Eur J Nutr

DOI 10.1007/s00394-017-1446-7

ORIGINAL CONTRIBUTION

An obesity‑preventive lifestyle score is negatively associated

with pediatric asthma
Constantina Papoutsakis1,2   · Eleni Papadakou2 · Maria Chondronikola3,4 ·
Georgios Antonogeorgos4 · Vasiliki Matziou2 · Maria Drakouli2 ·
Evanthia Konstantaki2 · Kostas N. Priftis5,6 

Received: 7 April 2016 / Accepted: 4 April 2017

© Springer-Verlag Berlin Heidelberg 2017

Abstract  target lifestyle behaviors that were either favorable or

Purpose  Lifestyle (diet and physical activity) may increase
asthma risk, but evidence in this area is lacking. The aims
of the present study were to calculate an obesity-preventive
lifestyle score comprising of eating and physical activity
behaviors and investigate the overall effect of lifestyle on
asthma in children.
Methods A cross-sectional case–control study was car-
ried out in 514 children (217 asthma cases and 297
healthy controls). Data were collected on medical his-
tory, anthropometry, dietary intake, and physical activity.
We constructed an overweight/obesity-preventive score
(OPLS) using study-specific quartile rankings for nine
Electronic supplementary material  The online version of this
article (doi:10.1007/s00394-017-1446-7) contains supplementary
material, which is available to authorized users.
* Constantina Papoutsakis
cpapoutsakis@eatright.org
1
Academy of Nutrition and Dietetics, 120 South Riverside
Plaza, Suite 2000, Chicago, IL 60606‑6995, USA
2
Department of Nursing, National and Kapodistrian
University of Athens, Athens, Greece
3
Department of Internal Medicine‑Nutritional Science,
Washington University School of Medicine, St. Louis, MO,
USA
4
Department of Nutrition and Dietetics, Harokopio University,
Athens, Greece
5
Pediatric Allergy and Pulmonology Units, Third Department
of Pediatrics, School of Medicine, University General
Hospital ‘Attikon’, National and Kapodistrian University
of Athens, Athens, Greece
6
Department of Allergy‑Pneumonology, Penteli Children’s
Hospital, Penteli, Greece
unfavorable in preventing obesity (i.e., screen time was
an unfavorable lifestyle behavior). The score was devel-
oped using the recommendations of the Expert Committee
of American Academy of Pediatrics. Score values ranged
from 0–18 points; the higher the score, the more protective
against high body weight.
Results  The OPLS was negatively associated with obesity
indices (BMI, waist circumference, and hip circumference),
(p < 0.05). Control children had a higher score when com-
pared to asthma cases (9.3 ± 2.7 vs. 8.6 ± 2.9, p = 0.007).
A high OPLS was protective against physician-diagnosed
asthma (OR 0.92; 95% CI 0.86–0.98, p = 0.014), adjusted
for several confounders. The OPLS was no longer protec-
tive after adjustment for BMI.
Conclusion Higher adherence to an obesity-preventive
lifestyle score—consistent with several behaviors for the
prevention of childhood overweight/obesity—is nega-
tively associated with obesity indices and lowers the odds
for asthma in children. Lifestyle behaviors that contribute
to a higher body weight may contribute to the obesity–
asthma link. These findings are hypothesis-generating and
warrant further investigation in prospective intervention
studies.
Keywords  Asthma · Obesity · Lifestyle · Children
Abbreviations
BMI Body mass index
FEV1 Forced expiratory volume in 1 s
FVC Forced expiratory vital capacity
METs Metabolic equivalents
OPLS Obesity-preventive lifestyle score
PEF Peak expiratory flow

