Obesity and asthma

Obesity and asthma

PAEDIATRIC RESPIRATORY REVIEWS (2006) 7, 223–228 REVIEW Obesity and asthma Susan Chinn* King’s College London, UK KEYWORDS asthma; incidence; obesit...

152KB Sizes 5 Downloads 15 Views



Obesity and asthma Susan Chinn* King’s College London, UK KEYWORDS asthma; incidence; obesity; prevalence

Summary There is a large literature on the possible association between obesity and asthma or asthma-like symptoms. However, no previous review has been confined to studies in children and adolescents. Studies vary in definitions of overweight or obesity and in whether outcomes were parent-reported symptoms or doctor-diagnosed asthma; there is no consistency in the findings. Three studies provide evidence for a greater incidence of diagnosed asthma in those obese at baseline compared to those not obese. The evidence for a stronger association in girls than in boys is weak. No study in a Western society has found objective evidence for the association, and results might be due to increased reporting of symptoms in obese children or to diagnostic bias. Intervention studies to prevent or reduce obesity could provide data to support or refute the association. ß 2006 Elsevier Ltd. All rights reserved.

INTRODUCTION Many epidemiological studies suggest an association between obesity and asthma or asthma-like symptoms in adults and children. There is some evidence that the association in children is of recent origin, whereas that in adults is more long standing. Hence, the underlying mechanisms might differ and a review concentrating on studies in children is warranted. Most of the evidence for an association comes from cross-sectional population surveys and, to a lesser extent, from case-control or cohort studies; these will be summarised. But first, definitions of overweight and obesity in children require some explanation.

DEFINITIONS OF OVERWEIGHT AND OBESITY IN CHILDREN Overweight in adults is defined as a body mass index (BMI; weight in kilograms divided by the square of height in metres) of 25 to 30 kg/m2, and obesity as 30 kg/m2 or * Division of Asthma, Allergy and Lung Biology, 5th floor Capital House, 42 Weston Street, London SE1 3QD, UK. Tel.: +44 20 7848 6607; Fax: +44 20 7848 6605. E-mail address: [email protected] 1526-0542/$ – see front matter ß 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.prrv.2006.04.007

more. Although BMI is not perfectly correlated with direct measures of fatness, it is not feasible to use direct measures in most population studies. The above definitions have been accepted since the early 1980s. There are two problems in children. One is that BMI changes with age and in mean value.1 Hence the cut-off points need to be age specific and the BMI must be transformed into a z-score (i.e. standard deviation score) if it is to be used as a continuous variable when data from children of more than one age are combined. In studies in the USA, the 95th centile of the US Centers for Disease Control (CDC) reference curves for BMI has usually been used as the definition of overweight/obese,2 with children at or above the 85th but below the 95th centile classified as at risk. In 2000, the International Obesity Task Force proposed ‘international’ definitions of overweight and obesity, based on the centiles corresponding to BMI 25 and 30 at age 18 in six countries.1 Although the averaging across countries has been criticised,3 these definitions have the merit of being less arbitrary than fixed centiles and have continuity with the adult definitions. Definitions based on the adult categories also discourage ‘drift’. As the prevalence of obesity has increased over time in most populations, the 85th and 95th centiles have also increased. If the


reference population is kept fixed this is not a problem but this might not be universally understood. Studies have been included in this review regardless of cut-off points used. The second problem is that, in children – even more so than in adults – BMI is only a proxy for fatness4 and the degree of correlation changes with age. Skinfold thicknesses have been used in a few studies but there is no agreed cutoff and most researchers have used BMI.

