Psychiatry Research ] (]]]]) ]]]–]]]

Contents lists available at SciVerse ScienceDirect

Psychiatry Research

0165-17

http://d

n Corr

Centre,

Australi

E-m

Pleasdistr

journal homepage: www.elsevier.com/locate/psychres

Association between serotonin transporter promoter polymorphismsand psychological distress in a diabetic population

Wilhelm Kay a,b,n, Gillis Inikaa, Reddy Jayaa,b,c, Mitchell Phillip B.a, Campbell Lesleyd,e,Dobson-Stone Carolf,g, Pierce Kerrie D.f, Schofield Peter R.f,g

a School of Psychiatry, Faculty of Medicine and the Black Dog Institute, University of New South Wales, Randwick, NSW 2031, Australiab Faces in the Street, Urban Mental Health and Wellbeing Institute, St Vincent’s Hospital, Sydney, NSW 2010, Australiac Department of Psychiatry, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysiad Diabetes Services, St. Vincent’s Hospital, Sydney, NSW 2010, Australiae School of Medicine, Faculty of Medicine, University of New South Wales, Sydney, NSW 2052, Australiaf Neuroscience Research Australia, Randwick, NSW 2031, Australiag School of Medical Sciences, University of New South Wales, Sydney, NSW 2052, Australia

a r t i c l e i n f o

Article history:

Received 4 May 2011

Received in revised form

4 July 2012

Accepted 12 July 2012

Keywords:

Serotonin transporter

SLC6A4

5-HTTLPR

rs25531

Anxiety

Depression

81/$ - see front matter Crown Copyright & 2

x.doi.org/10.1016/j.psychres.2012.07.008

espondence to: School of Psychiatry, Univers

St. Vincent’s Hospital, 390 Victoria Stree

a. Tel.: þ61 2 8382 1540; fax: þ61 2 8382 15

ail address: kwilhelm@stvincents.com.au (K.

e cite this article as: Wilhelm, K., etess in a diabetic population. Psychia

a b s t r a c t

Investigations into serotonin transporter and anxiety and depression have shown an association

between stress, depression onset and genotype. We investigated the relationship between 5-HTTLPR

genotype and depression and anxiety in a population with diabetes mellitus, a condition associated

with high rates of stress and depression. Participants were classified according to ‘S’ and ‘L’ alleles as

well as the modification of the single nucleotide polymorphism (SNP) rs25531. The 5-HTTLPR low-

expression genotype group (S or LG allele carriers) had significantly higher psychological distress (K10)

scores (N¼234, P¼0.047). Subsequent analysis revealed that the effect of genotype was related to

anxiety symptoms rather than depression symptoms. Furthermore, the main effect of genotype was not

observed when the modification of the SNP polymorphism was not taken into account. Findings suggest

that 5-HTTLPR/rs25531 genotype is associated with psychological distress in a sample of subjects with

diabetes.

Crown Copyright & 2012 Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction

The relationship between the common genetic polymorphism(5-HTTLPR) in the promoter region of the human serotonintransporter gene (SLC6A4) and response to stress, first reportedby Caspi et al. (2003), has led to considerable interest in examin-ing the relationship between variants of the serotonin transporterand proneness to stress adaptation and depression onset. Thefunctional 5-HTTLPR polymorphism has two primary alleles, short(S) and long (L), with the short allele being associated with lowertranscriptional efficiency and thereby decreased serotonin trans-porter (5-HTT) expression (Lesch et al., 1996) although additionalminor insertion/deletion alleles have been reported (Nakamuraet al., 2000). The S and L alleles are further modified by an A/GSNP polymorphism (rs25531) in the 5-HTTLPR that also affectsgene expression, such that the LG allele shows low expression,

012 Published by Elsevier Ireland

ity of NSW, Level 4, O’Brien

t, Darlinghurst, NSW 2010,

42.

Wilhelm).

al., Association between setry Research (2012), http:/

nearly equivalent to the S allele (Wendland et al., 2006; Hu et al.,2006).

The gene (serotonin transporter genotype)� environment(stressful life events or SLEs) interaction has been replicated insubsequent studies investigating a variety of populations andstressors (reviewed in Uher and McGuffin, 2008). However, debateover the robustness of the association was triggered by publicationof recent meta-analyses. Risch et al. (2009), in a meta-analysis of 14studies, concluded there was no evidence that 5-HTTLPR genotypealone, or in interaction with SLEs, was associated with elevated riskof depression in males, females alone or both sexes. Munafo et al.(2009) reported in a meta-analysis of five studies that the interac-tion between 5-HTTLPR genotype and SLEs on risk for depressionwas negligible. However, Karg et al. (2011) noted that the Caspi et al.(2003) study reported an interaction between 5-HTTLPR and twotypes of stressor: (a) SLEs such as employment, financial, housing,health and relationship stressors and (b) childhood maltreatmentstress. They conducted a meta-analysis of all 54 published studies,similarly stratified according to the type of stressor and reported astrong association for specific stressors such as childhood maltreat-ment and medical illness and a marginal association for SLEs. As noassociation was found when Karg et al. included only the studies

Ltd. All rights reserved.

rotonin transporter promoter polymorphisms and psychological/dx.doi.org/10.1016/j.psychres.2012.07.008

K. Wilhelm et al. / Psychiatry Research ] (]]]]) ]]]–]]]2

analysed by Munafo et al. and Risch et al., they concluded that thenegative findings were due to the exclusive focus on studies of SLE,as opposed to specific stressors. It was suggested that the weakerassociation for SLEs was due to potential biases in the SLE studydesign, including variable effects of the heterogeneous life eventsand inconsistent participant recall of stressors. Accordingly, Karget al. reported stronger evidence among studies which assessed SLEswith objective measures than those which used self-report (see alsoCaspi et al., 2010; Rutter, 2010, for review).

