obesity | VOLUME 16 NUMBER 8 | AUGUST 2008 1821
nature publishing group
Behavior and psychology
Familial Transmission of Eating Behaviors
in Preschool-aged Children
Dörte L. Jahnke1 and Petra A. Warschburger1
To examine weight-related differences in eating behaviors and nutrition of preschool-aged children, the influence
of maternal eating behavior on the child’s eating behavior, and sex-related differences in the transmission of eating
behaviors. A total of 142 mothers of children aged 3–6 years participated. Maternal and child’s eating behaviors
as well as child’s food consumption were assessed using questionnaires completed by mothers. Maternal BMI
and child’s standardized BMI (BMI-SDS) were also calculated. More than half of the mothers were obese. Multiple
regression analysis was used to predict eating behavior of the children by mothers’ variables. Overweight children
scored higher in external eating, food responsiveness, and speed of eating than normal-weight children, whereas
children of overweight mothers showed higher amounts of emotional eating than children of normal-weight mothers.
Maternal emotional eating (R2 = 0.19, P < 0.001) and mother’s BMI (R2 = 0.07, P < 0.05) positively predicted emotional
eating of sons. Maternal emotional eating (R2 = 0.19, P < 0.01) completely mediated the relation between mother’s
BMI and emotional eating of sons. For mother–daughter dyads, no such relation was found. The tested model shows
sex-related differences in the transmission of maternal eating behavior which is discussed as being related to the
development and maintenance of obesity.
Obesity (2008) 16, 1821–1825. doi:10.1038/oby.2008.255
Obesity in childhood is a growing problem in industrialized
countries. In the United States and Western Europe, the preva-
lence of overweight and obesity among children has dramati-
cally increased over the past years. According to the National
Health and Nutrition Examination Survey in 2003–2004,
17.1% of US children and adolescents are overweight (1). In
Germany, 9% of the preschool-aged children are overweight
and 3% are obese. The prevalence rates of overweight rise up to
16% for overweight and up to 7% for obesity by the time chil-
dren reach school age (2). Owing to the increased prevalence
at primary school age, this age seems to be a sensible period
for research exploring risk factors for the development of obe-
sity as well as for prevention activities. For children who are
already obese by the time they reach school age, the probability
for being obese in adulthood is very high (3).
Children of overweight parents also bear a greater risk to
develop overweight themselves (4,5). These familial patterns
indicate a genetic predisposition. The increasing prevalence
of obesity and overweight in children emphasizes the impor-
tance of environmental factors. Obesity occurs when a geneti-
cally predisposed person is exposed to certain environmental
factors. Influences of imbalanced energy intake and energy
expenditure, excess energy intake, and reduced activity levels
as well as eating behaviors were discussed in the development
of obesity (e.g., 6,7).
Several studies investigated the influence of parental eating
and weight characteristics on the child’s weight. Familial influ-
ences on child’s overweight differ according to parent’s and
child’s sex (8–10). A few studies found relations between paren-
tal and children’s weight, for mothers and their daughters only
(8,10), whereas Johannsen et al. (9) could show that both boys
and girls weight were related to mothers’ but not fathers’ BMI.
Furthermore, the evidence of associations between paternal
eating behaviors and children’s weight is also inconsistent. The
Framingham Children’s Study has shown that preschool-aged
children whose parents had higher dietary restraint scores pro-
spectively gained more body fat over a period of 6 years. The
highest increase in children’s body fatness over that period of
time was found in children whose parents described high lev-
els of dietary restraint and dietary disinhibition (11). But most
cross-sectional studies could not find a significant relationship
between parental eating behaviors and children’s weight at pre-
school age (9,10). Cutting et al. (8) showed that the maternal
BMI predicts overweight in daughters only, mediated by mater-
nal dietary disinhibition. There is also evidence for overeating
eating styles in children of obese parents. Wardle et al. (12)
compared children of lean and obese parents and showed that
1Department of Psychology, University of Potsdam, Potsdam, Germany. Correspondence: Dörte Jahnke (email@example.com)
Received 10 October 2007; accepted 7 March 2008; published online 15 May 2008. doi:10.1038/oby.2008.255
VOLUME 16 NUMBER 8 | AUGUST 2008 | www.obesityjournal.org
Behavior and psychology
offspring of obese families scored higher on food responsive-
ness and overeating in response to emotional cues.