13

Introduction
Pediatric asthma and overweight/obesity has increased
in recent years. These two major public health problems
may be causally linked, but associated mechanisms are
unclear. Epidemiologic evidence has established a sig-
nificant yet weak association between asthma and obesity
[1]. Prospective studies demonstrate obesity/overweight
likely precedes asthma, although other studies support
the opposite or that overweight/obesity and asthma may
develop in tandem during growth periods and maturation
[1]. Irrespective of which of the two conditions appears
first, two meta-analyses (a meta-analysis of seven stud-
ies of more than 300,000 adults and children [2], and a
pediatric meta-analysis of six prospective studies [3])
have both supported that BMI is associated with inci-
dent asthma. Overweight and obese individuals are at
increased risk of asthma in comparison to normal weight
counterparts. A westernized lifestyle (decreased energy
expenditure and increased energy consumption coupled
with intakes of nutrient poor food choices) may be in part
responsible, especially in inner city populations [4] and
developing countries undergoing rapid nutrition transi-
tion [5]. Investigations have shown associations between
diet and asthma [6] or physical activity and asthma sepa-
rately [7]. There is abundant evidence, showing the asso-
ciation between dietary quantity and quality and physi-
cal activity and overweight/obesity [8]. The combined
impact of an obesity-preventive lifestyle (diet and exer-
cise together) on asthma has not been studied. Here, we
explore the complex relationship between asthma, obe-
sity, diet, and exercise in children using a predefined
score [9]. Specifically, the aims of the present study are
to calculate an obesity-preventive lifestyle score compris-
ing of eating and physical activity behaviors to assess
adherence to pediatric obesity prevention guidelines [8]
and evaluate the association of this obesity-preventive
lifestyle score on asthma in children.
Methods
Study subjects
The study sample consisted of 514 children (217 cases
of physician-diagnosed asthma, and 297 gender-matched
healthy controls), aged 5–11 years. Recruitment took
place between November 2007 and September 2010 at the
Department of Allergy-Pneumonology, Penteli Children’s
Hospital, Penteli, Greece and two municipal multi-clinic
centers in Galatsi and Pefki, in the greater area of Athens,
13
Eur J Nutr
Greece. Permission to conduct the study was obtained
from the Institutional Review Board of Harokopio Uni-
versity and has, therefore, been performed in accordance
with the ethical standards laid down in the 1964 Decla-
ration of Helsinki and its later amendments. Parents and
caregivers were informed about the purpose and meth-
ods of the study and signed an informed consent form.
In addition, participating children provided verbal assent.
Each asthma patient received budesonide via a dry pow-
der inhaler 200 mcg twice daily or fluticasone propionate
via inhalation aerosol at dosages of 100 mcg twice daily
for at least 1 year and an average of 4.02 ± 2.63 years.
Exclusion criteria were any disease known to affect
growth (cystic fibrosis, Klinefelter syndrome, Turner syn-
drome, endocrine systemic, and skeletal disease). In addi-
tion, any child with clinically evident respiratory tract
infection in the previous 2 weeks or presence of wheezing
or current respiratory infection on physical examination
was excluded. Children with FEV1 <80% predicted were
also excluded. Spirometry was performed using a portable
spirometer (MicroLab, Micro Medical Ltd., Rochester,
UK). Three acceptable measurements were conducted and
the highest FEV1, FVC, and PEF were recorded as abso-
lute and % predicted values.
Asthma diagnosis was established by a pediatric pul-
monologist–allergist (KP) according to the Global Initia-
tive for Asthma Report, Global Strategy for Asthma Man-
agement and Prevention (www.ginasthma.org). Trained
interviewers recorded information on questionnaires and
obtained anthropometric measurements, as described
below. The recruitment rate was 86.8%. According to the a
priori power analysis that was conducted, the current study
has power greater than 80% to detect associations between
two variables with an odds ratio of 2.2 at a significance
level α = 0.05 between case and control groups.
Anthropometrics
Body weight was measured to the nearest 0.1 kg with
a scale (Seca Corp., Hanover, MD) with subjects in light
clothing and barefoot. In addition, standing height was
measured with a stadiometer (Holtain Ltd., Crymych,
Dyfed, UK) to the nearest 0.1 cm. Waist circumference (to
the nearest 0.1 cm) was measured in a standing position
using a soft measuring tape, midway between the lower rib
margin and the iliac crest, at the end of a gentle expiration.
During a single interview, measurements were repeated
three times and averaged for analyses. Physical measure-
ments of body weight and height were used to calculate
BMI [BMI = weight (kg) divided by height (m squared)].
BMI was used for subjects’ classification as normal weight,
overweight, or obese by sex and age [10]. Maturity status
Eur J Nutr
by Tanner stage was assessed with the previously validated
method of self-evaluation using standardized drawings in
the presence of a pediatrician [11].
Family and health history, and demographic
information
The approved Greek translation of the ISAAC question-
naire [12, 13] was used to obtain data on family history
of atopic disease (asthma and allergic rhinitis), housing
details, tobacco exposure, and early and perinatal-life con-
ditions. A supplementary questionnaire was used to record
demographic information regarding age, socioeconomic
status of the family (i.e., parental education, number of
cars in the family, and number of family members), and
information on the child’s past medical history, current and
past drug regimens, supplement use, history and duration
of breastfeeding, early feeding practices, birth weight, and
self-reported parental and siblings’ height and weight.
Dietary assessment
A validated 63-item semi-quantitative Food Frequency
Questionnaire (FFQ) was used to evaluate participants’
usual dietary intake for the last 12 months [14]. The par-
ent/caregiver and child were instructed to report on the
frequency of consumption (weekly or daily) of food items.
The FFQ examined the frequency of consumption of fruits,
vegetables, dairy and cheese, grains, nuts and seeds, snacks
(salty and sweet), traditional Greek foods, soda drinks, and
other beverages. Standardized portions were indicated by
typical serving sizes and household measures (e.g., ten cher-
ries, one slice of bread, two cups of beans, etc.). In addition,
frequency of eating breakfast, number of meals and snacks
per day, eating out, fast food consumption, cooking meth-
ods, snacking habits during school time, type of milk con-
sumed, type of fat/oil used for cooking, and salad dressing
use were recorded. The recorded dietary information from
the FFQ was further decoded to daily food group consump-
tion (in portions) to estimate daily energy intake (kcal/d).
Physical activity
Information on the frequency and duration of leisure time
and physical activity was evaluated using a previously vali-
dated questionnaire [15]. Questions inquired about the fre-
quency of sedentary and non-sports related activities during
free time, usual ways of transport, and frequency of par-
ticipation in structured physical activity in school or in free
time, as well as intensity and type of exercise. The daily
energy expenditure of participants was assessed by esti-
mating a physical activity score in metabolic equivalents
(METs) for the reported duration and type of each physical
activity.
Obesity‑preventive lifestyle score
For the purpose of this study, an overweight/obesity-pre-
ventive lifestyle score (OPLS) was developed. This is a
modified version of the obesity-preventive score described
elsewhere [9]. The score was constructed, so that a high
achieved score represents high adherence to an obesity-pre-
ventive lifestyle, a medium score represents medium adher-
ence, and a low score represents low adherence to an obe-
sity-preventive lifestyle. The score comprised of nine target
lifestyle behaviors. Each target behavior and the resulting
score (sum of nine target behaviors) was categorized into
tertiles to reflect the three categories: high, medium, and
low adherence. The nine target lifestyle behaviors that were
used to construct the OPLS are based on the recommen-
dations of the Expert Committee of American Academy
of Pediatrics for the prevention of pediatric obesity [8]:
consumption of fruit (servings/day), consumption of veg-
etables (servings/day), consumption of sugar sweetened
beverages (servings/day), consumption of breakfast (times/
week), eating food from restaurants/fast food (times/week),
screen time (h/week), moderate/intense physical activ-
ity (kcal kg−1 day−1), school sport physical activity (min/
week), and active commuting (kcal kg−1 day−1). To reflect
a child’s routine physical activity, the obesity-preventive
score included the types of physical activity that were
most likely to be scheduled or be part of a daily routine
(active commuting, school sport activities, and participa-
tion in moderate/intense activities). Each target behavior
had either a negative impact or positive impact on the score
calculation. Target behaviors described in the recommen-
dations [8] as preventing obesity or favorable behaviors
(consumption of fruits, consumption of vegetables, break-
fast, moderate/intense activities, school sport activities, and
active commuting) contributed positive points to the score
and vice versa (consumption of sugar sweetened beverages,
eating at restaurants/fast food, and screen time). Each target
behavior was categorized into tertiles and assigned ascend-
ing values (0, 1, 2) for favorable behaviors and descend-
ing values (2, 1, 0) for unfavorable behaviors. The itemized
scoring scheme appears in Supplementary Table 1. For each
subject, all points from all behaviors were summed and the
overweight/obesity-preventive score was calculated. Score
values ranged from 0–18 points with higher scores repre-
senting healthier lifestyles.
Statistical analysis
Continuous variables are presented as mean ± stand-
ard deviation, and categorical variables are presented as
13
 Eur J Nutr

