Rapid Responses to:

RESEARCH: John J Reilly, Louise Kelly, Colette Montgomery, Avril Williamson, Abigail Fisher, John H McColl, Rossella Lo Conte, James Y Paton, and Stanley Grant Physical activity to prevent obesity in young children: cluster randomised controlled trial BMJ 2006; 333: 1041 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

The important role of life style

Ada Majd   (11 October 2006)

BMI in the BMJ

Daniel J Green, Professor Tim Cable   (12 October 2006)

Childhood Obesity and Physical Activity - a National Campaign

Seyed Hamid Reza Naghavi MD   (13 October 2006)

The Activitystat Hypothesis

Terence J Wilkin, Brad S Metcalf, Linda D Voss   (15 October 2006)

Re: The Activitystat Hypothesis

David Bossano   (16 October 2006)

Negative outcome or protocol problem ?

Alastair Michell   (1 November 2006)

Childhood obesity prevention - What could be the winning formula?

Somasundari Gopalakrishnan, Miranda J Pallan, Clinical Research Fellow, Dept of Public health and epidemiology   (8 November 2006)

Physical activity to prevent obesity: author's reply

John J Reilly, James Y Paton, John H McColl, Avril Williamson   (15 November 2006)

Unsuccessful intervention is not the same as a negative outcome

Alastair R Michell   (15 November 2006)

Change in BMI is not a measure of change in obesity

Louise Parker   (18 November 2006)

Proves that single-interventions are not very fruitful

Allen P Ugargol   (20 November 2006)

Dietary and behavioral modifications in managing childhood obesity

S Kapoor   (21 November 2006)

Childhood Obesity:..No golden bullet.

Emmanuel Agogo   (21 November 2006)

Measuring Child Weight Status

Rory O'Conor   (22 November 2006)

Childhood obesity and gym membership

Ian P Rodd   (22 November 2006)

Role of dietary modification

Dr. Raghesh varot kangath   (22 November 2006)

Childhood obesity : Is exercise no good ?

Alka Thakur   (22 November 2006)

