Psychoneuroendocrinology 70 (2016) 70–84

Contents lists available at ScienceDirect

Psychoneuroendocrinology
journal homepage: www.elsevier.com/locate/psyneuen

Invited review

Symptoms of depression and anxiety in youth with type 1 diabetes: A

systematic review and meta-analysis
Barbara Buchberger a,∗ , Hendrik Huppertz a , Laura Krabbe a , Beate Lux a , Jessica T. Mattivi a ,
Aris Siafarikas b,c,d
a
Research Unit Health Technology Assessment and Systematic Reviews, Institute for Health Care Management and Research, Faculty of Economics and
Business Administration, University of Duisburg-Essen, Thea-Leymann-Straße 9, 45127 Essen, Germany
b
Department of Endocrinology and Diabetes, Princess Margaret Hospital, Roberts Rd, Subiaco WA 6008, Perth, Australia
c
School of Paediatrics and Child Health, University of Western Australia, 35 Stirling Hwy, Crawley, WA 6009, Nedlands, Australia
d
Institute for Health Research, University of Notre Dame, Fremantle, 32 Mouat St., Fremantle WA 6959, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: The interaction between psychosocial factors and type 1 diabetes is complex and screening
Received 24 December 2015 for psychosocial risk factors from diagnosis of type 1 diabetes has been recommended. This is a system-
Received in revised form 25 April 2016 atic review and meta-analysis to address the following questions: (1) How prevalent are symptoms of
Accepted 26 April 2016
depression and anxiety in children and adolescents with type 1 diabetes? (2) Is there an association of
symptoms of depression and anxiety with diabetes management and glycemic control?
Keywords:
Material and methods: We searched EMBASE, MEDLINE, The Cochrane Library, and PsycINFO in April 2014
Type 1 diabetes
with an update in May 2015. When possible, data were pooled to estimate summary effects.
Diabetes management
Adolescent
Results: 14 studies investigated symptoms of depression and anxiety in children and adolescents with type
Anxiety 1 diabetes. The pooled prevalence of depressive symptoms was 30.04%, 95% CI [16.33; 43.74]. There were
Depression correlations between symptom levels and glycemic control as well as three-way interactions between
Systematic review HbA1c, blood glucose monitoring frequency or diabetes-specific stress and depression. Symptoms of
anxiety were reported for up to 32% of patients. A negative impact on glycemic control was demonstrated.
Conclusions: Our analyses confirmed a high prevalence of symptoms of depression and anxiety in youth
with type 1 diabetes that potentially compromise diabetes management and glycemic control. In our
opinion these findings support recommendations for early screening for psychological comorbidity
and regular psychosocial assessment from diagnosis. Future prospective studies are warranted to fur-
ther explore the interaction of symptoms of depression and anxiety with type 1 diabetes and develop
evidence-based treatment models.
© 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2. Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.1. Data sources and searches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.2. Study selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.3. Data extraction and quality assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.4. Data synthesis and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.1. Quality assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.2. Study characteristics and outcome measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.3. Baseline characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

∗ Corresponding author at: University of Duisburg-Essen, Faculty of Economics and Business Administration, Institute for Health Care Management and Research, Research
Unit Health Technology Assessment and Systematic Reviews, Thea-Leymann-Str. 9, 45127 Essen, Germany.
E-mail address: barbara.buchberger@medman.uni-due.de (B. Buchberger).

http://dx.doi.org/10.1016/j.psyneuen.2016.04.019
0306-4530/© 2016 Elsevier Ltd. All rights reserved.
 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84 71

3.4. How prevalent are symptoms of depression and anxiety in children and adolescents with type 1 diabetes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.5. Is there an association of symptoms of depression and anxiety with diabetes management and glycemic control? . . . . . . . . . . . . . . . . . . . . . . . . . 74
4. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.1. Strengths and limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Conflicts of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ??
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

