1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information

Aim  The aim of this study was to investigate the prevalence of eating problems and their association with neurological and behavioural disabilities and growth among children born extremely preterm (EPC) at age 6 years.

Method  A standard questionnaire about eating was completed by parents of 223 children (125 males [56.1%], 98 females [43.9%]) aged 6 years who were born at 25 weeks’ gestation or earlier (mean 24.5wks, SD 0.7wks; mean birthweight 749.1g, SD 116.8g), and parents of 148 classmates born at term (66 males [44.6%], 82 females [55.4%]). All children underwent neurological, cognitive, and anthropometric assessment, and parents and teachers completed a behaviour scale.

Results  Eating problems were more common among the EPC than the comparison group (odds ratio [OR] 3.6, 95% confidence interval [CI] 2.1–6.3), including oral motor (OR 5.2, 95% CI 2.8–9.9), hypersensitivity (OR 3.0, 95% CI 1.6–5.6), and behavioural (OR 3.8, 95% CI 1.9–7.6) problems. Group differences were reduced after adjustment for cognitive impairment, neuromotor disability, and other behaviour problems. EPC with eating problems were shorter, lighter, and had lower mid-arm circumference and lower body mass index (BMI) even after adjusting for disabilities, gestational age, birthweight, and feeding problems at 30 months.

Interpretation  Eating problems are still frequent in EPC at school age. They are only partly related to other disabilities but make an additional contribution to continued growth failure and may require early recognition and intervention.

List of Abbreviations

Body mass index


Extremely preterm children

Surviving extremely preterm infants are at significantly increased risk of long-term cognitive, motor, or behavioural impairments.1–3 Preterm infants often have feeding problems while in neonatal care, including swallowing problems, oral sensory and motor dysfunction, or fatigue during feeding. Problems experienced after discharge include delayed feeding skill development, food refusal,4 and difficulty with the transition to solid foods, often accompanied by growth faltering.5 Reasons proposed for these difficulties include immaturity and treatment related to preterm birth including parenteral or tube feeding, adverse oral motor experiences,6 and postdischarge health problems, such as neurological or cognitive disabilities. All of these could complicate the child’s eating skill development and result in increased parental distress, depression, and lack of confidence in feeding.7

There are few estimates concerning the prevalence of feeding problems of extremely preterm infants, and none at school age,4 despite eating remaining a primary concern for parents post discharge.8 Furthermore, it is unclear whether eating problems are specific or can be fully or mostly accounted for by general cognitive deficits or neuromotor or other behavioural problems often found in EPC.1,3

In this report, we describe eating behaviour at age 6 years among children born at a gestational age of 25 weeks or less in a whole-population study. We aimed to answer three questions. First, is the prevalence of clinically relevant eating problems higher in EPC than in children born at term? Second, can eating problems be accounted for by comorbidity, including cognitive deficits or other neurological or behavioural disabilities? And, third, do eating problems in the clinical range have a significant impact on attained growth (height, weight, head circumference, mid-arm circumference, and body mass index) at age 6 years?


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information


The population comprised all surviving children in the UK and Ireland born at or before 25 weeks and 6 days of gestation during the period March to December 1995 (the EPICure study3). A total of 308 children were known to be alive at 30 months. Of these, the parents of 241 children (78%) consented to the study, of whom 223 (72%) completed the eating questionnaire. The majority of the children (204/241) were in mainstream education. For each child assessed in mainstream school, we sought an age- and sex-matched classmate as a comparison.3 The eating questionnaire was completed by the parents of 148 children in the comparison group.

This study was approved by the Trent Multicentre Research Ethics Committee and the local education authorities in Scotland.


Feeding and eating behaviour assessment

When the children were 30 months of age, their parents were asked, in a single item, whether their children had any feeding problems.9 At 6 years of age, the parents completed a specially developed eating questionnaire designed on the basis of a comprehensive review of the type of eating problems in studies of the general population.6 The scale included 19 items (see Appendix S1, supporting information published online). A principal component factor analysis with varimax rotation yielded four factors with distinct high loadings of 0.48 or higher and acceptable eigenvalues >1: refusal–faddy eating problems (seven items, e.g. refuses to eat, is a faddy eater); oral motor problems (six items, e.g. dribbles when drinking, has problems with biting crackers); oral hypersensitivity problems (two items, e.g. does not like things to be put in his/her mouth); and behavioural problems around meals (four items, e.g. makes a mess, has tantrum during meals). A total eating-difficulties score was also constructed (higher scores on each scale indicate more problems; see Appendix S1 for the Cronbach’s α of total eating problems and each subscale).