In an attempt to avoid such problems, recent studies haveexamined gene� environment (g� e) interactions in populationsexperiencing substantial homogenous stressors associated with highrates of depression. Positive associations have been reported inpopulations experiencing specific medical illnesses including Par-kinson’s disease, migraine, coronary disease, stroke, hip fracture, andinterferon treatment (Mossner et al., 2001; Lenze et al., 2005;Nakatani et al., 2005; Ramasubbu et al., 2006; Gonda et al., 2007;Otte et al., 2007; Kohen et al., 2008; Bull et al., 2009; see Karg et al.,2011, for review).

Depression is twice as common in persons with diabetes as innon-diabetic controls (Anderson et al., 2001) and meta-analysisindicates greater odds of depression (Ali et al., 2006) and general-ised anxiety disorder (Grigsby et al., 2002) amongst people withtype 2 diabetes than those without. Diabetes is characterised bydaily ‘hassles’ such as strict dietary and treatment regimens,which often include frequent injections, threat of hypoglycaemicevents, medical complications and frustration of poor controldespite strict adherence. The associated stress is thought tocontribute to high rates of anxiety and depression found in peoplewith diabetes (Aikens et al., 2009). Despite the high prevalence ofdepression in diabetes, to date no research has investigated therole of the serotonin transporter genotype.

The relationship between diabetes and depression is complex,whereby depression itself is a risk factor for the onset of type2 diabetes (Knol et al., 2006), potentially mediated via cortisoldysregulation and insulin resistance (Pan et al., 2008). However,diabetes-related distress has been shown to be predictive ofdepressive symptomatology one year after diagnosis, even whencontrolling for baseline depressive symptoms (Skinner et al.,2010), suggesting that the stress of the illness itself is a con-tributor to the higher incidence of depression, over and abovepre-existing risk factors. Furthermore, Aujla et al. (2010) reportedno significant association between impaired glucose regulationand depressive symptoms in people newly diagnosed with Type2 diabetes, suggesting that it is the stress of experiencing diabetesafter diagnosis, rather than the intrinsic impaired glucose regula-tion, that contributes to the higher incidence of depression.

The study described in this paper aimed to examine the relation-ship between serotonin transporter genotype and depression, anxi-ety and anxiety-related traits in a population with diabetes mellitus.Expression assays of 5-HTTLPR/rs25531 genotypes show co-domi-nant allele action and low, nearly equivalent expression for the LG

and S alleles (Hu et al., 2006). We hypothesised that in the presenceof a chronic stressor (i.e. having diabetes), individuals with the S andLG alleles would be more likely to report depressive and anxietydisorders as well as higher levels of symptoms of depression,neuroticism and psychological distress.

2. Method

2.1. Participants

Participants were recruited from Diabetes Clinics at St. Vincent’s Hospital and

the Prince of Wales Hospital, Sydney. Patients were approached in the waiting

room by study investigators and provided with a participant information sheet

and flyer. These patients were attending a clinic for people with established

Please cite this article as: Wilhelm, K., et al., Association between sedistress in a diabetic population. Psychiatry Research (2012), http:/

diabetes, often with complications, rather than newly diagnosed diabetes. Written

informed consent was obtained from patients, who were eligible to participate if

they had either type 1 or type 2 diabetes, were aged 18 or over and were able to

read and write in English. Exclusion criteria included cognitive impairment,

psychotic illness, severe depressive illness or the occurrence of a severe life event

in the past month.

Approximately 60% of those approached agreed to participate in the study. The

remainder fell roughly into three groups: those who did not meet the exclusion

criteria and two groups of ‘refusers’. The ‘refusers’ were more likely to be young or

middle-aged males who stated (i) they were already ‘‘involved with too many people

at the clinic’’ or (ii) they were ‘‘in a rush’’ to go to work or other commitments.

A total of 254 participants aged between 23 and 84 (M¼57.4, S.D.¼13.5) were

recruited between July 2006 and March 2008. One hundred and forty five (57.1%)

participants were male and 66 (26.3%) had Type 1 diabetes.

2.2. Measures

Diagnosis of current anxiety and depressive disorder was made using the

Patient Health Questionnaire (PHQ) (Spitzer et al., 1999). The PHQ is a 60-item

self-report diagnostic questionnaire used to diagnose eight current disorders using

DSM-IV criteria. These include Major Depressive Disorder (MDD), Panic Disorder

(PD) and Other Anxiety Disorder (OAD). Participants were classified as having an

anxiety disorder if they met criteria for PD or OAD. It is suggested that these

diagnoses are confirmed at a clinical interview (see below). The validity of the PHQ

as a measure of current MDD has been well established with structured clinical

interviews, including the Structured Clinical Interview for DSM disorders (SCID) as

the validation criterion (Grafe et al., 2004).

The PHQ-9 comprises the nine-item depression scale of the PHQ. Participants

are asked to rate whether nine symptoms of depression applied to them ‘not at all’,

‘several days’, ‘more than half the days’, or ‘nearly every day’ over the past two

weeks. Responses can be used in two ways: (i) to make a tentative diagnosis of

major depression, if a question about role dysfunction is endorsed and four or

more symptoms are rated as present ‘more than half the days’ or ‘nearly every day’

(except for suicidal ideation which is included if any response other than ‘not at

all’ is endorsed’) and (ii) to calculate a depression severity score, where ‘not at

all’¼0 and ‘nearly every day’¼3, to help select and monitor treatment. The

instrument is intended to be presented to the clinician for further clarification.