CROSS-SECTIONAL STUDIES OF ASTHMA OR ASTHMA-LIKE SYMPTOMS In the first extensive review, published in 2001, Tantisira and Weiss did not comment separately on children and adults.5 In 2003, Chinn summarised six studies in which a substantial proportion of participants were aged under 12 years.6–13 Each study provided some evidence for a positive association. Results from the National Health and Examination Survey (NHANES) III, carried out from 1988 to 1994, were reported in two papers,10,13 showing strong evidence for an association between asthma and agestandardised BMI. By contrast, NHANES II (which was conducted from 1976 to 1980) found no statistically significant trend in asthma with BMI, although the trend for wheeze was significant.9 Results from the 1994 UK National Study of Health and Growth (NSHG) showed strong evidence of an association with asthma in white boys and girls.11 This was in contrast to earlier findings from 1977,7 and suggests that the association arose after the early 1980s.14 Another UK study found little evidence of an association in data collected in 1965 and 1969.8 Chinn concluded that the association in children was of recent origin but found few studies in adults to provide corresponding information.6 One unpublished study suggested that the association was present in adults by the 1980s; since then, the Swedish Conscript Study of 17-year-old males has reported an association as early as 1952–1961.15 Two further cross-sectional studies of symptoms or asthma in children or adolescents under 17 years were referenced by Ford16–18 and another seven have also been identified.19–25 One of these was a re-analysis of NHANES III data,22 although with a slightly different age range to the previous reports.10,13 Of the remaining eight studies, three found a strong relation of current asthma with BMI with adjustment for confounding factors.17,19,21 Wickens et al. reported a strong relation of current asthma medication with BMI.25 Three found no significant relation,18,20,24 but Sulit et al. found a weak relation that persisted after additional adjustment for sleep-disordered breathing.23 Most researchers also analysed recent wheeze as an outcome, with three studies reporting a strong association,17,19,25 one a weak association23 and one a weak association only in girls and no relation in boys.18 Each statistically significant association was of increasing asthma or wheeze with greater BMI but authors


differed as to whether BMI was dichotomised, analysed as continuous or divided into quartile or quintile groups with a test for trend performed. Wickens et al. found a stronger relation of symptoms with BMI in 2000 than in 1989,25 although the difference was not statistically significant. They concluded that a diagnostic shift from wheezy bronchitis to asthma could explain a lack of association in early studies.

CASE-CONTROL STUDIES OF ASTHMA One of the three studies of asthmatic children compared to controls reported in earlier reviews showed a strong relation to obesity,26 a second showed very little difference between cases and controls27 and the third showed a relation of borderline significance,28 as did two recently published case-control studies.29,30

HETEROGENEITY OF CROSSSECTIONAL AND CASE-CONTROL STUDIES Ford considered that the cross-sectional and case-control studies in children and adolescents showed less consistent results than those in adults.16 Potential sources of variation in results include study size, period when the study was carried out, age range of the participants and cut-off point to define obesity. Significant odds ratios for associations between obesity and asthma ranged from 1.4 to 3.4. Small studies lack power to detect odds ratios of this magnitude but not all studies that found no association were small in size. The Canadian National Longitudinal Survey of Children and Youth included 5596 boys and 5501 girls aged 4– 11 years,24 finding little difference in current asthma between those in the highest BMI category and those in the lowest in either sex and evidently no trend. To et al. grouped all children above the 85th centile of the sample data ‘to ensure sufficient numbers within each category and to preserve the generally accepted definition of obesity as BMI  85th percentile in population-based research’.24 The corresponding BMI values were not given and the 85th centile is no longer ‘generally accepted’. It is possible that this definition of obesity encompasses children normally classified as overweight or at risk of overweight. This might be the reason for the lack of association but – alternatively – the association might not be present in Canadian children. Results from the Canadian National Population Health Survey found no statistically significant relation of asthma to BMI unadjusted for age in adolescents and young adults aged 12–24 years,31 although in females there was an association in adults aged 25 years and older. No pattern in the association with age range is apparent. The tendency of recent studies to show an association and for those carried out before 1980 to show little relation between obesity and asthma, was noted above. The Six



Cities Study, carried out in the US from 1974 to 1979, was an exception to this, finding a strong relation of asthma and wheeze to BMI in children aged 7–14 years, although BMI was not age adjusted.17

