International Journal of (2000) 24, 1445±1453 ß 2000 Macmillan Publishers Ltd All rights reserved 0307±0565/00 $15.00 www.nature.com/ijo The effects of a children's summer camp programme on , with a 10 month follow-up

PJ Gately1*, CB Cooke1, RJ Butterly1, P Mackreth1 and S Carroll1

1School of Leisure and Sports Studies, Leeds Metropolitan University, Fairfax Hall, Beckett Park, Leeds LS6 3QS, UK

OBJECTIVE: To assess the long-term effects of a multidisciplinary approach involving structured fun-type skill learning physical activities in the treatment of obese and children. DESIGN: A longitudinal investigation incorporating repeated measurements before and after the 8 week intervention and after the 44 week follow-up period. METHODS: The camp programme (Massachusetts, USA) utilised structured fun-based skill learning physical activities, moderate dietary restriction and behaviour modi®cation. The primary aims of the intervention were to reduce body mass and promote the maintenance of the reduction in body mass using an alternative to standard prescription. SUBJECTS: One-hundred and ninety-four children (64 boys and 130 girls, aged 12.6Æ 2.5 y) enrolled at a summer , of which 102 children (38 boys and 64 girls aged 13.6Æ 2.4 y) returned 1 y later. MEASUREMENTS: On commencement of the programme all children were assessed for body mass and stature. At follow-up, data was available on 102 subjects for body mass and stature. RESULTS: Over the 8 week intervention signi®cant reductions (P ˆ 0.00) in body mass were obtained. During the 44 week follow-up signi®cant increases (P ˆ 0.00) were noted in body mass, (BMI) and stature, but as expected there were large variations in the responses. One year after the initial measures had been taken mean body mass and BMI were lower than at the start of the intervention, BMI signi®cantly so (week 0, 32.9 Æ 7.4 kg=m2; week 8 29.1Æ 6.5 kg=m2; week 52, 30.05Æ 7.04 kg=m2); (P ˆ 0.00). Stature increased signi®cantly (week 0, 1.58Æ 0.12 m; week 52, 1.64Æ 0.11 m) (P ˆ 0.00) during this period, demonstrating a reduction in mean body mass over a 1 y period whilst subjects continued to increase in stature. When changes in BMI are analysed with the use of standard scores, there is a non-signi®cant increase (P ˆ 0.07) in BMI during the follow-up phase and 89% of children had a lower BMI than at week 0. CONCLUSIONS: These ®ndings suggest that the use of a structured fun-based skill learning programme may provide an alternative method of exercise prescription to help children prolong the effects of the 8 week intervention. Further investigations will help identify the key factors that are necessary for long-term lifestyle modi®cation. International Journal of Obesity (2000) 24, 1445±1452

