Please note: In 2003, the CTF updated its Grades of Recommendations to include an "I Recommendation" for situations where insufficient evidence exists to allow a recommendation to be made. (Formerly, these situations were captured under a "C Recommendation".) This change is not retroactive, and all "C Recommendations" made prior to 2003 have not been reevaluated in light of the new "I" recommendation grade. For a discussion of these recommendation grades, please link to the 2003 article in the Canadian Medical Association Journal here.
Overview
In 1979,
the Canadian Task Force on the Periodic Health Examination concluded that
there was no satisfactory justification for including childhood obesity
among conditions to be sought in a periodic health examination, based on
the lack of an effective program directed to treatment and prevention of
obesity in children. However, it was considered advisable to record height
and weight on standard growth charts and identify any deviations from normal
growth.
Since then, several studies have investigated the consequences of childhood obesity and several randomized controlled trials have been published on the effectiveness of treatment programs. In light of these developments, it was thought to be timely to reconsider our recommendations based on the new evidence. For more information on obesity on older age groups, consult Chapter 48.
Burden of
Suffering
Estimates of Canadian prevalence
rates for childhood obesity range from 7% to 43%, depending on whether
the basis is self-reports or objective measures, and on what measure of
obesity is used (e.g. body mass indices or skinfold thickness). The body
mass index (BMI) or Quetelet’s index, is weight divided by height squared
(kg/m2).
An inverse relationship has been shown between social class and prevalence of obesity in children aged 3-18 years, with rates ranging from 25% in low income families to 5% in high income families.
Recent adoption and twin studies suggest that family environment alone has no apparent effect on obesity but that genetic factors appear to predominate. The Swedish Adoption/Twin Study of Aging, combining the adoption and twin designs, showed that twins reared together were no more similar than twins reared apart, indicating that shared rearing environment had no effect on BMI. In a recent Canadian study of long-term overfeeding in 12 pairs of identical twins, the variance in weight gain and fatness gain between pairs was about six times greater than the variance within pairs. The authors concluded that genetic factors influence the effect environmental factors have on weight status.
Fewer than 5% of obese children have an underlying disease producing obesity – the majority simply have an imbalance between energy expenditure and intake. However, several disease conditions can cause obesity, including endocrinopathies, central nervous system diseases, and specific congenital syndromes.
Infants and children who are obese are at somewhat increased risk of becoming obese adults.<1,2> That is, relative weight and skinfold thickness tend to "track" over time. However, the correlations are low, and the probability that an individual obese infant will be an obese adult is not high. Furthermore, the longer the age interval, the weaker the relation.
There are few known physical health risks to children who are obese. There is some indication of an association between obesity and hypertension in children, although false high blood pressure readings are not uncommon, a result of using cuffs that are too small for oversized arms.<3>
Obese children have been observed to have elevated levels of low-density lipoprotein cholesterol, and decreased levels of high-density lipoprotein.<4> Obesity in adults however, is not a strong independent contributor to heart disease risk in the absence of other coronary risk factors. Nevertheless, the same circumstances which produce the obesity may also increase other risk factors. It is unclear, then, whether obesity is a cause or a correlate.
In the Bogalusa Heart Study, 3,503 subjects of varying weights aged 5-24 showed marked and highly significant (p<0.001) clustering of three cardiovascular disease risk factors: systolic blood pressure, fasting insulin, and ratio of low- to high-density lipoprotein cholesterol.<5> When lipid risk factors were correlated to subscapular skinfold thickness, the relationship was not clear-cut; for example, it was not statistically significant for white or black males aged 5-9 and 20-24 years.
An earlier report from the Bogalusa Heart Study also indicated that there was an association between obesity and cardiovascular risk factors. However, the children who were studied were only those whose weight category "tracked" over the 6 year period under investigation (18% of the sample) and the children also tended to have other risk factors: high blood pressure, triglyceride levels and cholesterol levels.
Evidence from the Muscatine Ponderosity Family Study is also somewhat ambiguous<6> providing data on consistently heavy children and a "heavy gain group". Some but not all risk factor levels were high (considering total cholesterol, LDL cholesterol, apolipoprotein A-1, apolipoprotein B, and diastolic blood pressure). The above studies do not prove that obesity increases the risk of cardiovascular disease or that prevention or treatment of childhood obesity decreases adult coronary artery morbidity and mortality.