LONGITUDINAL STUDIES OF ASTHMA AND OBESITY There have been far fewer longitudinal studies on obesity and asthma. The NSHG was a series of cross-sectional studies of children aged 4–11 years but, because the same schools were visited at each study occasion, the data also had a considerable longitudinal element, with most children followed up for at least 4 years. For children not recorded to have parent-reported wheeze or asthma at age 5 or 6 years, there was a considerably increased risk of asthma or wheeze 4 years later in those obese at baseline compared to those of normal weight, in both sexes.6,14 Three cohort studies in the US have analysed data on asthma or symptoms in relation to obesity or BMI.32–34 The Six Cities Study34 found an increasing incidence of parent-reported and doctor-diagnosed asthma with wheeze with increasing BMI z-score at baseline or annual increment in BMI z-score in girls aged 6–8 years at baseline; the relationship in boys was U shaped. The Children’s Health Study, carried out in Southern California, found increased incidence of reported physician-diagnosed asthma in those overweight or obese at baseline in 1993.33 In the Tucson study, new infrequent wheezing was more common in girls who became overweight or obese between age 6 and 11 years compared to those who did not.32 Corresponding results were not reported for boys as prevalence of wheeze at age 11 was unrelated to change in BMI. The incidence was not analysed in relation to baseline obesity. Estimates of incidence are dependent on the sensitivity of the criterion to exclude prevalent cases at baseline and it is known that bias in relative risks is also related to the sensitivity.35 An alternative analysis, which does not require

an exclusion criterion, is to analyse net change in prevalence in the whole sample with data on two occasions. Results from NSHG data, used previously to derive relative risks,6 are shown in Table 1. The baseline prevalence of overweight and obesity was slightly greater in boys than in girls. The baseline prevalence of underweight was higher in girls (column 2). Baseline prevalence of asthma attacks in the last year was greater in boys than in girls (column 3), and there was a larger net overall increase in boys than in girls. Only 1.3% of boys and 1.0% of girls were obese at baseline, so confidence intervals for the net increases in the prevalence of asthma attacks in these groups were wide, overlapping zero for boys. Overall, allowing for differences between boys and girls in prevalence and change in prevalence, there was a borderline significant association between net change and obesity group (P = 0.061) and a greater net increase in children obese at baseline compared to those of normal weight (P = 0.007). Even if the Tucson results show no statistically significant relation, overall there is convincing evidence for an overall relation of reported asthma incidence to baseline obesity. The studies rule-out reverse causality as the explanation, i.e. it is unlikely to be due to asthmatic children being more sedentary and hence developing overweight or obesity.

CONCOMITANT INCREASES IN ASTHMA AND OBESITY Parallel increases in the prevalence of asthma and obesity, together with the reported association, have led to speculation that the increase in obesity might be responsible for part of the rise in asthma;16 few studies have looked at this directly. The NSHG showed trends in BMI and prevalence of wheeze and asthma in UK children from the early 1980s to 1994 but the trend in BMI explained almost none of the trends in asthma or wheeze.14 A similar result was found in New Zealand from 1989 to 2000,25 and for young men in

Table 1 Net change over 4 years in prevalence of asthma attack in last 12 months in boys and girls aged 5 or 6 years at baseline, by obesity status at baseline, in the UK National Study of Health and Growth 1982 to 1994 % at baseline

% with asthma attacks at baseline

% with asthma attacks at follow-up

Net change in prevalence (%)

95% confidence interval for net change (%)

Boys (n = 3625) Underweight Normal weight Overweight Obese

22.1 68.1 8.5 1.3

5.8 5.6 5.5 8.1 8.3

8.7 8.4 8.4 10.1 20.8

2.7 2.9 1.9 12.5

1.1 1.9 1.3 1.1

to to to to

4.4 3.9 5.2 26.2

Girls (n = 3457) Underweight Normal weight Overweight Obese

28.6 63.1 7.3 1.0

3.7 2.9 4.1 3.6 2.9

5.1 4.4 5.1 6.3 13.9

1.4 1.0 2.8 11.1

0.2 0.1 0.2 0.8

to to to to

2.7 1.9 5.8 21.4

226 Sweden from 1969 and 1994.15 Inconsistency in the association of asthma and obesity over time might explain the NSHG results. Calculation has shown that the association is not strong enough to explain more than a small percentage of the rise in asthma.36