Relations between parental and child’s eating behavior were
also surveyed in laboratory settings. Faith et al. (13) com-
pared eating in the absence of hunger (EAH) of 5-year-old
children with high vs. low risk to develop obesity in a labora-
tory setting. They found a sex-related effect: boys at high risk
consumed more than twice amount of energy in the absence
of hunger than boys of low risk. The authors assumed that
obesity- promoting genes transmit their influence through
EAH in boys but not in girls. Franzen and Florin (14)
explored the transmission of dietary restraint between moth-
ers and daughters in preadolescence. In a laboratory setting,
7- to 14-year-old daughters of mothers with high dietary
restraint ate more in the absence of hunger (after preload) in
a free access to food procedure than in the condition with-
out preload. Daughters of mothers with low dietary restraint
ate more in the free access to food without preload condition
and ate less after preload. High parental dietary disinhibition
seems to be associated with child’s ability to regulate energy
intake in laboratory setting (15).
Research activities predominantly focused on the relation-
ship between parental eating behaviors and children’s body
fatness and weight, but little is known about the eating behav-
iors of children at preschool age which might be an impor-
tant factor in the familial transmission of obesity. Therefore,
we investigated the relationship between eating behaviors and
weight status in this age group. The aim of this study was to
examine weight-related differences in the eating behaviors and
nutrition of preschool-aged children. We expected offspring of
overweight mothers to score higher in emotional and exter-
nal eating behaviors as well as problematic food intake than
children of normal-weight mothers. Furthermore, we analyzed
the influence of maternal eating behaviors on the child’s eat-
ing behaviors. We also surveyed sex-related differences in the
transmission of eating behaviors.
Methods and Procedures
This cross-sectional study involved 142 mothers of preschool-aged chil-
dren. All mothers were recruited in clinics where they either escorted
their chronically ill child (e.g., child’s asthma) or received treatment
themselves (e.g., maternal depression, weight reduction). Mothers
were asked by the clinic staff (physicians or psychologists) to fill in a
set of questionnaires in case they fulfilled the defined inclusion criteria
(preschool-aged child; maternal or child overweight; low socioeconomic
status). The mothers came from different parts of Germany. The clinics
mainly recruited mothers with lower socioeconomic background. The
participation was voluntary. The research was granted prior approval by
the Ethics Board of our institution.
Mothers were asked for weights and heights to calculate their BMIs (in
kg/m2). A subsample of mothers were also measured for weight and
height data by doctors (N = 89). The objective data for this subsample
did not differ significantly from self-reported data (T = –0.47; P = 0.64).
Therefore, the self-reported data were used in all analyses.
Maternal eating behaviors were assessed using the Dutch Eating
Behaviour Questionnaire (DEBQ) (16). The DEBQ is a 30-item self-
report measure that includes three subscales: emotional eating, external
eating, and restrained eating. The DEBQ has good reliability and validity.
Several studies showed high Cronbach’s α of the scales (α = 0.82–0.93)
(e.g., refs. 17,18).
Child’s body weight was measured in light clothing to the nearest 0.1 kg
using a calibrated scale. Height was measured without shoes to the
nearest 0.1 cm using a vertical ruler. BMI was then calculated as weight
(kg)/square meters. In order to adjust data to age and sex, we calcu-
lated the standardized BMI (BMI-SDS) on the basis of percentiles (19).
Overweight was defined as BMI scores at or above the 90th percentile
for age and gender, obesity at or above the 97th percentile (19).
To evaluate the child’s eating behaviors, all questionnaires were adapted
asking mothers to evaluate the eating and nutritional behaviors of their
We assessed children’s eating behaviors using the DEBQ for chil-
dren (20). The DEBQ for children was adapted on the basis of the DEBQ
(16,17) in order to evaluate eating behavior in children. We used the
subscales emotional and external eating. With respect to the age of the
children, we excluded the restrained eating scale. In their original version,
these subscales showed good reliability with Cronbach’s α = 0.74–0.95
and satisfactory correlations with other measures (21,22). Both scales
were positively associated with spending time in front of television and
computer and parental pressure. External eating was positively associated
with more frequent consumption of sweet and snack-food (22).
The speed of eating was measured by four items concerning negative
eating behaviors (e.g., my child eats his or her food without any breaks).