1. Introduction Reynolds and Helgeson (2011) suggested that more recently

youth with diabetes are likely to experience less depressive symp-
Living with type 1 diabetes interferes with normal childhood toms than comparison groups. The aim of this article is to update the
activities and can lead to psychological distress (Reynolds and current knowledge base in this area. We systematically searched
Helgeson, 2011). The estimated prevalence of psychiatric disor- for high level evidence to address the following questions: (1) How
ders in children and adolescents with type 1 diabetes varies by prevalent are symptoms of depression and anxiety in children and
study design, the screening or diagnostic tool, diagnostic criteria adolescents with type 1 diabetes? (2) Is there an association of
and training of the interviewers. A ten year longitudinal analysis symptoms of depression and anxiety with diabetes management
identified depression as the most common psychiatric comorbid- and glycemic control?
ity, followed by anxiety and behavior disorders (Kovacs et al., 1997).
Main risk factors for depression in youth are female sex, family 2. Material and methods
dysfunction, low socioeconomic status, and stressful experiences,
which include being diagnosed with type 1 diabetes (Garrison This report follows the guidelines for Meta-Analyses and
et al., 1992; Grey et al., 2002; Hood et al., 2006). Individuals with Systematic Reviews of Observational Studies (MOOSE) and the Pre-
type 1 diabetes and concomitant psychiatric disorders were found ferred Reporting Items for Systematic Reviews and Meta-Analyses
to be less able to psychologically adapt to diabetes (Lernmark (PRISMA) where applicable (Moher et al., 2009; Stroup et al.,
et al., 1999). This negatively affects glycemic control and adher- 2000). It was prepared by the research unit “Health Technology
ence to treatment (Ducat et al., 2015). Long-term glycemic control Assessment (HTA) and Systematic Reviews”, Institute for Health
is reflected by the hemoglobin A1c (HbA1c). This refers to “glycated Care Management and Research, University of Duisburg-Essen,
hemoglobin” and identifies the average plasma glucose concen- Germany, working regularly in cooperation with German HTA
tration over 120 days (the average life span of an erythrocyte) agencies, and a clinical specialist in pediatrics and endocrinology.
(Braun et al., 2016; Diabetes.co.uk, 2016). Non-adherence to treat-
ment includes omitting insulin doses and reduced blood glucose
2.1. Data sources and searches
monitoring frequency (BGMF) and has been proposed as the link
between psychological symptoms and glycemic control (Herzer
We searched MEDLINE and EMBASE via Elsevier, PsycINFO via
and Hood, 2010; Hood et al., 2006). Depression associated with
Ovid, and The Cochrane Library in April 2014 and updated our
poorly managed diabetes leads to a higher risk of severe hypo-
searches in May 2015. A highly sensitive search strategy was used
glycemia (Rewers et al., 2002), hospitalization due to complications
that included both medical subject headings and free-text terms
like diabetic ketoacidosis, negative health outcomes (Stewart et al.,
(e.g. “diabetes mellitus”, “child”, “adolescent”, “mood disorders”,
2005) and reduced quality of life (QoL) in adolescents (Grey et al.,
“comorbidity”) (see Appendix Table A1 for complete search strat-
2002). Furthermore, suicidal ideation is more common in youth
egy). This search was initiated to provide an update for the 2009
with type 1 diabetes than in the general population (Grey et al.,
German clinical guideline “Diagnosis, therapy and control of dia-
2002). There is a strong association between suicidal thoughts,
betes mellitus in children and adolescents” (Deutsche Diabetes
psychiatric disorders and non-adherence to the treatment regi-
Gesellschaft 2009; Deutsche Diabetes Gesellschaft, 2015). This
men (Goldston et al., 1997). Fluctuating blood glucose levels seem
evidence-based guideline is part of the AWMF framework (German
to aggravate symptoms in depressive children and can eventually
working group of the scientific medical societies, Arbeitsgemein-
have adverse effects on their neurocognitive development (Johnson
schaft der Wissenschaftlichen Medizinischen Fachgesellschaften),
et al., 2013).
based on a systematic literature search. The previous guideline
A biological link between diabetes and depression has been pro-
included literature until December 2007. Following upon this we
posed: compensatory metabolic and inflammatory alterations to
run our searches from 2008 until today, limited to publications
the autoimmune-mediated destruction of beta-cells are thought
in German and English. In addition, we searched reference lists of
to aggravate depression (Kongkaew et al., 2014). Dyslipidemia
obtained articles.
appears to be a biological correlate of depression in youth with
diabetes (Hood et al., 2012).
This knowledge about the complex interaction between psy- 2.2. Study selection
chosocial factors and type 1 diabetes is reflected in international
guidelines. Recommendations include screening for psychoso- Two reviewers independently screened titles, abstracts, and
cial risk factors and psychological comorbidities from diagnosis full-texts and decided about the eligibility of articles. Any disagree-
of type 1 diabetes and integrated care by a multidisciplinary ments were resolved by discussion. We considered randomized
team that includes a diabetes educator, psychologist, psychi- and non-randomized controlled trials, cohort studies, case-control
atrist, social worker, dietitian and endocrinologist (American studies and cross-sectional studies reporting on children and ado-
Diabetes Association, 2014; Delamater et al., 2014). Comprehen- lescents up to 19 years diagnosed with type 1 diabetes as being
sive background knowledge is needed to improve evidence-based appropriate for our investigation. Case reports, case series, abstract
management. publications and unpublished studies were excluded. The main
72 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84
outcomes of interest were symptoms of depression and anxiety
and associations with diabetes management and glycemic control.
A complete list of included and excluded articles is available by
request from the corresponding author.
2.3. Data extraction and quality assessment
We extracted the following data: first author, year of publica-
tion, study location, sample size, participants’ characteristics, and
information on the presence of symptoms of depression and anx-
iety. In instances where there was a need, authors were contacted
for further information.
We assessed the internal validity of studies using the quality
assessment tool developed by Thomas et al. (2004) from the Effec-
tive Public Health Practice Project (EPHPP) which is recommended
by the Cochrane Collaboration for non-randomized studies (Higgins
and Green, 2011). The tool covers the following six domains:
selection bias, design, confounders, blinding, data collection meth-
ods, withdrawals and drop-outs. Each domain has to be rated as
“strong”, “moderate” or “weak” according to a detailed dictionary;
in case of no “weak” rating for all domains, the overall rating is
“strong”, in case of one “weak” rating “moderate”, and in case of two
or more “weak” ratings “weak”. Cross-sectional and non-blinded
study designs have to be rated as “weak”. We defined sex, age, edu-
cation, ethnicity, insulin treatment, body mass index (BMI), and
smoking as relevant confounders (Bot et al., 2013). The rating is
“strong” if at least 80%, “moderate” if 60–79% and “weak” if less
than 60% are controlled by study design, stratification, matching or
analyses. Withdrawals and drop-outs were rated “strong” if less
than 60%, “moderate” if 60–79% and “weak” if 80% and greater.
We interpreted missing or unclear information about confound-
ing (Hilliard et al., 2011) and drop-outs (Hood et al., 2011) as an
uncertain situation (“moderate”).
2.4. Data synthesis and analysis
Standard error (SE) was calculated from the prevalence using
√
the formula: SE = p* (1-p)/n and 95% CI = p ± 1.96 * SE; where
p = prevalence.
The meta-analysis was performed using Review Manager soft-
ware (Version 5.3) calculating heterogeneity by Tau2 , I2 and Chi2
statistics. Depending on the heterogeneity, random effects and
fixed effects models were used.
3. Results
Our search identified 2925 citations of which 12 publications
investigated the association of symptoms of depression and anxiety
with type 1 diabetes in children and adolescents: seven cross-
sectional studies, four cohort studies, and one case-control study
(Fig. 1). The update of systematic searches resulted in one fur-
ther cross-sectional study (Kristensen et al., 2014) and one further
cohort study (Baucom et al., 2015) fulfilling our inclusion criteria.
3.1. Quality assessment
The internal validity of the included clinical studies was affected
by major methodological flaws resulting in a high potential of bias.
The total quality rating for all studies was “weak” caused by at least
two “weak”-ratings (see Appendix Table A2). Only Kristensen et al.
(2014) analyzed data from a national registry. For the remainder of
studies selection bias must be assumed because information was
obtained from convenience samples, self-selected or samples from
clinical centers and thus not representative for the target popula-
tion. Everyday health care situations for children and adolescents
with type 1 diabetes are therefore not reflected adequately. In addi-
tion, the selection bias is increased by mainly single-center and
bicentric studies. Further limitations of the studies include small
sample sizes, and nationwide design, not permitting conclusions
to be drawn for other health care systems, or societies with differ-
ent attitudes, norms, and values. Additionally, self-report measures
such as the Children’s Depression Inventory (CDI) are at risk of dis-
tortion, especially in the study of Khan et al. (2013), where most of
the population was unable to understand English and the investi-
gators acted as mediators.
Four publications did not report conflicts of interest or state-
ments about funding or grants (Abdul-Rasoul et al., 2010; De Wit
and Snoek, 2011; Khan et al., 2013; Nardi et al., 2008).
3.2. Study characteristics and outcome measures
The included clinical trials reported nationwide investigations
from the USA (Baucom et al., 2015; Hilliard et al., 2011; Hood et al.,
2011; Herzer et al., 2011; Herzer and Hood, 2010; Ingerski et al.,
2010; McGrady and Hood, 2010; McGrady et al., 2009), Italy (Nardi
et al., 2008), the Netherlands (De Wit and Snoek, 2011), Denmark
(Kristensen et al., 2014), China (Guo et al., 2013), Kuwait (Abdul-
Rasoul et al., 2010) and Pakistan (Khan et al., 2013). The studies
were mostly single-center or bicentric clinical trials comprising
between 86 and 436 participants except for Kristensen et al. (2014)
comparing 786 children with type 1 diabetes to a normative sample
of children and adolescents from a large Danish registry. Abdul-
Rasoul et al. (2010) and Nardi et al. (2008) compared children with
type 1 diabetes to a healthy control group as well, all others were
single-arm studies (Table 1).
The main outcome measures were symptoms of depression and
anxiety, mostly reported by patients or caregivers and measured
by different measurement tools. Correlations and associations with
HbA1c and BGMF were analyzed using bivariate and multivariate
models.
Main exclusion criteria were the presence of a major psychiatric
or neurocognitive disorder, untreated attention-deficit hyperac-
tivity disorder, use of psychopharmacological medications or psy-
chotherapy, a medical disease other than type 1 diabetes, treated
thyroid or celiac disease and diabetes-related complications.
Measurement tools for symptoms of depression (Table 2)
include the CDI according to Kovacs et al. (1995) (De Wit and
Snoek, 2011; Herzer et al., 2011; Herzer and Hood, 2010; Hilliard
et al., 2011; Hood et al., 2011; Ingerski et al., 2010; Khan et al.,
2013; McGrady and Hood, 2010; McGrady et al., 2009), the Arabic
version of the CDI (ACDI) (Abdul-Rasoul et al., 2010), the Cen-
ter for Epidemiologic Studies-Depression (CES-D) scale (Baucom
et al., 2015; Ingerski et al., 2010), the Depression Self-Rating Scale
of Children (DSRS) (Guo et al., 2013), and the depression (BDI-
Y) and anxiety (BAI-Y) subscales of the Beck’s Youth Inventories
(BYI-II) (Kristensen et al., 2014). In three studies, the State-Trait
Anxiety Inventory (for Children) (STAI(C)) was used (Herzer et al.,
2011; Herzer and Hood, 2010; Hilliard et al., 2011). Because of
missing clinical cutoff points for the STAI(C), Herzer and Hood
(2010) and Hilliard et al. (2011) used one standard deviation above
the sample mean to denote clinically elevated anxiety symptoms.
Abdul-Rasoul et al. (2010) and Nardi et al. (2008) evaluated anxiety
symptoms using the Child Behavioral Check-list (CBCL).
3.3. Baseline characteristics
The mean age of patients ranged from 13.4 to 17.7 years, and
the gender distribution for female gender ranged between 21.4%
and 64.4%. Average duration of type 1 diabetes was 3.23–7.7 years
and mean HbA1c values ranged between 7.1% (54 mmol/mol) and
9.68% (82 mmol/mol) (Table 3).
 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84 73

Fig. 1. Flow diagram of literature search and study selection.