Table 1  Summary Asthma cases (n = 217) Control subjects (n = 297) p value

characteristics of asthma case
and control subjects Male gender 133 (61%) 164 (55%) 0.169
Age, y 7.7 ± 1.9 7.6 ± 1.8 0.490
BMI, kg/m2 18.56 ± 3.52 17.79 ± 2.90 0.007
Height, m 1.27 ± 0.15 1.27 ± 0.13 0.962
Waist circumference, cm 62.68 ± 9.13 60.79 ± 7.29 0.010
Overweight/obese 100 (46.1%) 108 (36.3%) 0.026
Lung function,  % predicted
 FEV1 98.4 ± 6.3 97.7 ± 5.1 0.634
 FVC 94.4 ± 5.7 93.8 ± 5.5 0.875
 PEF 92.0 ± 5.8 88.9 ± 5.7 0.289
Mother’s education, y 13.6 ± 3 13.9 ± 2.6 0.220
Father’s education, y 13.4 ± 3.3 13.8 ± 3.2 0.171
House, ­m2 110.6 ± 41.8 117.3 ± 46.7 0.09
Number of children 1.9 ± 0.7 2.0 ± 0.7 0.251
Smoking inside the house, yes/no 68 (31%) 107 (36%) 0.269
Breastfeeding, yes/no 165 (76%) 261 (88%) 0.001
Tanner pubic hair, ≥1/0 145 (67%) 166 (56%) 0.012
Energy intake, kcal/d 2091 ± 614 2142 ± 601 0.353
Physical activity, METs 10.9 ± 4.2 11.8 ± 4.0 0.004
Mother’s atopy, yes/no 89 (41%) 77 (26%) 0.000
Father’s atopy, yes/no 65 (30%) 62 (21%) 0.013

Data are presented as mean ± SD or no (%) unless otherwise indicated. Missing values in some variables
BMI body mass index, METs metabolic equivalents, SD standard deviation, FEV1 forced expiratory vol-
ume in 1 s, FVC forced expiratory vital capacity, PEF peak expiratory flow