The important role of life style 11 October 2006
Ada Majd, GP Tehran-Iran Send response to journal: Re: The important role of life style	You addressed a very important point. In fact the only real and long lasting, non-invasive cure for obesity, until now is changing lifestyle. Exercise by itself hardly makes any changes while other non healthy life styles like high carb diets, long time sittings and habits of eating non planed meal continues. I think more important than exercise can be educating people with the goal of changing behavior that ends to a change in the life style and promoting a healthy life style. Competing interests: None declared
BMI in the BMJ 12 October 2006
Daniel J Green, Professor of Cardiovascular Physiology Research Institute for Sport and Exercise Sciences, Liverpool John Moores University, L3 2ET, Professor Tim Cable Send response to journal: Re: BMI in the BMJ	In their recent study entitled “Physical activity to prevent obesity in young children …”, Reilly and colleagues conclude that preschool children (4.2 ± 0.2 yrs) require alternative interventions to physical activity to prevent obesity. In the abstract and discussion of their paper the authors appropriately refer to the lack of impact of their intervention on body mass index (BMI), their primary outcome measure. However, their emphasis on obesity throughout the paper, including in the title and conclusions, somewhat predictably led to the study being picked up internationally by the popular press and reported in the following or similar terms: “MORE EXERCISE DOES NOTHING TO STOP OBESITY IN YOUNGSTERS, STUDY FINDS” (Guardian Friday October 6th). This interpretation is obviously unfortunate, given that the energy balance equation states that energy storage equals energy intake (food) minus energy expenditure (exercise/physical activity). Putting this fundamental principle to one side, we believe the paper raises several important issues. Firstly, researchers, clinicians and the media should ask themselves what changes in body mass index (BMI) actually mean. In large cross- sectional population studies performed in adults, BMI arguably provides a useful surrogate index of obesity because it attempts to correct for individuals who are heavy by virtue of the fact that they are also tall. A higher BMI in a particular population might reasonably be assumed due to excess societal fat mass. Of course, BMI provides no information regarding the composition of the weight (more on this below), or its distribution. For the latter reason it is likely that waist circumference will ultimately prove a more predictive index of cardiovascular risk in adults. An important point is that changes in BMI within subjects following interventions are far more problematic to interpret. In adults, exercise training and diet do not, of course, alter body height. Hence, BMI reverts to a measure of body weight. Body weight is not a strong predictor of cardiovascular disease risk; if it was, we would be using body weight as a surrogate cardiovascular disease “risk factor”, rather than BMI or other measures that attempt to reflect fatness. We therefore submit that changes in BMI within subjects cannot be appropriately used as an index of change in cardiovascular risk, even in adults, let alone 4 year old children. Furthermore, in relation to the use of BMI as a risk factor in 4 year old children, the following questions are germane; Does BMI predict future cardiovascular disease risk? Does change in BMI in growing 4 year old children predict future cardiovascular risk? If not, is change in BMI a valid surrogate risk factor for cardiovascular risk in children? In addition, what does change in BMI mean in an age group in whom somatic growth patterns differ markedly and in whom height, as well as weight, are changing with time? How do differential changes in height between groups impact on the predictive accuracy of changes in BMI in children? We cannot tell much from the paper of Reilly et al. as there is no data in the paper on changes in height or weight. The absence of these components of the BMI calculation in growing children, in a journal with the international cache of the BMJ, is difficult to comprehend. The absence of follow-up mean BMI data, when this is the dependent variable, is mystifying. The arguments above raise serious issues about the use of changes in BMI as a surrogate for change in cardiovascular disease risk, especially in children. For reasons detailed below, we furthermore submit that changes in BMI as a result of exercise interventions should not even be used as a surrogate marker of changes in obesity. Exercise is an intervention which can increase skeletal muscle and lean body mass. Exercise studies in groups such as children, in whom somatic growth patterns are dynamic and variable in any case, require careful attention to changes in body composition, rather than simplistic measures of body weight or BMI. The point was illustrated to us in our own studies of supervised exercise training in obese children and adolescents.(1) Like Reilly et al., we observed no differences in body weight or BMI between the exercise training and inactive control conditions. However, dual energy x-ray absorptiometry (DEXA) scans revealed significant decreases in central fat mass, which were not apparent in terms of change in body mass or BMI because of increases in lean body mass in the lower limbs across the training period. Changes in central adiposity are, of course, particularly relevant in obese subjects and we also observed improvements in insulin sensitivity measured using euglycaemic hyperinsulinaemic clamps.(2) A review of the few well- controlled studies of obese children and adolescents in which DEXA scanning for body composition was undertaken revealed similar countervailing impacts of exercise training on fat and lean body mass.