Table 1
Study characteristics.

Reference (study type) No. (n) Country, centers Duration (mth) Relevant outcome parameter
(n)

Abdul-Rasoul et al., 2010 (CSS) 436 Kuwait, 5 ns Depressive symptoms (ACDI), anxiety symptoms (CBCL
T1D: 215CG: 221 subscale), BGMF, HbA1c, duration of type 1 diabetes
Baucom et al., 2015 (CS) 175 USA, ns 14 days Depressive symptoms (CES-D), HbA1c, duration of
type 1 diabetes, daily stress severity
De Wit and Snoek, 2011 (CSS) 233 The Netherlands, ns Depressive symptoms (CDI), self-reported HbA1c,
ns duration of type 1 diabetes
Guo et al., 2013 (CSS) 136 China, 17a ns Depressive symptoms (DSRS), HbA1c, duration of type
1 diabetes
Herzer et al., 2011 (CS) 147 USA, 1 9 Depressive symptoms (CDI), anxiety symptoms
(STAI(C)), BGMF, HbA1c, duration of type 1 diabetes,
diabetic-specific worry (PedsQL subscale), family
conflict (DFCS)b
Herzer and Hood, 2010 (CSS) 276 USA, 2 ns Depressive symptoms (CDI), anxiety symptoms
(STAI(C))b , BGMF, HbA1c, duration of type 1 diabetes
Hilliard et al., 2011 (CS) 150 USA, ns 12 Depressive symptoms (CDI), anxiety symptoms
(STAI(C)), BGMF, HbA1c, duration of type 1 diabetes,
QoL (PedsQL)c
Hood et al., 2011 (CS) 145 USA, 1 6 Depressive symptoms (CDI), BGMF, HbA1c, duration of
type 1 diabetes
Ingerski et al., 2010 (CSS) 261d USA, 2 ns Depressive symptoms (CDI, CES-Dc ), BGMF, HbA1c,
duration of type 1 diabetes, QoL (PedsQL), family
conflict (DFCS)c
Khan et al., 2013 (CSS) 86 Pakistan, 2 10 Depressive symptoms (CDI), HbA1c, duration of type 1
diabetes
Kristensen et al., 2014 (CSS) T1D: 786e Denmark, ns ns Depressive symptoms (BDI-Y)f , anxiety symptoms
CG: 3.200 (from (BAI-Y)f , HbA1c, duration of type 1 diabetes
registry)
McGrady and Hood, 2010 (CS) 144 USA, ns 6 Depressive symptoms (CDI), BGMF, HbA1c, duration of
type 1 diabetes
McGrady et al., 2009 (CSS) 276 USA, 2 ns Depressive symptoms (CDI), BGMF, HbA1c, duration of
type 1 diabetes
Nardi et al., 2008 (CCS) 140 Italy, 1 ns HbA1c, QoL (PedsQL)b , duration of type 1 diabetes,
T1D: 70 behaviour problems and competence (CBCL)c
CG: 70

ACDI: Arabic Children’s Depression Inventory; BAI-Y: Anxiety subscale of The Beck’s Youth Inventories; BDI-Y: Depression subscale of The Beck’s Youth Inventories; BGMF:
blood glucose monitoring frequency; CBCL: The Child Behavioral Check-list; CCS: case-control study; CDI: Children’s Depression Inventory; CES-D: The Center for Epidemio-
logic Studies-Depression scale; CG: control group; CS: cohort study; CSS: cross-sectional study; DFCS: Diabetes Family Conflict Scale; DSRS: Depression Self-Rating Scale of
Children; G: group; ns: not stated; mth: month/s; PedsQL: Pediatric Quality of Life Inventory; QoL: quality of life; STAI(C): State-Trait Anxiety Inventory (for Children); T1D:
type 1 diabetes mellitus group.
a
Two-thirds of the sample was recruited at one diabetes center.
b
Completed by patients and caregivers.
c
Completed by caregivers.
d
Patients were divided into four glycemic control and QoL groups (G1: HbA1c < 7.5, QoL ≥ 75.0; G2: HbA1c < 7.5, QoL < 75.0; G3: HbA1c ≥ 7.5, QoL ≥ 75.0; G4: HbA1c ≥ 7.5,
QoL < 75.0).
e
Patients were stratified by sex and age (1–4 yrs, 8–10 yrs, 11–14 yrs, 15–16 yrs, 17 yrs). For further details see Kristensen et al. (2014).
f
Subscale of The Beck’s Youth Inventories—second edition (BYI-II).
74 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

Table 2
Measurement tools.

Measure Items (n) Scoring Range Interpretation Cut-offa

Depressive symptoms
BDI-Y 20 0 = never, 1 = sometimes, 0-60 55 = average score, na
2 = often, 3 = always 55–59 = mildly elevated, 60-69
moderately elevated,
≥70 = extremely elevatedb
CDI 27 0 = no symptom to 2 = distinct 0-54 15–29 = mild, ≥13
symptom 30–43 = moderate, >43 = severe
CES-D 20 0 = rarely or none of the time to 0-60 0–15 = minimal, 16–23 = mild, ≥16
3 = most or all of the time 24–60 = moderate/severe
DSRS 18 0 or 2 = never, 1 = sometimes, 0 0-36 na > 15
or 2 = mostc

Anxiety symptoms
BAI-Y 20 0 = never, 1 = sometimes, 0-60 na na
2 = often, 3 = always
CBCL 118 0 = not true, 1 = sometimes na <67 = normal range, na
true, 2 = often true 67–70 = borderline clinical, >70
clinical rangeb
STAI(C)A-State A-Trait STAI(C)2020 1-3d 1–3d 20–6020–60 1-3d 1-3d nana
1 = hardly ever, 2 = sometimes, na
3 = often na

BAI-Y: Becks’s Anxiety Inventory-Youth; BDI-Y: Becks’s Depression Inventory-Youth; CBCL: The Child Behavioral Check-list; CDI: Children’s Depression Inventory; CES-D: The
Center for Epidemiologic Studies-Depression scale; DSRS: Depression Self-Rating Scale of Children; na: not available; STAI(C): State-Trait Anxiety Inventory (for Children).
a
Indicative of clinically elevated symptoms.
b
Raw scores translated into T-scores (according to age and gender).
c
Depending on the positive or negative polarity of the item.
d
The STAI state scale is constructed to ask children how they feel at a particular moment in time. A sample question is: “I feel very nervous, nervous, not nervous.” (Kirisci
et al., 1996).