In addition, we used the subscale food responsiveness (Child Eating
Behavior Questionnaire) (23) to examine child’s craving for food. This
subscale showed good reliability with Cronbach’s α = 0.80–0.82 and test-
retest-reliability r = 0.83 (23). No other subscales of Child Eating Behav-
ior Questionnaire were used.
For the child’s food intake, we asked mothers to rate the consumption
of problematic food (sweets, fast food, soft drinks, and unhealthy snacks)
as well as the consumption of fruits and vegetables of their children on a
six-point scale. High values represent a more frequent consumption (see
Figure 1). We created these scales in a criteria-orientated assessment.
According to the recommendation of nutrition experts, we included rel-
evant food for the analyzed age group and classified in problematic food
and unproblematic food (fruits and vegetables). In a previous study in
1- to 9-year-old children (N = 556), the scales showed moderate reliabil-
ity: problematic food: α = 0.62; fruits and vegetables: α = 0.73.
A total of 142 mothers of preschool-aged children aged 3–6 years par-
ticipated. Maternal mean age was 34.7 years (s.d. = 5.4). Children’s
mean age was 4.2 years (s.d. = 1.0) with 64% being boys. Families were
predominantly German (96%) and intact (86% married or marital
status). Of the 142 mothers, 22.3% had attended a maximum of 9 years
Figure 1 Children’s consumption of problematic food (sweets, fast
food, soft drinks, and unhealthy snacks) and fruits and vegetables rated
obesity | VOLUME 16 NUMBER 8 | AUGUST 2008 1823
Behavior and psychology
of school or had no school leaving certificate, 61.2% attended school
for 10 years, and 16.5% had Abitur (university entrance diploma) with
school attendance for 12 or 13 years. Thirty-one percent of the fami-
lies were defined as living below the national poverty threshold. More
than half of the mothers were overweight or obese (28.6% BMI ≥ 25 vs.
32.8% BMI ≥ 30). Regarding national cutoffs (19), 12% of the children
were overweight, and 7% were already obese (see Table 1). According
to the International Obesity Task Force criteria, the prevalence for over-
weight (>85th <95th percentile) was 12.7% whereas, the prevalence for
obesity (>95th percentile) was 10.6%. Table 2 shows children’s scores
on eating behavior and nutrition scales.
All variables were rated using five- or six-point Likert scale. For bet-
ter comparability, all scales were transformed to range from 0 to 100.
Therefore, scale means of all variables were divided by their number
of scale points and multiplied by 100. Higher scores indicate greater
presence of the factor. Data are presented as means (M) ± s.d. Significance
is determined by P < 0.05. A two (child’s weight status: overweight/obese
vs. normal weight/underweight) × two (maternal weight status: obese/
overweight vs. normal weight) multivariate ANOVA was conducted to
examine differences in child’s eating behaviors depending on child’s and
maternal weight status. We expected higher external and emotional eat-
ing, speed of eating, food responsiveness, and problematic food intake in
overweight children and children of overweight mothers. Correlational
relationships between eating behaviors and BMI of mothers and their
children were tested. Multiple regression analysis was used to predict
eating behaviors of the children by mothers’ variables. We expected a
significant influence of maternal BMI and maternal eating behavior on
children’s eating behavior whereas the influence of maternal BMI was
supposed to be mediated by maternal eating behavior. Therefore, we
independently analyzed the influence of mothers’ BMI and eating behav-
ior on children’s eating behavior as well as the prediction of maternal eat-
ing behavior by mothers’ BMI. In addition, the mediation of maternal
BMI and maternal emotional eating was tested by entering both predic-
tors. All analyses were performed using SPSS 14.0 (SPSS, Chicago, IL).
Weight-related differences in child’s eating behaviors
We compared children’s eating and nutrition-related behav-
iors depending on their weight status. As Figure 2 shows,
overweight and obese children score higher on external eating
(F = 6.56; P = 0.01, food responsiveness (F = 12.42; P = 0.001),
and speed of eating (F = 5.41; P = 0.02). Unexpectedly, children
with higher weight status eat less problematic food (F = 8.09;
P = 0.005).