Keywords: ; physical activity; multidisciplinary approach; weight loss

Introduction of physical activity) and follow-up measurements are rarely carried out. Most studies have been conducted Obesity is recognized as a major cause of morbidity on adult subjects and have shown that intervention and mortality and has been identi®ed as a major factor strategies are dif®cult to maintain.8,9 This lack of in a range of hypokinetic diseases.1,2 Trends described success is attributed to adjustments in dietary intake from the National Health and Nutrition Examination when subjects return to their normal environment,10 or Surveys (NHANES) indicate a dramatic increase in poor adherence to exercise.11 overweight prevalence in the US adult and paediatric Many treatments of obesity have used dietary populations.3,4 interventions to effect a negative energy balance.12 Numerous studies have reported short-term success However, researchers have identi®ed that some sub- in the treatment of obesity.5±7 However, these studies jects undertaking dietary interventions may become contain methodological problems (small sample, lack susceptible to over-eating and success in dietary of control over energy intake, problems in assessment programmes is limited.7,8 It has been suggested that obesity may be due to a lack of physical activity.13,14 Researchers have sug- *Correspondence: PJ Gately, School of Leisure and Sports gested that physical activity is necessary to increase Studies, Leeds Metropolitan University, Fairfax Hall, Beckett energy expenditure, decrease appetite, positively Park, Leeds LS6 3QS, UK. Received 31 March 1999; revised 21 January 2000; accepted affect eating behaviour, improve cardiovascular func- 3 April 2000 tioning and psychological well being.14,15 Despite Children's weight loss programme PJ Gately et al 1446 these ®ndings, little is known about what factors affect The aim of this study was to evaluate the overall adherence. Parker16 has suggested that structured fun- effects of a multifactorial intervention for reducing type exercise is a form of intervention which can be body mass in children using moderate calori®c restric- used to promote adherence and motivation to develop tion, educational and behavioural modi®cation and a habitual physical activity which will lead to an structured fun-based skill learning physical activity increase in energy expenditure and positive changes programme. The acute effects of the 8 week pro- in body mass. An alternative approach to structured gramme were assessed pre- and post-treatment with a physical activity prescription has been proposed by follow-up after a further 44 weeks. Epstein,17 which involves the reduction of sedentary behaviour which has a positive effect on physical activity levels. This research illustrates that alterna- Methods tives to structured physical activity prescription should be developed to investigate their potential Eight-week programme for improving adherence to physical activity One-hundred and ninety-four children (64 boys and interventions. 130 girls aged 12.6Æ 2.5 y) enrolled on the camp Approaches which have used and exercise to programme. The programme is a traditional residen- effect reductions in body mass are more successful than tial summer camp in a rural setting (Massachusetts, diet or exercise alone.5,18 ± 21 Many programmes, how- USA) with a speci®c aim of weight loss. The very ever, are short-term in nature and do not educate the nature of the programme prohibits the opportunity of participants or provide them with the necessary skills to identifying an appropriate control group. In addition, maintain their reduced body mass. In general, results this type of programme is restricted to children of have been disappointing, with typical treatments result- higher socio-economic status, although a number of ing in a 5 ± 20% reduction in body mass, only to return children were supported by assisted places or social to original levels of body mass within 5 y.8 services funding. Some studies have produced some success during follow-up programmes. Pavlou's15 study on Boston policemen in a work-centred programme demon- Follow up strated a signi®cant reduction in body mass using a One-hundred and two children (38 boys and 64 girls combination of dietary restriction and exercise and aged 13.6Æ 2.4) re-enrolled on the camp programme illustrated a non-signi®cant change in body mass in 1 y later. All subjects were assessed for body mass and the subjects who continued to exercise during the 18 stature. All measures were taken by the same month follow-up phase. researcher during the initial programme and the There have been numerous studies undertaken on follow-up. No signi®cant differences were found in the treatment of obesity in the adult population. In any of the variables between the returners and non- contrast, there has been little work carried out on the returners. As part of the camp programme a selection paediatric population.10,11 This would seem to be of non-returners were contacted to investigate reasons problematic, given that levels of inactivity and obesity for non-return. A variety of reasons were identi®ed by persist into adulthood.22 ± 26 Indeed evidence continues the parents or children for not returning to the camp. to mount with regard to the risks associated with These included families who reported that their child obesity during childhood, signi®cant morbidity being had continued to lose weight and did not need to the immediate consequence and increases in adult return to the camp and families who reported that their mortality also being highlighted by a number of child did not enjoy their experience at camp the studies.27,28 Many practitioners see obesity as a dif®- previous year. There seemed to be no common cult problem to treat, as there is limited success from responses or reasons for children not returning. It interventions.7 Without research evidence on appro- must be noted that this information was purely obser- priate interventions for children the likelihood of vational data recorded by the camp programme rather developing appropriate treatments for this population than the current investigators, so caution must be is low. taken when interpreting this information. Despite this, Epstein17 has proposed the use of One confounding factor in the analysis of long- behavioural interventions to increase participation in itudinal data for children is growth, which will sig- physical activity, rather than using regimented exer- ni®cantly affect body mass during the intervention cise to affect reductions in body mass. In addition, and follow-up. The contribution of growth to the these interventions offer the opportunity of establish- change in body mass over the follow-up therefore ing longitudinal effects, as they have been followed has to be accounted for within the present study. The up over 5 and 10 y, with data provided by Epstein29 use of population data for comparison is problematic, illustrating that 34% of the sample decreased percen- due to the sample of obese children not being compar- tage overweight by 20% or more and 30% were no able to the normal population data provided by the longer obese. These longitudinal studies demonstrate National Center for Health Statistics (NCHS). Data on the importance of multifactorial methodologies using differences in the patterns of growth for obese and diet, physical activity and behaviour modi®cation. non-obese children have been reported by Epstein.30

