Cystic fibrosis (CF) is the most common life-shortening autosomal recessive disease of childhood. The clinical manifestations of CF include progressive suppurative lung disease and pancreatic exocrine insufficiency.1 In Australia, most infants with CF are identified through newborn screening by 4 weeks of age.2 Newborn screening for CF was introduced in the United Kingdom in 2008 and, in the United States, 40 states were expected to have CF newborn screening programs by the end of 2008. The proven benefits of newborn screening for CF include improved nutrition, normal fat soluble vitamin levels, improved cognitive skills, and avoidance of early complications.3^–5 Newborn screening also offers the opportunity for reproductive decision making for subsequent pregnancies. Screening may reduce parental anxiety, negativity and lack of confidence in health professionals, all of which have been associated with a delay in diagnosis.4 5 Early diagnosis, however, may increase the risk of parental depression.6 Other potential problems of early diagnosis have generally not been well explored.

The early diagnosis of CF, in most cases, will lead to earlier treatment that includes a rigorous daily regimen of airway clearance, pancreatic enzyme supplementation and high calorie meals. There is potential for conflict between the primary caregiver and the child, which may lead to harsh parenting (eg, increased yelling and smacking). Harsh parenting is the single most important, modifiable risk factor for later child externalising behavioural problems (eg, aggression, oppositional defiance and conduct problems).7 8 By contrast, early diagnosis may lead to over-involved and protective parenting, a known risk factor for later child internalising behavioural problems (eg, emotional problems such as anxiety, depression and withdrawal).9

The limited research on the behaviour of young children with CF has focused on mealtime behaviour and calorie intake. Observational studies suggest there are lower levels of family functioning at mealtimes10 and that parents engage in higher rates of ineffective mealtime behaviour management strategies compared to healthy controls.11^–13 The prevalence of internalising and externalising problems in preschool children with CF has not been measured. Broader CF care issues, such as adherence with airway clearance and medication regimens, have also not been well studied and apart from physiological parameters of sleep, such as hypoxia, there are no data regarding sleep behaviours in children with CF. Furthermore, the available research is based on unscreened populations. Finally, there is little information about parenting styles (eg, harsh, inconsistent and overprotective parenting) and mental health problems in parents of children with CF.

We therefore aimed to describe, in a sample of preschool children with CF: (a) the prevalence of child behavioural problems including therapy adherence, sleep, mealtime problems and internalising and externalising behavioural problems; (b) the prevalence of primary caregiver mental health problems and poor sleep quality; and (c) the associations between child problems and parenting styles. We also asked caregivers about their willingness to receive parenting advice about common problems.

METHODS

Setting and participants

Primary caregivers of children with CF, identified by newborn screening, aged 6 months to 5 years at the time of recruitment (September–November 2007), who attended a CF outpatient clinic at one of the participating hospitals, were eligible to participate. Participating hospitals included The Royal Children’s Hospital, Melbourne, Victoria (RCH), Sydney Children’s Hospital, New South Wales (SCH) and Monash Medical Centre, Victoria (MMC). These hospitals provide care for all the children with CF in Victoria and approximately 40% of children with CF in New South Wales. Children were identified from clinic databases. Primary caregivers with insufficient English to complete the questionnaire were excluded from the study.

Recruitment and consent

Participants from RCH and MMC were initially contacted by telephone for verbal consent. Following this, an information statement, consent form and questionnaire were mailed to each family. Participants from SCH were mailed the information statement, consent form and questionnaire without prior telephone consent. If there was more than one eligible child in the home, then a questionnaire was sent for each child. The difference in recruitment was due to requirements from the participating hospital ethics committees. The study was approved by the Human Research and Ethics Committee at each centre: RCH (HREC27061A), MMC (09074C) and SCH (07/199).

Measures

Primary caregivers completed either a 90-item questionnaire only (for children aged 6 months to 1.5 years) or a 99-item questionnaire together with the Achenbach Child Behavior Checklist (for children aged 1.5–5 years).14 Outcome and primary exposure measures are summarised in tables 1 and 2.15^–20 Covariates were child gender (male/female), age in months and primary caregiver highest level of education attained (did not complete high school, completed high school only or completed tertiary/postgraduate studies). In addition, primary caregivers reported on their willingness to participate in, and the preferred structure of, a parenting program to address common problems identified by the study.