3.4. How prevalent are symptoms of depression and anxiety in
children and adolescents with type 1 diabetes?
Six studies reported symptoms of depression as measured by
CDI and were included in a meta-analysis (Abdul-Rasoul et al.,
2010; De Wit and Snoek, 2011; Herzer and Hood, 2010; Hood et al.,
2011; Khan et al., 2013; McGrady and Hood, 2010) all referring to
one reference (Kovacs, 2003) except of Abdul-Rasoul et al. (2010)
who used the ACDI and Khan et al. (2013) who refer to a dif-
ferent reference (Kovacs, 1980). All studies mention that the CDI
consists of 27 items with scores ranging from 0 to 54 (Table 2).
With the exclusion of Abdul-Rasoul et al. (2010) and De Wit and
Snoek (2011) who set cutoff points of ≥15 and ≥16, all other studies
used ≥13 points. The results (Fig. 2, random effects model) showed
an overall prevalence of depressive symptoms of 30.04% (95% CI
[16.33; 43.74]) with high heterogeneity (I2 = 97%); similar results
were achieved using the fixed effects model.
Using the DSRS for depressive symptoms, Guo et al. (2013)
identified symptoms of depression in 17.6% of patients (Table 4).
School attendance, family annual revenue, peer relationship and
the parent-child relationship were significant predictors. Baucom
et al. (2015) found moderate to severe symptoms of depression in
25.1% of adolescents using the CES-D. Kristensen et al. (2014) used
the BDI-Y, and in the largest subgroup with children aged 11–14
years, there were statistically significant differences in the fre-
quency of slightly elevated scores between healthy boys and boys
with type 1 diabetes, and in the frequency of moderately elevated
scores between healthy girls and girls with type 1 diabetes. Nardi
et al. (2008) showed a low but statistically significant correlation
of longer disease duration with the subscale “anxious/depressed”
for the CBCL.
We included five studies examining the association between
anxiety as analyzed using the STAI(C) and type 1 diabetes in chil-
dren and adolescents (Abdul-Rasoul et al., 2010; Herzer et al., 2011;
Herzer and Hood, 2010; Hilliard et al., 2011; Kristensen et al., 2014)
(Table 4). Herzer and Hood (2010) reported 13.4% and 17% of partic-
ipants with state and trait anxiety scores (respectively) falling one
standard deviation above the sample mean denoting clinically ele-
vated symptoms. Using the BAI-Y, Kristensen et al. (2014) showed
statistically significant differences between both female and male
11–14 year old controls and patients with type 1 diabetes in the
frequency of slightly elevated scores and only for females in the fre-
quency of moderately elevated scores. Abdul-Rasoul et al. (2010)
found higher anxiety scores (CBCL) in diabetic children than in
healthy controls (p < 0.001).
Marked and statistically significant correlations of anxiety
(STAI(C)) with depressive symptoms (CDI) were found by Herzer
and Hood (2010) and Hilliard et al. (2011) (Table 4). Kristensen et al.
(2014) reported a strong and statistically significant correlation
between the BYI-II subscales for anxiety (BAI-Y) and depression
(BDI-Y).
3.5. Is there an association of symptoms of depression and
anxiety with diabetes management and glycemic control?
A positive correlation of statistical significance between symp-
toms of depression as measured by CDI and HbA1c was found in five
studies (Abdul-Rasoul et al., 2010; De Wit and Snoek, 2011; Herzer
et al., 2011; Hilliard et al., 2011; McGrady and Hood, 2010) but was
strong only in the study of Abdul-Rasoul et al. (2010) (Table 4).
This correlation was present for as long as six, nine (Herzer et al.,
2011; McGrady and Hood, 2010) and twelve months (Hilliard et al.,
2011). Hood et al. (2011) observed an inverse correlation between
CDI and HbA1c: decrease in CDI raw scores from 7.9 ± 7.1 at base-
line to 6.1 ± 5.3 at six months (p < 0.001) and increase in HbA1c
from 8.8 ± 1.9% to 9.1 ± 2.1% (p < 0.01), but with modest (Cohen’s
d = 0.34) respectively small magnitude (d = 0.20). The change in
CDI scores was not attributable to a higher frequency of medi-
cal or mental health visits over that time (p > 0.05). Ingerski et al.
(2010) assigned patients to subgroups according to HbA1c and
median-split QoL scores. Fewer depressive symptoms (CDI) were
reported by the group with optimal glycemic control (HbA1c <7.5%
(<58 mmol/mol)) and high QoL scores (≥75), as well as for patients
with suboptimal control (HbA1c >7.5% (>58 mmol/mol)) and high
QoL compared to those with suboptimal control and low QoL (<75).
Table 3
Baseline characteristics.

Reference (study type) No. (n) Age (years) Female gender (n) Duration of type 1 HbA1c (%) BGMF (times daily) Depression Anxiety
diabetes (years)

Abdul-Rasoul et al., 2010 (CSS) 436 7–9 yrs: T1D: 105 (48.8%) Mean (range) Mean (range) ≥ 2: – –
T1D: 215 T1D: n = 59 (27.4%) CG: 105 (47.5%) T1D: 5.1 (3.5-12.5) T1D: 8.5 n = 148 (68.8%)
CG: 221 CG: n = 62 (28%) (6.4-9.2)
10–12 yrs:
T1D: n = 68 (31.6%)

B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

CG: n = 66 (29.9%)
> 12 yrs:
T1D: n = 88 (40%)
CG: n = 93 (42.1%)
Baucom et al., 2015 (CS) 175 17.7 ± 0.38 113 (64.6%) 7.46 ± 3.76 8.19 ± 1.6 – CES-D –a
16.56 ± 12.43
De Wit and Snoek, 2011 (CSS) 233 15.5 ± 2.2 163 5.9 ± 4.4 8.1 ± 1.6 (n = 181) – CDI: 9.4 –
Guo et al., 2013 (CSS) 136 13.5 ± 3.54 86 (63.2%) Mean (range) 9.68 ± 2.86 – – –
3.23 yrs
(6mth–10yrs)
Herzer et al., 2011 (CS) 147 15.5 ± 1.4 51.7% 6.0 ± 3.8 At mth 9b : 3.0 ± 1.7 CDI at mth 6b : STAI state anxiety
9.1 ± 2.1 6.1 ± 5.3 baseline at mth
6b , c : 29.7 ± 4.8
Herzer and Hood, 2010 (CSS) 276 15.63 ± 1.39 47.5% 6.61 ± 4.01 8.91 ± 1.81 4.0 ± 1.75 CDI: 7.3 ± 6.4 STAI state anxiety:
29.8 ± 5.0
STAI trait anxiety:
32.1 ± 7.0
STAI trait anxietyc :
36.1 ± 9.3
Hilliard et al., 2011 (CS) 150 15.5 ± 1.4 51.3% 6.0 ± 3.9 8.8 ± 1.9 3.8 ± 1.7 CDI: 8.0 ± 7.1 STAI state anxiety:
30.3 ± 5.2
Hood et al., 2011 (CS) 145 15.5 ± 1.4 52.4% 5.9 ± 3.8 8.8 ± 1.9 3.5 ± 1.5 CDI: 7.9 ± 7.1 –
Ingerski et al., 2010 (CSS) 261 G1d : 15.4 ± 1.6 G1d : 37.0% G1d : 4.8 ± 2.9 G1d : 6.8 ± 0.6 G1d : 5.3 ± 1.7 CDI: –
G2d : 15.7 ± 1.6 G2d : 56.3% G2d : 4.2 ± 3.0 G2d : 7.1 ± 0.2 G2d : 4.9 ± 2.4 G1d : 4.5 ± 5.1
G3d : 15.5 ± 1.4 G3d : 42.6% G3d : 7.7 ± 3.9 G3d : 9.0 ± 1.4 G3d : 4.2 ± 1.6 G2d : 9.4 ± 7.0
G4d : 15.8 ± 1.3 G4d : 51.8% G4d : 7.1 ± 3.8 G4d : 9.7 ± 1.9 G4d : 3.3 ± 1.5 G3d : 5.1 ± 5.4
G4d : 10.0 ± 6.8
CES-D:
G1d : 8.6 ± 7.6
G2d : 12.1 ± 12.1
G3d : 10.2 ± 9.6
G4d : 10.4 ± 7.9

