the statements printed on the second page of the measure. This left a matched sample size of 69 (31 boys), with the following age-range distribution: 1 to 4 years (n = 22), 5 to 8 years (n = 26), and 9 to 12 years (n = 21).

Table 2 shows that parents of children with CF recorded significantly higher DCB and IPR FSs than those of healthy control subjects. They also identified significantly more DCBs and IPRs as problematic. Further analyses indicated that parents in the CF group identified seven DCBs and six IPRs as being problematic to them, significantly more than those in the control group (Table 3).





Given the wide age range of the sample, we believed it important to explore any differences between the CF and control groups on a developmental age basis (Table 4). This shows significant differences in the number of DCBs and IPRs between the CF and control groups for children aged 5 to 8 years and 9 to 12 years but not for those aged 1 to 4 years.



Back to Top | Article Outline


Despite the optimal use of pancreatic enzyme supplements, high energy expenditure and inadequately controlled malabsorption results in increased energy requirements for many children with CF (6). Energy intakes are often compromised by poor appetite. Feeding behavior problems (FBP), which include DCB and IPR during mealtimes, may also contribute to low energy intake (16,19).

The prevalence of DCBs and IPRs identified by parents in this study was similar to previous reports (15), but the mean number of behaviors and responses perceived as “problematic” was considerably lower. Despite nearly three quarters (70%) of parents indicating that “CF made them more anxious about food intake” and more than one quarter (28%) feeling “unhappy about their child's growth,” only a fraction (6.5%) requested intervention. This suggests that parents acknowledge the importance of nutritional intake and are concerned about it but do not perceive FBPs as important factors. Consequently, there is a need for parents to be taught specifically about the impact of FBPs on nutrition (13,16,18) and given effective strategies for dealing with them. Furthermore, in view of our results, such education should continue throughout childhood and vary according to the child's developmental capacities.

In our group, the mode of presentation of CF, age at diagnosis, and parental marital and employment status had no effect on the frequency of DCBs or IPRs during mealtimes or on the number of FBPs considered problematic. However, parents of boys reported significantly more DCBs and IPRs than those of girls. This may reflect the general trend that boys display more behavior difficulties than girls (28).

When compared with a group of healthy control subjects, preadolescent children with CF were perceived by their parents as having a higher frequency and number of FBPs. Significantly more parents in the CF group indicated that their child displayed a range of DCBs during mealtimes compared with those of healthy control subjects, including no enjoyment of eating, poor appetite, reluctance to come to mealtimes, preferring to drink rather than eat, eating snack food but not eating at mealtimes, and trying to negotiate foods to be eaten. These behaviors seem to lead to frustration and unhappiness for parents and to beliefs that the eating patterns of some children with CF do not promote good health. Such feelings may result in ineffectual and counterproductive strategies during mealtimes, such as coaxing their child to eat. If used excessively, such responses can unwittingly reward (“reinforce”) unwanted feeding behavior in children of certain ages as opposed to promoting “desired” conduct during mealtimes. Previous studies have shown that DCBs vary with age and are most frequently reported in younger children (14–16). Consequently, it follows that parental responses during mealtimes will also vary.

Parents of children aged 5 to 8 years reported a significantly higher frequency of DCBs during mealtimes than parents of those in the other age groups. The frequency of IPRs was also shown to be significantly higher in this group of children than in those aged 9 to 12 years and 13 to 17 years, but not in those aged 1 to 4 years. Contrary to previous findings, parents of preschool children with CF in this sample reported an almost identical prevalence of FBPs to those of healthy children. Parents of older children with CF aged 5 to 8 years and 9 to 12 years reported a significantly higher frequency of DCBs and IPRs than their counterparts in the control group.

There may be several explanations for these findings. First, dietitians in our unit tend to focus advice regarding the prevention of FBPs and coping strategies on parents of preschool children. Therefore, the parents of this group of children may feel more able to cope. Second, in younger children, the disease may be less visible with potentially fewer exacerbations (e.g., chronic infection or hospitalization). This may lead parents to be less anxious about feeding their children than those of older patients. Third, it is plausible that the BPFAS-UK may be insensitive to distinguishing normal as opposed to abnormal feeding patterns in young children. Some statements (e.g., “takes longer than 20 minutes to finish a meal,” “whines and cries at feeding time,” and “spits out food”) may be developmentally appropriate for this age group. Fourth, behavior problems in children in middle childhood are common (29).

Although this study adds to existing knowledge about FBPs in children with CF, there are limitations for interpreting the results. We found the modified version of the BPFAS (15) to be a useful and practical screening tool for FBPs; however, it may be inappropriate for use with an adolescent population and insensitive to detecting abnormal feeding patterns in young children. Some statistical work has now taken place on the factor structure of the measure (30). However, further statistical analyses may lead to improved validity of the measure, identification of the core components of problematic mealtime behavior, and developmentally sensitive questions.

The present results suggest that many parents of patients with CF in middle childhood experience mealtime behavior difficulties and engage in unhelpful responses. Consequently, preventative and reactive nutritional interventions must continue throughout childhood, and not focus solely on the preschool population. To be successful, these must be tailored to the developmental capacities of the child. Strategies may include providing parents with information to help them understand their child's behavior rather than perceiving it as deviant. In addition, children should be involved in their own nutritional management. They should be provided with age-appropriate literature, and their participation in nutritional education sessions should be encouraged.