International Journal of Obesity Children's weight loss programme PJ Gately et al 1447 In addition, Van Lenthe31 has identi®ed that rapidly Behavioural and educational intervention maturing children have a signi®cantly higher body Children took part in an educational programme mass index (BMI) during adulthood. This research incorporating aspects of health, nutrition and physical illustrates the problems associated with assessing ®tness. All sessions were conducted by quali®ed obese and overweight children over a prolonged professionals on a twice weekly basis. The purpose period of time. Thus, the analysis of the data has of the sessions was to educate the children so that they taken a number of forms to clearly present the data could develop the necessary knowledge to help mini- and to attempt to account for the confounding effect of mize recidivism upon completion of the programme. growth on body mass. Each form has its limitations The sessions were informal, and included question but the general pattern of the analysis may help and answer elements as well as fun-type activities illustrate the pattern of change during the follow-up such as quizzes and competitions. period. The forms of analysis include change in body A social cognitive model of behavioural modi®ca- mass, change in BMI, comparison of change in body tion was incorporated within the camp programme. mass and stature with NCHS data and change in Fox32 identi®ed that children's motives for participa- standardised BMI scores. tion in physical activity should be considered during the development of appropriate interventions. There- fore, the programme included the following key ele- The physical activity programme ments; fun-type activity, skill-based sessions to The physical activity programme was based on activ- develop competence in a range of activities, choice ities designed to provide children with a wide range and a strong social component. These factors have of fun-type skill learning experiences, and to improve been identi®ed by researchers as key motives for physical ®tness and con®dence within a sporting and participation in physical activity.33 ± 35 This beha- exercise environment. Each session was conducted vioural approach to physical activity was used to by a specialist instructor. The ®rst 2 weeks of the improve self-ef®cacy and therefore increase participa- programme were speci®cally aimed at developing tion in physical activity. In addition, during the 8 children's sports skills, cardiorespiratory ®tness, week programme parents attended a weekend of exercise tolerance and the prevention of injuries classes involving training in nutrition, physical activ- through appropriate awareness and body preparation. ity and long-term . Parents were The following 6 weeks involved further development instructed to act as role models to help reinforce the of children's sports skills and exercise tolerance. behaviours that had been developed during the camp During this part of the programme, the newly programme. acquired sporting skills were developed further offer- ing an opportunity for children to be involved in competitive activities, thus, providing children with a Measures number of positive experiences to help them maintain Upon arrival all children were assessed for body mass their levels of activity on return to their normal and stature. All measures were taken according to the environment. The programme consisted of ®ve one- guidelines suggested by Lohman.36 and-a-half hour sessions per day. These activities The primary tools used for analysis during the included: one aerobic (eg orienteering), one water- intervention and follow-up were body mass, BMI based (eg water polo), one circuit based (eg resis- and standardised BMI. BMI is a simple measure of tance training), and two games based sessions (eg overweight and has been identi®ed by a number of football, basketball). A wide range of activities were researchers as an appropriate tool for identifying provided so that children could experience a variety overweight in children and offering the opportunity of new activities. to predict overweight and health risks in adult- hood.28,37 Hammer38 has suggested `for any index of overweight to be meaningful in the paediatric age Dietary restriction group, there must be standards de®ned by age and All children underwent calori®c restriction throughout sex'. Therefore, NCHS39 standards were used to the 8 week residential period. The calori®c restriction develop age and sex-related standardized BMI was 5860 kJ=day (1400 kcal) provided as three meals scores so that the children enrolled on the camp (breakfast, lunch and dinner) and two snacks in the programme could be compared to national reference afternoon and early evening. In addition, the children data. These data were also used to identify if there were observed during meals to ensure they did not were any differences between the expected increase in abstain from eating and also that they did not consume stature and body mass of the normal population and more calories than permitted. The aim of the dietary the actual increase in stature and body mass of the intervention was to cause suf®cient caloric restriction sample during the 44 week follow-up period. to effect a reduction in body mass without causing Body mass was recorded on a weekly basis during malnourishment or interruption in growth and matura- the 8 week programme. All measures were taken by tion, but with suf®cient energy content so that the same researchers throughout the study. Children children could sustain the physical activities. were required to have a medical examination prior to

International Journal of Obesity Children's weight loss programme PJ Gately et al 1448 acceptance at the camp and written informed consent was also obtained from the parents.