(3) We concluded this review by emphasising the importance of consideration of changes in body composition, rather than gross measures of body weight, in studies involving exercise; change in BMI is a particularly poor measure of change in obesity in such studies. Finally, it must be noted that the intervention used by Reilly et al. apparently did not lead to differences in physical activity or sedentary behaviour levels between the experimental groups. This raises the fundamental question of whether their intervention was effective in manipulating the independent variable. Why would one expect a physical activity intervention which has no impact on physical activity to alter energy balance or “obesity”? In summary, the study of Reilly et al. illustrates some pitfalls associated with the use of a surrogate risk factor which, whilst arguably valid in cross-sectional comparisons of adults, should not be used in within-subjects studies of change, particularly those conducted in children. When the nature of the intervention further compromises the validity of using changes in BMI to detect changes in adiposity, a study entitled “Physical activity to prevent obesity ....” becomes more than difficult to interpret. It becomes misleading. Professors Daniel J Green and N. Timothy Cable Research Institute for Sport and Exercise Sciences Liverpool John Moores University 1. Watts K, Beye P, Siafarikas A, et al. Exercise training normalises vascular dysfunction and improves central adiposity in obese adolescents. JACC 2004;43:1823-7. 2. Thompson AM, Naylor L, Ratnam N, et al. Resistance training improves insulin sensitivity in insulin resistant obese adolescents (Abstract). Australasian Society for the Study of Obesity, 14th Annual Scientific Meeting, October 28-30 2005. 3. Watts K, Davis E, Jones T, et al. Effect of exercise training in obese children and adolescents. Sports Med. 2005;35:1-18. Competing interests: None declared
Childhood Obesity and Physical Activity - a National Campaign 13 October 2006
Seyed Hamid Reza Naghavi MD, Specialist Registrar in Occupational Medicine Hammeresmith hospital W12 0HS Send response to journal: Re: Childhood Obesity and Physical Activity - a National Campaign	EDIROR- I read with great interest the article by John J Reilly[1] discussing the physical activity for prevention of childhood obesity. Obesity has been described as “Global Epidemic” by World Health Organisation (WHO). In the UK there are about 1 million obese children who are less than 16 years of age[2]. In England by 2001, 8.5% of 6 year olds and 15% of 15 year olds were obese.[3] Diet and physical activity are two main factors in childhood obesity. Childhood Obesity can cause several adverse health impacts for example it rises the incidence of diabetes. Overweight adolescents have a 70% chance of becoming overweight or obese adults.[3] Therefore immediate steps should be taken to address the life style of children in the country especially with regards to healthy diet and physical activity. In this study the authors could not find significant association between physical activities and reduce in BMI (body mass index). They targeted the nurseries to implement the exercise program which seems to have limited impact. It is important that for changing the attitude and behaviour toward the physical activity a wider approach is needed and the role families and communities are essential. 'Walk to School' is a national scheme to encourage parents, children and teachers to promote physical activity by walking to and from school so far as is practicable. October is International Walk to School Month.[4] The authors have used the BMI to measure the outcome of physical activity. Ellen W. Demerath et al suggest that BMI percentile changes may not accurately reflect changes in adiposity in children over time.[5] Therefore other outcomes like FMI (fat mass index) can be considered. References: 1 John J Reilly et al. Physical activity to prevent obesity in young children: cluster randomised controlled trial BMJ, doi:10.1136/bmj.38979.623773.55 (published 6 October 2006) 2 Preventing childhood obesity British Medical Association, June 2005 3 Childhood Obesity; Parliamentary Office of Science and Technology Number 205 September 2003 4 http://www.walktoschool.org.uk/index.php 5 Ellen W. Demerath et al. Do Changes in Body Mass Index Percentile Reflect Changes in Body Composition in Children? Data From the Fels Longitudinal Study Pediatrics 2006;117;487-495 Competing interests: None declared
The Activitystat Hypothesis 15 October 2006
Terence J Wilkin, Prof Endocrinology and Metabolism Peninsula Medical School, Plymouth campus, Brad S Metcalf, Linda D Voss Send response to journal: Re: The Activitystat Hypothesis	Sir, In their study of nursery school children,(1) Reilly and colleagues hypothesised that an intervention of three thirty-minute sessions per week of enhanced physical activity would raise total activity. It did not, and the authors conclude that their intervention may not have been sufficiently intense. However, there is an alternative, possibly more compelling, explanation – that the physical activity of children is regulated internally by an ‘activitystat’, and not externally by opportunity. There is a growing literature to support the ‘Activitystat Hypothesis’,(2,3,4) but the authors do not refer to it nor test it in their children. It could explain their results. The Activitystat Hypothesis has been tested both in and out of school. In a study of more than 200 children wearing MTI accelerometers in three schools providing very different opportunity, school 1 (S1) provided 9.0h per week of timetabled PE, S2 provided 2.2h and S3 1.8h.(2) Activity in school was greater among the children attending S1 compared with S2 or S3, but the reverse was true out of school, such that total daily activity was the same for all three (difference less than 0.3%). Indeed, <1% of the four-fold variation in activity of the children could be explained by the five-fold variation in opportunity provided. In a second study, the energy cost of being driven to school amounted to 16% during the hours of the ‘school run’ (8-9am and 3-4pm), but total activity was the same between those who walked and those who did not (difference less than 0.1%).