75
 76
Table 3 (Continued)

Reference (study type) No. (n) Age (years) Female gender (n) Duration of type 1 HbA1c (%) BGMF (times daily) Depression Anxiety
diabetes (years)

Khan et al., 2013 (CSS) 86 7–11.5 yrs: 45 (52.4%) 1–3 yrs: Control – – –
n = 41 (47.6%) n = 24 (27.9%) Good (< 8):
> 11.5-15 yrs: > 3-8 yrs: n = 13 (15.1%)
n = 45 (52.3%) n = 29 (33.7%) Mean (8–10):
> 8 yrs: n = 10 (11.6%)

B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

n = 33 (38.3%) Poor (> 10):
n = 63 (73.2%)
Kristensen et al., 2014e (CSS) T1D: 786 13.4 ± 2.6 51.5% 5.24 ± 3.28 8.04 ± 1.13f – Median (25th/75th Median (25th/75th
CG: 3.200 8.06 ± 1.14g percentile) percentile)
(from Danish BDI-Y: BAI-Y:
registry) T1D Girls (n = 277): T1D Girls (n = 168):
5.0 (2.0/12.0) 7.0 (3.0/12.0)
CG Girls (n = 169): CG Girls (n = 273):
11.0 (7.0/16.0) 13.0 (9.0/18.0)
T1D Boys (n = 167): T1D Boys (n = 167):
3.0 (1.0/7.0) 4.0 (2.0/9.0)
CG Boys (n = 292): CG Boys (n = 294):
8.0 (4.0/12.0) 10.0 (6.0/14.0)
McGrady and Hood, 2010 (CS) 144 15.45 ± 1.39 75 (52%) 5.94 ± 3.78 3.57h ± 1.52 3.91 ± 1.64 CDI: 7.92 ± 7.14 –
McGrady et al. (2009) (CSS) 276 15.6 ± 1.4 47.5% 6.6 ± 1.81 8.9 ± 1.8 3.83 ± 1.45 CDI: 7.3 ± 6.4 –
Nardi et al., 2008 (CCS) 140 T1D: 13.8 ± 3.0 T1D: 31 T1D: 6.1 ± 3.6 T1D: 8.6 ± 1.3 – – –
T1D: 70 CG: 13.7 ± 2.9 CG: 30
CG: 70

BGMF: blood glucose monitoring frequency; CCS: case-control study; CDI: Children’s Depression Inventory; CES-D: The Center for Epidemiologic Studies-Depression scale; CG: control group; CS: cohort study; CSS: cross-sectional
study; DSS: diabetes-specific stressful events; G: group; mth: month/s; GS: general stressful events; STAI(C): State-Trait Anxiety Inventory (for Children); T1D: type 1 diabetes mellitus group; yrs: years
Data given as mean ± standard deviation, otherwise stated.
a
Other baseline characteristics: Mean GS severity: 2.89 ± 0.76, Mean DSS severity: 2.16 ± 0.68.
b
Data obtained at baseline, 6 mth, 9 mth.
c
Completed by caregivers.
d
Glycemic control and QoL groups (G1 (n = 27): HbA1c < 7.5, QoL ≥ 75.0; G2 (n = 16): HbA1c < 7.5, QoL < 75.0; G3 (n = 108): HbA1c ≥ 7.5, QoL ≥ 75.0; G4 (n = 110): HbA1c ≥ 7.5, QoL < 75.0).
e
Patients were stratified by sex and age (1–4 yrs, 8–10 yrs, 11–14 yrs, 15–16 yrs, 17 yrs). For further details see Kristensen et al., 2014.
f
Based on submitted blood samples.
g
Based on registry data (used for analyses, if participants did not provide a blood sample for the study).
h
The authors were asked about the conspicuous value but didn’t answer.
Table 4
Results depressive and anxiety symptoms.

Reference (study type) No. (n) Depressive symptoms Anxiety symptoms

Abdul-Rasoul et al. 436 ACDI CBCL

(2010) (CSS) T1D: 215 CG: 221
• Symptoms of mild depression: T1D: 63%, CG: 21%, p < 0.05
• Symptoms of moderate depression: similar results
• Correlation with HbA1c: r = 0.87, p < 0.001
- Similar results in subgroups HbA1c > 8.5% vs. HbA1c < 8.5%,
p < 0.001
• Anxiety higher in T1D vs. CG, p < 0.001
• Associations
- Poor glycemic control (HbA1c > 8.5%) with higher frequency of
borderline: p < 0.001 and clinical anxiety: p < 0.01

Baucom et al. (2015) 175 CES-D –

B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

(CS)
• Minimal range (0–15): 58.3% (n = 102); mild range (16–23):
16.6% (n = 29); moderate/severe range (24–60): 25.1% (n = 44)
• Correlation
- With HbA1c: r = 0.27, p < 0.001
• With mean adherence (SCI): r = 0.33, p < 0.001
• With mean GS severity: r = 0.31, p < 0.001
• With mean DSS severity: r = 0.36, p < 0.001
• Associations three-way interaction
- Greater depressive symptoms (p < 0.001) and more severe mean
DSS (p < 0.002) with higher HbA1c, but more severe mean GS
was not significantly with HbA1c (p = 0.961)

De Wit and Snoek 233 CDI –

(2011) (CSS)
• CDI scores ≥16: n = 40 (17.2%); girls: n = 30; boys: n = 10, p < 0.01
• Mean scores: girls: 10.3 ± 8.0 vs. boys: 7.4 ± 6.3, p = 0.003
• Correlation with HbA1c: r = 0.24, p < 0.001

Guo et al. (2013) (CSS) 136 DSRS –

• 9.51 ± 5.19, Score > 15: 17.6% of patients

Herzer et al. (2011) 147 CDI STAIC

(CS)
• Correlation
- At baseline with family conflicta at baseline: r = 0.31, p < 0.001
• At 6 mth with family conflicta at baseline: r = 0.17, p < 0.05
• At 6 mth with HbA1c at 9 mth: r = 0.21, p < 0.01
• Baseline family conflictb predicted 6 mth CDI, p < 0.01
• Correlation
- At baseline with family conflicta at baseline: r = 0.30, p < 0.001
• At 6 mth with family conflicta at baseline: r = 0.39, p < 0.001
• At baseline with HbA1c at 9 mth: r = 0.17, p < 0.05
• At 6 mth with HbA1c at 9 mth: r = 0.27, p < 0.001
• Anxiety mediated and accounted for 20% of the family
conflict-glycemic control link
• Baseline family conflictb predicted 6 mth STAIC, p < 0.001

77
 78
Table 4 (Continued)

Reference (study type) No. (n) Depressive symptoms Anxiety symptoms

Herzer and Hood 276 CDI STAIc , d

(2010) (CSS)
• 21% with CDI ≥13 • Mean state anxiety score: 29.82 ± 4.99
• Correlation of CDI with STAI: state anxiety: r = 0.55, p < 0.001; • Mean trait anxiety score: 32.15 ± 7.01
trait anxiety: r = 0.72, p < 0.001 • Correlation of state anxiety
- With HbA1c: r = 0.25, p < 0.001
• With BGMF: r = −0.25, p < 0.001
• Associations
- Higher STAI with less BGMF (r2 = 0.25, p < 0.0001) and
suboptimal glycemic control (r2 = 0.32, p < 0.0001)e
• STAI symptoms were correlates of BGMF and glycemic control

B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

independent of depressive symptoms
• Correlation of trait anxiety
- With BGMF: r = −0.17, p = 0.005