It has been found to be useful in the detection of major depression (Spitzer et al.,

1999) and assessment of depression severity and has established criterion,

construct and external validity (Kroenke et al., 2001).

A clinical interview was conducted with one of the two study psychiatrists

(JR or KW) in the diabetes clinic immediately after completion of self-report

measures. Responses to the PHQ were reviewed at this interview and participants

were then asked a series of structured questions and shown a list of psychotropic

medications to act as a memory prompt. The aim was to: (i) assess current

diagnoses and treatment, (ii) clarify physical health issues and (iii) assess past

mental health and treatment history.

Neuroticism was measured with the relevant sub-scale from the 60 item NEO

Five Factor Inventory (NEO-FFI) (Costa and McCrae, 1992). Participants are asked

to rate the applicability of statements to themselves on a five point Likert type

scale where 1¼strongly disagree and 5¼strongly agree.

The 10-item Kessler psychological distress scale (K10) (Kessler et al., 2002)

was used to measure psychological distress experienced ‘over the past 30 days’. It is

suitable for screening for cases in the general population and there is a strong

association between a high K10 score and current Composite International

Diagnostic Inventory (CIDI) diagnosis of anxiety or affective disorder (Andrews

and Slade, 2001).

2.3. DNA extraction and genetic analysis

Genomic DNA was extracted from cheek swabs according to the method

described in Wilhelm et al. (2006). Phase-known 5-HTTLPR and rs25531 geno-

types were determined by a modification of the procedure by Wendland et al.

(2006). Genomic DNA was amplified using the QIAGEN Multiplex PCR Kit (QIAGEN

Pty Ltd., Victoria, Australia) with the following primers: forward, 50-

TCCTCCGGTTTGGCGCCTCTTCC-30; reverse, 50-TGGGGGTTGCAGGGGAGATCCTG-30 .

PCR amplification conditions were as follows: 94 1C for 15 min; 38 cycles of 94 1C

for 30 s, 66 1C for 45 s and 72 1C for 60 s; followed by 72 1C for 10 min. Amplicons

were digested with HpaII, and fragments were separated by agarose gel electro-

phoresis. The forward primer introduces an additional HpaII site to enable efficient

digestion of the amplicon. Fragment sizes (in bp) for each allele were therefore as

follows: LA, 506þ6; LG, 396þ110þ6; SA, 463þ6; SG, 396þ67þ6. All genotypes

were scored independently by two researchers.

To account for the effect of the A/G SNP rs25531 polymorphism, we grouped

people according to their predicted levels of 5-HTT expression: ‘Low’ (SASA, SASG,

SALG, SGLG and LGLG genotypes), ‘Medium’ (SALA, SGLA and LALG genotypes), or

‘High’ (LALA genotypes). Of the 254 participants, 72 (28.3%) were classified as

‘Low’, 129 (50.8%) as ‘Medium’ and 49 (19.3%) as ‘High’. Four participants (1.6%)

had ambiguous genotypes and were not included in the statistical analyses.

rotonin transporter promoter polymorphisms and psychological/dx.doi.org/10.1016/j.psychres.2012.07.008

K. Wilhelm et al. / Psychiatry Research ] (]]]]) ]]]–]]] 3

To enable comparison of results with existing literature, we also carried out

separate analyses using the ‘S’ and ‘L’ alleles, without the added modification of

rs25531. When using 5-HTTLPR only genotype, two participants were excluded

from the analyses because of ambiguous genotypes.

2.4. Statistical analysis

All analyses used PASW version 18 for Windows. An alpha level of 0.05 was set

for all statistical tests. Differences in baseline characteristics between participants

were determined using chi-square (w2) goodness-of-fit tests for categorical

variables and one-way analysis of variance (ANOVA) for continuous variables.

Sequential logistic regression was performed to assess the contribution of

genotype to diagnoses of depressive and anxiety disorder, with gender entered

as a covariate. Sequential multiple regression was performed to assess the

contribution of genotype to scores on the PHQ-9 (depression severity), NEO

Neuroticism and K10 (psychological distress) scales, with gender entered as a

covariate.

Table 2Sequential logistic regression analyses of DSM depressive and anxiety disorders as

a function of genotype with gender as a covariate.

95% Confidenceintervalfor odds ratio

B Wald test(z-ratio)

Oddsratio

Lower Upper

5-HTTLPR/rs25531 genotype categorisationDepressive

disorderGender 0.46 1.55 1.59 0.77 3.28

Genotype �0.06 0.06 0.94 0.56 1.57

Anxietydisorder

Gender 0.24 0.31 1.27 0.55 2.9

Genotype �0.31 1.04 0.73 0.4 1.33

5-HTTLPR only genotype categorisationDepressive

disorderGender 0.46 1.53 1.58 0.77 3.28

Genotype 0.04 0.03 1.05 0.61 1.8Anxiety

disorderGender 0.28 0.43 1.32 0.58 3.04

Genotype �0.37 1.35 0.69 0.38 1.29

3. Results

The haplotypes comprising the SA, SG, LA and LG alleles were inHardy–Weinberg equilibrium (w2

¼0.02, d.f.¼2, P¼0.99). Therewere significantly more males with the G variant of the SNP thanfemales (w2

¼4.12, d.f.¼2, P¼0.04). Similarly, for the 5-HTTLPR,the S/S, S/L and L/L genotypes were in Hardy–Weinberg equili-brium (w2

¼3.26, d.f.¼2, P¼0.07). There were no gender differ-ences in allele frequency (w2

¼1.98, d.f.¼1, P¼0.16).There were no differences in diabetes type, age, gender or rate

of depression or anxiety disorder observed between the threegroups when analysed according to the ‘‘Low’’, ‘‘Medium’’ and‘‘High’’ groupings or by S/S, S/L and L/L genotypes (Table 1).