(3) Again, the difference was made up at other times during the week. These two studies, and other data reported by the EarlyBird study,(4) suggest that the activity of children follows a set-point that is individual to the child. The four-fold range of activity observed among young children appears to represent their range of set-points, not their environments. Most activity is unstructured, and children seem to use it to compensate for the impact of structured or imposed activity. It is important to recall that many studies are able to show a temporary rise in physical activity following exercise intervention that does not last long- term, nor once the intervention has been withdrawn. In both the studies that we cite in support of the Activitystat Hypothesis, the children were in the 'exposure group' for months or even years before the activity recordings were taken and were therefore in ‘steady-state’. Reilly and colleagues noted a response to their short-term pilot intervention of 12 weeks that was not (or possibly no longer) apparent at the end of the full, longer-term study (24 weeks). Reilly and colleagues will be able to identify the intervention days (where the physical activity recorded will be higher than controls, provided the intervention was effective) and should be able to show when the excess was compensated for. The data could be made available to these correspondence columns, for it is important to know whether their results were indeed due to failure of the intended intervention at the time allotted, or to the operation of an activitystat at other times. The data that would most help would be 1. The mean activity (at all intensities) recorded among the test group during nursery-time on the intervention days, compared with the corresponding period for the controls, and for both groups on non-intervention days. 2. The mean daily activity (at all intensities) recorded out-of-nursery in both groups on intervention days and non-intervention days. The key issue is whether structured intervention in-nursery had any effect on in-nursery activity and, if it did, where the subsequent compensation occurred. Sincerely, Terence Wilkin MD, Brad Metcalf BSc, Linda Voss PhD References (1) Reilly JJ, Kelly L, Montgomery C, Williamson A, Fisher A, McColl JH et al. Physical activity to prevent obesity in young children:cluster randomized controlled trial. Brit Med J;Oct 6, 2006:Epub (2) Mallam KM, Metcalf BS, Kirkby J, Voss LD, Wilkin TJ. Contribution of timetabled physical education to total physical activity in primary school children: cross sectional study. BMJ. 2003;327:592-3. (3) Metcalf BS, Voss LD, Jeffery AN, Perkins J, Wilkin TJ. Physical activity cost of the school-run: impact on schoolchildren of being driven to school (EarlyBird 22). Brit Med J 2004;329:832-3. (4) Wilkin TJ, Mallam KM, Metcalf BS, Jeffery AN, Voss LD. Variation in Physical Activity lies with the Child, not his Environment: Evidence for an ‘Activitystat’ in Young Children (EarlyBird 16) Int J Obesity (Lond). 2006;30:1050-55. Competing interests: None declared
Re: The Activitystat Hypothesis 16 October 2006
David Bossano, GP St Leonards Practice, Exeter Send response to journal: Re: Re: The Activitystat Hypothesis	This is an interesting hypothesis which I have not encountered before. I have 2 comments: The first is anecdote, regarding my 4 year old daugher. We frequently talk about how much "energy" she has. She often discerns two types of "energy", not having enough "energy" for walking or climbing the stairs to bed but having lots of "energy" for running, swinging on monkey bars or bouncing on the bed. However sometimes she does choose not to walk, prefering to reserve "energy" for more enjoyable and vigorous activities! I am grappling with the activitystat concept. It seems to me that childrens' activity levels must be distributed (normally) around a population mean (by definition). The activitystat is perhaps more closely related to the SPREAD of activity around the mean than to the LEVEL of the mean activity? If so this is important because it suggests that intervention would be most usefully directed at shifting the mean level of activity (probably by social or population methods) not the spread of the activity. This follows the line of argument of Rose in shifting the curve versus treating the high risk individuals. Geoffrey Rose Sick individuals and sick populations Int. J. Epidemiol. 30: 427-432. http://ije.oxfordjournals.org/cgi/content/full/30/3/427 Competing interests: Parent; Tutor at Peninsula Medical School
Negative outcome or protocol problem ? 1 November 2006
Alastair Michell, Professor of Comparative Medicine University of London, Barts Hospital [Biochemical Pharmacol] Harvey Institute, London EC1M 6BQ. Send response to journal: Re: Negative outcome or protocol problem ?	While the BMJ text emphasises that adherence to protocol was good, an interview with Prof Reilly on the Today programme [7.20 am, 6.10. 06] cited the difficulty in getting children to increase their activity as much as required ie not a negative outcome, but a problem in achieving sufficient increase to make the intervention effective. Which was the case ? Competing interests: None declared
Childhood obesity prevention - What could be the winning formula? 8 November 2006
Somasundari Gopalakrishnan, Honorary Research Fellow, Department of Public Health and epidemiology University of Birmingham, B15 2TT, Miranda J Pallan, Clinical Research Fellow, Dept of Public health and epidemiology Send response to journal: Re: Childhood obesity prevention - What could be the winning formula?	