absolute and relative frequencies. Histograms and normal-
ity plots were used to evaluate the assumption of normal-
ity of the distribution of the continuous variables. Inde-
pendent sample Student’s t tests assessed the differences
in continuous variables between asthma cases and healthy
controls. Chi-square tests were used to assess the differ-
ences between categorical variables. Pearson’s correla-
tion coefficients were determined to identify significant
correlations between continuous variables of interest. A
one-way ANOVA was used to test for differences in the
OPLS among the three BMI categories (normal weight,
overweight, obese). Logistic regression models were con-
ducted to estimate the association between asthma and the
obesity-preventive score adjusting for potential confound-
ers (sex, age, maternal and paternal atopy, energy intake,
Tanner stage, breastfeeding, and BMI). These variables
were selected, because they have been reported in the lit-
erature as important confounders or because they were
significantly different between cases and controls. Data
are presented as odds ratios (ORs) and 95% confidence
intervals (CI). The Hosmer–Lemeshow statistic was calcu-
lated to evaluate the model’s goodness-of-fit. For all logis-
tic regression models, multicollinearity was checked using
Tolerance and the Variance Inflation Factor (VIF), with
cut-off values of 0.1 or less for tolerance, and 10 or more
for VIF, for each independent variable of the model. Linear
regression models were used to study the effect of obesity
indices (BMI, waist circumference, and hip circumference)
on the score adjusting for potential confounders (sex, age,
and puberty status). All statistical tests were two-sided, and
p < 0.05 was required to reject the null hypothesis. Statisti-
cal analysis was performed using SPSS 21.0 for Windows
(SPSS, Chicago, IL).
Results
Approximately 28.8% (n  = 148) of children were over-
weight and 11.3% were obese. Children with asthma
had a higher BMI and waist circumference than healthy
subjects (18.56 ± 3.52 vs. 17.79 ± 2.9, p  = 0.007 and
62.7  ± 9.1 cm vs 60.8 ± 7.3 cm, p  = 0.01, respectively)
(Table 1). The prevalence of overweight/obesity was almost
10% higher in the asthma group when compared to controls
(46.1% vs. 36.3%, p = 0.026). Energy intake did not differ
between the two groups (2091 ± 614 vs. 2142 ± 601 kcal,
p  = 0.353); however, asthma cases had a lower physical
activity score compared to healthy children (10.9 ± 4.2 vs.
11.8 ± 4.0, p = 0.004). Breastfeeding during infancy (yes
vs no) was lower in asthma cases than in healthy controls

13
Eur J Nutr
(76% vs. 88%, p = 0.001). History of maternal and pater-
nal atopy was reported in higher frequencies in the asthma
group than the control group (41% vs. 26%, p = 0.000 and
30% vs. 21%, p  = 0.013, respectively). All measures of
lung function (FEV1, FVC, and PEF) were similar between
asthma cases and healthy controls (Table 1).
The average measure of each of the nine diet and physi-
cal activity behaviors used to construct the OPLS and the
resulting score were calculated for the entire sample and
by study group. When examined individually, target behav-
iors showed no difference between the two groups except
for moderate/intense activities and school sport activities
both of which were lower in children with asthma com-
pared to controls (10.8 ± 4.27 vs. 11.88 ± 4 kcal kg−1 d−1,
p  = 0.004 and 56.82 ± 30 vs. 62.58 ± 27.51 min/week,
p = 0.026).
Asthma cases also displayed a significantly lower
OPLS when compared to controls (8.6 ± 2.9 vs. 9.3 ± 2.7,
p  = 0.008). In addition, logistic regression models were
conducted to estimate the association between asthma and
each target behavior of the OPLS, and the resulting OPLS
adjusting for potential confounders (sex, age, maternal and
paternal atopy, energy intake, Tanner stage, breastfeeding,
and BMI). This was done to describe the individual effects
vs the combined effects of the lifestyle factors investi-
gated (Table 2). With the exception of two physical activity
parameters, each lifestyle factor alone was not significant
in reducing the odds for asthma, but the combined lifestyle
factors (OPLS) were significant (Table 2).
A one-way ANOVA was used to test for differences in
the OPLS among the three BMI categories (normal weight,
overweight, obese). Scores differed across the three cat-
egories (F  = 4.542, p  = 0.011). Additional Bonferroni
post hoc comparisons showed that the obese children had a
lower score [8.0 ± 2.8, 95% CI (7.23–8.77)] than their nor-
mal weight [9.06 ± 2.8, 95% CI (8.74–9.39)], p = 0.032,
Table 2  Logistic regression for Target behavior
the evaluation of the association
between each target behavior of Fruits (servings/day)
the obesity-preventive lifestyle
Vegetables (servings/day)
score and physician-diagnosed
asthma Sugar sweetened beverages (servings/day)
Breakfast (times/week)
Restaurants/fast food (times/week)
Screen time (h/week)
Moderate/intense activities (kcal kg−1 day−1)
School sport activities (min/week)
Active commuting (kcal kg−1 day−1)
Model 1 adjusted for gender, age
Model 2 adjusted for gender, age, maternal atopy, paternal atopy, breastfeeding, and Tanner stage
OR odds ratio, CI confidence interval
ª p < 0.05
and overweight counterparts [9.33 ± 2.79, 95% CI (8.89–
9.79)], p  = 0.009. The difference between the normal
weight and the overweight group was not significant at
p < 0.05.
The OPLS was negatively associated with obesity
indices: BMI (b  =  −0.09, p  = 0.041), waist circumfer-
ence (b  =  −0.019, p  = 0.001), and hip circumference
(b = −0.044, p = 0.009) after adjustment for gender, age,
and Tanner stage.
The OPLS was negatively associated with physician-
diagnosed asthma (OR 0.927, 95% CI 0.864–0.995)
(Table 3). For every one point increase of the OPLS, sub-
jects were almost 7% less likely to have asthma even after
adjustment for several confounders (gender, age, maternal
and paternal atopy, history of breastfeeding, and puberty
status). Further adjustment for BMI rendered the associa-
tion nonsignificant.
Discussion
To the best of our knowledge, this is the first study to
investigate the relation between pediatric asthma and life-
style (dietary and physical activity behaviors combined).
We found that a 1-unit increase in the Obesity-Preventive
Lifestyle Score (OPLS), (meaning a lifestyle closer to rec-
ommended guidelines for the prevention of pediatric over-
weight/obesity) [8] was associated with a 7% decrease in
the odds of physician-diagnosed asthma after adjustment
for several confounders.
Various hypotheses have been proposed to explain the
link between overweight/obesity and asthma including a
common genetic background, comorbidities, mechani-
cal changes associated with high body weight, changes in
airway hyperresponsiveness, abnormal lipid and glucose
metabolism, low-grade systemic inflammation, shared
Model 1 OR (95% CI) Model 2 OR (95% CI)
0.996 (0.849–1.170) 0.955 (0.843–1.175)
0.894 (0.744–1.075) 0.955 (0.744–1.090)
0.920 (0.725–1.169) 1.005 (0.784–1.289)
1.083 (0.983–1.194) 1.083 (0.978–1.199)
1.024 (0.889–1.180) 0.970 (0.834–1.127)
1.006 (0.955–1.060) 1.002 (0.949–1.057)
0.920 (0.877–0.965)ª 0.922 (0.878–0.969)ª
0.992 (0.986–0.998)ª 0.992 (0.986–0.999)ª
0.991 (0.978–1.003) 0.995 (0.982–1.008)
13