Sequential logistic regression was performed to assess thecontribution of genotype to diagnoses of depressive and anxietydisorder, with gender entered as a covariate (depressive/anxietydisorder¼b0þb1 [gender, 0¼female, 1¼male]þb2 [5-HTTLPR/rs25531 genotype, 0¼Low, 1¼Medium and 2¼High]). Table 2shows regression coefficients, Wald statistics, odds ratios and 95%confidence intervals after entry of both independent variables, fordepressive and anxiety disorder. For depressive disorder, afterstep one, the test of gender against a constant-only model was notstatistically significant, w2(1, N¼222)¼1.605, P¼0.205, indicatingthat gender did not reliably distinguish between participantswith and without an anxiety disorder. Adding genotype to themodel did not result in a significant increment in prediction ofdepressive disorder, w2(1, N¼222)¼0.057, P¼0.812. For anxietydisorder, after step one, the test of gender against a constant-only model was not statistically significant, w2(1, N¼240)¼0.321, P¼0.571. Adding genotype to the model did not resultin a significant increment in prediction in anxiety disorder,w2(1, N¼240)¼1.060, P¼0.303.

When using the same model with gender as a covariate, withthe ‘S’ and ‘L’ alleles (without the rs25531 modification), genotype

Table 1Characteristics of patients with diabetes mellitus (DM) by genotype, using both genoty

5-HTTLPR/rs25531 genotype (n¼250)n

Variables for genotype groups(n¼250/252)n

Low 5-HTT expression(n¼72) n (%)

Med 5-HTT expressi(n¼129) n (%)

DM Type I 18 (25.0) 33 (25.6)

Male 43 (59.7) 69 (53.5)

Depressive disorder 11 (28.2) 22 (17.1)

Anxiety disorder 11 (40.7) 11 (8.5)

Mean (S.D.) age in years 57 (12.7) 57.9 (14.1)

n For 5-HTTLPR/rs25531 genotype, four participants were excluded due to ambig

ambiguous genotypes, thus N=250 and 252 respectively.

Please cite this article as: Wilhelm, K., et al., Association between sedistress in a diabetic population. Psychiatry Research (2012), http:/

did not result in a significant increment in prediction of depres-sive disorder, w2(1, N¼224)¼0.029, P¼0.864 or anxiety disorder,w2(1, N¼242)¼1.357, P¼0.244.

Sequential multiple regression was performed to assess thecontribution of genotype to scores on the PHQ-9, NEO Neuroti-cism and K10 scales, with gender entered as a covariate. Table 3displays correlations, unstandardised regression coefficients (B),standardised regression coefficients (b), semipartial correlations(sri

2), R, R2, and R2 change after entry of both independentvariables. Neither gender nor genotype significantly predictedscores on the PHQ-9 scale. Neither gender nor genotype signifi-cantly predicted scores on the NEO Neuroticism scale. For K10score, after step one, gender alone was not a significant predictor,R2¼0.003, F(1, 228)¼0.638, P¼0.425. After step two, addition of

genotype to the equation resulted in a significant increment in R2,R2¼0.020, F(2, 227)¼2.321, P¼0.047. When using only the ‘S’ and

‘L’ alleles, without the modification of rs25531, addition ofgenotype to the ‘gender only’ model did not result in a significantincrement in R2 for scores on the PHQ-9 scale, R2

¼0.010,F(2, 227)¼1.180, P¼0.626, the NEO Neuroticism scale, R2

¼0.00,F(2, 177)¼0.051, P¼0.950 or the K10 scale, R2

¼0.011, F(2, 229)¼1.310, P¼0.161. Mean K10 scores for each group, with andwithout the modification of rs25531, are presented in Tables 4aand 4b.

4. Discussion

This study investigated the relationship between two geneticvariants (5-HTTLPR and rs25531) that impact the expressionlevels of the serotonin transporter gene and the incidence of

pe categorisations.

5-HTTLPR only genotype (n¼252)n

on High 5-HTT expression(n¼49) n (%)

S/S (n¼44)n (%)

S/L (n¼140)n (%)

L/L (n¼68)n (%)

15 (30.6) 11 (25.0) 37 (26.4) 18 (26.5)

31 (63.3) 23 (52.3) 75 (53.6) 45 (66.2)

6 (12.2) 4 (9.1) 26 (18.6) 9 (13.2)

5 (10.2) 8 (29.6) 12 (44.4) 7 (25.9)

56.2 (13.6) 58.7 (12.9) 57.9 (13.9) 55.6 (13.0)

uous genotypes; for 5-HTTLPR genotype, two participants were excluded due to

rotonin transporter promoter polymorphisms and psychological/dx.doi.org/10.1016/j.psychres.2012.07.008

Table 3Sequential multiple regression analyses for gender and genotype predicting scores on PHQ-9 (depression severity) and K10 (total, depression and anxiety) scales.