To derive clinical categories, each scale was dichotomized into normal versus clinical (score above the 90th centile or near according to the comparison group), as suggested by other standard behaviour scales.1 If the child scored below the 90th centile, the eating behaviour was considered normal (no eating difficulty). Parents were also asked whether or not they felt that their child had an eating problem (mild or severe) and if the eating difficulties upset or distressed their child (a little, quite a lot, or a great deal).

Pervasive behaviour problems

Child behaviour was reported by parents and teachers.1 Using the Strengths and Difficulties Questionnaire,10 Child behaviour was classified as follows: if the child scored below the 90th centile in both the parent and the teacher report, the behaviour was considered normal (no behaviour difficulty); mild difficulty refers to the classification of the child in the clinical range (above the 90th centile) as reported by either the parent or the teacher; and clinical pervasive behaviour problems refers to the classification of the child in the clinical range by both the parent and the teacher.1

Cognitive ability

Children were assessed with the Kaufman Assessment Battery for Children at 6 years of age.11 The Mental Processing Component provides an ‘overall cognitive score’.2 Cognitive impairment was categorized according to conventional SD-banded cut-offs using the scores of the comparison group as reference data (mild, 1–2SD below the mean; moderate, 2–3SD below the mean; severe, more than 3SD below the mean).

Disability classification (neurological assessment)

Mild disability included neurological signs with minimal functional consequences. Moderate disability included reasonable independence and ambulant CP. Severe disability included non-ambulant CP (see Table II).3

Growth parameters

The weight of individuals was measured on identical weighing scales (Salter Housewares Ltd, Kent, UK), height using a standard stadiometer (Child Growth Foundation, London, UK), and maximum occipitofrontal head circumference and mid-arm circumference using a LASSO-O tape (Child Growth Foundation). Each measure was taken twice and the mean value computed. Body mass index (BMI) was computed as weight/height2 (kg/m2).

Developmental panel

The children were assessed by seven experienced developmental paediatricians and eight psychologists, who received formal training. Assessment sessions of every second child were videotaped and randomly quality checked by the senior assessment paediatrician or psychologist (agreement >90%).

Statistical analysis

Analysis of variance (anova) was used to compare the EPC and comparison groups and to compare males with females in each group (with effect sizes reported as η2). Categorical outcomes were evaluated with χ2 tests for trends or Fisher’s exact test as appropriate (Statistical Package for the Social Sciences 15.0; SPSS Inc., Chicago, IL, USA). All statistical tests were two-sided. Odds ratios and 95% confidence intervals (CIs) are reported for the comparison of EPC and classmates and for the comparison of males and females in each group. The 95% CIs were obtained using bootstrapping (20 000 bootstrap samples, using the bias-corrected and accelerated method in MatLab R2009a [the Statistics Toolbox for MatLab]12,13). Odds ratios approximated risk ratios when the incidence of the outcome in the study population (comparison group) was low (<0.10%).14 Selective dropout was determined by comparing neonatal, 12-month, and 30-month follow-up data of those assessed at 6 years and those who were lost to follow-up. To test for the presence of specific eating problems, logistic regressions adjusted for cognitive disability (no/mild vs moderate/severe), neuromotor disability (no vs others), or pervasive behaviour disability (no vs mild/severe) were computed. In the EPC group logistic regressions were also performed to test whether eating difficulties predict parents’ acknowledgement of eating problems and of their child’s distress in the EPC group. Correlations and partial correlations (adjusting for disabilities, gestational age, birthweight, and feeding problems at 30mo) were performed to test the relationship between total eating problems and growth measures. Graph Pad Prism 5 software (GraphPad Software Inc., San Diego, CA, USA) was used to design the graphs.


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information

Children lost to follow-up (dropouts)

Compared with children who were assessed, dropouts (maximum n=85) were more likely to be of non-white ethnic origin (34.1/100 vs 18.9/100; p=0.004), to have young mothers (≤21y age: 24.7/100 vs 9.4/100; p=0.001), to live in overcrowded homes (49.4/100 vs 21.5/100; p<0.001), to have experienced more than one serious life event by 30 months (48.2/100 vs 23.3/100, p<0.001), to suffer from cerebral palsy (CP) at 30 months (30.8/100 vs 15.6/100; p=0.007), to have a lower psychomotor development index score (PDI mean 78.7 vs 84.8; p=0.015), to have more feeding problems (42.4/100 vs 30.4/100; p=0.049), and to be diagnosed with overall severe disability (40/100 vs 25/100; p=0.014) at 30 months of age. No differences were found in any of the assessed nine neonatal complications (e.g. prenatal steroid treatment), five other socio-economic factors (e.g. mother is single or separated), and six developmental and growth parameters (e.g. weight, height) up to 30 months.1,3

Eating behaviour differences

Total eating problems were more common in the EPC group than in the comparison group according to the total eating problems scale (p<0.001; effect size=0.080; Table I). This was also true for the oral motor (p<0.001; effect size=0.099), refusal–faddy (p=0.026; effect size=0.016), behavioural (p<0.001; effect size=0.076), and hypersensitivity (p<0.001; effect size=0.052) subscales. Males were found to have more oral motor difficulties (p=0.001; effect size=0.124) and behavioural problems (p=0.013; effect size=0.087) than females (Table I). The interaction between group and sex was not significant.