We read with interest the study titled, 'Physical activity to prevent obesity...' by Reilly J et al(1). Obesity is now considered a global epidemic. UK research suggests that the prevalence of overweight and obesity amongst children of all ages is increasing (2-4). The authors have targeted an obesity prevention intervention at children of 4 years of age in this study. It is widely known that about 50% of obese children in the absence of any intervention will become obese adults (5), and so will be at risk of the many health related consequences of obesity (6). Obesity is also known to have an important impact on the psychological well being with many children developing a negative self image and experiencing low self-esteem (7). It is therefore of utmost importance to try and find an effective intervention that can prevent the development of obesity in children. The behaviours that influence the development of obesity ( diet and physical activity ) begin in early life, and therefore, the authors approach of targeting young children makes good sense. However, the authors comment that the intervention in this study only targets physical activity and not dietary intake, because interventions targeting one or two behaviours have been the most promising. In reality, despite much research in this area, there has been little success in finding effective interventions to prevent childhood obesity. We propose that a more comprehensive approach of both these behaviours, and hence tackling both sides of the energy intake and energy expenditure equation, is a more logical way forward. The study group involved here had a transition period from nursery to primary school. The primary and secondary outcomes were assessed through this period. The primary outcome measure (BMI) was done at baseline, 6 and 12 months and the secondary outcome measures were done at baseline and at 6 months only. It may be that the intervention was effective in influencing behaviour in the short term, but this was lost in the transition from nursery to school. It becomes obvious to point out here that interventions in children must have a sustained effect to prevent the epidemic of obesity. We are currently working on a study aiming to develop and pilot a childhood obesity prevention intervention ( targeting 5 to 7 year olds). We aim to use lay knowledge from a range of people, including parents and teachers, as well as existing literature to develop a multifaceted intervention that influences both diet and physical activity. We believe that a complex intervention is required for a complex problem, and due attention needs to be given to the development of the intervention. References: 1.Reilly JJ, Kelly L, Montgomery C, Williamson A. Fisher A, McColl JH et al. Physical activity to prevent obesity in young children: cluster randomised controlled trial. BMJ, Oct 2006; doi:10.1136/bmj.38979.623773.55 2.Reilly JJ et al. Epidemic of obesity in UK children. Lancet, 1999; 354. 3.Reilly JJ et al. Prevalence of overweight and obesity in British children: cohort study. BMJ, 1999;319:1039. 4.Rudolf MCJ et al. Increasing prevalence of obesity in primary school children: cohort study. BMJ, 2001;322:1094-5.Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF. Byers T. Do obese children become obese adults? A review of literature. Prev Med 1993;22:167-77. 6.National Audit Office. Tackling obesity in England. Report by the controller and the Auditor General London: The stationery office.2001. 7.Strauss r. Childhood obesity and self esteem. Paediatrics, 2000;105:46- 53. Competing interests: Miranda Pallan is a co-investigator in a study looking at childhood obesity prevention in the UK with a focus on South Asian children, funded by the National Prevention and Research Initiative.
Physical activity to prevent obesity: author's reply 15 November 2006
John J Reilly, Professor of Paediatric Energy Metabolism University of Glasgow Division of Developmental Medicine, Yorkhill Hospitals, Glasgow G3 8SJ, James Y Paton, John H McColl, Avril Williamson Send response to journal: Re: Physical activity to prevent obesity: author's reply	Authors reply to Green & Cable; Michell The correspondents (1,2) raise important issues in relation to the interpretation of our ‘Movement and Activity Glasgow Intervention in Children’ (MAGIC) trial (3). Michell (1) asked whether our ‘negative’ outcomes were the result of inadequate implementation of the physically active play intervention (it was not implemented and so failed), or, alternatively, the failure of a well implemented programme (it was implemented but did not produce the intended outcomes). We trained nursery staff in the intervention and monitored their provision of the sessions. We reported good implementation of the intervention in nursery based on recording the number of play sessions offered and the attendance of children at the sessions (3). We emphasise that these indicators were not measures of the ‘quality’ or ‘quantity’ of the physical activity achieved during the sessions. While we were satisfied with the provision of sessions, and with children’s attendance, ie the programme was well implemented when seen as a pragmatic trial, the level and duration of physical activity attained during sessions may have been less than intended, and less than achieved in the pilot study (4). Since attendance at activity sessions was high it seems likely that insufficient activity was achieved within sessions to increase overall levels of physical activity. Alternatively, children may have ‘compensated’ for increased physical activity within nursery with reduced physical activity when out of nursery (5), though we did not observe any such compensation in our pilot study (4). Green and Cable (2) asked for clarification on the proposed mechanism for any effect of increased physical activity/ reduced sedentary behaviour. We tested whether our intervention would significantly increase objectively measured habitual physical activity and reduce time spent sedentary: pre-school children in Glasgow typically spend around 80% of their waking time sedentary, ie with no trunk movement, and less than 30 minutes per day in moderate-vigorous physical activity (6), a lifestyle characterised by low levels of energy expenditure and many opportunities to consume energy (6). We hypothesised that increases in physical activity would increase energy expenditure (7) and we might reduce opportunities for eating by reducing sedentary time (4). Green and Cable correctly point out that we were unable to increase habitual physical activity significantly, ruling out any effect on the more distal outcome of body mass index (BMI). This emphasises the importance of objective outcome measures, particularly