Hilliard et al. (2011) 150 CDI STAIC

(CS)
• Correlation
- With HbA1c: at baseline: r = 0.22, p < 0.01; 12 mth: r = 0.21,
p = 0.01
• With lower BGMF at baseline: r = −0.22, p < 0.01
• Predictor of 12 mth BGMFf , p < 0.05g
• For every 20 point rise in CDI scores, BGMF decreased by one
check per day
• Correlation of CDI with STAIC state anxiety: at baseline: r = 0.58,
p < 0.0001; 12 mth: r = 0.67, p < 0.0001
• Correlation
- With HbA1c: at baseline: r = 0.30, p < 0.001; 12 mth: r = 0.25,
p < 0.01
• With less frequent BGM at baseline: r = −0.19, p < 0.05
• Higher STAIC predictor of HbA1c: p < 0.05
• 14-point increase in anxiety screener scores associated with
clinically meaningful rise of 1% in HbA1cf

Hood et al. (2011) (CS) 145 CDI –

• Participants with CDI ≥ 13: baseline: 22.8%; 6 mth: 15.2%

• Baseline: 7.9 ± 7.1, 6 mth: 6.1 ± 5.3, p < 0.001h
• Predictor of HbA1c change: CDI change score, p < 0.001i
• Three-way interaction between CDI change after 6 mth, BGMF at
baseline and HbA1c at baseline: p < 0.01

Ingerski et al. (2010) 261 CDI –

(CSS)
• Correlation
- With BGMF: r = −0.16, p < 0.001
• With CES-D: r = 0.27, p < 0.001
• G1j vs. G4j : fewer depressive symptoms: OR = 1.11, 95% CI
[1.001; 1.23], p < 0.05k
• G3j vs. G4j : fewer depressive symptoms: OR = 1.13, 95% CI [1.06;
1.20], p < 0.001k
Table 4 (Continued)

Reference (study type) No. (n) Depressive symptoms Anxiety symptoms

Khan et al. (2013) (CSS) 86 CDI > 13 –

• n = 29 (33.7%) vs. not depressed n = 57 (66.2%)

• Groups by age (yrs): 7–11.5: n = 9 (31%); >11.5–15: n = 20 (69%)
• Groups by duration of diabetes (yrs): 1–3: n = 4 (13.7%); >3–8:
n = 9 (31%); >8: n = 16 (55.1%)
• Groups by glycemic control/HbA1c (%): HbA1c < 8: n = 4 (13.7%);

B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

HbA1c 8–10: n = 4 (13.7%); HbA1c > 10: n = 21 (72.4%)

Kristensen et al. T1D: 786 BDI-Y BAI-Y

(2014)l (CSS) CG: 3.200m
11-14 yrs: • Subgroup 11–14 yrs • Subgroup 11–14 yrs
T1D Girls: 169 - Slightly elevated BDI-Y Scores: T1D Girls: 5.3%, CG Girls: 9.4%, - Slightly elevated BAI-Y Scores: T1D Girls: 4.2%, CG Girls: 15.8%,
CG Girls: 277 p = 0.12; T1D Boys: 3.6%, CG Boys: 8.6%, p = 0.04 p = 0.0002; T1D Boys: 6.0%, CG Boys: 15.6%, p = 0.002
T1D Boys: 167 • Moderately elevated BDI-Y Scores: T1D Girls: 6.5%, CG Girls: • Moderately elevated BAI-Y Scores: T1D Girls: 5.4%, CG Girls:
CG Boys: 292 13.0%, p = 0.03; T1D Boys: 6.6%, CG Boys: 14.4%, p = 0.01 11.0%, p = 0.04; T1D Boys: 6.0%, CG Boys: 10.2%, p = 0.12
• Extremely elevated BDI-Y Scores: T1D Girls: 5.3%, CG Girls: 5.4%, • Highly elevated BAI-Y Scores: T1D Girls: 3.6%, CG Girls: 5.1%,
p = 0.97; T1D Boys: 6.0%, CG Boys: 4.8%, p = 0.58 p = 0.45; T1D Boys: 3.0%, CG Boys: 5.8%, p = 0.17
• Correlation (n = 786) with HbA1c: r = 0.16, p < 0.01 • Correlation (n = 786) with HbA1c: r = 0.08, p < 0.05
• Correlation of BDI-Y with BAI-Y: r = 0.83, p < 0.01

McGrady and Hood 144 CDI –

(2010) (CS)
• Baseline: 23% with CDI ≥13
• Correlation at baseline
- With HbA1c: r = 0.26n , p < 0.01
• With BGMF: p < 0.001
• HbA1c with subscales: negative mood: p < 0.01, interpersonal
problems: p < 0.05, ineffectiveness: p < 0.0001
• BGMF with subscales: negative mood: p < 0.01, interpersonal
problems: p < 0.05, ineffectiveness: p < 0.01, anhedonia: p < 0.01,
negative self-esteem: p < 0.01
• Correlation at 6 mth
- With BGMF: p < 0.05
• BGMF with subscales: negative mood, ineffectiveness, negative
self-esteem: p < 0.05, respectively

79
 80
Table 4 (Continued)

Reference (study type) No. (n) Depressive symptoms Anxiety symptoms

McGrady et al. (2009) 276 CDI –

(CSS)
• Associations
- Lower levels of BGMF with more depressive symptoms: p = 0.02
• Higher HbA1c with more depressive symptoms: p = 0.05

B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

• BGMF mediated and accounted for 37.5% of the depressive
symptoms-glycemic control link, p < 0.05

Nardi et al. (2008) 140 • Correlation of CBCL subscale anxious/depressed with longer
(CCS) T1D: 70 disease duration: r = 0.20, p < 0.05
CG: 70