Table I.   Total eating difficulties and the subscales assessed at age 6 years for 223 children born at 25 weeks’ gestation or less and 148 age-matched comparison children
Eating difficultiesComparison groupExtremely preterm groupOdds ratio for clinical range (95% CI)a
Number/number with informationMean (SD)Clinical range (%)Number/number with informationMean (SD)Clinical range (%)
  1. Mean (SD) values are shown together with the proportion of the clinical range of the comparison group by sex. Overall results in bold. a95% confidence intervals (CIs) based on 20 000 bootstrap samples, using the bias-corrected and accelerated method. bp<0.001; cp<0.01; dp<0.05 for differences between extremely preterm and comparison groups.

Total eating difficulties19/1486.1 (4.9)b12.8b76/21810.0 (7.4)b34.9b3.6 (2.1–6.5)
 Males8/666.1 (5.1)12.1b43/12210.5 (7.3)35.2b3.9 (1.7–9.0)
 Females11/826.1 (4.7)13.4c33/969.4 (7.5)34.4c3.4 (1.6–7.2)
Oral motor difficulties13/1480.4 (1.0)b8.8b72/2151.8 (2.5)b33.5b5.2 (2.8–10.3)
 Males9/660.6 (1.3)c13.6b49/1202.2 (2.7)c40.8b4.4 (1.9–9.6)
 Females4/820.3 (0.6)c4.9b23/951.4 (2.3)c24.2b6.2 (2.1–18.9)
Refusal–faddy problems16/1483.9 (3.5)d10.838/2234.9 (4.0)d17.01.7 (0.9–3.3)
 Males7/663.6 (3.4)10.619/1254.8 (3.7)15.21.5 (0.6–3.8)
 Females9/824.3 (3.5)11.019/985.3 (4.3)19.401.9 (0.8–4.6)
Behavioural problems14/1481.4 (1.4)b9.5b52/2192.3 (1.7)b23.7b3.0 (1.6–5.8)
 Males6/661.6 (1.4)d9.1c31/1232.5 (1.8)d25.2c3.4 (1.3–8.5)
 Females8/821.3 (1.4)d9.8d21/962.1 (1.7)d21.9d2.6 (1.1–6.2)
Hypersensitivity problems11/1480.1 (0.5)b7.4b50/2130.5 (1.0)b23.5b3.8 (1.9–8.0)
 Males6/66`0.1 (0.4)9.1c35/1200.6 (1.1)29.2c4.1 (1.6–10.4)
 Females5/820.1 (0.5)6.1d15/930.4 (0.9)16.1d2.9 (1.0–8.5)

Total eating difficulties in the clinical range were also more frequent in the EPC group than in the comparison group (odds ratio [OR] 3.6; 95% CI 2.1–6.3; p<0.001), as were oral motor problems (OR 5.2; 95% CI 2.8–9.9; p<0.001), behavioural problems (OR 3.0; 95% CI 1.6–5.6; p<0.001), and hypersensitivity problems (OR 3.8; 95% CI 1.9–7.6; p<0.001; Table I).

Both males and females in the EPC group had more total eating difficulties (males p<0.001; females p=0.001), oral motor problems (p<0.001), behavioural problems (males p=0.005; females p=0.023), and hypersensitivity problems (males p=0.001; females p=0.031) than their sex-matched counterparts in the comparison group (Table I). Within the EPC group, males were more likely than females to have oral motor problems (OR 2.2; 95% CI 1.2–3.9; p=0.007) and more often had hypersensitivity problems (OR 2.1; 95% CI 1.1–4.2; p=0.019).

Gestation, disability, and eating difficulties

In the EPC group, there were significant associations between gestation at birth and total eating difficulties (p=0.003) and hypersensitivity problems (p=0.038; Table II). Cognitive impairment and neuromotor disability were associated with an increased prevalence of clinical oral motor problems (p=0.022, p<0.001 respectively) and hypersensitivity problems (p<0.001, p<0.001 respectively). Pervasive behaviour difficulties showed significant associations with eating problems on all scales: total eating difficulties (p<0.001), oral motor problems (p=0.001), refusal–faddy problems (p=0.010), behavioural eating problems (p<0.001), and hypersensitivity problems (p=0.001) were all in the clinical range (Table II).

