The study was conducted at Al Faiha polyclinic (primary care clinic) during 2009. For entry into the study, adolescents, ages 10 to 14 years had to be obese (body mass index (BMI) above the 95th percentile
), have at least one parent who expressed a willingness to attend the intervention described below if allocated to it, and have no serious underlying medical condition that might be either a cause or consequence of their obesity. Adolescents were recruited from their schools after BMI and health screening was conducted by one of the researchers (SAB). The schools and residences of all participants were located in the capital, and both the venue for the intervention (Al Faiha polyclinic) and the venue for the control primary care clinic were also in the capital to ensure that distance was not an issue hindering attendance. Ethical approval was obtained from the Ethics Committee for Medical Research, Ministry of Health of Kuwait (ref MPH/112), and written informed consent was obtained from both parents and adolescents.
Randomisation and allocation concealment
Participating adolescents attended a research clinic where all baseline measures (see below) were taken, then assigned a unique study code prior to random allocation into the treatment or control group. To ensure concealment of allocation, codes were sent electronically to a statistician (JHM) who produced a computer generated randomisation list that allocated participants to intervention or control group, with participants balanced for gender in blocks of 10. The statistician informed the researcher responsible for delivering the intervention (SAB) of the allocation, and families were invited to intervention or control groups as appropriate.
In brief, the intervention was intended as a relatively low intensity (6 sessions, 1 hour contact time per session, delivered as a group session) programme, which might be readily generalisable if evidence of feasibility and efficacy was obtained from the present study. The intervention was delivered to the adolescents and their parents in group-discussion fashion over a 24-week period by a physician with specialist training in Nutrition (SAB) and the study dietician. The programme was adapted from the Scottish Childhood Obesity Treatment Trial (SCOTT), which tested a ‘good practice’ treatment intervention in Scotland
[8, 9]. Parents were provided with treatment materials that were adapted from those used in SCOTT
. The intervention is described here as a ‘good practice’ intervention on two grounds. First, because it focused on changing the behaviours recommended as the key targets in recent evidence-based management guidelines
[10–12] for the treatment of adolescent obesity (reduction in sedentary behaviour, particularly screen-media use; diet, using a modified version of the ‘traffic light diet’ system
; and promotion of physical activity). Second, the intervention incorporated theoretically based behaviour change techniques to all three of the targeted behaviours
: exploration of the pros and cons of changes in diet, physical activity, and sedentary behaviour; exploration of motivation to change diet, physical activity, and sedentary behaviour; self-monitoring of sedentary behaviour (recording of screen time in diaries), diet, and physical activity (recording of walking, sport, and physically active play in a diary); identifying the main barriers to behaviour change and problem solving in relation to these barriers; goal setting in relation to diet, physical activity, and sedentary behaviour; and relapse prevention.
The intervention group was further divided into boys (n = 21, each attending with at least one parent) and girls (n = 20, each attending with at least one parent) groups in accordance with cultural norms of the Kuwaiti population, and their sessions were delivered on two consecutive days. Any adolescent who attended the intervention session alone was welcomed, although ideally at least one parent should have been present.
Primary care-based treatment of child and adolescent obesity in Kuwait is somewhat limited, as in many other countries, but it was felt ethically and scientifically appropriate to use referral to primary care as a control condition in the present study. Adolescents, and their parents, who were allocated randomly to the control group were therefore informed that they were obese and advised to attend primary care.
Outcome measures and blinding
BMI Z scores were calculated based on US CDC 2000 reference data
 using the software available at
http://stokes.chop.edu/web/zscore/index.php. Outcome measures were made at baseline and again at 6 months (26 weeks) after the start of the intervention by the same trained research assistants who were blinded to group allocation and were not involved in delivery of the treatment intervention. Blood pressure was measured when the participant was sitting quietly in the upright position, with the correct cuff size applied to the right arm. The reading was repeated three times and the average of the three readings was taken.
The primary study outcome measure was change in BMI Z score. Weight was measured to 0.1 kg in light indoor clothing with children not wearing shoes, and height was measured to 0.1 cm with a portable stadiometer (Leicester Height Measure, SECA, London, UK) and adolescents not wearing shoes. Secondary outcomes were waist circumference and blood pressure.
Sample size, power, and statistical analysis
No local data were available upon which to base a power calculation. The present study was therefore powered using BMI data from the Scottish SCOTT RCT
, which was used to develop the treatment intervention. With a between-group difference in the change in BMI Z-score of -0.25 at 6 months (which is a small change in BMI Z score, as discussed below) and a SD of change in BMI Z score of 0.21, giving a delta of 1.15, a sample size of around 30 adolescents per arm at 6 months would give 90% power at the 0.05 significance level. Dropout from the trial could not be predicted, but it was hoped that entering around 90 adolescents would make sufficient allowance for attrition during the 6-month study to leave around 30 participants per arm at the end of the trial.
Outcomes were analysed in two ways. First, changes in outcome variables within each group (intervention and control) between baseline and 6-month follow-up are presented. The issue of whether changes in outcome variables differed significantly between groups (intervention versus control) was examined using independent sample t-tests. The analysis was intention-to-treat, where we used data from all adolescents for whom data were available on the basis of the group to which they were allocated, regardless of their adherence to the protocol (attendance).
Feasibility of the trial: treatment intervention and control conditions
Since the present study was the first of its kind in Kuwait, it was not considered as a ‘definitive trial’
 but rather as the initiation of a process which should lead to a more definitive trial subsequently
. Feasibility of the trial and feasibility of the trial interventions were measured by summarising the extent of sample attrition (dropout) over the 6-month period, the extent of missing data, and recording attendance at intervention and control treatments.