Table 3  Multivariate logistic
regression for the evaluation
of the association between the Gender
obesity-preventive lifestyle
Age
score and physician-diagnosed
asthma Obesity-preventive lifestyle score
Maternal atopy, yes/no
Paternal atopy, yes/no
Breastfeeding, yes/no
Tanner stage pubic hair, ≥1/0
OR odds ratio, CI confidence interval
a
  Model 1 adjusted for gender, age
b
  Model 2 adjusted for gender, age, maternal atopy, paternal atopy, breastfeeding, and Tanner stage
c
  p < 0.05
psychosocial factors, and undesirable lifestyle changes in
diet or physical activity [16–19].
There is an increasing interest in the role of lifestyle
in the development of asthma, since lifestyle is subject to
manipulation. The majority of studies have evaluated the
role of specific nutrients, food groups, or physical activity
on asthma measures in children separately [20, 21]. Some
studies have shown a positive association between a west-
ernized diet (low consumption of fruits and vegetables,
high fast food, and soft drink consumption) or a sedentary
behavior (long duration of screen time and reduced physi-
cal activity) and asthma or asthma symptoms, but others
have not [7, 22]. The variability of results on the asso-
ciation between diet and asthma or physical activity and
asthma may be influenced by the definitions and categori-
zations of these entities. Epidemiologic studies differ in the
measures of dietary/physical activity exposures and in the
use of asthma definitions.
The relationships between diet, exercise, sedentarism,
and asthma are complex and studying these behaviors in
isolation may fail to reveal a possible synergy. A holistic
approach that assesses various lifestyle aspects together
has not been investigated. In this study, a general lack of
association was noted between individual target behaviors
(Table  2) and asthma. The only individual measures that
were different between cases and control children were two
physical activity-related measures (moderate/intense activi-
ties and school sport activities). We also found that school
sports activities were significantly higher by about 6 min/
week in normal weight than overweight/obese children
(data not shown). These findings may be attributed to the
fact that it might be easier to accurately report organized/
scheduled physical activity than other aspects of lifestyle
[23]. An initial observation might be that it is not surprising
children with asthma are less active; yet, it should be noted
that in this study, asthma in cases is well controlled given
their similar respiratory measures (FEV1, FCV, and PEF)
to control children. Although these two physical activity
13
Eur J Nutr
Model ­1a OR (95% CI) Model ­2b OR (95% CI)
1.279 (0.87–1.864) 1.3 (0.884–1.914)
1.012 (0.914–1.120) 0.947 (0.842–1.065)
0.910 (0.851–0.973)c 0.927 (0.865–0.994)c
1.613 (1.077–2.416)c
1.625 (1.057–2.499)c
0.440 (0.269–0.722)c
1.650 (1.058–2.573)c
measures are different, effect sizes were relatively small.
The school activity difference is about 10% or 6 min per
week which does not appear clinically significant. These
results support that not any one individual lifestyle measure
may be influential on its own and that there are complex
interactions between them. When we assessed lifestyle as a
composite score which combined diet and physical activity,
we showed that adherence to an obesity-preventive lifestyle
may have a protective effect in reducing the odds of asthma
in children (Table 3). The difference in the OPLS between
the two groups is about 0.7 (8.6 ± 2.9 asthma cases, vs.
9.3 ± 2.7 healthy controls, p = 0.008).
It is intriguing that such a small difference contributes
to a detectable influence on the odds of asthma in children.
The small difference is consistent with evidence support-
ing a weak yet significant association between obesity and
asthma [1].
A ‘healthy diet’ has been defined in several different
ways in the form of dietary recommendations across coun-
tries. A frequently used approach to quantify adherence to
a healthy diet is to construct a composite score that catego-
rizes individuals according to the intakes of foods or nutri-
ents that have been associated with diseases or risk factors
[24–26]. The widely recognized concept of a dietary score
was extended to that of a lifestyle score. The OPLS was
negatively associated with indexes of overweight/obesity
(BMI, waist circumference, and hip circumference) after
adjustment for gender, age, and puberty status.
Multiple regression analysis showed that the OPLS
decreases the odds of asthma. However, this finding was
no longer significant after adjustment for BMI. This sup-
ports that associations between asthma development and
lifestyle behaviors are mediated by BMI, and that possibly
a normal weight BMI in and of itself is protective against
asthma. Thus, the conclusion that overweight/obese chil-
dren are more likely to have asthma than those of normal
weight remains. A question that is still open is the exact
mechanism of the association between obesity and asthma
Eur J Nutr
as well as the management of this association. Our results
are hypothesis-generating and lend support that a high
body weight due to lifestyle behaviors influences the odds
of asthma in children. Thus, adopting an obesity-preven-
tive lifestyle as a means to a healthier body weight status
may have a role as a complementary non-pharmacologi-
cal intervention [1]. The OPLS encompasses key behav-
iors to focus on during weight-management/weight loss