Differences in child’s eating behaviors depending on maternal
weight status were also analyzed. The results shown in Figure 3
imply that children of overweight or obese mothers show higher
amounts of emotional eating (F = 7.27; P = 0.008) than children
of normal-weight mothers. There was no significant interaction
between mother’s and child’s weight status regarding the eating
behaviors of the children (emotional eating: F = 0.99; P = 0.32;
external eating: F = 0.31; P = 0.58; food responsiveness: F = 0.15;
P = 0.70; speed of eating: F = 0.29; P = 0.59; problematic food: F =
0.01; P = 0.93; fruits and vegetables: F = 0.27; P = 0.43).
Children do not differ in their BMI-SDS (F = 0.14; P = 0.71)
or in their eating behavior (emotional eating: F = 0.07; P = 0.79;
external eating: F = 1.87; P = 0.17; food responsiveness: F = 0.08;
P = 0.78; speed of eating: F = 0.13; P = 0.72; problematic food:
F = 0.01; P = 0.93; fruits and vegetables: F = 3.92; P = 0.06)
depending on maternal socioeconomic status.
table 1 Weight characteristics of children and mothers
(N = 142)
Child (M; s.d.) Mother (M; s.d.)
BMI16.2 (2.0)28.0 (6.6)
Normal weight 76.8%30.0%
BMI-SDS, standardized BMI.
table 2 children’s scores on eating behavior and nutrition
Emotional eating8.14 (0–55)12.83
External eating48.88 (0–85)16.0
Food responsiveness18.39 (0–90)20.6
Speed of eating31.0 (0–81)18.9
Problematic food38.32 (0–7 0)13.92
Speed of eating
102030 40506070 80
* * *
Figure 2 Differences in children’s eating behaviors depending on their
weight status (mean scores) *P < 0.05; **P < 0.01; ***P < 0.001.
01020 3040 506070 80
Speed of eating
Figure 3 Differences in children’s eating behaviors depending on the
weight status of their mothers (mean scores) **P < 0.01.
VOLUME 16 NUMBER 8 | AUGUST 2008 | www.obesityjournal.org
Behavior and psychology
BMI and emotional eating. Mothers’ BMI (R2 = 0.07, P < 0.05)
and maternal emotional eating (R2 = 0.19, P < 0.001) indepen-
dently predicted emotional eating of their sons. Mothers’ BMI
(R2 = 0.30, P < 0.001) significantly predicted maternal emotional
eating. Furthermore, we tested the mediation of maternal BMI
and maternal emotional eating by entering both predictors. As
Table 4 indicates, maternal BMI was no longer a significant pre-
dictor of emotional eating of their sons but maternal emotional
eating (R2 = 0.19, P < 0.01) mediates the relation between moth-
ers’ BMI and the emotional eating of their sons. The Sobel test
(T = 3.55; P < 0.01) confirmed this mediation. Models involving
mothers and their daughters do not show a significant relation
(see Table 5). We could not find a significant relation between
maternal BMI and child’s external eating.
Children who are already overweight at preschool age bear a
high risk to be obese at school-age. The probability for being
obese in adulthood is very high for obese children (3). In this
study, 19% of the children were already overweight or obese.
Research activities predominantly focused on the relationship
between parental eating behaviors and children’s body fatness
and weight, but little is known about the eating behaviors of
children at preschool age which might be an important factor
in the familial transmission of obesity. It is known that chil-
dren of obese families tend to overeat in response to emotional
cues (12). In the past years, several studies examined the influ-
ence of parental eating and weight characteristics as well as feed-
ing practices on children’s body fatness and weight. The familial
influences on child overweight differ according to parent’s
and child’s sex (8,10,11,14). Whereas Hood et al. (11) showed
higher gain in weight of children whose parents report higher
dietary restraint and dietary disinhibition, Cutting et al. (8)
found this effect only for mothers and their daughters. In this
study, we were not able to find any relations between mater-
nal eating behaviors and child’s BMI-SDS. We assume that the
relation between parental eating behaviors and child’s weight is
very small in the preschool age. We rather suppose the familial
transmission of weight-related eating behaviors as an essential
factor in the onset of obesity. Eating behaviors triggered by
relations in eating behaviors and weight between
mother and child
To analyze sex differences with respect to maternal eating
behaviors, BMI and children’s eating behaviors and BMI-SDS,
separate correlations for boys and girls were tested. Boys and
girls did not differ in age, BMI-SDS or socioeconomic status. As
shown in Table 3, emotional eating is merely correlated between
mothers and their sons. Additionally, boys’ emotional eating is
moderately associated with maternal BMI. For mothers and
their daughters, no such relation was found. Maternal external
eating is related to external eating of daughters and sons.