IS THERE A DIFFERENCE IN THE ASSOCIATION BETWEEN BOYS AND GIRLS? Weiss and Shore speculated that ‘clear-cut sex differences in the epidemiology of obesity and its impact suggest that female sex hormones may be contributing to the increased risk of asthma in obesity’.37 However, Ford pointed out that the evidence for a sex difference was inconsistent and considered the issue far from settled.16 Few studies were confined to prepubertal children. The study of 5- and 6year-old children in Bavaria reported a statistical relation in girls but not in boys.12 However, no test of interaction was given, without which no real conclusion can be drawn.38 In second-grade children in Israel a stronger relation was found in boys than in girls.19 Again, no test of the difference was reported. However, Oddy et al. – in their study of children in western Australia21 – found no significant interaction between BMI and gender on asthma at 6 years. Hence the results are either inconsistent or – overall – suggest no difference in the asthma–obesity relation at young ages. All other cross-sectional studies included children of 10 years or older. Most, but not all, reported results by sex, others reported non-significant tests for interaction. Slightly larger associations were found in 4- to 11-yearold girls than in same-aged boys in the NSHG representative sample,11 although the differences were not statistically significant. However, in the inner-city sample, which included a high proportion of children of African–Caribbean and Indian subcontinent origin, there was evidence for associations in girls but not in boys. NHANES III showed no effect modification by ethnicity or sex in children aged 4–17 years.13 Gold et al. studied sex and race differences in prevalence of respiratory illness and environmental exposures in 7- to 14-year-olds,17 but do not appear to have looked at differences in the crosssectional association of BMI and asthma. Schachter et al. found a significant trend in wheeze and cough with increasing BMI in girls but not in boys aged 7–12,18 but did not test the interaction, and trends in ever-diagnosed asthma and recent asthma were not statistically significant in either boys or girls. In 2000, in New Zealand, associations of wheeze and asthma outcomes with BMI appeared stronger in 11- to 12-year-old girls than in same-aged boys.25 Outcomes analysed included hospital admission, two medication variables, bronchial hyperresponsiveness (BHR) and exercise, and differences in relations with BMI between boys and girls were not significantly different when P-values were adjusted for nine multiple compar-


isons. Mai et al. found a statistically significant relation of wheeze with BMI in Swedish boys aged 11 to 13, but not in girls; however, the interaction was not significant.29 No relevant information was available from the other crosssectional studies. Overall, even in older children, the evidence for a stronger association in girls than in boys is weak. The relative risk for incidence of asthma over 4 years for obese compared to normal-weight children was similar for boys and girls in the NSHG.6 These findings are borne out by the net change results shown in Table 1. Gold et al. divided children into quintile group categories of BMI at baseline and change in BMI.34 Comparison of the highest quintile group with that of lowest incidence appears to show that the incidence rate ratios were similar for boys and girls. In the Children’s Health Study, the relation of incidence to overweight or obesity seemed confined to boys33 but the formal test of interaction of gender and obesity was of borderline significance (P = 0.09). Results from the Tucson study have often been quoted as evidence for a gender difference in children32 but are difficult to interpret. The outcome variable was ‘wheezing’, divided into mutually exclusive categories of none, infrequent and frequent. Outcome data were available at ages 6, 8, 11 and 13 years. BMI at 6 and 11 was used to categorise children as overweight/obese or not. The cross-sectional relation of wheeze category to overweight/obese age 11 years was stronger in girls than in boys but no interaction was reported. Relations at other ages were not statistically significant for boys or girls. Based on the weakly significant greater prevalence of either wheezing category at age 11 in those who became overweight or obese in girls, but not in boys, incident wheeze was analysed only in girls. It seems unlikely, in this relatively small study of 600 children at age 11, that any gender–obesity interaction was statistically significant. Overall, there seems to be no convincing evidence for a gender difference in the asthma–obesity relation in children or adolescents.

