fpsyg-08-01361
August 7, 2017
Time: 15:7
# 1
ORIGINAL RESEARCH
published: 09 August 2017
doi: 10.3389/fpsyg.2017.01361
Edited by:
Leigh Gibson,
University of Roehampton,
United Kingdom
Reviewed by:
Simone Munsch,
University of Fribourg, Switzerland
Ulrich Schweiger,
Lübeck University Medical School,
Germany
*Correspondence:
Vanja Rozenblat
vanja@rozenblat.net
Specialty section:
This article was submitted to
Eating Behavior,
a section of the journal
Frontiers in Psychology
Received: 28 March 2017
Accepted: 26 July 2017
Published: 09 August 2017
Citation:
Rozenblat V, Ryan J, Wertheim EH,
King R, Olsson CA and Krug I (2017)
Investigating Direct Links between
Depression, Emotional Control,
and Physical Punishment with
Adolescent Drive for Thinness
and Bulimic Behaviors, Including
Possible Moderation by the Serotonin
Transporter 5-HTTLPR
Polymorphism.
Front. Psychol. 8:1361.
doi: 10.3389/fpsyg.2017.01361
Investigating Direct Links between
Depression, Emotional Control, and
Physical Punishment with
Adolescent Drive for Thinness and
Bulimic Behaviors, Including
Possible Moderation by the
Serotonin Transporter 5-HTTLPR
Polymorphism
Vanja Rozenblat
1
*, Joanne Ryan
2
, Eleanor H. Wertheim
3
, Ross King
4
, Craig A. Olsson
2,4,5
and Isabel Krug
1
1
School of Psychological Sciences, Faculty of Medicine, Dentistry and Health Sciences, The University of Melbourne,
Parkville, VIC, Australia,
2
Murdoch Childrens Research Institute, Royal Children’s Hospital Melbourne, Parkville, VIC,
Australia,
3
School of Psychology and Public Health, Faculty of Health, La Trobe University, Melbourne, VIC, Australia,
4
School of Psychology, Faculty of Health, Centre for Social and Early Emotional Development, Deakin University, Geelong,
VIC, Australia,
5
Department of Paediatrics, The Royal Children’s Hospital Melbourne, Faculty of Medicine, Dentistry and
Health Sciences, The University of Melbourne, Parkville, VIC, Australia
Objectives: To examine the relationship between psychological and social factors
(depression, emotional control, sexual abuse, and parental physical punishment) and
adolescent drive for Thinness and Bulimic behaviors in a large community sample, and
to investigate possible genetic moderation.
Method: Data were drawn from the Australian Temperament Project (ATP), a
population-based cohort study that has followed a representative sample of 2443
participants from infancy to adulthood across 16 waves since 1983. A subsample
of 650 participants (50.2% female) of Caucasian descent who provided DNA were
genotyped for a serotonin transporter promoter polymorphism (5-HTTLPR). Adolescent
disordered eating attitudes and behaviors were assessed using the Bulimia and Drive
for Thinness scales of the Eating Disorder Inventory-2 (15–16 years). Depression and
emotional control were examined at the same age using the Short Mood and Feelings
Questionnaire, and an ATP-devised measure of emotional control. History of sexual
abuse and physical punishment were assessed retrospectively (23–24 years) in a
subsample of 467 of those providing DNA.
Results: EDI-2 scores were associated with depression, emotional control, and
retrospectively reported parental physical punishment. Although there was statistically
significant moderation of the relationship between parental physical punishment and
bulimic behaviors by 5-HTTLPR (p = 0.0048), genotypes in this subsample were not in
Hardy–Weinberg Equilibrium. No other G×E interactions were significant.
Frontiers in Psychology | www.frontiersin.org
1
August 2017 | Volume 8 | Article 1361
fpsyg-08-01361
August 7, 2017
Time: 15:7
# 2
Rozenblat et al.
5-HTTLPR and Disordered Eating
Conclusion: Findings from this study affirm the central importance of psychosocial
processes in disordered eating patterns in adolescence. Evidence of moderation by 5-
HTTLPR was not conclusive; however, genetic moderation observed in a subsample
not in Hardy–Weinberg Equilibrium warrants further investigation.
Keywords: 5-HTTLPR, gene–environment interactions, disordered eating, parental physical punishment,
depression, emotional control
INTRODUCTION
Eating disorders (EDs) are believed to have a substantial heritable
component (
Bulik et al., 2016
), with estimates from twin
studies ranging from 40 to 60% (
Yilmaz et al., 2015
). Thus
far, research examining molecular genetic mechanisms that may
increase risk for eating pathology has largely investigated whether
certain genetic polymorphisms (e.g., serotonin transporter linked
polymorphism,
5-HTTLPR) are found in different frequency in
those with a clinical ED compared to controls. These studies have
largely produced inconsistent findings (
Calati et al., 2011
;
Solmi
et al., 2016
), supporting the notion that genetic risk operates
in a manner more complex than simple association. One line
of research receiving increasing attention is the possibility that
certain polymorphisms may induce differential risk, depending
upon exposure to certain environmental factors, via gene by
environment (G×E) interaction.
