Study design and recruitment
A pre-post design evaluated a home-based lifestyle support intervention conducted in 2008. It was hypothesized that an individualised home-based intervention would be a feasible child obesity treatment approach. Children aged 4–12 years who were overweight or obese according to the International Obesity Task Force (IOTF) Body Mass Index (BMI) cut-points [18] and who were living in metropolitan Adelaide were recruited via referrals to the Department of Paediatrics and Child Health, Flinders Medical Centre for management of obesity. Exclusion criteria were medical conditions affecting weight or growth or being enrolled in any other structured weight management program, however no child was excluded on these grounds. Children could continue to be managed by their general practitioner or paediatrician during the intervention. Written informed parent consent and child assent was obtained. The study was approved by the Flinders University Social and Behavioural Research Ethics committee.
The study sequence is shown in Fig. 1. At baseline and follow up, participants attended a 1-h assessment at which anthropometry and questionnaire data were obtained. This was followed by a 75-min home visit to assess the home environment. The three home visitors were trained via a 3 h workshop covering the socioecological basis of childhood obesity, dietary and physical activity guidelines, and how to work with families in non-judgemental partnerships. Comprehensive written instructions and a training site visit ensured a standardised procedure was followed.
Measures
At baseline a socio-demographic questionnaire was completed by parents, and child weight, height and waist circumference were measured to the nearest 0.1 kg and 0.1 cm, using standardised scales and stadiometer. These measurements were undertaken in the paediatric outpatient department, Flinders Medical Centre, using the same set of scales and stadiometer for each child and at both baseline and follow-up. The research assistant undertaking these measurements was not involved in the home visit intervention and had been trained in anthropometry. Specifically, children were weighed on portable digital scales wearing light clothing, with shoes and socks removed. Height was measured using a wall-mounted stadiometer with shoes and socks removed, heels touching the wall and head in the ‘Frankfort’ plane. Waist circumference was measured using a standard metric tape measure held horizontally at a level midway between the lower rib margin and iliac crest (approximately in line with the umbilicus). BMI was calculated (weight, kg/height, m2) and weight status determined by applying the IOTF cut-points [18]. BMI and waist circumference z-scores were calculated using LMSGrowth Excel add-in which is based on 1990 United Kingdom reference data [19, 20].
Parents completed the 24-item Children’s Dietary Questionnaire (CDQ) [21] which generates four continuous scores (i.e. fruit and vegetables, non-core foods, fat from dairy, sweetened beverages). Non-core foods are defined as foods that are not essential to meet nutrient requirements and contain excess energy, fat, sugar and salt [22]. The CDQ shows acceptable reliability (ICC 0.5–0.9) and relative validity at the group level for the fruit and vegetables and non-core food scores (Cronbach alpha 0.62–0.72) [21].
Child activity patterns were measured using an adapted version of The Outdoor Play and Small Screen Entertainment Checklist [11, 23]. Parents reported the average time children spent per weekday and weekend day (in hours and minutes) (1) playing outdoors, (2) doing moderate-vigorous physical activity, and (3) using electronic media such as television and computers for both. Six items were added to the original checklist to explore average time spent per day (4) participating in physical activity and (5) using small screen entertainment. Times were averaged (weighted for type of day) to calculate (1) daily activity (vigorous physical activity + outdoor play) and (2) daily sedentary activity.
The Home Environment Inventory provides an assessment of 70+ family home characteristics associated with children’s dietary, physical activity, or sedentary behaviours [11]. Items cover the outdoor environment to support children’s play, parental role modelling, accessibility of small screen entertainment, parent behaviours around food preparation and availability of healthy and non-core foods and beverages in the home and child’s lunchbox. Inventory items were assessed by either direct observation or by parental report (details described elsewhere [11]).
Study intervention
The study intervention was based on ecological theory [3, 24, 25]. The intervention was undertaken in family’s homes by trained visitors. Three home visits and two follow-up phone calls were offered to each family. Three home visitors were trained in the methodology with particular emphasis on being nonjudgmental, supportive and sensitive to socioeconomic determinants of obesity. Home visitors were chosen for their interpersonal skills and in this study all three were in the process of completing bachelor degrees in either nutrition, occupational therapy or human movement. The family was contacted by telephone by the home visitor who introduced themselves, explained the process and made a suitable time for the first home visit (Visit 1). At Visit 1 the home visitor clarified that the carer (in all instances this was the mother) understood the study and evaluated the family’s home environment using the Home Environment Inventory. The completed Home Environment Inventory was used to identify changes the family needed to make to improve the home environment. These changes were set as the intervention goals. To provide advice for each family, these goals were used to individualise a report template containing 25 strategies, along with ideas of how each strategy could be implemented. The ten most relevant strategies for each family were highlighted at the beginning of the report. This report was presented to the family at a second home visit (Visit 2), conducted approximately 2 weeks (±1 week) after Visit 1. The purpose of Visit 2 was to go through the report with the parent and to discuss ways of achieving change that the parent felt would be manageable for their family situation. The strategies were based on key evidence for improving healthy eating and physical activity in children [5, 22, 25–28]. Any additional issues relevant to achieving the desired outcomes that mother/family/carer identified during the home visit were noted for example housing, budgeting and child behaviour. The impact of these socioeconomic factors on the ability of families to institute change became more obvious once the study progressed and by the fact that visits were undertaken in the family home. The importance of this aspect of obesity management is illustrated in the case study provided in Fig. 2.
At Visit 2, time was made for the first follow-up phone call approximately 1 month after Visit 2. The purpose of this phone call was to gauge the family’s progress, provide ongoing support and encouragement with regards to the strategies and to provide further advice and support if requested. If families requested additional information, standard resources were posted to them (for example label reading) or a suggestion was provided about where they could access assistance (for example a child psychologist). The phone calls also provided opportunity to discuss any strategies that families felt were not working, and make alternative suggestions. At the end of the call, a time was made for the second follow-up phone call which was to occur approximately a month after the first phone call. And at the end of this call, a time for the final home visit (Visit 3) was made (aiming to be approximately 4 months after Visit 1). At Visit 3, the inventory was repeated. Encouragement for continuing to make lifestyle changes was also given to the families at this final visit.
Statistical analysis
Analyses were conducted using SPSS for Windows version 17.0. Descriptive statistics are presented as frequencies or mean (standard deviation)/median (IQ range). Wilcoxin signed-rank tests and paired t tests were used to assess changes from baseline to follow up in subscale scores for the CDQ, Physical Activity Questionnaire and anthropometric data. Categorical data from the Home Environment Inventory were analysed using Chi square tests for independence. Statistical significance was set at p < 0.05.