Table II.   Gestational age at birth, severity and type of disability, and total eating difficulties score and the subscales scores at age 6 years among extremely preterm children, n (%)
 Total eating difficulties (n=218)Oral motor problems (n=215)Refusal–faddy problems (n=223)Behavioural problems (n=219)Hypersensitivity problems (n=213)
NormalClinical rangeNormalClinical rangeNormalClinical rangeNormalClinical rangeNormalClinical range
  1. Significant values in bold. ap<0.01; bp<0.05; cp<0.001 for differences between gestation ages and disability groups. CP, cerebral palsy; SDQ, Strengths and Difficulties Questionnaire.

Gestational age (wks)
 ≤2313 (59.1)9 (40.9)a12 (60)8 (40)19 (86.4)3 (13.6)14 (63.6)8 (36.4)18 (81.8)4 (18.2)b
 2434 (50)34 (50.0)39 (59.1)27 (40.9)57 (83.8)11 (16.2)50 (74.6)17 (25.4)41 (65.1)22 (34.9)
 2595 (74.2)33 (25.8)92 (71.3)37 (28.7)109 (82)24 (18)103 (79.2)27 (20.8)104 (81.3)24 (18.8)
Overall cognition
 No disability (score >94)44 (68.8)20 (31.3)46 (71.9)18 (28.1)b54 (84.4)10 (15.6)53 (85.5)9 (14.5)57 (90.5)6 (9.5)c
 Mild disability (score 82–94)49 (71)20 (29)51 (75)17 (25)57 (80.3)14 (19.7)52 (75.4)17 (24.6)54 (80.6)13 (19.4)
 Moderate disability (score 70–81)27 (60)18 (40)28 (62.2)17 (37.8)34 (73.9)12 (26.1)31 (67.4)15 (32.6)35 (81.4)8 (18.6)
 Severe disability (score ≤69)22 (55)18 (45)18 (47.4)20 (52.6)40 (95.2)2 (4.8)31 (73.8)11 (26.2)17 (42.5)23 (57.5)
 No disability113 (66.1)58 (33.9)125 (74)44 (26)c141 (82)31 (18)134 (79.8)34 (20.2)135 (81.6)30 (18.2)c
 Abnormal signs14 (66.7)7 (33.3)10 (47.6)11 (52.4)18 (81.8)4 (18.2)13 (59.1)9 (40.9)14 (70)6 (30)
 CP ambulatory10 (66.7)5 (33.3)7 (50)7 (50)15 (100)0 (0)10 (66.7)5 (33.3)10 (66.7)5 (33.3)
 CP, non-ambulatory5 (45.5)6 (54.5)1 (9.1)10 (90.9)11 (78.6)3 (21.4)10 (71.4)4 (28.6)4 (30.8)9 (69.2)
Pervasive total difficulties (SDQ)
 No disability89 (80.9)21 (19.1)c85 (78.7)23 (21.3)c100 (90.9)10 (9.1)b103 (93.6)7 (6.4)c94 (85.5)16 (14.5)a
 Mild disability36 (50.7)35 (49.3)40 (57.1)30 (42.9)56 (75.7)18 (24.3)43 (60.6)28 (39.4)49 (74.2)17 (25.8)
 Severe disability17 (47.2)19 (52.8)18 (50)18 (50)28 (75.7)9 (24.3)20 (55.6)16 (44.4)20 (55.6)16 (44.4)

Group differences in hypersensitivity and behavioural problems between EPC and comparison children became non-significant after adjustment for cognitive abilities, neuromotor disability, and pervasive behavioural difficulties together (Table III). In contrast, even after adjustment for all variables, total eating difficulties and oral motor problems still differed between EPC and comparison children, and are thus only partly explained by these disabilities.

Table III.   Odds ratios for clinical scores for total eating difficulties and the subscales before and after adjustment for general cognitive scores and neuromotor and/or pervasive total behavioural difficulties (SDQ)
  Unadjusted (95% CI) Adjusted for categorized MPCa (95% CI) Adjusted for neuromotor (95% CI)Adjusted for pervasive total behaviour difficulties – SDQa (95% CI) Adjusted for all variables (95% CI)b
  1. 95% confidence intervals based on 20 000 bootstrap samples, using the bias-corrected and accelerated method. a≤2SD vs ≥2SD (no and mild vs moderate and severe). bCognitive, neuromotor and pervasive behaviour difficulties. cp<0.001, dp<0.01, ep<0.05 for differences between extremely preterm and comparison groups. CI, confidence interval; MPC, mental processing component; SDQ, Strengths and Difficulties Questionnaire.