The questionnaire also measured socio-demographic characteristics (child age, gender, ordinal number, primary caregiver age, marital status, education and employment status, and partner age and education and employment status) and the number of hospital admissions in the past 12 months. Socio-economic status (SES) was measured using the Socio-Economic Indexes for Areas (SEIFA) for each family’s residential postcode.21 SEIFA values are standardised scores compiled from census data that numerically summarise the social and economic conditions of Australia by postcode (national mean 1000, SD 100; lower values represent greater disadvantage).

Statistical analysis

The distributions of the child and family characteristics and each of the sample outcomes were summarised using appropriate summary statistics. The prevalence of outcome variables was compared to published Victorian or Australian normative data (preferred) or international normative data if no local data were available (as described in tables 1 and 2). Unadjusted and adjusted logistic regression analyses were carried out with parenting style (harsh, inconsistent or overprotective) as the binary primary exposure and child problem behaviour (externalising, internalising, eating, sleep and treatment adherence) as the outcome. All regressions were adjusted for child gender, age in months and primary caregiver highest level of education attained. We present odds ratios and 95% confidence intervals with p values only for statistically significant results (p<0.05). Data obtained about caregiver willingness to participate in, and preferred structure of, a parenting program were summarised using appropriate summary statistics. Analyses were conducted using Stata v 10.0 (StataCorp, College Station, TX).

RESULTS

Participant characteristics

There were 145 eligible children with CF attending the three centres. Two were excluded (non-English speaking) and four did not consent. This left 139 participants (RCH, n = 72; MMC, n = 22; SCH, n = 45), of whom 117 (84.2%) completed the questionnaire. Mothers comprised 92.3% of the primary caregivers. Demographic data are presented in table 3. Responders did not differ from non-responders in terms of SES (SEIFA value 1023 vs 998, p>0.1), child age or child gender (all p>0.05).

Prevalence of child and primary caregiver problems

The prevalence of child problem behaviours are presented in table 4. Caregivers reported a sleeping problem (mild to severe) in 61/114 (53.5%) children across all ages, with 22/36 (61.1%) children aged 4–6 years reported to have a sleep problem. Overall, moderate and severe sleep problems were more common in children with CF than in healthy children. Children with CF were twice as likely to experience four or more eating problem behaviours (40.2% (41/102) vs 21%) compared to normative data.22 Caregivers reported that physiotherapy adherence was a problem in 59/117 (50.4%) children. The prevalence of internalising and externalising problems was comparable to normative data.

Primary caregiver mental health was poor (table 4). The numbers of caregivers reporting symptoms above the clinical cut-point indicative for depression were 39/117 (33.3%), for anxiety 19/116 (16.4%) and for stress 40/117 (34.2%). More than a third (44/177, 37.6%) reported that their own sleep quality was “fairly” or “very” bad, similar to community data.

Associations between child behaviours and parenting styles

Unadjusted and adjusted logistic regression for parenting styles and child behaviours are presented in table 5. Children whose caregivers reported harsh parenting were almost four times more likely to display internalising problem behaviours than children whose caregivers did not report harsh parenting (95% CI 1.15 to 13.85).

Parenting program

When caregivers were asked whether they would like information about child behaviour via a parenting program, 86/116 (74.1%) responded “yes”, 11/116 (9.5%) responded “no” and 19/116 (16.4%) were unsure. Of those who responded “yes” or who were unsure, 25/105 (23.6%) preferred a group program and 19/105 (17.9%) an individual program, with the remainder unsure. The majority of caregivers preferred that a parenting program be available before the onset of child behavioural problems (83/106, 78.3%) and that information be made available in a face-to-face session (65/106, 61.3%). Other acceptable methods of receiving parenting information included written handouts (72/106, 67.9%) and DVDs (54/106, 50.9%). Of those who preferred a face-to-face information session, 56/65 (86.1%) indicated that they would want a health professional to present it.

DISCUSSION

This is the first study to report the prevalence of child sleep, mealtime, adherence and behavioural problems in Australian preschool children with CF. Our sample captured all Victorian and approximately 40% of NSW children under 6 years of age with CF, so our findings are likely to be generalisable to most Australian children with CF. Compared to healthy children, children with CF were twice as likely to have mealtime and sleep problems and their primary caregivers were three times more likely to report symptoms indicative of depression and stress. Adherence to physiotherapy regimens was also problematic. Harsh parenting was significantly associated with internalising behavioural problems, but other parenting styles were not associated with child outcomes as anticipated, possibly due to a lack of power.