important because subjective measurements of physical activity interventions in children are prone to bias ( 8). Green and Cable also question the use of BMI as an outcome measure, and suggest that it may not take account of changes with age and growth in children. Use of absolute BMI could have had the flaw suggested by Green and Cable, but this does not apply to our study since we interpreted BMI relative to reference data. The BMI changes with normal growth and development, and so BMI in children must be interpreted relative to reference data: we did this by calculating SD scores for BMI relative to UK 1990 reference data. A high BMI for age and sex (high BMI SD score) is a highly specific and moderately sensitive index of excessive fatness in children (9), and recent UK studies, including our trial (tables 1 and 2 give mean BMI SD scores; 3) show that as well as increased prevalence of high BMI for age (obesity), the mean/median BMI SD score of children is increasing, reflecting a secular trend to increased fatness across the distribution of body fatness (3,9). A high BMI SD score is also clinically meaningful since it denotes children at high risk of a number of co- morbidities of obesity (9). Other paediatric obesity prevention trials (e.g. 10) have successfully examined differences in the trajectory of increasing BMI SD score between groups as a means of testing the efficacy of the preventive intervention. The BMI SD score is also a practical proxy for body fatness in field studies and is measured with high precision. We accept that measuring proxies for body composition is imperfect. However, field measures of body composition in children have poor validity relative to criterion 3 or 4 component models (11), and are usually less practical and less precise than proxy measures such as BMI SD score. We did include body composition estimates from foot-foot impedance as an outcome in our trial- this method is of unknown validity in young children, but did not suggest any differences between intervention and control groups in our multilevel models. We also included waist circumference SD score as an index of fat distribution, again with no significant difference between groups. In summary, the BMI SD score is an appropriate outcome measure for paediatric field trials because it is practical, precise, valid as a proxy for fatness, and biologically meaningful. John J Reilly, University of Glasgow Division of Developmental Medicine James Y Paton, University of Glasgow Division of Developmental Medicine JH McColl, University of Glasgow Department of Statistics A Williamson, Glasgow City Council Education Department References 1.Michell A. Negative outcome or protocol problem ? (letter in response to Reilly et al). 2. Green DJ, Cable T. BMI in the BMJ. (letter in response to Reilly et al). 3. Reilly JJ, Kelly LA, Montgomery C et al. Physical activity to prevent obesity in young children: cluster randomised controlled trial. Br Med J (online 6th October 2006; print version 17th November 2006). 4.Reilly JJ, McDowell ZC. Physical activity interventions in the prevention and treatment of childhood obesity. Proc Nutr Soc 2002; 62: 611 -619. 5. Wilkin TJ, Mallam KM, Metcalf BS et al. Variation in physical activity lies with the child, not his environment: the activitystat hypothesis. Int J Obes 2006; 30: 1050-1055. 6.Reilly JJ, Jackson DM, Montgomery C et al. Total energy expenditure and physical activity in young Scottish children. Lancet 2004; 363: 211- 212. 7. Montgomery C, Reilly JJ, Jackson DM et al. Relation between physical activity and energy expenditure in a representative sample of young children. Am J Clin Nutr 2004; 80: 591-596. 8.Reilly JJ. Tackling the obesity epidemic: new approaches. Arch Dis Child 2006; 91: 724-726. 9.Reilly JJ. Diagnostic accuracy of the BMI for age in paediatrics. Int J Obes 2006; 30: 595-597. 10. Robinson TN. Reducing children’s television viewing to prevent obesity: a randomised controlled trial. JAMA 1999; 282: 1561-1567. 11. Wells JC, Fuller NJ, Dewit O et al. Four component model of body composition in children: density and hydration of fat free mass and comparison with simpler models.Am J Clin Nutr 1999 69: 904-12. Competing interests: None declared
Unsuccessful intervention is not the same as a negative outcome 15 November 2006
Alastair R Michell, Professor of Comparative Medicine Univ. of London, Bart's Hospital [Biochem. Pharmacol], Harvey Inst., EC1M6BQ Send response to journal: Re: Unsuccessful intervention is not the same as a negative outcome	Professor Reilly's response [BMJ Nov 15th]confirms the thrust of his 'Today' interview, as noted in my earlier response ie the study demonstrated the difficulty of achieving the intended increase in physical activity. But that sits uncomfortably with the emphasis on 'rigorous implementation' and the conclusion that 'physical activity...did not reduce body mass index'. Had this been a drug trial in which the children attended a clinic to receive three tablets a day, but doubts remained whether they actually took the full dose required by the protocol, we could not conclude that the drug was ineffective in reducing obesity. Competing interests: None declared
Change in BMI is not a measure of change in obesity 18 November 2006
Louise Parker, Professor of Pediatrics and Community Health and Epidemiology Dalhousie University, Halifax, Canada B3K 6R8 Send response to journal: Re: Change in BMI is not a measure of change in obesity	Reilly et al conclude that, since their exercise regimen for 4 year olds did not result in a fall in BMI, exercise in youngsters will not reduce obesity. However they used BMI as their measure of obesity (a measure of fatness) and BMI does not measure fatness. BMI is a composite measure of weight for height which includes muscle as well as fat mass. It is entirely possible that the exercising children gained muscle mass and lost fat - ie became leaner and less obese, without any change in BMI. Indeed intense exercise can lead to increases in BMI even in the face of falling fat mass due to accumulation of muscle. To determine whether an intervention is effective against obesity it is important to measure body fat. Using BMI as a surrogate measure of fat mass can easily lead to the wrong conclusions. Competing interests: None declared