Total eating difficulties3.6 (2.1–6.5)c3.0 (1.6–5.5)c3.5 (1.9–6.4)c2.7 (1.5–5.0)d2.5 (1.3–4.8)d
Oral motor problems5.2 (2.8–10.3)c3.7 (1.8–7.8)c3.6 (1.8–7.4)c4.1 (2.2–8.2)c2.7 (1.3–5.7)d
Refusal–faddy problems1.7 (0.9–3.3)1.8 (0.9–3.6)1.8 (0.9–3.6)1.3 (0.7–2.6)1.6 (0.8–3.3)
Behavioural problems3.0 (1.6–5.8)d2.4 (1.2–5.0)e2.4 (1.2–5.0)d1.7 (0.8–3.5)1.6 (0.7–3.6)
Hypersensitivity problems3.8 (1.9–8.0)c2.5 (1.1–6.0)e2.8 (1.3–6.3)d3.0 (1.5–6.4)d1.9 (0.8–4.7)

Table IV shows that in the EPC group total eating difficulties and the subscales significantly predicted parents’ judgement of significant eating difficulties and their child’s distress.

Table IV.   Odds ratios for total eating difficulties and the subscales predicting parents’ acknowledgement of eating problems and the distress caused to their extremely preterm children
 Parents’ acknowledgement of eating problems (95% CI) Distress (95% CI)
  1. 95% confidence intervals based on 20 000 bootstrap samples, using the bias-corrected and accelerated method. ap<0.001; bp<0.05; cp<0.01.

Total eating difficulties1.4 (1.3–1.6)a1.3 (1.1–1.4)a
Oral motor problems1.6 (1.3–1.9)a1.5 (1.2–2.0)a
Refusal–faddy problems1.6 (1.4–1.8)a1.2 (1.01–1.4)b
Behavioural problems1.8 (1.5–2.2)a1.5 (1.1–2.2)b
Hypersensitivity problems2.2 (1.7–3.1)a1.8 (1.1–3.1)c

Eating problems and growth parameters

Attained growth was significantly poorer in EPC without eating problems (normal range) than in the comparison children without eating problems in terms of weight (mean difference 2.6kg; 95% CI 1.6–3.7kg; p<0.001; Fig. 1), height (mean difference 3.1cm; 95% CI 1.6–4.6cm; p<0.001), head circumference (mean difference 1.4cm 95% CI 1.03–1.8cm; p<0.001), and mid-arm circumference (mean difference 1.2cm 95% CI 0.7–1.7cm; p<0.001). In addition, BMI was lower in the EPC group (mean difference 1.2; 95% CI 0.8–1.7; p<0.001; see appendix S2, supporting information published online). Among children with eating problems (clinical range), mean differences between EPC and comparison children were even greater for height (4.2cm; 95% CI 1.2–7.1cm; p=0.006) and head circumference (1.7cm; 95% CI 0.8–2.6cm; p<0.001; Appendix S2). In contrast, differences in weight (2.1kg; 95% CI 0.5–3.7kg; p=0.013; Fig. 1), mid-arm circumference (0.8cm; 95% CI 0.04–1.6cm; p=0.049), and BMI (0.7; 95% CI: 0.07–1.3; p=0.029) were similar between the two groups without eating problems.


Figure 1.  Measured weight (kg) parameters at age 6 years for extremely preterm and comparison children with (clinical) and without (normal) eating problems.

Download figure to PowerPoint

Among the EPC group, children with eating problems weighed less (mean difference 1.1kg; 95% CI 0.09–2.1kg; p=0.033 Fig. 1) and had smaller heads (mean difference 0.6cm; 95% CI 0.2–1.1cm; p=0.009) and lower BMI (mean difference 0.5; 95% CI 0.04–0.96; p=0.032) than children without eating problems (Appendix S2), but no significant difference in height and mid-arm circumference was found. In contrast, among the comparison group, only BMI was lower in those with eating problems than in those without (mean difference 1.1; 95% CI 0.1–2.0; p=0.028).