The correlations in Table 3 only show a significant relation
between emotional eating in children and maternal BMI for boys
but not for girls. To determine the prediction of emotional eat-
ing in boys, we independently analyzed the influence of mothers’
table 5 linear regression analysis predicting emotional
eating in girls (N = 51)
Maternal BMI0.40 0.2126.96.36.199
table 4 linear regression analysis predicting emotional
eating in boys (N = 91)
0.21 0.060.41 3.50.001
table 3 correlations (r) between eating behaviors and BMI of the mothers and eating behaviors and BMI-sds in daughters and
sons significant correlations are marked bold; (*P < 0.05; **P < 0.01)
Restrictive eatingEmotional eating External eatingBMI
Children’s variables DaughtersSonsDaughtersSonsDaughters Sons DaughtersSons
Emotional eating0.030.16 0.10
Speed of eating0.05
Problematic food–0.090.100.110.130.180.23*–0.17 0.01
Fruits/vegetables0.21–0.06 0.070.01 –0.23–0.01 –0.01–0.02
BMI-SDS, standardized BMI.
obesity | VOLUME 16 NUMBER 8 | AUGUST 2008 1825
Behavior and psychology
1. Ogden CL, Flegal KM, Carroll MD, Johnson CL. Prevalence and trends
in overweight among US children and adolescents, 1999–2000. JAMA
2. Kurth BM, Schaffrath-Rosario A. Die Verbreitung von Übergewicht und
Adipositas bei Kindern und Jugendlichen in Deutschland. Ergebnisse
des bundesweiten Kinder- und Jugendgesundheitssurveys (KIGGS).
Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 2007;50:
3. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity
in young adulthood from childhood and parental obesity. N Engl J Med
4. Bourchard C, Pérusse L. Genetics of Obesity. Annu Rev Nutr 1993;13:
5. Sørensen TI, Holst C, Stunkard AJ, Skorgaard LT. Correlations of body mass
index of adult adoptees and their biological and adoptive relatives. Int J
Obes Relat Metab Disord 1992;16:227–236.
6. Eck LH, Klesges RC, Hanson, CL, Slawson, D. Children at familial risk for
obesity: an examination of dietary intake, physical activity and weight status.
Int J Obes Relat Metab Disord 1992;16:71–78.
7. Hill JO, Peters JO. Environmental contributions to the obesity epidemic.
8. Cutting TM, Fisher JO, Grimm-Thomas K, Birch LL. Like mother, like
daughter: familial patterns of overweight are mediated by mothers’ dietary
disinhibition. Am J Clin Nutr 1999;69:608–613.
9. Johannsen DL, Johannsen NM, Specker BL. Influence of parents’ eating
behaviors and child feeding practices on children’s weight status. Obesity
(Silver Spring) 2006;14:431–439.
10. Whitaker RC, Deeks CM, Baughum AE, Specker BL. The relationship of
childhood adiposity to parent body mass index and eating behavior. Obes
11. Hood MY, Moore LL, Sundarajan-Ramamurti A, Singer M, Cupples LA,
Ellison RC. Parental eating attitude and the development of obesity in
children. The Framingham Children’s Study. Int J Obes Relat Metab Disord
12. Wardle J, Sanderson S, Guthrie CA, Rapoport L, Plomin R. Parental feeding
style and the intergenerational transmission of obesity risk. Obes Res
13. Faith MS, Berkowitz RI, Stallings VA et al. Eating in the absence of hunger:
a genetic marker for childhood obesity in prepubertal boys? Obesity
14. Franzen S, Florin I. Familiale Transmission von gezügeltem Essverhalten.
Z Kl Psych Psychoth 1995;6:116–122.
15. Johnson SL, Birch LL. Parents’ and children’s adiposity and eating style.
16. Van Strien T, Frijters JE, Bergers GP, Defares PB. The Dutch Eating Behavior
Questionnaire (DEBQ) for assessment of restrained, emotional and external
eating behaviour. Int J Eat Disord 1986;5:295–315.