Statistical analysis Both a paired t-test and a one-way analysis of variance weeks 0 ± 52

with repeated measures were performed on the data for Percentage change, the 8 week camp period and 44 week follow-up, respectively. Statistical procedures were carried out using SPSS. All results were considered signi®cant at 2.6** 8 1.3 ns 2 0.06** 4 15.2 ± 13.5)

(P < 0.05), but actual computer-printed P values are 7 7 ‡ Change, 35.8 ± 37.19) 7 weeks 0 ± 52

quoted in the results. A Scheffe post hoc test was used 7 for the one-way anova where appropriate. Finally, a independent t-test was carried out to assess if there were any differences between the returners and non-returners. weeks 8 ± 52

Results Percentage change,

Eight-week intervention In the sample of 194 children (64 boys 130 girls, aged 1.8** 6 12.6Æ 2.5 y) highly signi®cant reductions (P ˆ 0.000) 10.0** 14 ‡ 11.1 ± 19.6) ( 20.4 ± 52.2) ( ‡ were achieved in body mass (Table 1). Change, 7 7 weeks 8 ± 52 7.0 0.11 Follow up period 25.9 Æ Æ During the follow-up phase for the sample of 102 Æ children (38 boys and 64 girls aged 13.6Æ 2.4 y) signi®cant increases (P ˆ 0.000) were noted. Stature increased by 4%, body mass by 14% and BMI by 6% (Table 1). Independent t-tests were carried out on the returners and non-returners. This analysis showed that

there were no signi®cant differences between groups 102) on body mass, BMI, WHR and the sum of nine circumferences ˆ weeks 0 ± 8 Week 52 in terms of body mass (P ˆ 0.061), and the change in n

body mass (P ˆ 0.809). Percentage change, 0.6) 0.9) (20.2 ± 62.3) ( 3.2) 2.0) (48.1 ± 201.9) ( 7 7 7 Overall changes 7 3.8**4.4** 12 13 30.1

Results from the 8 week intervention and the follow- 10.8**11.2** 13 13 82.2 7 7 Change, 7 7 10.7 to 10.7 to 25.4 to up phase for the 102 children are shown in Table 1. As 27.4 to weeks 0 ± 8 7 7 7 might be expected, the 8 week intervention produced 7 desirable changes of a large magnitude which were followed by a reversal during the follow-up period. 6.5 6.4 22.5 23.6 Æ Æ

However, the reversal that occurred during the follow- 194) and the 44 week follow-up ( Æ Æ

up period did not produce a return to the initial mean ˆ values recorded at week 0. n Figure 1 illustrates the variation of responses to the 44 week follow-up period. It shows that 10% of 7.47.2 29.1 28.3 0.12 1.64 25.826.6 74.4 72.3

subjects continued to reduce body mass during the Æ Æ Æ Æ Æ follow-up period and 53% of subjects had a lower (20.2 ± 67.6)(20.2 (18.2 ± ± 67.6) 59.6) ( (18.2 ± 58.9) ( (1.30 ± 1.88) (1.36 ± 1.93) (0.01 ± 0.10) (36.3 ± 214.1) (33.1 ± 188.7) ( body mass than when they had enrolled on the camp (46.7 ± 214.1) (39.5 ± 186.69) ( programme 1 y previously. With regard to BMI,

Figure 1 shows that 26% of subjects continued to 194) 85.2 102) 83.5 ˆ ˆ n n reduce BMI during the 44 week follow-up period and 0.01. 194)102) 32.9 32.7 102) 1.58 < ˆ ˆ 83% of subjects had a lower BMI than when they ˆ P n n n ( ( were enrolled onto the programme. 2 2 Effects of the 8 week programme ( m m = One attempt to account for the confounding = effects of growth was to compare the present sample 0.05; ** < P BMI, kg Data presented as mean,* standard deviation and the range. Table 1 Stature, m ( VariablesBody mass, kg ( Body mass, kg ( Baseline Week 8 with the 95th percentile for 13 ± 14 y-old children BMI, kg

International Journal of Obesity Children's weight loss programme PJ Gately et al 1449

Figure 1 Percentage of sample categorised according to changes in each variable during the 44 week follow-up period with respect to baseline values. using data from the NCHS percentiles.27 The 95th weight can be accounted for by growth. A further percentile for stature and body mass has been chosen problem with the use of NCHS is that the majority of as this is a more representative standard for this the children in the present sample are signi®cantly sample. The change in boys mean stature over a 1 y above the 95th percentile. period for this sample was 7.00 cm, which is similar to The 85th percentile has been identi®ed as an the NCHS data where the mean increase in stature was appropriate tool for the de®nition of obesity.40 ± 42 6.9 cm. Within the girls sample, the change in mean Therefore, BMI scores were standardised in an stature was 5 cm, which is higher than the NCHS attempt to account for the confounding effects of mean change of 3.2 cm. The change in boys' mean growth. This form of analysis is carried out using body mass over the 1 y period for this sample was the 50th percentile and the 85th percentile, as an 11.78 kg compared to the expected change of 7.11 kg approximation of the population mean and standard from the NCHS data. The girls' change in mean body deviation. The use of standard scores allows the mass over the 1 y period was 8.9 kg compared with an observation of the children over the one year period expected gain of 5.78 kg from the NCHS data. These while they are growing. Figure 2 shows the BMI data illustrate that a large amount of the increase in expressed as standard scores at weeks 0, 8 and 52.