counseling. These behaviors include increasing fruit and
vegetable intake, limiting sugar sweetened beverage con-
sumption, limiting screen time, eating breakfast, limiting
eating out, and increasing physical activity. In our study,
obese children had a lower (worse) score [8.0 ± 2.8, 95%
CI (7.23–8.77)] than their normal weight [9.06 ± 2.8, 95%
CI (8.74–9.39)], p  = 0.032 and overweight counterparts
[9.33 ± 2.79, 95% CI (8.89–9.79)], p = 0.009. In addition,
we found that overweight children had a higher (better)
score than the normal weight children, yet the difference
between the score of overweight and normal weight chil-
dren was not significant. It is important to highlight that the
cohort we studied was 28.8% overweight and 11.3% obese.
The increased prevalence of overweight/obesity in the pre-
sent study is in line with levels reported in other young
Greek populations [27, 28]. The alarmingly high preva-
lence of overweight/obese children underscores the need
for effective interventions.
BMI is a widely used measure of weight status in epide-
miologic and clinical studies; yet, in children, a higher BMI
may not always represent higher fat mass but increased lean
mass instead. This is important, because a higher lean mass
is associated with better respiratory function [1]. Still, in
children, when BMI was compared to other body composi-
tion measures like percent body fat measurement, and sum
of skinfold thicknesses, it provided similar results about the
relationship between overweight/obesity and asthma [29,
30].
Some caveats are warranted when interpreting the pre-
sented data. Our observations are based on a modest size
sample. A basic assumption is that the dietary and physi-
cal activity factors most relevant for influencing asthma
diagnosis were included in the OPLS and that these fac-
tors were adequately measured based on the previous vali-
dation studies [14, 15]. Self-report measures of physical
activity and dietary intake are subject to recall bias, and the
respondents’ willingness/ability to report accurately.
In the present paper, we have applied a predefined index
based on diet and lifestyle that has previously been pub-
lished as a composite measure [9]. Several such indexes
have been proposed and validated by relating a composite
index score to a health outcome in children [9, 31, 32]. This
approach is an ‘a priori’ procedure, because the index is
developed based on current scientific knowledge (i.e., rec-
ommended evidence-based guidelines).
Such an approach is limited by the content of existing
guidelines that define optimal lifestyle patterns. In any
investigation, utilizing a composite score, it should be
clear what the score intends to accomplish. Is it to measure
absolute diet and lifestyle contribution or to evaluate over-
all adherence to guidelines? The aim of the present study
is the latter. The American Academy of Pediatrics (AAP)
has underlined the need for research aimed at developing
methods to comprehensively assess all behaviors that could
prevent chronic disease risk [8]. Most existing indexes are
able to predict health outcome to an extent, but the associa-
tions are generally moderate for all scores, creating doubts
about their usefulness. Nevertheless, scores can be useful
to measure the extent to which individuals adhere to guide-
lines, and these scores like the one we have applied need
to be used and interpreted with caution. Further research is
needed to evaluate this index prospectively and in diverse
populations, and improvements to this index may be
needed.
Observed associations between the OPLS and BMI,
waist circumference, and hip circumference are biologi-
cally plausible thus serving, to an extent, as evidence of
construct validity. The odds reduction of asthma attrib-
uted to the OPLS was modest. Although this is consist-
ent with most studies of associations between dietary fac-
tors or physical activity factors, and asthma, our results
should be interpreted with consideration of their clinical
importance. The presented evidence does not provide any
information about the impact of improved lifestyle in the
short term, a point that is important to consider in design-
ing intervention studies and/or prevention programs. The
sample of Greek children we studied is fairly homogene-
ous in ethnic and socioeconomic background. The relation-
ship between asthma and lifestyle may be underestimated,
because children from different ethnicities or impoverished
backgrounds were not represented in this sample. Finally,
limitations of this study are those inherent to case–control
studies, and hence, the possibility of residual confounding
cannot be ruled out. The results of this study cannot be used
to prove causality due to the lack of temporal sequence of
events and reverse causality cannot be excluded.
Our case–control study had several strengths. Cases
included children with physician-diagnosed asthma accord-
ing to previously established criteria, minimizing diagnostic
misclassification. Weight, height, and waist circumference
were objectively measured. We measured lung function
(FEV1, FVC, and PEF) at rest which may not be the best
marker, yet lung function at rest is frequently used in typi-
cal clinical settings such as ours. The three lung function
measures (FEV1, FVC, and PEF) were similar between
cases and healthy controls. This may be attributed to well-
controlled asthma in cases. Because cases were treated
with low-dose inhaled steroids, the pharmacotherapy is not
13