Variables DV Gender B SE B b

5-HTTLPR/rs25531 genotype categorisationPHQ-9 Depression severity scale Gender 0.1 – 1.1 0.75 0.1 R2

¼0.01, Adj. R2¼0.00, R¼0.1, sri

2¼0.00, N¼232

Genotype �0.04 0.02 �0.32 0.54 �0.04

K10 Total Gender 0.05 – 0.80 0.95 0.06 R2¼0.02, Adj. R2

¼0.01, R¼0.14, sri2¼0.02*, N¼229

Genotype �0.13 0.02 �1.36 0.68 �0.13

K10 Anxiety Gender 0.05 – 0.28 0.36 0.05 R2¼0.03 Adj. R2

¼0.02, R¼ .016, sri2¼0.02*, N¼229

Genotype �0.15 0.02 �0.58 0.26 �0.15

K10 Depression Gender 0.05 – 0.52 0.66 0.05 R2¼0.01, Adj. R2

¼0.01, R¼0.12, sri2¼0.01, N¼229

Genotype �0.11 0.02 �0.78 0.47 �0.11

5-HTTLPR only genotype categorisationPHQ-9 Depression severity scale Gender 0.1 – 1.13 0.75 0.1 R2

¼ .01, Adj. R2¼ .00, R¼ .10, sri

2¼ .00, N¼232

Genotype �0.02 0.10 �0.28 0.56 �0.03

K10 Total Gender 0.05 – 0.90 0.96 0.06 R2¼ .01 Adj. R2

¼ .003, R¼ .11, sri2¼ .02, N¼229

Genotype �0.09 0.10 �1.01 0.72 �0.09

K10 Anxiety Gender 0.05 – 0.32 0.36 0.06 R2¼0.01, Adj. R2

¼0.01, R¼0.12, sri2¼0.01, N¼229

Genotype �0.10 0.10 �0.44 0.27 �0.11

K10 Depression Gender 0.05 – 0.65 0.67 0.06 R2¼0.01, Adj. R2

¼�0.00, R¼0.09, sri2¼0.01, N¼229

Genotype �0.07 0.10 �0.76 0.48 �0.10

n Po0.05.

Table 4bMean (S.D.) for total K10 scores for each of the 5-HTTLPR/rs25531 genotypes.

Genotype K10 Score M (S.D.) n

Low LGLG 14.0 (4.2) 2

SALG 19.2 (8.7) 23

SGLG 36.0 (0.0) 1

SASA 18.2 (7.3) 39

SASG 18.0 (9.9) 2

SGSG – 0

Medium LALG 16.8 (8.1) 17

SALA 16.5 (6.6) 102

SGLA 10.0 (0.0) 1

High LALA 15.4 (6.8) 43

Table 4aNumbers of participants and mean (S.D.) K10 scores in each genotype categorisation.

5-HTTLPR/rs25531 genotypecategorisation M (S.D.)

5-HTTLPR only genotypecategorisation M (S.D.)

Low (n¼67) 18.7 (7.9) S/S (n¼41) 18.2 (7.2)

Medium (n¼120) 16.5 (6.8) S/L (n¼129) 17.0 (7.2)

High (n¼43) 16.2 (6.8) L/L (n¼62) 16.3 (7.0)

K. Wilhelm et al. / Psychiatry Research ] (]]]]) ]]]–]]]4

depression and anxiety, as well as depression severity, neuroti-cism and psychological distress in a population with diabetes.The 5-HTTLPR/rs25531 genotype was not predictive of diagnosisof depression or anxiety disorder, nor was it a significantpredictor of depressive symptoms, as measured by the PHQ-9scale, nor neuroticism, as measured by the NEO Neuroticismscale. However, in the face of diabetes, as a chronic stressor,genotype was a significant predictor of psychological distresswhereby those classified as ‘Low’ expressing (S and LG alleles) hadsignificantly higher psychological distress (K10) scores. Whenonly the 5-HTTLPR S and L alleles were used to classify partici-pants, there was no effect of gender or genotype on any of theoutcome measures.

We are not aware of any previous reports of an associationbetween 5-HTTLPR/rs25531 and K10 scores. It is not clear why

Please cite this article as: Wilhelm, K., et al., Association between sedistress in a diabetic population. Psychiatry Research (2012), http:/

there was an association with the K10 (a measure of non-specificpsychological distress) but no such association with depressionand neuroticism scores. We cannot distinguish between thisbeing a genuine association between this polymorphism and‘‘distress’’ (as measured by the K10) or an artefactual type I errorbut were prompted to look a little further into the K10.

The K10 was designed for use in the US National HealthInterview Survey as a measure of a core dimension of nonspecificpsychological stress (Kessler et al., 2002) following a review ofscreening scales of psychological distress (Dohrenwend et al.,1980; Link and Dohrenwend, 1980) that concluded that previousscales typically included too heterogeneous a set of cognitive,behavioural, emotional and psychophysiological items. Our find-ing is consistent with the interaction reported by Grabe et al.(2005) between chronic disease burden and serotonin transportergenotype and scores on a measure of nonspecific general mentaland physical distress similar to the K10 (including irritability,nervousness, sleeplessness, dizziness and fatigue).

This finding is somewhat consistent with previous studieswhich have identified a relationship between genotype-depen-dent serotonin transporter expression and self-reported psycho-logical distress in populations experiencing chronic medicalillness, albeit using the 5-HTTLPR categories alone. For example,Parkinson’s disease patients with the S/S genotype had signifi-cantly higher self-reported depressive symptoms than those withthe L/L genotype (Mossner et al., 2001). A study of 1803 patientsfollowing acute myocardial infarction found a significantlygreater frequency of depressive symptoms in those with the Sallele (Nakatani et al., 2005), whilst Kohen et al. (2008) reportsignificantly higher odds of meeting criteria for post-strokedepression amongst stroke survivors with the S/S genotype, thanthose without. In a prospective study of 98 chronic hepatitis Cpatients undergoing interferon-alpha and ribavirin treatment,Bull et al. (2009) reported significantly fewer symptoms ofdepression amongst those with the L/L genotype than those withthe S allele.