Figure 2 Standard scores for BMI of children weeks 0, 8 and 52.

International Journal of Obesity Children's weight loss programme PJ Gately et al 1450 These data illustrate population BMI at the 50th can be fun and can be developed by the individual in a percentile (0.00) the 85th percentile (1.00) and the variety of ways, thus empowering them to participate mean standard score at weeks 0, 8 and 52. During the more fully in recreational exercise and sport away 8 week intervention there was a signi®cant from the intervention situation. This is supported by (P ˆ 0.000) reduction in the mean standard score for Dietz, who has suggested that `alternatives to the BMI (3.61Æ 1.9 week 0 to 2.43Æ 1.64 week 8). conventional approach to therapy must be consid- During the follow-up period there was a non-signi®- ered',24 and Epstein,17 who has suggested the reduc- cant (P ˆ 0.071) increase in mean BMI standard score tion of sedentary behaviour rather than the adoption of (2.43Æ 1.64 week 8 to 2.73Æ 1.93 week 52). In regimented physical activity as an important factor in addition, the 44 week follow-up score for standardised long term weight management. BMI was signi®cantly lower (P ˆ 0.000) than the Calori®c restriction in this study was not as severe standard BMI score at week 0. The individual stan- as in other studies, ie 1400 kcal=day in the present dard scores provide a similar picture to the other study compared to 1000 kcal=day in Pavlou's study.15 forms of analysis, indeed, the analysis with standard This was an attempt to avoid the problems associated scores provide further positive data on the camp with loss of lean muscle tissue and interference with programme and the 44 week follow-up period. growth and maturation, which can occur with severe Figure 1 shows that, during the 44 week follow-up calori®c restriction.43,44 The emphasis in this element period with respect to standardised BMI, 40% of the of the programme was to give children the necessary subjects continued to reduce their standardized score nutrition for maintenance of their physical activity for BMI and 91% of the subjects had a lower standard together with the skills and understanding of portion score for BMI at week 52 compared to week 0. control to modify their energy intake on return to their normal environment. The mean reductions in body mass reported during the 8 week programme are similar to those reported Discussion during other studies. In a review by Miller20 mean reductions in body mass were 11 kg for diet and Many studies have indicated that obesity is associated exercise programmes which averaged 13.4 weeks in with increased morbidity, mortality and a variety of duration. This weight loss is similar to current study disease including coronary heart disease (CHD), where the mean reduction in body mass was 10.8 kg; and .1,2 The Framingham study however, the duration of the present study was 8 has attempted to quantify the effect of a successful weeks. Caution must be taken with this analysis as reduction in body mass on these diseases. The results the review by Miller was carried out on programmes suggested that heart disease could be reduced by 25 ± involving adults. Unfortunately, very little directly 35% if optimal weights were attained.1 comparable data exists on children's weight loss It would seem that strategies such as dietary inter- programmes, so it is dif®cult to compare the outcomes vention, exercise and behavioural modi®cation all of this intervention with others. Researchers that have contribute to short-term success in the treatment of reported success in the treatment of obese children obesity.7 However, each has problems associated with include: Epstein et al 45 used programmed exercise poor long-term success. Adopting a strategy which and dietary restriction to obtain a signi®cant reduction uses these tools collectively may help address (P < 0.01) in BMI of 1.37 kg=m2 in a sample of obese the large reversal of body mass seen in many children during an 8 week programme; and Sasaki et programmes. al 46 used dietary restriction and exercise during a 2 y This study used an approach that involved a multi- programme and noted signi®cant reductions factorial strategy which included calori®c restriction, (P < 0.001) of 55% (boys) and 48% (girls) in obesity physical activity, behavioural modi®cation and educa- index. Clearly differences in age of children, length of tion. Furthermore, the physical activity programme programme, modality of programme and form of utilized differed greatly from previously reported assessment used, make direct comparisons between studies, as the camp programme incorporated a struc- these interventions dif®cult. tured fun-based skill-learning approach to physical During the follow-up period in this study there was activity and sport. The aim was to improve children's a signi®cant increase (P ˆ 0.000) in mean body mass. skills in a variety of sporting situations so that they This was to be expected as subjects were unlikely to would improve in self con®dence and performance fully maintain the levels of physical activity or dietary and therefore be more likely to maintain these activ- restriction experienced on the camp programme and ities when they returned to their normal environments. also this is a time of rapid growth and development in This is different from conventional approaches which this age group. In spite of this, the mean BMI was have used exercise prescriptions based on ®xed inten- signi®cantly lower (P ˆ 0.000) at week 52 than at the sity, duration and frequency and focused only on start of the programme. reducing body mass.5,15,17,23 Such interventions may Obesity has been de®ned as a BMI above the 85th have been unsuccessful because they did not teach percentile for age and sex.27 ± 29 This de®nition was participants that physical activity, exercise and sport used to produce standard scores so that the effects of