Correlations of total eating problems and growth parameters amongst EPC indicated that, with increasing eating problems, EPC were lighter (Spearman’s ρ=−0.242, n=223, p<0.001) and shorter (ρ=−0.232, n=220, p<0.001) and had smaller head circumference (ρ=−0.211, n=221, p=0.002), smaller mid-arm circumference (ρ=−0.165, n=221, p=0.014), and lower BMIs (ρ=−0.164, n=220, p=0.015). The partial correlations remained significant between total eating problems and weight (r=−0.204, n=203, p=0.003), height (r=−0.149, n=203, p=0.033), mid-arm circumference (r=−0.156, n=203, p=0.025), and BMI (r=−0.162, n=203, p=0.021), even when adjusting for gestational age, birthweight, feeding problems at 30 months, and cognitive, neuromotor, and pervasive behaviour disabilities, but were no longer significant for head circumference (r=−0.087, n=213, p=0.213).


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information

In this whole population cohort of EPC, we found a considerable excess of eating problems. These difficulties continue to cause significant distress and are perceived as significant problems by parents. Total eating problems at 6 years of age in EPC are only partly explained by other disabilities; in particular, differences with comparison children in total and oral motor problems remain after adjustment for disabilities. The eating problems in EPC significantly correlate with poorer attained growth at 6 years of age beyond the prediction afforded by disabilities, gestation, birthweight, and early feeding problems.

Among EPC, both males and females were more likely to have eating problems than their classmates. However, as previously shown,2,3 among EPC, cognitive and neurological problems were twice as frequent in males as in females, which partly explains the sex difference in oral hypersensitivity. Males in the EPC group were also more likely than females to suffer oral motor problems, an indicator of generally more delayed development in males than in females. Our findings also show that oral motor dysfunction persists beyond infancy7 in one-third of EPC. Furthermore, oral hypersensitivity and behavioural eating problems are still found in one-quarter of EPC at early school age, whereas food refusal or faddy eating is only slightly more common. Infants and toddlers with neurological impairments4,15 are more likely to experience eating problems and have more difficulty in dealing with highly textured food. Our findings indicate that EPC with neurological or cognitive disabilities, which often occur together,3 contribute to, but do not fully account for, eating problems observed.15 In particular, oral motor problems and oral hypersensitivity are increased in children with CP15 (found in 12/100 of EPC).3 Furthermore, learning disabilities,* as indicated by low cognitive scores, are associated with overall, and specifically oral motor and hypersensitivity, eating problems but do not explain them.16 Although impaired oral motor eating skills are transient and resolve in some extremely preterm infants, in others they may be early indicators of neurodevelopmental impairment17 due to brain damage,18 with both the neurodevelopmental and eating problems persisting. The more preterm the infant, the longer the dependency on tube feeding. As shown here, those infants born at extremely low gestation with neurodevelopmental problems are at greatest risk of developing tactile defensiveness and oral hypersensitivity.6,19

In contrast, refusal to eat, the most frequent problem encountered in general population samples of infants,20 was only slightly more common in EPC than in children in the comparison group. Refusal to eat is often related to difficult and irregular temperament and negative emotionality,21 and frequently leads to higher levels of conflict, non-contingency, and maternal intrusiveness during feeding interactions,22 thus resulting more often in secondary problems, including hyperactive behaviour23 or distress.24 The behavioural eating problems were explained by general behavioural problems of the children both at home and at school, rather than confined to the eating situation. This is another indicator that the pervasive and multiple problems common among EPC,1,2 rather than parenting difficulties, are a major contributor to eating problems in these children. Nevertheless, the eating and, in particular, the behavioural and oral hypersensitivity problems were perceived by parents as difficulties and as distressing. Thus, eating problems put an additional burden on the families of EPC with impairment.25

EPC (with and without eating problems) were smaller than comparison children.26 However, EPC with eating problems were significantly smaller and had less muscle mass than those without eating problems. In particular, poor weight gain and the development of muscle was lowest among EPC with eating problems, as reported previously.4,9 The relationship of eating problems to poor growth in our study was not explained by other disabilities, gestation, birthweight, or more physical activity (i.e. hyperactivity). Thus, although disabilities partly contribute to eating problems, it appears that the poor nutritional intake associated with eating problems in EPC explains some of the growth deficits in these children. Eating problems in infants are highly persistent throughout childhood27 and have been found to increase the risk of other cognitive, behavioural, and psychosocial problems28 in general population samples owing to malnutrition. Thus, even in children with significant disability, rectifying eating problems may potentially improve their growth and possibly their cognitive development.29

Overall, considering the stress caused to the child and caregiver, eating problems are not trivial for the families.30 Interventions suggested include oral motor therapy to reduce oral motor deficits and oral hypersensitivity.31 Furthermore, Fucile et al.32 have shown that intervention before the transition from tube to oral feeding may prevent some of the ‘early’ feeding difficulties in preterm children. On the other hand, behavioural therapy can be effective in treating behavioural eating problems.15 Early advice and support to parents whose preterm infants experience eating problems may reduce perceived eating difficulties33 and increase parental confidence in parenting and nurturing their children.