ACDI: Arabic Children’s Depression Inventory; BAI-Y: Anxiety subscale of The Beck’s Youth Inventories; BDI-Y: Depression subscale of The Beck’s Youth Inventories; BGMF: blood glucose monitoring frequency; CBCL: Rescorla’s
Child Behaviour Checklist; CCS: case-control study; CDI: Children’s Depression Inventory; CES-D: The Center for Epidemiologic Studies-Depression scale; CG: control group; CI: confidence interval; CS: cohort study; CSII:
continuous subcutaneous insulin infusion; CSS: cross-sectional study; DSS: diabetes-specific stressful events; G: group; GS: general stressful events; mth: month(s); OR: odds ratio; SCI: Self-Care Inventory; STAI(C): State-Trait
Anxiety Inventory (for Children); T1D: type 1 diabetes mellitus group; yrs: years.
Data given as mean ± standard deviation, otherwise stated.
a
Completed by caregivers.
b
Reported by adolescents.
c
STAIC does not provide clinical cutoffs to denote elevated levels of anxiety suggestive of further evaluation according to Herzer and Hood (2010).
d
Completed by patients and caregivers.
e
Lower BGMF associated with older age (p < 0.0001), lower caregiver educational attainment (p = 0.008), participation at the Midwestern site (p = 0.01), insulin delivery via injections versus CSII (p < 0.001), higher state anxiety
(p = 0.03); higher HbA1c associated with single-caregiver marital status (p = 0.01), participation at the Northeastern site (p = 0.04), longer diabetes duration (p = 0.02), less BGMF (p < 0.0001), higher STAI (p = 0.002).
f
Adjusted for all measured demographic and medical data.
g
Without inclusion of mental health visit frequency into the model.
h
CDI raw scores.
i
Adjusted for age, gender, ethnicity, diabetes duration, mode of insulin delivery, caregiver education level and marital status, and family insurance status.
j
Glycemic control and QoL groups (G1 (n = 27): HbA1c < 7.5, QoL ≥ 75.0; G2 (n = 16): HbA1c < 7.5, QoL < 75.0; G3 (n = 108): HbA1c ≥ 7.5, QoL ≥ 75.0; G4 (n = 110): HbA1c ≥ 7.5, QoL < 75.0).
k
Adjusted for adolescent age, disease duration, BGMF, gender, ethnicity, mode of insulin delivery, family insurance status, caregiver marital status and educational level, adolescent depressive symptoms and negative affect
around BGMF, caregiver depressive symptoms, and caregiver-reported family conflict.
l
Patients were stratified by sex and age (1–4 yrs, 8–10 yrs, 11–14 yrs, 15–16 yrs, 17 yrs). For further details see Kristensen et al. (2014).
m
Data from a Danish registry.
n
Methods for bivariate correlation not stated, probably use of Pearson’s correlation coefficient.
 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84
Fig. 2. Forest plot prevalence of depressive symptoms (%) measured by CDI.
Guo et al. (2013) used the DSRS to identify symptoms of depres-
sion and showed a low positive correlation with the HbA1c without
statistical significance. Kristensen et al. (2014) found a correlation
between depressive symptoms (BDI-Y) and HbA1c in the entire
type 1 diabetes group which was low but statistically significant.
Baucom et al. (2015) used the CES-D and described low but statis-
tically significant multiple correlations of depression with HbA1c,
general stress symptoms (GS), and diabetes-specific stress symp-
toms (DSS): greater depressive symptoms and more severe mean
DSS were significantly associated with higher HbA1c levels, but
more severe mean GS was not significantly associated with HbA1c.
Negative correlations of CDI and BGMF were reported to be statisti-
cally significant (Hilliard et al., 2011; Ingerski et al., 2010; McGrady
and Hood, 2010) but low (Hilliard et al., 2011; Ingerski et al., 2010).
When using the CES-D there was a negative correlation between
depression and BGMF (Hilliard et al., 2011).
In multivariate analyses McGrady and Hood (2010) and
McGrady et al. (2009) showed that lower BGMF and higher HbA1c
were separately and significantly correlated with higher CDI scores,
whereas Herzer and Hood (2010) could not find a significant
association of depression (CDI) with BGMF or HbA1c. A complex
interaction between CDI change, BGMF and HbA1c at baseline was
found by Hood et al. (2011): an increase of 1 point on CDI from
baseline to six months was associated with a 0.11% rise in HbA1c.
In the study of Hilliard et al. (2011), CDI was a significant predictor
of 12 months BGMF when calculating the model without men-
tal health visit frequency; for every 20 point rise in CDI scores,
BGMF decreased by one check per day. Khan et al. (2013) pub-
lished descriptive analyses for subgroups of patients describing the
interaction of age, duration of diabetes and glycemic control with
symptoms of depression.
Looking at symptoms of anxiety, correlations of STAI(C) with
HbA1c and BGMF were low but statistically significant (Herzer et al.,
2011; Herzer and Hood, 2010; Hilliard et al., 2011) (Table 4). Hilliard
et al. (2011) showed that STAI(C) scores were significant predictors
of HbA1c, and a 14-point increase in anxiety scores was associ-
ated with a clinically meaningful rise of 1% in HbA1c. Using the
CBCL, Abdul-Rasoul et al. (2010) found a higher frequency of bor-
derline (T-scores 67–70) and symptomatic anxiety (T-score above
70) in patients with an HbA1c > 8.5% (>69 mmol/mol) (p < 0.001 and
p < 0.01). Using the BAI-Y, Kristensen et al. (2014) demonstrated a
low but statistically significant positive correlation between anxi-
ety symptoms and HbA1c.
4. Discussion
This is a systematic review of recent studies investigating the
association between type 1 diabetes and the frequency of symp-
toms of depression and anxiety and their impact on diabetes
management and glycemic control.
81
Using the CDI, the prevalence of depressive symptoms ranged
from 17.2 to 63%. There were correlations between symptom lev-
els and glycemic control as well as three-way interactions between
HbA1c, BGMF or DSS and depression. Symptoms of anxiety were
reported for up to 32% of patients. A negative impact on glycemic
control was demonstrated. In previous meta-analyses (Johnson
et al., 2013; Reynolds and Helgeson, 2011), the prevalence of symp-
toms of depression in youth with type 1 diabetes was lower than in
our analysis or the difference was inconclusive if compared to nor-
mal controls (Reynolds and Helgeson, 2011). This is an important
finding as it contradicts a previous statement that the prevalence of
symptoms of depression and anxiety in youth with type 1 diabetes
has decreased in comparison to control groups. Looking at different
age groups, Baechle et al. (2015) identified at least one symptom of
depression in 43% of female and 33% of male 18–21 year old patients
with type 1 diabetes in a nationwide study. Comorbid depression is
present in 12% of adults with type 1 diabetes (Roy and Lloyd, 2012).
The prevalence of anxiety in type 1 diabetes is less well analyzed.
Cameron et al. (2007) and Kovacs et al. (1997) reported a 10-year
overall point-prevalence of psychiatric disorders in diabetic youth
of 37% and 47%. An increased risk of psychological comorbidities
has been attributed to adult chronic health conditions in general:
heart disease, stroke, asthma, arthritis and osteoporosis are asso-
ciated with a prevalence of depression ranging from 20 to 80%,
anxiety is present in 10–70% (Clarke and Currie, 2009; Overman
et al., 2012). There are no comparable detailed analyses for children
and adolescents.
In the context of screening for psychological comorbidity accu-
rate evaluation of symptoms is essential, as initial symptoms are
often first indicators of an evolving pathology (Raymond, 2015).
Prospective studies in this area are underway (Eilander et al., 2015).
Diabetes societies have recommended screening for psychosocial
pathology and promote an integrated health care model from diag-
nosis of type 1 diabetes (American Diabetes Association, 2014;
Delamater et al., 2014). This approach has a potential for significant
cost-savings in the health sector (Egede et al., 2002): early initia-
tion of support potentially reduces non-adherence and thus further
diabetes-related complications and can address and prevent sui-
cidal ideation and even suicide (American Diabetes Association,
2014; Cameron et al., 2007; Delamater et al., 2014).
4.1. Strengths and limitations
It has to be considered that the quality assessment with EPHPP
(Thomas et al., 2004) was “weak” for all studies included. Main
reasons were selection bias, low evidence level of study design,
and further bias caused by uncontrolled confounding variables. To
address this, for our data analysis confounding factors were cho-
sen according to Bot et al. (2013). This is why some factors were
not included for the nationwide comparison of the current review.
Examples are single factors with differing social importance like the
82 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84