International Journal of Obesity Children's weight loss programme PJ Gately et al 1451 growth may be quanti®ed. Figure 2 shows the change addition, these data showed a non-signi®cant increase in standardised BMI scores from weeks 0, 8 and 52. (P ˆ 0.071) in the standardised BMI scores during the There was a non-signi®cant difference between 44 week follow-up. body mass at week 0 and week 52, which illustrates One obvious problem with the study design was the that overall the subjects did not gain extra body mass. absence of a control group. This poses a number of This is even more encouraging given there was a problems with regard to generalising the results of this signi®cant increase (P ˆ 0.000) in stature over the intervention. However, this form of intervention pre- same period. sents problems with the assignment of a control group These ®ndings were further supported when com- due to it being a special residential programme. parison was made with the NCHS data to illustrate the Indeed, the multifactorial nature of this intervention expected change in body mass due to growth over this increases the problems experienced when trying to period. This crude use of the NCHS data shows that control for the variety of changing variables. growth may account for as much as 65% in girls and Despite the lack of a control group, the problems of 60% in boys of the total change in body mass during controlling exactly for growth, and the lack of support the follow-up period. The comparison of change in that is available to the children throughout the year, stature of the children and the NCHS data illustrate the evidence presented here suggests that the camp that the change in the sample of boys is similar to the programme may have had some role in the mainte- change in NCHS reference data; however, the change nance of body mass over 1 y whilst the subjects have in stature of the sample of girls was higher than the continued to grow. Furthermore, BMI has reduced NCHS reference data. This greater change in stature over the same time period. It would appear that a of the sample of girls may indicate the attainment of camp programme which concentrates on increasing peak height velocity during this period which would the skills associated with sports performance and have a favourable affect on the results. Unfortunately other recreational physical activities and not just on there is no growth data available on these children reducing body mass, is to some degree successful in further than the 1 y period and the evidence that obese producing some lifestyle modi®cations in the children. children have different growth rates to the normal Given the lack of support that exists for the 44 week population makes these results dif®cult to interpret. period these results are encouraging. Therefore, caution must be taken when viewing the Further work is needed to establish some local ®ndings in this context. It should be noted these data support in order to increase the likelihood of greater were used as an observation to illustrate that chil- success for long-term change. In addition, children dren's increased body mass may in part be attributed should be monitored during the follow-up phase to to growth, rather than just their inability to manage investigate if they have employed lasting lifestyle their body mass. changes. It would also be useful to quantify the An important and interesting factor is the large improvements in sporting skills made by the children variation in the responses during the 44 week so that valid statements can be made about the follow-up period, when children were unsupported improvements in their ability to perform sports in and subject to a range of factors likely to in¯uence their normal environments. Qualitative analysis is their ability to maintain their reduced body mass required to identify any common factors in the during the treatment period. The camp programme responses of the participants to the camp programme. can be seen to be successful given that 10% of Finally, reference data for overweight and obese subjects continued to reduce body mass and a further children's growth patterns would be useful to enable 43% of subjects had a lower body mass than when the assessment of treatments and long-term follow-ups they enrolled on the programme. In addition, 26% of within this population. At present it is dif®cult to the subjects reduced their BMI during the 44 week quantify changes in this population as there is no follow-up. The variation in responses are important, applicable data to measure treatments against. This and may highlight the need for individualized support form of reference data would therefore be useful to during the follow-up period. Currently little data identify appropriate references of treatment and exists with regard to how best to follow-up partici- change. pants involved in interventions for the treatment of obesity. Overall there was an 6% increase in BMI during the 44 week period but 80% of subjects had a lower BMI References 1 Hubert HB, Feinleib M, McNara PM, Castelli WP. Obesity as than when they enrolled on the camp programme. an independent risk factor for a 26 y When individualized standard scores are calculated a follow up of participants in the Framingham Heart study. similar picture is seen in the data. However, the Circulation 1983; 67: 452 ± 458. standard scores illustrate an improved analysis with 2 Sjonstrom LV. Morbidity of severely obese subjects. Am J the data suggesting 40% of subjects continue to Clin Nutr 1992; 55: 508S ± 515S. 3 Kuczmarski RJ, Flegal KM, Campbell SM, Johnson CL. reduce their standardised BMI during the 44 week Increasing prevalence of overweight among US adults. The follow-up and 89% of subjects had a lower standar- National Health and Nutrition Examination Surveys 1960 to dised BMI than at the start of the programme. In 1991. JAMA 1994; 272: 205 ± 211.