Back to Top | Article Outline


1. Dodge JA, Morison S, Lewis PA, et al. Cystic fibrosis in the United Kingdom incidence, population and survival: update to 1995. Arch Dis Child 1997; 77:493–6.
2. Kraemer R, Rudeberg A, Hadorn B, et al. Relative under weight in cystic fibrosis and its prognostic value. Acta Paediatr Scand 1978; 67:33–7.
3. Corey M, Mclauchlin FJ, Williams M. A comparison of survival, growth and pulmonary function in patients with cystic fibrosis in Boston and Toronto. J Clin Epidemiol 1988; 41:588–91.
4. Cystic Fibrosis Foundation. Patient Registry 1998 Annual Data Report, Bethesda, Maryland, September 1999.
5. Department of Health. Dietary reference values for food, energy and nutrients for the United Kingdom. Report No 41. London: HMSO; 1993.
6. Buchdahl RM, Fullylove C, Marchant JL, et al. Energy and nutrient intakes in cystic fibrosis. Arch Dis Child 1989; 64:373–8.
7. Ellis JA, Bond SA, Wootton SA. Energy and protein intakes of patients with cystic fibrosis. J Hum Nutr Diet 1992; 5:333–42.
8. Morrison JM, O'Rawe A, McCracken KJ, et al. Energy intakes and losses in cystic fibrosis. J Hum Nutr Diet 1994; 7:39–46.
9. Kawchak DA, Zhao H, Scanlin TF, et al. Longitudinal, prospective analysis of dietary intake in children with cystic fibrosis. J Pediatr 1996; 129:119–29.
10. Thompson RJ, Hodges K, Hamlett KW. A matched comparison of adjustment in children with cystic fibrosis and psychiatrically referred and non-referred children. J Pediatr Psychol 1990; 15:745–59.
11. Fischer-Fay A, Goldberg S, Simmons RJ, et al. Chronic illness and infant-mother attachment. J Dev Behav Pediatr 1988; 9:266–70.
12. Goldberg S, Morris P, Simmons RJ, et al. Chronic illness in infancy and parenting stress: a comparison of three groups of mothers. J Pediatr Psychol 1990; 15:347–58.
13. Quittner AL, DiGirolamo AM, Michel M, et al. Parental response to cystic fibrosis: a contextual analysis of the diagnostic phase. J Pediatr Psychol 1992; 17:683–704.
14. Stark LJ, Jelalian E, Mulvihill MM, et al. Eating in pre-school children with cystic fibrosis and healthy peers: a behavioral analysis. Pediatrics 1995; 95:210–5.
15. Crist W, McDonnell P, Beck M, et al. Behavior at mealtimes and the young child cystic fibrosis. Dev Behav Pediatr 1994; 15:157–61.
16. Sanders MR, Turner KM, Wall CR, et al. Mealtime behavior and parent-child interaction: a comparison of children with cystic fibrosis, children with feeding problems and non-clinic controls. J Pediatr Psychol 1997; 22:881–900.
17. Stark LJ, Jelalian E, Powers SW, et al. Parent and child mealtime behavior in families of children with cystic fibrosis. J Pediatr 2000; 136:195–200.
18. Stark LJ, Powers SW, Jelalian E, et al. Modifying problematic mealtime interactions of children with cystic fibrosis and their parents via behavioral parent training. J Pediatr Psychol 1994; 19:751–68.
19. Stark LJ, Knapp L, Bowen AM, et al. Behavioral treatment of calorie consumption in children with cystic fibrosis: replication with two year follow-up. J Appl Behav Anal 1993; 26:435–50.
20. Stark LJ, Mulvihill MA, Powers SW, et al. Behavioral intervention to improve calorie intake of children with cystic fibrosis: treatment vs wait-list controls. J Pediatr Gastroenterol Nutr 1996; 23:240–53.
21. Singer L, Nofer J, Benson-Szekely, et al. Behavioral assessment and management of food refusal in children with cystic fibrosis. J Dev Behav Pediatr 1990; 12:115–20.
22. Jelalian E, Stark LJ, Reynolds L, et al. Nutritional intervention for weight gain in cystic fibrosis: a meta analysis. J Pediatr 1998; 132:486–92.
23. Ramsey B, Farrell P, Pencharz P. Nutritional assessment and management in cystic fibrosis: consensus conference. Am J Clin Nutr 1992; 55:108–16.
24. Piaget J. The Language and Thought of the Child. New York: Humanities Press; 1952.
25. Bandura A. Social Learning Theory. Englewood Cliffs, NJ: Prentice-Hall; 1977.
26. McDonnell P, Crist W, Green M, et al. Assessing common behavioral difficulties at mealtime with the preschool child. Presented at the 52nd meeting of the Canadian Psychological Association, Calgary, Alberta, June 1991.
27. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika 1951; 16:297–334.
28. Cohen P, Cohen J, Kasen S, et al. An epidemiological study of disorders in late childhood and adolescence. 1. Age and gender specific prevalence. J Child Psychol Psychiatry 1993; 34:851–67.
29. Carr A. The Handbook of Child and Adolescent. Clinical Psychology: A Contextual Approach. London: Routledge; 1999.
30. Crist W, Napier-Phillips A. Mealtime behavior of young children: a comparison of normative and clinical data. J Developmental & Behav Pediatrics 2001; 22:279–86.

Cystic fibrosis; Feeding; Behavior; Nutrition; Prevalence

© 2003 Lippincott Williams & Wilkins, Inc.