Proves that single-interventions are not very fruitful 20 November 2006
Allen P Ugargol, MPH Scholar Achutha Menon Centre for Health Science Studies, SCTIMST, Trivandrum - 695011 Send response to journal: Re: Proves that single-interventions are not very fruitful	I must congratulate the authors of this article on these good findings, a landmark in sorts. Only a well-structured and well organised study can bring out such results. This shows that mere exercise cannot reduce the BMI of children and paves the way for further research that would have to look into two or three modalities that can lead to a successful reduction in the BMI of the children. It is a path-breaking study as exercise was being touted as the sure shot way of reducing BMI and improving fitness. The positive effect of exercise on motor skills of the children is noteworthy and this can be explored further in the aid of children with deficient motor skills and other handicaps. I feel this study will pave the way for a new school of thinking that does not think of exercise as the sole way of reducing BMI, but a holistic approach including diet, food habits, genetic makeup, etc. would need to be looked at in arriving at a regimen to reduce BMI in children. Competing interests: None declared
Dietary and behavioral modifications in managing childhood obesity 21 November 2006
S Kapoor, Res. physician UIC, Chicago, IL 60612 Send response to journal: Re: Dietary and behavioral modifications in managing childhood obesity	The article “Physical activity to prevent obesity in young children: cluster randomized controlled trial” by John Reilly was highly interesting and provided conclusive evidence that other interventions besides enhanced physical activity are necessary for decreasing the BMI (Body Mass Index). The management of childhood obesity needs a multi dimensional approach involving dietary modifications, behavioral modifications and physical exercise. Only in conjunction with the former two will physical activity have a significant impact in reducing obesity. Dietary modifications include avoiding eating in restaurants, decreasing soft drink intake (1), decreasing portion sizes, avoiding dried and calorie rich foods and increasing the fiber content of diet. Behavioral modification strategies include educating children and parents about healthy diets, encouraging children to keep food diaries and avoiding habits such as eating while watching television. Reinforcement of these strategies along with regular physical exercise is likely to produce significant results rather than using one approach exclusively. The management of childhood obesity is especially important to prevent complications such as low self esteem (2), hypertension, hyperlipidemia, sleep apnea (3), slipped femoral epiphyses (4) and diabetes mellitus (5). Besides it needs to be remembered that though rare, there are genetic causes of obesity such as Alstrom syndrome (6) and Prader Willi Syndrome as well as endocrine causes such as hypothyroidism that need to be excluded before the above mentioned approaches are used. 1. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 2001; Feb 17;357(9255):505-8. 2. Davison KK, Birch LL. Weight status, parent reaction, and self- concept in five-year-old girls. Pediatrics 2001; Jan;107(1):46-53. 3. Mallory GB,Jr, Fiser DH, Jackson R. Sleep-associated breathing disorders in morbidly obese children and adolescents. J Pediatr 1989; Dec;115(6):892-7. 4. Kelsey JL, Acheson RM, Keggi KJ. The body build of patients with slipped capital femoral epiphysis. Am J Dis Child 1972; Aug;124(2):276-81. 5. Ludwig DS, Ebbeling CB. Type 2 diabetes mellitus in children: primary care and public health considerations. JAMA 2001; Sep 26;286(12):1427-30. 6. Russell-Eggitt IM, Clayton PT, Coffey R, Kriss A, Taylor DS, Taylor JF. Alstrom syndrome. Report of 22 cases and literature review. Ophthalmology 1998; Jul;105(7):1274-80. Competing interests: None declared
Childhood Obesity:..No golden bullet. 21 November 2006
Emmanuel Agogo, SHO Peadiatrics Good Hope hospital, Sutton Coldfield,Birmingham Send response to journal: Re: Childhood Obesity:..No golden bullet.	Obesity is an emerging global epidemic with clear evidence of increasing prevalence across all age groups in the United Kingdom. The most recent estimates from the National Health survey in 2004, suggest that about one in five children aged 2-15 years is obese. 1 There was a 10% increase in prevalence of obesity since 1995. One in 4 boys (24%) and girls (26%) aged 11-15 years were obese. In younger children, 16% of boys and 11% of girls aged 2-10 years were classified as obese. If figures for overweight are also considered this increases the numbers of children at a weight that poses risk to their health to 1 in 3 boys (32.6%) and girls (34.1%). 1 The definition and management of obesity in children is quite difficult because of the rapid physiological changes and consequent nutritional requirement that would ensure optimum growth and development. The current definition of obesity based on the Body Mass Index (BMI) is influenced by age, height and sex of the child. However for clinical purposes, obesity is defined as a BMI greater than the 98th percentile of the UK 1990 reference charts for age and sex (greater than the 91st percentile is classified as overweight). 2 The effects of obesity are both physiological (increased blood pressure, adverse lipid profiles, changes in left ventricular mass, hyperinsulinaemia) and psychological (negative self image and low self esteem). 3 The combination of healthier eating, increasing habitual physical activity and reducing physical inactivity with active parental involvement are suggested as strategies to achieve weight loss or weight maintenance. 