International Journal of Obesity Children's weight loss programme PJ Gately et al 1452 4 Center for Disease Control. Update: prevalence of overweight 26 Freedman DS, Shear CL, Burke GL, Srinivasan SR, Webber among children, adolescents and adults Ð United States, LS, Harsha DW, Berenson GS. Persistence of juvenile onset 1988 ± 1994. JAMA 1997; 277: 1111. obesity over eight years. The Bogalusa Heart study. AJPH 5 Cohen CJ, McMillan CS, Samuelson DR. Long-term effects of 1987; 77: 588 ± 592. a lifestyle modi®cation exercise programme on the ®tness of 27 Must A, Strauss RS. Risks and consequences of childhood and sedentary, obese children. J Sports Med Phys Fitness 1991; 31: adolescent obesity. Int J Obes Relat Metab Disord 1999; 23: 183 ± 188. S2 ± S11. 6 Hagan RD, Upton SJ, Wong L, Whittam J. The effects of 28 Must A, Jacques PF, Dallal GE, Bajema CJ, Dietz WH. Long- aerobic conditioning and=or caloric restriction in overweight term morbidity and mortality of overweight adolescents. men and women. Med Sci Sports Exercise 1986; 18: 87 ± 94. New Engl J Med 1992; 327: 1350 ± 1355. 7 Hill JO. The effects of exercise and food restriction on body 29 Epstein LH, Valoski AM, Wing RR, McCurley J. Ten-year composition and metabolic rate of obese women. Am J Clin outcomes of behavioural family-based treatment for childhood Nutr 1987; 46: 622 ± 627. obesity. Health Psychol 1994; 13: 373 ± 383. 8 Garner DM, Wooley SC. Confronting the failure of beha- 30 Epstein LH, McCurley J, Valoski A, Wing R. Growth in vioural and dietary treatments for obesity. Clin Psychol Rev obese children treated for obesity. Am J Dis Child 1990; 1991; 11: 729 ± 780. 140: 1360 ± 1364. 9 Wardle J. Obesity and behaviour change: matching problems 31 Van Lenthe FJ, Kemper CG, Van Mechelen W. Rapid matura- to practice. Int J Obes Relat Metab Disord 1996; 1: S1 ± S8. tion in adolescence results in greater obesity in adulthood: the 10 Depres JP, Bouchard C, Savard A. The effect of a twenty week Amsterdam Growth and Health Study. Am J Clin Nutr 1996; endurance training programme on morphology 64: 18 ± 24. and lipolysis in men and women. Metabolism 1984; 33: 32 Fox K, Biddle S. The child's perspective in physical educa- 235 ± 239. tion. Part 2: Children's Participation Motives. Br J Phys Educ 11 Bar-Or O, Foreyt J, Bouchard C, Brownell K, Dietz WH, 1988; 19: 79 ± 82. Ravussin E, Salbe AD, Schwenger S, St. Joer S, Torun B. 33 Gould D. Psychosocial development and children's sport. In: Physical activity, genetic and nutritional considerations in Thomas JR (ed). Motor Development During Childhood and childhood weight management. Med Sci Sports Exerc 1998; Adolescence. Burgess: Minneapolis, MN 1984. 30: 2 ± 10. 34 Wankel LM, Kreisel PS. Factors underlying enjoyment of 12 Brown MR, Klish WJ, Hollander J. A high protein, low calorie youth sports: sport and age group comparisons. J Sports liquid diet in the treatment of very obese adolescents: long Psychol 1985; 7: 51 ± 64. term effect on . Am J Clin Nutr 1983; 38: 35 Thompson CE, Wankel LM. The effects of perceived activity 20 ± 31. choice upon frequency of exercise behaviour. J Appl Soc 13 Prentice AM, Jebb SA. Obesity in Britain; gluttony or sloth. Psychol 1980; 10: 436 ± 443. Br Med J 1995; 311: 437 ± 439. 36 Lohman TG, Roche AF, Martorell R. Anthropometric Stan- 14 Bouchard C, Depres JP, Tramblay A. Exercise and obesity. dardization Reference Manual. Human Kinetics: Champaign, Obes Res 1993; 1: 133 ± 147. Il, 1989. 