Many published studies have assessed the contribution of the5-HTTLPR S and L alleles to proneness to depression and distress,without taking into account the effect of the SNP rs25531 A/Gpolymorphism. We used the 5-HTTLPR/rs25531 genotype in our

rotonin transporter promoter polymorphisms and psychological/dx.doi.org/10.1016/j.psychres.2012.07.008

K. Wilhelm et al. / Psychiatry Research ] (]]]]) ]]]–]]] 5

analyses but also re-analysed the data without taking the rs25531SNP into account, to allow comparability with the existingliterature. The differences between the results based on the twomethods of genotype classification may be considered in light ofthe recent debate over the robustness of the g� e interaction(Munafo et al., 2009; Risch et al., 2009; Caspi et al., 2010; Karget al., 2011). Specifically, our analyses suggest that if the rs25531genotype is not taken into account, the predicted serotonintransporter expression of a proportion of the population studied(i.e., those bearing the LG allele) will be misclassified. This is likelyto obscure the effect of 5-HTTLPR on an associated phenotype, theeffect being dependent on the number of people bearing the LG

allele that have been reclassified. Inspection of the means in thecurrent sample suggested that those with the S/LG genotype weremore comparable to those in the low-expressing group. Mean-while, those with the LGLA genotype were comparable to both thehigh- and medium-expressing groups, which were similar to eachother. It is difficult to generalise with respect to those with theLGLG genotype, as there were only two in our sample. Both thechange in statistical power whereby the number in the low-expressing group increased from 41 to 67, along with the refinedability to predict serotonin transporter gene expression, poten-tially led to the difference in the findings.

5. Limitations

The study was undertaken on a clinical group in an outpatientsetting, without trying to induce too high a burden on partici-pants. Future studies could assess differences in stress levelsamong patients with diabetes by assaying stress hormone levels,for example, salivary cortisol. A control group of healthy partici-pants without diabetes would have enabled assessment of theeffects of genotype in the absence of a chronic stressor. A long-itudinal study with repeated measures over time may have been abetter means of establishing chronicity of stress and assessmentof the relationship between levels of chronic stress and develop-ment of depressive or anxiety disorders.

6. Implications

Diabetes has been shown to compromise monoaminergicneurotransmission in two studies on rodents. The first (Petrisicet al., 1997) found that diabetes contributes to decreased seroto-nin turnover in the central nervous system, while the second(Miyata et al., 2007) featuring streptozocin-induced Type 1 dia-betes demonstrated decreased serotonin response to stress in theprefrontal cortex. These studies suggest a potential role for theserotonin transporter genotype in stress regulation amongpatients with diabetes. Further investigation is required featuringhealthy control populations before any conclusions implicatingserotonin transporter moderated stress pathways specific todiabetes can be drawn.

It is possible that variations in previous findings may beattributable to individuals with the LG allele being misclassifiedin the L/L or S/L groups, when in fact serotonin expression inindividuals with S/LG and LGLG or LGLA genotypes was comparableto those in the S/S or S/L groups respectively. It would be valuableif previous studies were to re-analyse their data taking thers25531 SNP into account, to determine whether this is in factthe case.

Furthermore, the study by Szily et al. (2008), of the 5-HTTLPRand emotional appraisal, found that participants with the S alleleexperienced negative emotions more intensely and felt less ableto cope with stressors. Searching for endophenotypes between

Please cite this article as: Wilhelm, K., et al., Association between sedistress in a diabetic population. Psychiatry Research (2012), http:/

the ‘Low’ expressing alleles and emotional reactivity will enhanceunderstanding of the mechanisms related to emotional regulationand designing of appropriate interventions focusing on dealingwith the predictable stressors related to diabetes as well as SLEs(e.g., using mindfulness meditation) to meet the needs of this sub-population.

Identification of diabetic sub-populations according to geno-type and measuring of physiological stress markers in particularcan help guide the trial of emotional regulation interventions inhighly susceptible genetically predisposed individuals.

Ethics committee approval

This study was granted approval by the University of NewSouth Wales Human Research Ethics Committee. Both hospitalsinvolved are teaching hospitals related to the University of NSW.

Acknowledgements

Adam Finch, Liesbeth Geerligs, Elaine Royal, Lucinda Wedg-wood, Tara Showyin and Jennifer Siegel were involved in datacollection. Funding from the NHMRC Programme Grant no.222708, the Project Grant no. 510216, the ARC Discovery ProjectGrant no. DP0774248, and an Infrastructure Grant from theCentre for Mental Health, NSW Department of Health. We thankKevin Bird and Dusan Hadzi-Pavlovic for statistical input andadvice; the Diabetes Services at St Vincent’s and Prince of WalesHospitals, Sydney for access to their patients and the patientsthemselves, for their time, interest and provision of material forgenotyping.

References

Aikens, J.E., Perkins, D.W., Lipton, B., Piette, J.D., 2009. Longitudinal analysis ofdepressive symptoms and glycemic control in type 2 diabetes. Diabetes Care32, 1177–1181.

Ali, S., Stone, M.A., Peters, J.L., Davies, M.J., Khunti, K., 2006. The prevalence ofco-morbid depression in adults with Type 2 diabetes: a systematic review andmeta-analysis. Diabetic Medicine 23, 1165–1173.

Anderson, R.J., Clouse, R.E., Freedland, K.E., Lustman, P.J., 2001. The prevalence ofcomorbid depression in adults with diabetes. Diabetes Care 24, 1069–1078.