This study has a number of strengths, including its large sample size and the inclusion of a comparison group from the same neighbourhood. The logistic regressions before and after adjusting for disability factors included bootstrapping to determine 95% CIs, and they provided very similar values to those produced by SPSS but safer estimates of effect size ranges. Limitations are that we were unable to recruit a comparison child for each preterm child in mainstream classes.3 Previous research34 has shown that eating disorder questionnaires and interviews are highly correlated but may slightly overestimate the rate of eating problems.35 A structured feeding assessment including direct observation or structured testing of oral motor skills36 would have been desirable but was not feasible within a half-day comprehensive psychological and medical examination. Furthermore, those participants who did not attend the assessments were more likely to be from socially disadvantaged families. Although there was no difference regarding medical variables, early feeding behaviour, or growth, CP and overall disability were more frequent in those lost to follow-up than those assessed.1,3 Thus, the rate of eating problems reported here, often associated with another disability or social deprivation,15 is likely to be an underestimate of the true rate in the total EPC population.37


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information

In conclusion, at school age, EPC still have a two to five times increased risk of eating problems, and these are only partly accounted for by coexisting neurological, developmental, or pervasive behavioural impairments. Eating problems in EPC increase the risk of growth problems. Clinicians should be aware of the distress caused to the children and families, and early intervention, identification, and referral may alleviate some of the problems for the children and caregivers.38

  • *

    North American usage: mental retardation.