15 Pavlou KN, Krey S, Steffee WP. Exercise as an adjunct to 37 Guo SS, Roche AF, Chumlea WC, Gardner JD, Siervogel RM. weight loss and maintenance in moderately obese subjects. Am The predictive value of childhood body mass index values for J Clin Nutr 1989; 49: 1115 ± 1123. overweight at age 35 y. Am J Clin Nutr 1994; 59: 810 ± 819. 16 Parker DL, Bar-Or O. Juvenile obesity. The importance of 38 Hammer LD, Kraemer HC, Wilson DM, Ritter PL, Dornbusch exercise and getting children to do it. T Physician Sportsmed SM. Standard percentile curves of body-mass index for chil- 1991; 19: 113 ± 125. dren and adolescents. Am J Dis Child 1991; 145: 259 ± 263. 17 Epstein LH, Valoski AM, Vara LS, Mccurley J, Wisniewski L, 39 Hamill PV, Drizd TA, Johnson CL, Reed RB, Roche AF, Kalarchian MA, Klein KR, Shrager LR. Effects of decreasing Moore WM. Physical growth: National Center for Health sedentary behavior and increasing activity on weight change in Statistics percentiles. Am J Clin Nutr 1979; 32: 607 ± 629. obese children. Health Psychol 1995; 14: 109 ± 115. 40 Dietz WH, Robson TN. Assessment and treatment of child- 18 Ballor DL, Poehlman ET. A meta-analysis of the effects of hood obesity. Pediat Rev 1993; 14: 337 ± 344. exercise and=or dietary restriction on resting metabolic rate. 41 Must A, Dallal E, Dietz WH. Reference data for obesity: 85th Eur J Appl Physiol 1995; 71: 535 ± 542. and 95th percentiles of body mass index (wt=ht2) and triceps 19 Garrow JS, Summerbell CD. Meta-analysis: effect of exercise, skinfold thickness. Am J Clin Nutr 1991; 53: 839 ± 846. with or without , on the body composition of over- 42 World Health Organisation. Obesity, preventing and managing weight subjects. Eur J Clin Nutr 1995; 49: 1 ± 10. the global epidemic. Report of a WHO consultation on 20 Miller WC, Koceja DM, Hamilton EJ. A meta-analysis of the obesity. WHO: Geneva, 1997. past 25 y of weight loss research using diet, exercise or diet 43 Rowland T. Exercise and Children's Health, 2nd edn. Human plus exercise intervention. Int J Obes Relat Metab Disord Kinetics: Champaign, Il, 1990. 1997; 21: 941 ± 947. 44 Saris W. The role of exercise in the dietary treatment of 21 Ready AE, Fitzpatrick DW, Boreskie SL, Hrycaiko DW. The obesity. Int J Obes Relat Metab Disord 1993; 17: S17 ± S21. response of obese females to low impact exercise and diet 45 Epstein L, Wing RR, Koeske R, Ossip DJ, Beck S. A counselling. J Sports Med Phys Fitness 1991; 31: 587 ± 595. comparison of lifestyle change and programmed aerobic 22 Armstrong N. The challenge of promoting physical activity. exercise on weight and ®tness changes in obese children. J R Soc Health 1995; 6: 187 ± 192. Behav Ther 1982; 13: 651 ± 665. 23 Malina RM. Tracking of physical activity and physical ®tness 46 Sasaki J, Shindo M, Tanaka H, Ando M, Arakawa K. A long- across the lifespan. Res Q Exercise Sport 1996; 67:48±57 term aerobic exercise program decreases the obesity index and (Suppl 3). increases the high density lipoprotein cholesterol concentra- 24 Dietz WH. Therapeutic strategies in childhood obesity. Horm tion in obese children. Int J Obes 1987; 11: 339 ± 345. Res 1993; 39: 86 ± 90. 25 Stark O, Atkins E, Wolff OH, Douglas JWB. Longitudinal study of obesity in the national survey of health and develop- ment. Br Med J 1981; 281: 13 ± 17.

International Journal of Obesity