ALTERNATIVE MEASURES OF OBESITY As noted above, BMI is a very rough and ready measure of fatness in children but few studies have included other assessments. In NHANES II, Schwartz et al. found wheeze to be associated with increasing BMI and triceps skinfold thickness but current doctor-diagnosed asthma with neither.9 In the 1994 NSHG representative white sample, asthma and wheeze were strongly associated with increasing BMI but not with the sum of triceps and subscapular skinfolds, whereas in the multi-ethnic inner-city sample there were strong associations with each in girls but with neither in boys.11 All other studies used a definition based on BMI alone. Thus the limited information as to whether the observed relations with BMI are found with fatness is conflicting.


ASTHMA OR ASTHMA-LIKE SYMPTOMS AND OTHER OUTCOMES Many studies had questions on wheeze as well as recent asthma attacks or doctor-diagnosed asthma. Results were not always consistent. As already noted, in NHANES II there was an association between wheeze and BMI but not asthma and BMI.9 In the NSHG in 1994, odds ratios for persistent wheeze were greatest in the representative sample and largest for asthma attacks in inner-city girls.11 In children in Bavaria, the relation with BMI category was stronger for lifetime doctor-diagnosed asthma than for wheeze in the last 12 months.12 In NHANES III, there was a clear trend with increasing BMI for asthma outcomes but not for wheeze in the last 12 months, which did not increase monotonically with BMI.13 Gold et al. found similar odds ratios with BMI for asthma in the past year and for wheeze apart from colds or for most days or nights.17 Schachter et al. found stronger evidence for trends in wheeze with increasing BMI than for asthma outcomes.18 Bibi et al. found associations with obesity for asthma and wheeze without a cold.19 Sulit et al. found a stronger relation of asthma than of wheeze to obesity.23 Wickens et al. found strong evidence for trends in wheeze and asthma with BMI category in girls.25 Results are inconsistent for wheeze and asthma in different directions. Wheeze is not specific to asthma. Increased wheeze in obese children could be related to the mechanics of breathing,5 whereas asthma might be diagnosed more frequently in obese children.19 Lack of consistency either way can be argued to indicate that the association is not one of true asthma with obesity. Only an objective measure, such as BHR, can provide further evidence. Schachter et al. found no trend of increasing BHR to histamine with greater BMI, in boys or in girls.18 Wickens et al. found no trend of BHR to exercise, even in girls who showed strong relations of wheeze and asthma to BMI.25 In teenagers in Taiwan there was no relation of BHR to methacholine to BMI in boys or girls after adjustment for other variables.39 Mai et al. found no difference in BHR to hypertonic saline between children above and below the 75th centile of BMI, in those with or without wheeze. However, this case-control study, with selection on current wheeze, gave no overall result for the relation of BHR to BMI.29 A case-control study in South Africa selected on BHR to exercise found a strong relation between BHR and BMI.40 However, only 0.6% of the African children were obese according to UK criteria, and 4.5% overweight. This study, therefore, does not provide evidence of a relation of BHR to BMI in the context of the majority of studies. The relation between BHR to exercise and BHR to histamine is not strong.41 This might also have some bearing on the results of Calvert and Burney40 and on the lack of relation found by Wickens et al. in contrast to other outcomes.25 Kaplan and Montana found a greater exercise-induced fall in


forced expiratory flow between 25% and 75% of forced vital capacity in non-asthmatic obese children compared to non-asthmatic non-obese, with obesity defined using triceps skinfold thickness.42

INTERVENTION STUDIES One paper reported a resolution of co-morbid conditions including asthma in four surgically treated morbidly obese adolescents, aged 17 to 19 years,43 but it was unclear whether all had asthma at the outset and surgical intervention was not carried out in younger patients. No nonsurgical intervention study in obese asthmatic children reporting change in asthma or symptoms has been identified. There have been a number of surgical and non-surgical intervention studies in adults but a study that measured BHR before and after the intervention found no relation between change in weight and change in BHR.44 All other studies reported subjective outcomes and changes in weight were generally associated with dietary changes, so these studies are not direct evidence for a causal relation between asthma and obesity.6

CONCLUSIONS Schachter et al. concluded that increased symptoms in obese children were unlikely to be due to asthma18; others have also expressed scepticism.19 There is no convincing evidence to support a real relation. There are many good reasons to reduce obesity in children and intervention studies that include respiratory outcomes, and BHR could provide better evidence for or against an association between asthma and obesity.