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information
  • 1
    Samara M, Marlow N, Wolke D. Pervasive behaviour problems at 6 years of age in a total-population sample of children born at <=25 weeks of gestation. Paediatrics 2008; 122: 56273.
  • 2
    Wolke D, Samara M, Bracewell M, Marlow N. Specific language difficulties and school achievement in children born at 25 weeks of gestation or less. J Paediatr 2008; 152: 25662.
  • 3
    Marlow N, Wolke D, Bracewell MA, Samara M. Neurologic and developmental disability at six years of age after extremely preterm birth. N Engl J Med 2005; 352: 919.
  • 4
    Cerro N, Zeunert S, Simmer KN, Daniels LA. Eating behaviour of children 1.5–3.5 years born preterm: parents’ perceptions. J Paediatr Child Health 2002; 38: 728.
  • 5
    Gewolb IH, Vice FL. Maturational changes in the rhythms, patterning, and coordination of respiration and swallow during feeding in preterm and term infants. Dev Med Child Neurol 2006b; 48: 58994.
  • 6
    Reilly S, Skuse D, Wolke D. The nature and consequences of feeding problems in infancy. In: CooperPJ, SteinA, editors. Childhood Feeding Problems and Adolescent Eating Disorders. London: Routledge, 2006: 740.
  • 7
    Mathisen B, Worrall L, O’Callaghan MO, Wall C, Shepherd RW. Feeding problems and dysphagia in six-month-old extremely low birth weight infants. Advances in Speech Language Pathology 2000; 2: 917.
  • 8
    Pridham K, Saxe R, Limbo R. Feeding issues for mothers of very low-birth-weight, premature infants through the first year. J Perinat Neonatal Nurs Res 2004; 18: 1619.
  • 9
    Wood NS, Costeloe K, Gibson AT, Hennessy EM, Marlow N, Wilkinson AR. The EPICure study: growth and associated problems in children born at 25 weeks of gestational age or less. Arch Dis Child Fetal Neonatal Ed 2003; 88: F492500.
  • 10
    Goodman R. Psychometric properties of the strengths and difficulties questionnaire. J Am Acad Child Adolesc Psychiatry 2001; 40: 133745.
  • 11
    Kaufman A, Kaufman N. Kaufman Assessment Battery for Children. Circle Pines, MN: American Guidance Service, 1983.
  • 12
    Efron B. Nonparametric estimates of standard error: the jackknife, the bootstrap and other methods. Biometrika 1981; 3: 58999.
  • 13
    Efron B, Tibshirani R. An Introduction to the Bootstrap. Boca Raton, FL: Chapman & Hall/CRC, 1993.
  • 14
    Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes JAMA 1998; 280: 16901.
  • 15
    Schädler G, Süss-Burghart H, Toschke A, Von Voss H, Von Kries R. Feeding disorders in ex-prematures: causes – response to therapy – long term outcome. Eur J Paediatr 2007; 166: 80308.
  • 16
    Bosma JF. Postnatal ontogeny of performances of the pharynx, larynx, and mouth. Am Rev Respir Dis 1985; 131: 5105.
  • 17
    Palmer MM, Heyman MB. Developmental outcome for neonates with dysfunctional and disorganized sucking patterns: preliminary findings Infant-Toddler Intervention. Transdisciplinary J 1999; 9: 299308.
  • 18
    Braun M, Palmer MM. A pilot study of oral-motor dysfunction in ‘at risk’ infants. Paediatrics 1985; 5: 1325.
  • 19
    Dodrill P, McMahon S, Ward E, Weir K, Donovan T, Riddle B. Long-term oral sensitivity and feeding skills of low-risk pre-term infants. Early Hum Dev 2004; 76: 2337.
  • 20
    Wolke D, Skuse D, Reilly S. The management of infant feeding problems. In: CooperPJ, SteinA, editors. Childhood Feeding Problems and Adolescent Eating Disorders. London: Routledge, 2006: 4191.
  • 21
    McDermott BM, Mamun AA, Najman JM, Wiliams GM, O’Callaghan MJ, Bor W. Preschool children perceived by mothers as irregular eaters: physical and psychosocial predictors from a birth cohort study. J Dev Behav Paediatr 2008; 29: 197205.
  • 22
    Lindberg L, Bohlin G, Hagekull B, Palmerus K. Interactions between mothers and infants showing food refusal. Infant Ment Health J 1996; 17: 33447.
  • 23
    Dahl M, Sundelin C. Feeding problems in an affluent society. Follow-up at four years of age in children with early refusal to eat. Acta Paediatr 1992; 81: 5759.
  • 24
    Jacobi C, Agras WS, Bryson S, Hammer LD. Behavioural validation, precursors and concomitants of picky eating in childhood. J Am Acad Child Adolesc Psychiatry 2003; 42: 7684.
  • 25
    Mathisen B, Worrall L, Masel J, Wall C, Shepherd RW. Feeding problems in infants with gastrooesophageal reflux disease: a controlled study. J Paediatr Child Health 1999; 35: 1639.
  • 26
    Elliman A, Bryan E, Walker J, Harvey D. The growth of low-birthweight children. Acta Paediatr Scand 1992; 81: 3114.
  • 27
    Dahl Rydell AM, Sundelin C. Children with early refusal to eat: follow-up during primary school. Acta Paediatr 1994; 83: 548.
  • 28
    Esparo G, Canals J, Jane C, Ballespi S, Vinas F, Domenech E. Feeding problems in nursery children: prevalence and psychosocial factors. Acta Paediatr 2004; 93: 6638.
  • 29
    Heird WC. Determination of nutritional requirements in preterm infants, with special reference to ‘catch-up’ growth. Semin Neonatol 2001; 6: 36575.
  • 30
    Blissett J, Meyer C, Haycraft E. Maternal mental health and child feeding problems in a non-clinical group. Eat Behav 2007; 8: 31118.
  • 31
    Carroll L, Reilly S. The therapeutic approach to the child with feeding difficulty. II. Management and treatment. In: SullivanPB, RosenbloomL, editors. Feeding the Disabled Child. London: Mac Keith Press, 1996; pp 11732.
  • 32
    Fucile S, Gisel EG, Lau C. Effect of an oral stimulation program on sucking skill maturation of preterm infants. Dev Med Child Neurol 2005; 47: 15862.
  • 33
    Satter E. The feeding relationship: problems and interventions. J Pediatr 1990; 117: S1818.
  • 34
    Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJV. Validity of the Eating Disorder Examination Questionnaire (EDE-Q) in screening for eating disorders in community samples. Behav Res Ther 2004; 42: 55167.
  • 35
    Wilfley DE, Schwartz MB, Spurrell EB, Fairburn CG. Assessing the specific psychopathology of binge eating disorder patients: interview or self-report? Behav Res Ther 1997; 35: 11519.
  • 36
    Reilly S, Skuse D, Mathisen B, Wolke D. The objective rating of oral-motor functions during feeding. Dysphagia 1995; 10: 17791.
  • 37
    Wolke D, Söhne B, Ohrt B, Riegel K. Follow-up of preterm children: important to document dropouts. Lancet 1995; 345: 447.
  • 38
    Wolke D. Frequent problems in infancy and toddler years: excessive crying, sleeping and feeding difficulties. In: BergmannKE, BergmannRL, editors. Health Promotion and Disease Prevention in the Family. Berlin: De Gruyter, 2003: 4488.

Supporting Information

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  8. Supporting Information
DMCN_3512_sm_AppendixS1.doc28KSupporting info item
DMCN_3512_sm_AppendixS2.doc62KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.