Studies examining whether G×E interactions play a role
in ED etiology have largely focussed on
5-HTTLPR, with the
short (s) allele associated with lower serotonin transcription
activity compared to the long (l) allele (
Heils et al., 1996
).
Serotonin plays a role in mood regulation, appetite, and weight
(
Blundell, 1984
;
Leibowitz and Alexander, 1998
;
Kalra et al.,
1999
;
Ruhé et al., 2007
), all known to be involved in eating
pathology. Serotonin is also involved in the stress-response
system (
Gotlib et al., 2008
;
van Eekelen et al., 2012
), and as
5-HTTLPR is a functional polymorphism it may conceivably
play a role in EDs, directly or indirectly through interaction
with other environmental stressors. However, in other fields
of psychiatry, the role of
5-HTTLPR in moderating the effects
of environmental stressors remains controversial; for example,
in depression, where independently conducted meta-analysis
continue to contradict one another (cf.
Munafò et al., 2009
;
Risch
et al., 2009
;
Karg et al., 2011
). Lack of consensus stems from a
range of methodological limitations, such as insufficient sample
size, inappropriate statistical techniques and multiple testing,
as well as substantial publication bias favoring significant G×E
findings (
Duncan and Keller, 2011
;
Duncan et al., 2014
;
Dick
et al., 2015
;
de Vries et al., 2016
).
To date, seven publications have investigated the role of G×E
interactions in the ED field involving
5-HTTLPR (
Rozenblat
et al., 2017
). Systematic review and meta-analysis of these studies
suggested that
5-HTTLPR may moderate the risk relationship
between experiencing both sexual and physical abuse and bulimic
symptomatology (combined
N = 1,096), and traumatic life events
and ED symptomatology (
N = 909). This was not the case,
however, for risk relationships between depressive and bulimic
symptomatology (
N = 1254) or impulsivity and disordered
eating (
N = 1122). Findings from this review suggest that risk
associated with
5-HTTLPR may be intensified under increasingly
severe social stress, but not psychological distress.
However, findings from this work were based on a combined
sample that was derived by summing across small highly
heterogeneous samples [e.g., two community samples,
N = 369,
Akkermann et al., 2012
;
N = 623,
van Strien et al., 2010
; a
clinical sample,
N = 89,
Richardson et al., 2008
; and a discordant
sister-pair sample,
N = 168 from European cross-institutional
data set used in
Karwautz et al. (2011)
]. Testing interactions in
one large, homogenous sample would be preferable (
Cochran,
1954
). This, for example, may explain why
5-HTTLPR was found
to moderate the effects of depression on eating outcomes in
two of the original studies (
van Strien et al., 2010
;
Mata and
Gotlib, 2011
) but not in the combined data analysis (
Rozenblat
et al., 2017
). Furthermore, the lack of significant interaction
between impulsivity and
5-HTTLPR in the combined-sample
may be partly due to the analysis of impulsivity as an overall
construct, rather than separately testing the particular facets
of impulsivity that have previously been associated with EDs,
such as negative urgency (
Racine et al., 2009, 2013
). Negative
urgency refers to the tendency to act rashly or feel strong impulses
when experiencing negative affect (
Whiteside and Lynam, 2001
),
and, along with the broader ability to regulate one’s emotions,
has wide empirical support for a role in EDs, particularly
bulimia nervosa symptomatology (
Fischer et al., 2003
;
Claes
et al., 2005
), with some evidence linking emotional regulation
and
5-HTTLPR function (
Hariri and Holmes, 2006
). From a
theoretical perspective, lowered emotional control may lead to
greater eating pathology, as individuals may attempt to control
their emotional states via altered food intake (e.g., binge eating
or restricted intake;
Haynos and Fruzzetti, 2011
;
Pearson et al.,
2015
). Meanwhile, the other psychological factor that has been
analyzed in a G×E framework, depressed mood, is believed
to precipitate bulimic behaviors under a number of key ED
models (e.g., Dual Pathway Model;
Stice, 2001
) with support
from longitudinal investigations of high-risk samples (
Stice
et al., 2017
), although there is evidence suggesting depressed
mood may also arise as a consequence of eating pathology
(
Puccio et al., 2016
).
While many prior studies investigating G×E interactions have
focussed on patients with clinical EDs (
Rozenblat et al., 2017
),
analysis of disordered eating in community samples is of equal, if
not greater importance. Developing a better understanding of the
correlates and risk factors for pre-clinical eating pathology, which
may later develop into a ‘full blown’ ED (
Herpertz-Dahlmann
et al., 2013
), can help promote prevention at the earliest possible
opportunity to reduce ED incidence (
Stice et al., 2007
). From
Frontiers in Psychology | www.frontiersin.org
2
August 2017 | Volume 8 | Article 1361