likely to account for the differences in overweight/obese
indices. Finally, the possible role of Tanner stage, and other
known confounders (age, sex, breastfeeding history, and
maternal and paternal atopy) was accounted for by includ-
ing them in the multiple logistic regression models.
Conclusion
Adhering to an obesity-preventive lifestyle may decrease
the odds of asthma in children, yet this relationship is
dependent upon BMI. These findings are hypothesis-gen-
erating and warrant further investigation in the form of pro-
spective/intervention studies.
Acknowledgements The authors thank all faculty, students of Ath-
ens University, Harokopio University, and all clinical staff of Penteli
Children’s Hospital and the municipal multi-clinic centers in Galatsi
and Pefki for their invaluable assistance. The authors also thank the
participants of these studies and their families.
Author contribution  Dr. CP conceptualized and designed the study,
drafted the initial manuscript, and approved the final manuscript as
submitted. Dr. EP carried out the initial analyses, co-drafted parts
of the initial manuscript, reviewed and revised the manuscript, and
approved the final manuscript as submitted. Ms. MC conducted
data collection and analyses, critically reviewed the manuscript, and
approved the final manuscript as submitted. Dr. GA conducted data
analyses, critically reviewed the manuscript, and approved the final
manuscript as submitted. Dr. VM coordinated and supervised data
collection, critically reviewed the manuscript, and approved the final
manuscript as submitted. Ms. MD conducted data collection, criti-
cally reviewed the manuscript, and approved the final manuscript
as submitted. Ms. EK conducted data collection, critically reviewed
the manuscript, and approved the final manuscript as submitted. Dr.
KNP contributed to the design of the study, analyses, and drafting the
article.
Compliance with ethical standards  14. Antonogeorgos G, Grigoropoulou D, Papadimitriou A, Priftis
Conflict of interest  On behalf of all authors, the corresponding author
states that there is no conflict of interest.
Funding Dr. Papoutsakis was supported by a post-doctoral grant,
Greek State Scholarship Foundation. Dr. Chondronikola was sup-
ported by a scholarship grant, Onassis Foundation.
References
1. Papoutsakis C, Priftis KN, Drakouli M, Prifti S, Konstantaki E,
Chondronikola M, Antonogeorgos G, Matziou V (2013) Child-
hood overweight/obesity and asthma: is there a link? A system-
atic review of recent epidemiologic evidence. J Acad Nutr Diet
113(1):77–105
2. Beuther DA, Sutherland ER (2007) Overweight, obesity, and
incident asthma: a meta-analysis of prospective epidemiologic
studies. Am J Respir Crit Care Med 175(7):661–666
13
Eur J Nutr
3. Egan KB, Ettinger AS, Bracken MB (2013) Childhood body
mass index and subsequent physician-diagnosed asthma: a sys-
tematic review and meta-analysis of prospective cohort stud-
ies. BMC Pediatr 13:121
4. Platts-Mills TA, Erwin E, Heymann P, Woodfolk J (2005)
Is the hygiene hypothesis still a viable explanation for the
increased prevalence of asthma? Allergy 60(Suppl 79):25–31
5. Asher MI, Montefort S, Bjorksten B, Lai CK, Strachan DP,
Weiland SK, Williams H (2006) Worldwide time trends in
the prevalence of symptoms of asthma, allergic rhinocon-
junctivitis, and eczema in childhood: ISAAC Phases One and
Three repeat multicountry cross-sectional surveys. Lancet
368(9537):733–743
6. Boeing H, Bechthold A, Bub A, Ellinger S, Haller D, Kroke A,
Leschik-Bonnet E, Muller MJ, Oberritter H, Schulze M, Ste-
hle P, Watzl B (2012) Critical review: vegetables and fruit in
the prevention of chronic diseases. Eur J Nutr 51(6):637–663.
doi:10.1007/s00394-012-0380-y
7. Eijkemans M, Mommers M, Draaisma JM, Thijs C, Prins MH
(2012) Physical activity and asthma: a systematic review and
meta-analysis. PLoS ONE 7(12):e50775
8. Barlow SE (2007) Expert committee recommendations regarding
the prevention, assessment, and treatment of child and adolescent
overweight and obesity: summary report. Pediatrics 120(Suppl
4):S164–S192
9. Ntalla I, Yannakoulia M, Dedoussis GV (2016) An Overweight
Preventive Score associates with obesity and glycemic traits.
Metabolism 65(1):81–88. doi:10.1016/j.metabol.2015.09.021
10. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH (2000) Establish-
ing a standard definition for child overweight and obesity world-
wide: international survey. BMJ 320(7244):1240–1243
11. Duke PM, Litt IF, Gross RT (1980) Adolescents’ self-assessment
of sexual maturation. Pediatrics 66(6):918–920
12. Papadopoulou A, Hatziagorou E, Matziou VN, Grigoropoulou
DD, Panagiotakos DB, Tsanakas JN, Gratziou C, Priftis KN
(2011) Comparison in asthma and allergy prevalence in the two