REFERENCES 1. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1240–1243. 2. Flegal KM, Ogden CL, Kuczmarski RL, Johnson CL. Prevalence of overweight in US children: comparison of US growth charts from the Centers for Disease Control and prevention with other reference values for body mass index. Am J Clin Nutr 2001; 73: 1086–1093. 3. Chinn S, Rona RJ. International definitions of overweight and obesity for children: a lasting solution? Ann Hum Biol 2002; 29: 306–313. 4. Dietz WH, Robinson TN. Use of the body mass index (BMI) as a measure of overweight in children and adolescents. J Paediatr 1998; 132: 191–193. 5. Tantisira KG, Weiss ST. Complex interactions in complex traits; obesity and asthma. Thorax 2001; 56(suppl II): ii64–ii74. 6. Chinn S. Obesity and asthma: evidence for and against a causal relation. J Asthma 2003; 40: 1–16. 7. Somerville SM, Rona RJ, Chinn S. Obesity and respiratory symptoms in primary school. Arch Dis Child 1984; 59: 940–944. 8. Kaplan BA, Brush G, Mascie-Taylor CGN. The relationship of childhood asthma and wheezy bronchitis with height, weight and body mass index. Hum Biol 1987; 59: 921–931. 9. Schwartz J, Gold D, Dockery DW et al. Predictors of asthma and persistent wheeze in a national sample of children in the United States:




12. 13.

14. 15.

16. 17.

18. 19.





24. 25.


associations with social class, perinatal events, and race. Am Rev Respir Dis 1990; 142: 555–562. Epstein LH, Wu Y-WB, Paluch RA et al. Asthma and maternal body mass index are related to pediatric body mass index and obesity: results from the third national health and examination survey. Obes Res 2000; 8: 575–581. Figueroa-Mun˜oz JI, Chinn S, Rona RJ. Association between obesity and asthma in 4 to 11 year old children in the UK. Thorax 2001; 56: 133–137. von Kries R, Hermann M, Grunert VP, von Mutius E. Is obesity a risk factor for childhood asthma? Allergy 2001; 56: 318–322. von Mutius E, Schwartz J, Neas LM et al. Relation of body mass index to asthma and atopy in children: the National Health and Nutrition Examination Study III. Thorax 2001; 56: 835–838. Chinn S, Rona RJ. Can the increase in body mass index explain the rising trend in asthma in children? Thorax 2001; 56: 845–850. Bra˚ba¨ck L, Hjern A, Rasmussen F. Body mass index, asthma and allergic conjunctivitis in Swedish conscripts - a national cohort study over three decades. Respir Med 2005; 99: 1010–1014. Ford ES. The epidemiology of obesity and asthma. J Allergy Clin Immunol 2005; 15: 897–909. Gold DR, Rotnitzky A, Damokosh AI et al. Race and gender differences in respiratory illness prevalence and their relationships to environmental exposures in children 7 to 14 years of age. Am Rev Respir Dis 1993; 148: 10–18. Schachter L, Peat JK, Salome CM. Asthma and atopy in overweight children. Thorax 2003; 58: 1031–1035. Bibi H, Shoseyov D, Feigenbaum D et al. The relationship between asthma and obesity in children: is it real or a case of over diagnosis? J Asthma 2004; 41: 403–410. Huang S-L, Lin K-C, Pan W-H. Dietary factors associated with physician-diagnosed asthma and allergic rhinitis in teenagers: analyses of the first nutrition and health survey in Taiwan. Clin Exper Allergy 2001; 31: 259–264. Oddy WH, Sherriff JL, de Klerk NH, Kendall GE. Breastfeeding, body mass index, and asthma and atopy in children. Adv Exp Med Biol 2004; 554: 387–390. Rodriguez MA, Winkleby MA, Ahn D et al. Identification of population subgroups of children and adolescents with high asthma prevalence. Arch Pediatr Adolesc Med 2002; 156: 269–275. Sulit LG, Storfer-Isser A, Rosen CL et al. Associations of obesity, sleepdisordered breathing, and wheezing in children. Am J Respir Crit Care Med 2005; 171: 659–664. To T, Vydykhan TN, Dell S et al. Is obesity associated with asthma in young children? J Pediatr 2004; 144: 162–168. Wickens K, Barry D, Friezema A et al. Obesity and asthma in 11–12 year old New Zealand children in 1989 and 2000. Thorax 2005; 60: 7–12. Gennuso J, Epstein LH, Paluch RA, Cerny F. The relationship between asthma and obesity in urban minority children and adolescents. Arch Pediatr Adolesc Med 1998; 152: 1197–1200.