17. Grunert SC. Ein Inventar zur Erfassung von Selbstaussagen zum
Ernährungsverhalten. Diagnostica 1989;35:167–179.
18. Pook M, Tuschen-Caffier B. Sensitivity to change of scales assessing
symptoms of bulimia nervosa. Psychiatry Res 2004;128:71–78.
19. Kromeyer-Hauschild K, Wabitsch M, Kunze D et al. Perzentile für den
Body-Mass-Index für das Kindes- und Jugendalter unter Heranziehung
verschiedener deutscher Stichproben. Monatsschr Kinderheilkd
20. Franzen S, Florin I. Der Fragebogen zur Messung der Angst vor dem
Dickwerden (FAD). Ein Fragebogen zum Einsatz bei Kindern. Kindh Entwickl
21. Franzen S, Florin I. Der Dutch Eating Behavior Questionnaire für Kinder
(DEBQ-K)—Ein Fragebogen zur Erfassung gezügelten Eßverhaltens. Kindh
22. Van Strien T, Oosterveld P. The children’s DEBQ for assessment of
restrained, emotional, and external eating in 7- to 12-year-old children.
Int J Eat Disord 2008;1:72–81.
23. Wardle J, Guthrie CA, Sanderson S, Rapoport L. Development of the
Children’s Eating Behaviour Questionnaire. J Child Psychol Psychiatry
emotional and external stimuli contribute to the development
and maintenance of obesity. Our results are consistent with
this finding as we could show that overweight children scored
higher on external eating behavior, food responsiveness, speed
of eating, and offspring of overweight mothers scored higher on
emotional eating than children of normal-weight mothers.
Little is known about the influence of parental eating behaviors
on child’s eating behaviors in preschool age. Research activities
on this topic in laboratory settings or in older childhood or ado-
lescence suggest a familial transmission of eating habits (14,15).
Our research implies the maternal transmission of emotional
eating behavior in sons but not in daughters. This result corre-
sponds with Faith et al. (13) who also found a limited effect in
boys. Boys at high risk to develop obesity consumed more than
twice amount of energy in the absence of hunger (EAH) than
boys with low risk. The authors assumed that obesity- promoting
genes transmit their influence through EAH in boys but not in
girls. Recent studies have revealed sex-related transmission in
eating behaviors with maternal disinhibition independently pre-
dicting daughters’ free access intakes and daughters’ overweight
but not for boys (8). Faith et al. (13) assumed that parental
restriction of child food intake may lead to the development of
EAH especially in daughters being less relevant in sons.
Owing to the cross-sectional nature of our study, the inter-
pretation of the results is limited. Beyond, we used mother’s
evaluation to assess the eating behavior of the preschool-aged
children. This could influence the results, as mothers rated
their own eating behavior as well as the eating behavior of
their children. Furthermore, the assessment of eating behavior
is a very sensitive topic; hence, self-serving biases cannot be
excluded. The assessment of father’s influence in the transmis-
sion of eating behavior was not realized due to the low readi-
ness of fathers to participate in our study. Our research suggests
that the familial transmission of emotional eating behavior
may lead to eating and weight disorders. Prospective studies
are necessary to analyze whether high emotional and external
eating in children causes obesity or whether obesity leads to
the changes in the eating behaviors. Nevertheless, our findings
emphasize to integrate the transmission of eating behaviors as
an important topic in programs to prevent childhood obesity.
This study was supported by BMBF/DLR, Förderkennzeichen GFELO
1054304 (to P.W.). We thank all mothers and children who participated in
our study as well as the following clinics for their support (in alphabetical
order): DRK-Mutter-Kind-Kurzentrum Carolinensiel, Ms Menzel;
Inselklinik Sylt, Ms Kötter; Klinik “Nordseedeich,” Ms Petry; Klinik “Santa
Maria,” Mr Gulyas; Klinik “Sonnenalm,” Mr Polan; Klinik “Waldfrieden,”
Ms Patzwall; Klinik “Werraland,” Ms Hasselmann; Ostseeklinik Königshörn,
Mr Schlichting; Südstrandklinik Fehmarn, Mr Bräter and Ms Groos. Special
thanks to Ms Kröller, Ms Richter, and Ms Berger for their helpful advice.
The authors declared no conflict of interest.
© 2008 The Obesity Society