Abraham and colleagues conducted a retrospective cohort study examining the relationship between childhood weight status and adult morbidity.<7> They located 902 adult males for whom childhood weight and health records were available. Examination showed no association between weight status in childhood and adult cholesterol status, adult cardiovascular/renal disease, atherosclerosis or diabetes. The highest risk for hypertensive vascular and cardiovascular renal disease was found in those overweight adults who moved from a below average childhood weight category to an overweight adult group. Smoking status of adults was not assessed. It is possible that adult onset obesity is associated with more cigarette smoking and it is the smoking which is related to cardiovascular disease.
Studies have shown an association between infant or childhood obesity and increased incidence of acute respiratory infections, even when data is adjusted for factors such as social class, smoking, and overcrowding.<8,9>
Javier-Nieto et al<10> reviewed over 13,000 files on persons who were between 5-18 years old between 1933-1945. These were linked to census and death certificate records. The heaviest 20% during pre-puberty and adolescence had an adult mortality odds ratio adjusted for education and smoking status that was significantly greater than that of the lowest weight 20%. Must et al<11> did a follow-up of 508 people who were adolescents between 1922-1935. Overweight in adolescence (BMI top 25%) was associated with an increased risk of mortality from all causes and disease-specific mortality among men but not among women.
Severe obesity has been linked to increased mortality in adults, but there is no association between moderate childhood obesity and increased mortality, unless other risk factors are present.<4>
However, obese children may suffer significant social and psychological difficulties. In the Western world, there is strong cultural prejudice against obesity and a cultural obsession with thinness, which can lead to stigmatization and discrimination in academic and work environments, as well as in other social settings.<12> Obese children may be teased and harassed by their peers, may be socially ostracized, and may be treated as stupid and inferior by teachers and other adults.<13> Children share these negative cultural concepts about obesity, acquiring them very early, possibly as young as age 7.<14> Obese children therefore often have a very negative body image and generally low self-esteem.<4,13> One consequence of this poor self-image may be the development of eating disorders such as anorexia nervosa in adolescence.<13,14>
Maneuver
The assessment of obesity
in children requires the use of reliable and valid measures of obesity.
The use of body weight alone fails to allow for differences in height and
stature. For adults there are acceptable tables of desirable weight for
height based on the lowest mortality rates in an insured population e.g.
the Metropolitan Life weight charts. However, there is no equivalent for
children. There are some tables using average weight for age, height and
sex, and several indices have been developed to attempt to standardize
for height, such as the BMI.
The measurement of triceps or subscapular skinfold thickness to assess subcutaneous fat is a more direct measure of adiposity, and is better able than BMI to distinguish adiposity from muscularity and very large or very small stature. The correlation of multiple skinfold measures with total body fatness is in the range of 0.7-0.8.<3> However, the accuracy of skinfold measures decreases among very obese children. The Eid Index is the child’s weight compared to the 50th percentile weight for the age at which the child’s height is in the 50th percentile expressed as a percentage.
Effectiveness
of Screening and Treatment
The benefits of early detection
depend upon the availability of effective means to reduce body fat, and
the feasibility of maintaining the loss and a resultant improvement in
measures of mortality and/or morbidity.
The more drastic methods of fat reduction, such as surgery, pharmacotherapy, and gastric balloons are not appropriate for children, and very low calorie diets may impair their growth and development.
There is no evidence that traditional means of fat reduction – low calorie diets and/or increased exercise – produce successful results which are maintained on a long-term basis.<15> In one of the few intervention trials with adequate follow-up time, 21 children who were at least 30% overweight and as much as 144% overweight (n=21) lost an average of 33% on very low calorie diets.<16> On re-examination 4-10 years later, ten had maintained a reduced percentage overweight, by 3-61%; the other 11 children had gained back all the weight they had lost plus an additional 5-86% excess weight.
Two types of exercise programs have been evaluated, life-style and aerobic. The evidence about whether diet plus exercise of either type improves results over diet alone (level II-2) is ambiguous.<16-19> In one study Epstein and his colleagues found it did,<19> but in an earlier study the same researchers found it did not.<20> However, the Task Force recommends the regular practice of moderate physical activity for the maintenance of a healthy body weight (Chapter 47).
Epstein and his colleagues have conducted a series of trials investigating the effectiveness of behavioral modification for weight reduction in children. Two of these studies were with children under the age of eight. In one trial in 5-8 year old girls (n=19), those in a family-based treatment program decreased their percentage overweight almost twice as much as control girls in a health education program.<21> In a later study, 17 children aged 1-6 years showed significant change in percentage overweight in a 12-month family-based treatment program.<22> At no point did the group mean go below obesity status (20% overweight).