It is a novel, but not surprising, finding that the greater daily demands on caregivers of children with CF present more parenting challenges. The children need to comply with an increased calorie requirements, pancreatic enzyme, vitamin and salt replacement, daily airway clearance treatments and disruptive hospital visits. The high rates of problems with eating and physiotherapy adherence for airway clearance in our sample are therefore not unexpected. Pleasingly, most caregivers would welcome additional parenting support to manage these problem behaviours.

The high prevalence of reported mealtime problems is consistent with results from previous studies. A UK case–controlled study of 108 children with CF aged 1–7 years, using the same mealtime problem measure (Behavioral Pediatrics Feeding Assessment Scale), found a similar mean problem frequency score of 58.2 (SD 12.8) compared to 54.2 (SD 15.5) in this study.23 Additional controlled studies report that children with CF display more mealtime problem behaviours,12 15 23 parents view their child’s behaviours as more problematic15 24 and the number of problem behaviours correlates inversely with caloric intake.15

There is some evidence that programs focusing on behavioural management are beneficial in a CF population. A randomised clinical trial of preschool children with CF, using behavioural management training and nutrition counselling as the intervention, resulted in children meeting their recommended caloric intake.25 Uncontrolled studies also show that short-term behavioural intervention programs, focusing on parent training, positive reinforcement of food acceptance and extinction of negative behaviours, result in improve caloric intake, both initially and at 9-month follow-up.26^–28 While there are no studies specifically reporting treatment adherence rates in preschool children with CF, one study, using a multi-method approach to measure treatment adherence in primary school aged children, reported overall adherence rates well below 50%.29 Although not directly comparable, this is in line with the high reported rates of physiotherapy adherence problems in our study.

There are no previous data regarding internalising and externalising problems in preschool children with CF. However, two case–controlled studies (n = 144 school aged children with 177 controls and n = 43 children aged 7–14), found an increased prevalence of anxiety in CF children versus control children.30 31 This is in contrast to our findings and there could be a number of reasons for this. Preschool children with CF in our population are generally well and rarely experience prolonged hospitalisation. Also, compared to school aged children, preschool children are less likely to understand the longer term implications of CF.

Sleep problems in this population have not been well studied. Data from a large, nationally representative study showed that sleep problems in Australian healthy infants and preschool children are common.20 However, moderate or worse sleep problems in our CF population were even more common than those in the national study, suggesting that CF could have a negative impact on sleep in preschool children.

Our finding of relatively high rates of caregiver mental health problems is consistent with the limited data available. One study (n = 45) found that parents of infants with CF may be more vulnerable to depression than matched controls.6 Two further studies have shown that mothers of preschool and primary school aged children with CF had poorer mental health than control groups and the general population.32 33

There were several strengths to this study. First, we approached a substantial proportion of the eligible CF families from the participating hospitals with a high response rate. Second, non-responders were similar to responders with respect to socio-economic status and child age and gender, suggesting that results could generalise to the wider group of preschool CF children in Australia. Finally, we used validated measures for most outcomes and compared them with Australian normative or community data. Although a case–control study design may have strengthened our comparisons, case–control studies are subject to their own biases, such as selection bias, and we did not have adequate funds to cover the increased costs that such a design would incur.

Our study had some limitations. It was a cross-sectional survey and as such causality cannot be determined. As there was no matched control group, we chose not to compare our outcomes statistically with the population data. The study involved a primary caregiver report, which may be subject to bias compared to direct observation of child behaviours. Primary caregivers reported on their own parenting as well as their child’s behaviour, so the association between harsh parenting style and internalising child behaviours needs to be interpreted with caution. However, for sleep problems, caregiver report correlates well with observational methods such as over-night infra-red camera recordings.34 Primary caregiver report is also likely to reflect clinically meaningful problems as it takes into account the impact of the child’s behaviour on the family and parent help-seeking practices. Finally, despite being a large sample for a CF population, a lack of power may have contributed to the lack of statistically significant relationships between other parenting styles and child behaviours.

This study has a number of clinical and research implications. Clinicians should be aware of, and inquire about, child sleep, mealtime and airway clearance problems during regular consultations for children with CF. Routine patient care should also take into account caregiver mental health and sleep. A multidisciplinary approach, encompassing patient and family wellbeing and appropriate referral for caregiver mental health and parenting assistance, should be part of the management of children with CF. Future research should focus on a parenting intervention to prevent or treat the common problems identified in this study.