2 Notwithstanding the growing public health significance of obesity, there are no proven evidence based strategies for prevention and management of obesity in children. This informs the need for research and evaluation of current practice. Apart from the Scottish Intercollegiate Guidelines Network (SIGN) guidelines, there are no established national or Royal College guidelines for managing these patients in primary or secondary care in the United Kingdom. The NICE guidelines are due to be issued in early 2007. It is important to identify the goal of management on a case by case basis. It should be clear whether to work towards weight maintenance or aim for weight reduction. This can be achieved by discussing with the patients (where possible) and their parents/guardians and agreeing on target weights and outlining strategies to achieve concordance. It should be obvious that for a multi-factorial and multi-dimensional problem like obesity, effective interventions have to be multi- disciplinary. The importance of involving other health professionals like dieticians, play specialists, physiotherapists and psychologists cannot be overemphasised. Even more important is the co-operation of the family; otherwise the war will always be lost before it is started. Evidence gathered ,so far, shows that there is no 'golden bullet' in the fight against Obesity. Competing Interests: None declared References Health Survey for England 2004 http://www.ic.nhs.uk/pubs/hsechildobesityupdate Obesity in children and young people: an evidence based guideline. Scottish Intercollegiate Guidelines Network (SIGN). SIGN 69. 2003. www.sign.ac.uk The prevention and treatment of childhood obesity. Effective Health Care. NHS Centre for Reviews and Dissemination. Vol.7, Number 6, 2002. Competing interests: None declared
Measuring Child Weight Status 22 November 2006
Rory O'Conor, Consultant in Public Health Wakefield District PCT Send response to journal: Re: Measuring Child Weight Status	My assessment of this trial was that the intervention used did not demonstrably increase the level of physical activity in the intervention group, and hence no demonstrable effect on weight status. My interpretation of the outcome measure: change in BMI SD score is that they used an appropriate measure, although some of the text is confusing. However, this is difficult for many to understand. I would suggest that in looking at child weight status and interventions to impact on that weight status, that a more intelligible measure would be: changes in adult-adjusted body mass index (AABMI) where the AABMI is the BMI that the child would have if you (mathematically) follow the BMI centile line up to the age 18 years. To my understanding the AABMI is mathematically very similar to the BMI SD score used, but much more intelligible. Rory Competing interests: None declared
Childhood obesity and gym membership 22 November 2006
Ian P Rodd, Consultant Paediatrician Royal Hampshire County Hospital, Winchester SO21 5DG Send response to journal: Re: Childhood obesity and gym membership	Sir I commend the authors of this paper on their strategy – namely to try and look at an individual piece of the jigsaw that is obesity in a well constructed and conducted trial. Sadly, as everyone who has bought gym membership as a single intervention to lose weight would have told them, they were always on a losing wicket. If we assume that a pound of fat requires roughly 3 500 kCals to be burnt off unreplaced, and that the most that many adults can burn off is no more than 10 kCals/min during cardiovascular exercise, one pound of fat’s worth of energy takes at least 6 hours of decent intensity cardiovascular exercise to lose if that is all you do different. I know there are a host of assumptions and fudge factors here, but the principle works. I have no idea what the equivalent total would be for a group of overweight 4 year olds in a group at nursery, but I struggle to believe that we would expect 36 hours of this (assuming they attended and co- operated with every session and didn’t up their food intake afterwards) would be even close to enough to see a significant difference. Consequently, I don’t think this tells us anything we didn’t already know ie there are no quick fixes, exercise over a period of a lifetime is a good thing, and obesity is a lifestyle issue, encompassing diet, exercise, and a whole family approach to healthier living. Tough but true Competing interests: None declared
Role of dietary modification 22 November 2006
Dr. Raghesh varot kangath, lecturer Dr SMCSI MCH, karakonam Send response to journal: Re: Role of dietary modification	The regulations in diet have a major role in reducing obestiy and probably BMI.Exercising without dietary modification could prove to be of little benefit. Children get easily addicted to junk food. Introduction of dietary modification at an early age can bring longstanding and markedly evident effects on obesity and general well being. Competing interests: None declared
Childhood obesity : Is exercise no good ? 22 November 2006
Alka Thakur, SpR Paediatrics, Royal London Hospital UB1 1BB Send response to journal: Re: Childhood obesity : Is exercise no good ?	Regarding the aticle by Reilly et al, I would like to make the following comments: The primary objective of the study was to evaluate the impact of a physical activity rogramme on Body Mass Index (BMI). Given that the programme did not result in any difference in physical activity, it would be incorrect to conclude that physical activity does not reduce BMI. Some flaws are evident in the way the study was designed and conducted: It is unclear how the randomisation schedule was generated and how participants were allocated to the intervention or control groups. Socioeconomic status and children undergoing physical activity as part of an existing curriculum, two potential confounding factors were not considered in the baseline characteristics of the subjects. No intention to treat analysis was done on infants lost to follow up (12%at 6 months and 7% at 12 months) which dilutes the validity of the findings. The study deals with a topic which is of significant importance to the community and it is therefore extremely important that it is interpreted correctly. Competing interests: None declared