27. Brenner JS, Kelly CS, Wenger AD et al. Asthma and obesity in adolescents: is there an association? J Asthma 2001; 38: 509–515. 28. Luder E, Melnik TA, DiMaio M. Association of being overweight with greater asthma symptoms in inner city black and Hispanic children. J Pediatr 1998; 132: 699–703. 29. Mai X-M, Nilsson L, Axelson O et al. High body mass index, asthma and allergy in Swedish schoolchildren participating in the international study of asthma and allergies in childhood: phase 2. Acta Paediatr 2003; 92: 1144–1148. 30. Vignolo M, Silvestri M, Parodi A et al. Relationship between body mass index and asthma characteristics in a group of Italian children and adolescents. J Asthma 2005; 42: 185–189. 31. Chen Y, Dales R, Krewski D, Breithaupt K. Increased effects of smoking and obesity on asthma among female Canadians: the national population health survey, 1994–1995. Am J Epidemiol 1999; 150: 255–262. 32. Castro-Rodriguez JA, Holberg CJ, Morgan WJ et al. Increased incidence of asthmalike symptoms in girls who become overweight or obese during the school years. Am J Respir Crit Care Med 2001; 163: 1344–1349. 33. Gilliland FD, Berhane K, Islam T et al. Obesity and the risk of newly diagnosed asthma in school-age children. Am J Epidemiol 2003; 158: 406–415. 34. Gold DR, Damokosh AI, Dockery DW, Berkey CS. Body-mass index as a predictor of incident asthma in a prospective cohort of children. Pediatr Pulmonol 2003; 36: 514–521. 35. Pekkanen J, Sunyer J, Chinn S. Non-differential disease misclassification may bias incidence risk ratios away from the null. J Clin Epidemiol 2006; 59: 281–289. 36. Chinn S. Concurrent trends in asthma and obesity. Thorax 2005; 60: 3–4. 37. Weiss ST, Shore S. Obesity and Asthma. Am J Respir Crit Care Med 2004; 169: 963–968. 38. Matthews JNS, Altman DG. Statistics Notes: interaction 2: compare effect sizes not P-values. BMJ 1996; 313: 808. 39. Huang S-L, Shiao G-M, Chou P. Association between body mass index and allergy in teenage girls in Taiwan. Clin Exper Allergy 1999; 29: 323– 329. 40. Calvert J, Burney P. Effect of body mass on exercise-induced bronchospasm and atopy in African children. J Allergy Clin Immunol 2005; 116: 773–779. 41. Haby MM, Peat JK, Mellis CM et al. An exercise challenge for epidemiological studies of childhood asthma: validity and repeatability. Eur Respir J 1995; 8: 729–736. 42. Kaplan TA, Montana E. Exercise-induced bronchospasm in nonasthmatic obese children. Clin Pediatr 1993; 32: 220–225. 43. Stanford A, Glascock JM, Eid GM et al. Laparoscopic Roux-en-Y gastric bypass in morbidly obese adolescents. J Pediatr Surgery 2003; 38: 430–433. 44. Aaron SD, Fergusson D, Dent R et al. Effect of weight reduction on respiratory function and airway reactivity in obese women. Chest 2004; 125: 2046–2052.