Andrews, G., Slade, T., 2001. Interpreting scores on the Kessler PsychologicalDistress Scale (K10). Australian and New Zealand Journal of Public Health 25,494–497.

Aujla, N., Skinner, T.C., Khunti, K., Davies, M.J., 2010. The prevalence of depressivesymptoms in a white European and South Asian population with impairedglucose regulation and screen-detected Type 2 diabetes mellitus: a compar-ison of two screening tools. Diabetic Medicine 27, 896–905.

Bull, S.J., Huezo-Diaz, P., Binder, E.B., Cubells, J.F., Ranjith, G., Maddock, C.,Miyazaki, C., Alexander, N., Hotopf, M., Cleare, A.J., Norris, S., Cassidy, E.,Aitchison, K.J., Miller, A.H., Pariante, C.M., 2009. Functional polymorphisms inthe interleukin-6 and serotonin transporter genes, and depression and fatigueinduced by interferon-a and ribavirin treatment. Molecular Psychiatry 14,1095–1104.

Caspi, A., Hariri, A.R., Holmes, A., Uher, A., Moffitt, T.E., 2010. Genetic sensitivity tothe environment: the case of the serotonin transporter gene and its implica-tions for studying complex diseases and traits. American Journal of Psychiatry167, 509–527.

Caspi, A., Sugden, K., Moffitt, T.E., Taylor, A., Craig, I.W., Harrington, H., McClay, J.,Mill, J., Martin, J., Braithwaite, A., Poulton, R., 2003. Influence of life stress ondepression: moderation by a polymorphism in the 5-HTT gene. Science 301,386–389.

Costa, P.T., McCrae, R.R., 1992. Revised NEO Personality Inventory (NEO-PI-R)and NEO Five-Factor Inventory (NEO-FFI) Professional Manual. PsychologicalAssessment Resources, Odessa, FL.

Dohrenwend, B.P., Shrout, P.E., Ergi, G.E., Mendelsohn, F.S., 1980. Measures of non-specific psychological distress and other dimensions of psychopathology in thegeneral population. Archives of General Psychiatry 37, 1229–1236.

Grafe, K., Zipfel, S., Herzog, W., Lowe, B., 2004. Screening for psychiatric disorderswith the Patient Health Questionnaire (PHQ). Results from the Germanvalidation study. Diagnostica 50, 171–181.

rotonin transporter promoter polymorphisms and psychological/dx.doi.org/10.1016/j.psychres.2012.07.008

K. Wilhelm et al. / Psychiatry Research ] (]]]]) ]]]–]]]6

Gonda, X., Rihmer, Z., Juhasz, G., Zsombok, T., Bagdy, G., 2007. High anxiety andmigraine are associated with the s allele of the 5HTTLPR gene polymorphism.Psychiatry Research 149, 261–266.

Grabe, H.J., Lange, M., Wolff, B., Volzke, H., Lucht, M., Freyberger, H.J., John, U.,Cascorbi, I., 2005. Mental and physical distress is modulated by a polymorphismin the 5-HT transporter gene interacting with social stressors and chronicdisease burden. Molecular Psychiatry 10, 220–224.

Grigsby, A.B., Anderson, R.J., Freedland, K.E., Clouse, R.E., Lustman, P.J., 2002.Prevalence of anxiety in adults with diabetes. A systematic review. Journal ofPsychosomatic Research 53, 1053–1060.

Hu, X., Lipsky, R.H., Zhu, G., Akhtar, L.A., Taubman, J., Greenberg, B.D., Xu, K.,Arnold, P.D., Richter, M.A., Kennedy, J.L., Murphy, D.L., Goldman, D., 2006.Serotonin transporter promoter gain-of-function genotypes are linked toobsessive-compulsive disorder. American Journal of Human Genetics 78,815–826.

Karg, K., Burmeister, M., Shedden, K., Sen, S., 2011. The serotonin transporterpromoter variant (5-HTTLPR), stress and depression meta-analysis revisited.Evidence of a genetic moderation. Archives of General Psychiatry 68, 444–454.

Kessler, R.C., Andrews, G., Colpe, L.J., Hiripi, E., Mroczek, D.K., Normand, S.L.T.,Walters, E.E., Zaslavsky, A.M., 2002. Short screening scales to monitor popula-tion prevalences and trends in non-specific psychological distress. Psycholo-gical Medicine 32, 959–976.

Knol, M.J., Twisk, J.W.R., Beekman, A.T.F., Heine, R.J., Snoek, F.J., Pouwer, F., 2006.Depression as a risk factor for the onset of type 2 diabetes mellitus. A meta-analysis. Diabetologia 49, 837–845.

Kohen, R., Cain, K.C., Mitchell, P.H., Becker, K., Buzaitis, A., Millard, S.P., Navaja, G.P.,Teri, L., Tirschwell, D., Veith, R., 2008. Association of serotonin transportergene polymorphisms with poststroke depression. Archives of General Psy-chiatry 65, 1296–1302.

Kroenke, K., Spitzer, R.L., Williams, J.B.W., 2001. The PHQ-9. Validity of a briefdepression severity measure. Journal of General Internal Medicine 16,606–613.

Lenze, E.J., Munin, M.C., Ferrell, R.E., Pollock, B.G., Skidmore, E., Lotrich, F., Rogers,J.C., Quear, T., Houck, P., Reynolds, C.F., 2005. Association of the serotonintransporter gene-linked polymorphic region (5-HTTLPR) genotype withdepression in elderly persons after hip fracture. American Journal of GeriatricPsychiatry 13, 428–432.