There have also been several studies on preadolescents, children about eight to twelve years old. At least three of these have been controlled trials, where controls got no treatment. In all cases, the controls gained weight while children receiving treatment lost 6-20% excess weight.<20,23,24> The control group in each case was composed of children put on a waiting list to get into the treatment program – there is no way to know if that had any effect on the control children’s weight status.
Some studies have examined the effect of including parents in treatment programs. There is some evidence that including parents in the program improves the success of both weight loss and maintenance of the loss. However, this is not unequivocal. In one trial, Epstein et al examined the importance of child self-control vs. parent control of the therapeutic process, randomly assigning 37 obese children to either of two groups (level I).<25> There was no significant difference in weight status after 5 years; all groups had returned to approximately baseline levels. In another small study, (n=30) Kirschenbaum et al found no evidence of superior results from including parents in the treatment programs.<24>
In a series of trials using behavioral modification therapy, with the "Traffic Light Diet" (green = low calories, yellow = moderate calories, red = high calories), Epstein et al found including parents does improve maintenance of weight loss.<26-28> They have recently published the results of a ten-year follow-up for this approach.<28> Children aged 6-12 years from intact, predominantly white, middle-class families (75 out of 185 applicant families met the study entrance requirements) were randomized to each of the three treatment groups (diet and exercise information plus behaviour modification for the child alone, for the parent and child, or information without behaviour modification). All groups attended 8 weekly meetings, and 6 additional meetings over the next 6 months.
Follow-up of 61 families was achieved at 10 years; six children had developed psychiatric problems and were excluded (final n=55). While initial amounts of weight lost were similar, after five years and after ten years members of the group in which both parents and children were targeted for behaviour modification were significantly less obese than members of the information only group. Not that obesity was completely prevented – even the parent-child information and behaviour modification group was about 35% overweight at the 10 year follow-up. This is the first evidence that the effects of treatment for obesity in childhood can persist into young adulthood. Moreover, the treatment does not appear to have had adverse effects, especially on ultimate height. Given the intensity of the intervention and the high motivation of these families caution in generalizing these results is indicated; they are nonetheless encouraging.<28>
There is little evidence regarding potential risks of treatment – it is generally assumed that weight control efforts are safe for children and infants. There is, however, some evidence that dietary restrictions at this age level may produce adverse effects. Mallick reviews several early studies which indicate that dietary restriction in infants and young children may result in retardation of growth and development, both physically and mentally, and these adverse effects may be permanent.<29> Other recent reviews also suggest that adolescents may not be safely subjected to calorie restriction at any point during their growth spurt.<30,31> Epstein<32> has argued that this previous research may have failed to consider the height of the children upon entry into weight loss programs.
Recommendations
of Others
The American Academy of Pediatrics
recommends that children’s height and weight measurements be taken throughout
infancy, annually from ages 1 to
6, and bi-annually thereafter. The U.S. Preventive Services Task Force
recommends height and weight measurements be taken regularly and plotted
on a growth chart throughout infancy and childhood;<33> this recommendation
is currently being reviewed. Although the U.S. Task Force makes recommendations
for nutrition and exercise counselling for adults, there are no specific
recommendations for children.
Conclusions
and Recommendations
It is understood that plotting
the height and weight of infants and children during a periodic health
examination will continue to be done, primarily to identify children who
are failing to thrive. There is insufficient evidence that screening children
for obesity is of value; nor is there evidence that screening for obesity
is harmful (C Recommendation). (Note: For highly motivated obese children
and their families, there is good evidence that intensive diet, exercise,
and family behaviour management counselling is successful in lowering the
degree of obesity 10 years later, but these obese children were not identified
by screening.
There is insufficient evidence to include or exclude counselling about nutrition and exercise from the routine treatment of severely obese children (C Recommendation). However, the regular practice of moderate physical activity, taking into account current fitness levels, is recommended for all Canadians to maintain a healthy body weight (Chapter 47). There is fair evidence for excluding very low calorie diets from the routine treatment of pre-adolescent obese children (D Recommendation). Evidence is conflicting regarding the inclusion or exclusion of exercise from the routine treatment of obese children. (C Recommendation)
Unanswered Questions (Research Agenda)
Full Citation
Feldman W. and Beagen B.L. Screening for childhood
obesity. In: Canadian Task Force on the Periodic Health Examination. Canadian
Guide to Clinical Preventive Health Care. Ottawa: Health Canada,
1994; 334-44.