major cities in Greece: the ISAAC phase II survey. Allergol
Immunopathol (Madr) 39(6):347–355
13. Jenkins MA, Clarke JR, Carlin JB, Robertson CF, Hopper JL,
Dalton MF, Holst DP, Choi K, Giles GG (1996) Validation of
questionnaire and bronchial hyperresponsiveness against respira-
tory physician assessment in the diagnosis of asthma. Int J Epi-
demiol 25(3):609–616
KN (2011) Validation of food frequency questionnaire designed
for children 10–12 years: the PANACEA-FFQ. In: 52nd annual
meeting of the European Society for Paediatric Research, New-
castle, UK
15. Avgerinos A, Argiropoulou EC, Almond I, Michalopoulou M
(2000) A new instrument for evaluating energy expenditure: con-
vergent validity and reliability of the physical activity and life-
style questionnaire (PALQ). Sport Perform Health 4:281–300
16. Farah CS, Salome CM (2012) Asthma and obesity: a known asso-
ciation but unknown mechanism. Respirology 17(3):412–421
17. Black MH, Anderson A, Bell RA, Dabelea D, Pihoker C, Say-
dah S, Seid M, Standiford DA, Waitzfelder B, Marcovina SM,
Lawrence JM (2011) Prevalence of asthma and its association
with glycemic control among youth with diabetes. Pediatrics
128(4):e839–e847
18. Cottrell L, Neal WA, Ice C, Perez MK, Piedimonte G (2011)
Metabolic abnormalities in children with asthma. Am J Respir
Crit Care Med 183:441–448
19. Hasler G, Gergen PJ, Ajdacic V, Gamma A, Eich D, Rossler
W, Angst J (2006) Asthma and body weight change: a 20-year
prospective community study of young adults. Int J Obes
30(7):1111–1118. doi:10.1038/sj.ijo.0803215
Eur J Nutr
20. McCloud E, Papoutsakis C (2011) A medical nutrition therapy
primer for childhood asthma: current and emerging perspectives.
J Am Dietetic Assoc 111(7):1052–1064
21. Anthracopoulos MB, Fouzas S, Papadopoulos M, Antono-
georgos G, Papadimitriou A, Panagiotakos DB, Nicolaidou P,
Priftis KN (2012) Physical activity and exercise-induced bron-
choconstriction in Greek schoolchildren. Pediatr Pulmonol
47(11):1080–1087
22. Nurmatov U, Devereux G, Sheikh A (2011) Nutrients and
foods for the primary prevention of asthma and allergy: sys-
tematic review and meta-analysis. J Allergy Clin Immunol
127(3):724–733
23. Telama R, Leskinen E, Yang X (1996) Stability of habitual phys-
ical activity and sport participation: a longitudinal tracking study.
Scand J Med Sci Sports 6(6):371–378
24. Hu FB (2002) Dietary pattern analysis: a new direction in nutri-
tional epidemiology. Curr Opin Lipidol 13(1):3–9
25. Nguyen HT, Bertoni AG, Nettleton JA, Bluemke DA, Levitan
EB, Burke GL (2012) DASH eating pattern is associated with
favorable left ventricular function in the multi-ethnic study of
atherosclerosis. J Am Coll Nutr 31(6):401–407
26. Abiemo EE, Alonso A, Nettleton JA, Steffen LM, Bertoni AG,
Jain A, Lutsey PL (2013) Relationships of the Mediterranean
dietary pattern with insulin resistance and diabetes incidence in
the Multi-Ethnic Study of Atherosclerosis (MESA). Br J Nutr
109(8):1490–1497
27. Kotanidou EP, Grammatikopoulou MG, Spiliotis BE, Kanaka-
Gantenbein C, Tsigga M, Galli-Tsinopoulou A (2013) Ten-
year obesity and overweight prevalence in Greek children: a
systematic review and meta-analysis of 2001-2010 data. Hor-
mones (Athens, Greece) 12(4):537–549
28. Papoutsakis C, Vidra NV, Hatzopoulou I, Tzirkalli M, Farmaki
AE, Evagelidaki E, Kapravelou G, Kontele IG, Skenderi KP,
Yannakoulia M, Dedoussis GV (2007) The Gene-Diet Attica
investigation on childhood obesity (GENDAI): overview of the
study design. Clinl Chem Lab Med 45(3):309–315. doi:10.1515/
cclm.2007.080
29. Garcia-Marcos L, Valverde-Molina J, Ortega ML, Sanchez-Solis
M, Martinez-Torres AE, Castro-Rodriguez JA (2008) Percent
body fat, skinfold thickness or body mass index for defining obe-
sity or overweight, as a risk factor for asthma in schoolchildren:
which one to use in epidemiological studies? Matern Child Nutr
4(4):304–310. doi:10.1111/j.1740-8709.2008.00144.x
30. Vangeepuram N, Teitelbaum SL, Galvez MP, Brenner B,
Doucette J, Wolff MS (2011) Measures of obesity associ-
ated with asthma diagnosis in ethnic minority children. J Obes
2011:517417. doi:10.1155/2011/517417
31. Manios Y, Kourlaba G, Grammatikaki E, Koubitski A, Siatitsa
PE, Vandorou A, Kyriakou K, Dede V, Moschonis G (2010)
Development of a lifestyle-diet quality index for primary
schoolchildren and its relation to insulin resistance: the Healthy
Lifestyle-Diet Index. Eur J Clin Nutr 64(12):1399–1406.
doi:10.1038/ejcn.2010.172
32. Manios Y, Kourlaba G, Grammatikaki E, Androutsos O, Mos-
chonis G, Roma-Giannikou E (2010) Development of a diet-
lifestyle quality index for young children and its relation to obe-
sity: the Preschoolers Diet-Lifestyle Index. Public Health Nutr
13(12):2000–2009. doi:10.1017/s1368980010000698
13