Lesch, K.P., Bengel, D., Heils, A., Sabol, S.Z., Greenberg, B.D., Petri, S., Benjamin, J.,Muller, C.R., Hamer, D.H., Murphy, D.L., 1996. Association of anxiety-relatedtraits with a polymorphism in the serotonin transporter gene regulatoryregion. Science 274, 1527–1531.

Link, B.G., Dohrenwend, B.P., 1980. Formation of hypotheses about the truerelevance of demoralization in the United States. In: Dohrenwend, B.P.,Dohrenwend, B.S., Gould, M.S., Link, B., Neugebauer, R., Wunsch-Hitzig, R.(Eds.), Mental Illness in the United States: Epidemiological Estimates. Praeger,New York, pp. 114–132.

Miyata, S., Yamada, N., Hirano, S., Tanaka, S.-I., Kamei, J., 2007. Diabetes attenuatespsychological stress-elicited 5-HT secretion in the prefrontal cortex but not inthe amygdala of mice. Brain Research 1147, 233–239.

Mossner, R., Henneberg, A., Schmitt, A., Syagailo, Y.V., Grassle, M., Hennig, T.,Simantov, R., Gerlach, M., Riederer, P., Lesch, K.P., 2001. Allelic variation of

Please cite this article as: Wilhelm, K., et al., Association between sedistress in a diabetic population. Psychiatry Research (2012), http:/

serotonin transporter expression is associated with depression in Parkinson’sdisease. Molecular Psychiatry 6, 350–352.

Munafo, M.R., Durrant, C., Lewis, G., Flint, J., 2009. Gene� environment interac-tions at the serotonin transporter locus. Biological Psychiatry 65, 211–219.

Nakamura, M., Ueno, S., Sano, A., Tanabe, H., 2000. The human serotonintransporter gene linked polymorphism (5-HTTLPR) shows ten novel allelicvariants. Molecular Psychiatry 5, 32–38.

Nakatani, D., Sato, H., Sakata, Y., Shiotani, I., Kinjo, K., Mizuno, H., Shimizu, M., Ito,H., Koretsune, Y., Hirayama, A., Hori, M., 2005. Influence of serotonin trans-porter gene polymorphism on depressive symptoms and new cardiac eventsafter acute myocardial infarction. American Heart Journal 150, 652–658.

Otte, C., McCaffery, J., Ali, S., Whooley, M.A., 2007. Association of a serotonintransporter polymorphism (5-HTTLPR) with depression, perceived stress, andnorepinephrine in patients with coronary disease: the heart and soul study.American Journal of Psychiatry 164, 1379–1384.

Pan, A., Ye, X., Franco, O.H., Li, H., Yu, Z., Zou, S., Zhang, Z., Jiao, S., Lin, X., 2008.Insulin resistance and depressive symptoms in the middle-aged and elderlyChinese: findings from the Nutrition and Health of Aging Population in ChinaStudy. Journal of Affective Disorders 109, 75–82.

Petrisic, M.S., Augood, S.J, Bicknell, R.J., 1997. Monoamine transporter geneexpression in the central nervous system in diabetes mellitus. Journal ofNeurochemistry 68, 2435–2441.

Ramasubbu, R., Tobias, R., Buchan, A.M., Bech-Hansen, N.T., 2006. Serotonintransporter gene promoter region polymorphism associated with poststrokemajor depression. The Journal of Neuropsychiatry and Clinical Neurosciences18, 96–99.

Risch, N., Herrell, R., Lehner, T., Liang, K., Eaves, L., Hoh, J., Griem, A., Kovacs, M.,Ott, J., Merikangas, K.R., 2009. Interaction between the serotonin transportergene (5-HTTLPR), stressful life events, and risk of depression: a meta-analysis.Journal of the American Medical Association 301, 2462–2471.

Rutter, M., 2010. Gene-environment interplay. Depression and Anxiety 27, 1–4.Skinner, T.C., Carey, M.E., Cradock, S., Dallosso, H.M., Daly, H., Davies, M.J., Doherty,

Y., Heller, S., Khunti, K., Oliver, L., 2010. Depressive symptoms in the first yearfrom diagnosis of Type 2 diabetes: results from the DESMOND trial. DiabeticMedicine 27, 965–967.

Spitzer, R.L., Kroenke, K., Williams, J.B.W., the Patient Health QuestionnairePrimary Care Study Group, 1999. Validation and utility of a self-report versionof PRIME-MD: the PHQ primary care study. Journal of the American MedicalAssociation 282, 1737–1744.

Szily, E., Bowen, J., Unoka, Z., Simon, L., Keri, S., 2008. Emotion appraisal ismodulated by the genetic polymorphism of the serotonin transporter. Journalof Neural Transmission 115, 819–822.

Uher, R., McGuffin, P., 2008. The moderation by the serotonin transporter gene ofenvironmental adversity in the aetiology of mental illness: review andmethodological analysis. Molecular Psychiatry 13, 131–146.

Wendland, J.R., Martin, B.J., Druse, M.R., Lesch, K-P., Murphy, D.L., 2006. Simulta-neous genotyping of four functional loci of human SLC6A4, with a reappraisalof 5-HTTLPR and rs25531. Molecular Psychiatry 11, 224–226.

Wilhelm, K., Mitchell, P.B., Niven, H., Finch, A., Wedgwood, L., Scimone, A., Blair,I.P., Parker, G., Schofield, P.R., 2006. Life events, first depression onset and theserotonin transporter gene. British Journal of Psychiatry 188, 210–215.

rotonin transporter promoter polymorphisms and psychological/dx.doi.org/10.1016/j.psychres.2012.07.008