caregiver’s marital status that might be less relevant for children’s
psychological well-being in liberal compared to conservative soci-
eties. Without doubt, the family situation plays a central role in the
management of juvenile type 1 diabetes (Delamater et al., 2014).
Nevertheless, inconsistent definitions of variables did not permit
a special consideration. Interestingly neither bodyweight nor BMI
were evaluated or discussed by any of the authors as potential
confounding factors and could not be included.
None of the studies used psychometric interviews to establish
a diagnosis of depression or anxiety. Despite of the frequent use
of psychometric measurement tools for the evaluation of anxiety
and depression in children, references of studies clearly confirm-
ing their validity and reliability were in most cases neither stated
by the authors nor otherwise procurable. However, this might be
due to poor reporting quality or publication bias rather than an
aspect of study quality. Therefore “data collection methods” were
assessed as “strong” although “unclear” might be more appropriate.
The prevalence of depressive symptoms could even be under-
estimated due to selection bias or more precisely healthy user
bias.
We went ahead with data pooling to improve the precision of
our analysis by aggregating the results of the included studies.
As we identified national and mostly single center studies homo-
geneity of the populations could not be guaranteed. We decided
to conduct a meta-analysis based on the following considerations:
1) the outcome was assessed by the same measurement tool (CDI)
throughout all studies included, 2) a cross-sectional study design
was used in most cases, 3) the outcome parameter cannot be
regarded predominantly as safety or efficacy data rather than an
aspect of QoL, therefore pooling is appropriate because no imme-
diate life threatening effect will be obscured (however, it needs to
be kept in mind that depressive symptoms may lead to suicidal
thoughts with potentially lethal consequences); 4) data pooling is
driven by the specific need to call attention to depressive symp-
toms in youth with type 1 diabetes due to the rising incidence of
the disease and relevance for Public Health, 5) an underestimation
of prevalence can be assumed because of selection bias, precisely
healthy user bias, and thus, data pooling will not increase the risk
of bias, 6) already existing evidence should be used first before ini-
tiating new studies in a vulnerable population such as children and
adolescents with a chronic condition and 7) the methods used are
described in detail and thus made completely transparent and open
for discussion.
Importantly our analysis could not assess whether the risk of
youth with type 1 diabetes to manifest symptoms of depression
and anxiety is different from youth not affected by type 1 diabetes
or affected by other chronic conditions as the included studies did
not provide these data. Only two studies (Abdul-Rasoul et al., 2010;
Kristensen et al., 2014) included control groups.
5. Conclusions
We confirmed a high prevalence of symptoms of depression and
anxiety in youth with type 1 diabetes and their negative impact
on diabetes management and glycemic control. We acknowledge
that our results cannot be generalized. However, their information
is vital because of the rising incidence of type 1 diabetes and
its relevance for Public Health in combination with an increased
prevalence of both, depression and anxiety. In our opinion these
findings support the need for regular psychosocial assessment
from diagnosis of type 1 diabetes. Given the vulnerability of
youth, well designed future prospective studies and randomized
controlled trials are needed to further explore the interaction of
symptoms of depression and anxiety with diabetes management
and glycemic control and to develop evidence-based screening
and treatment models.
Contribution statement: BB, HH and JTM established the inclu-
sion and exclusion criteria. HH and JTM searched the databases and
screened titles, abstracts and fulltext publications according to the
inclusion criteria. For the update, BB and LK searched the databases
and screened titles, abstracts and full-text publications according
to the inclusion criteria. BB, HH, LK, BL and JTM extracted the data.
BB and LK conducted the meta-analysis. BB and AS conceptualized
and wrote the draft of the manuscript. All authors contributed to
the critical revision of the article and approved the final manuscript
for publication.BB is guarantor of the work.

Table A1
Search strategy (21.03.2014, update 18.05.2015).

Searches Terms/Thesaurus (for Medline/Embase via Elsevier, The Cochrane Library and PsychINFO via Ovid)

1 Type 1 diabetes
2 Children and Adolescents
3 Behavior disorders
4 Mood disorders
5 Psychological comorbidities psych* NEAR/1 comorbidit*; psych* NEAR/1 begleiterkrankung*; psych* NEAR/1 komorbidität*;
6 Inadequate metabolic
adjustment
Query
’diabetes mellitus’; diabetes NEAR/1 ‘type 1’; diabetes NEAR/1 ‘typ 1’; ‘insulin dependent diabetes’; DM; diabet*; zuckerkrankheit;
E10-E14;
EMTREE: ‘insulin dependent diabetes mellitus’/syn
MeSH: ‘diabetes mellitus, type 1’ explode
SH: diabetes mellitus.sh.
child*; adolescen*; youth*; pediatr*; juvenile; infan*; newborn*; kind*; jugendlich*; neugeboren*; säugling*;
EMTREE: ‘childhood’/syn; ‘adolescence’/syn; ‘newborn period’/syn
MeSH: ‘child’ explode; ‘infant’ explode; ‘adolescent’ explode
behavior*; behaviour*; verhalten*;
EMTREE: ‘behavior disorder’/syn
MeSH: ‘adolescent behavior’ explode; ‘child behavior’ explode
SH: behavior.sh.
affective NEAR/1 disorder*; affective NEAR/1 disturbance*; mood NEAR/1 disorder*; depress*; bipolar*; affektive* NEAR/1 störung*;
EMTREE: ‘mood disorder’/syn
MeSH: ‘mood disorders’ explode
SH: affective disorders.sh.; affective disturbances.sh.
EMTREE: (psych* AND ‘comorbidity’/syn)
MeSH: ((’mental disorders’ explode OR psych*.mp.) AND ‘comorbidity’ explode)
glucose NEAR/1 metabol*; dysglyc*; hypoglyc*; hyperglyc*; ‘blood glucose level’; HbA1c; ‘glycosylated hemoglobin’;
glukosestoffwechsel; hypoglykämie*; hyperglykämie*; stoffwechsel*; dysglykämie*; blutzucker; blutglukose; ‘glykosyliertes
hämoglobin’
EMTREE: ‘glucose blood level’/syn; ‘hypoglycemia’/syn; ‘hyperglycemia’/syn
MeSH: ‘blood glucose’ explode; ‘hypoglycemia’ explode; ‘hyperglycemia’ explode; ‘metabolism’ explode
1 AND 2 AND (3 OR 4 OR 5) AND 6

E10-E14: ICD-10 classification; MeSH: Medical Subject Headings; SH: subheading; syn: synonym.
 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84 83

Table A2
Quality assessment according to EPHPP assessment tool for quantitative studies (Thomas et al., 2004).

Reference (study type) Selection Bias Design Confounders Blinding Data collection Withdrawals Total
methods and drop-outs

Abdul-Rasoul et al. (2010) (CSS) WEAK WEAK WEAK WEAK STRONG Not applicable WEAK
Baucom et al. (2015) (CS) WEAK MODERATE WEAK WEAK STRONG STRONG WEAK
De Wit and Snoek (2011) (CSS) WEAK WEAK WEAK WEAK STRONG Not applicable WEAK
Guo et al. (2013) (CSS) WEAK WEAK WEAK WEAK STRONG Not applicable WEAK
Herzer et al. (2011) (CS) WEAK MODERATE MODERATE WEAK STRONG STRONG WEAK
Herzer et al. (2010) (CSS) WEAK WEAK MODERATE WEAK STRONG Not applicable WEAK
Hilliard et al. (2011) (CS) WEAK MODERATE MODERATE WEAK STRONG STRONG WEAK
Hood et al. (2011) (CS) WEAK MODERATE MODERATE WEAK STRONG MODERATE WEAK
Ingerski et al. (2010) (CSS) WEAK WEAK MODERATE WEAK STRONG Not applicable WEAK
Khan et al. (2013) (CSS) WEAK WEAK Not applicable WEAK STRONG Not applicable WEAK
Kristensen et al., 2015 (CSS) STRONG WEAK WEAK WEAK STRONG Not applicable WEAK
McGrady et al. (2010) (CS) WEAK MODERATE WEAK WEAK STRONG STRONG WEAK
McGrady et al. (2009) (CSS) WEAK WEAK MODERATE WEAK STRONG Not applicable WEAK
Nardi et al. (2008) (CCS) WEAK MODERATE WEAK WEAK STRONG Not applicable WEAK

STRONG = No WEAK ratings, MODERATE = One WEAK rating, WEAK = ≥ 2 WEAK ratings.
CCS: case-control study; CS: cohort study; CSS: cross-sectional study.

Conflicts of interest
None declared.
Funding
There was no dedicated funding.
Acknowledgement
None.
Appendix A.
See Tables A1 and A2.
References
Abdul-Rasoul, M., Hajiyya, J., Alshawaf, F., Abdal, Y., Al-Azmni, A., Al-Mahdi, M.,
2010. Psychological status of children with type 1 diabetes in Kuwait. Kuwait
Med. J. 42, 139–145.
American Diabetes Association (ADA), 2014. Executive summary: standards of
medical care in diabetes—2014. Diabetes Care 37, S5–S13.
Baechle, C., Lange, K., Stahl-Pehe, A., Castillo, K., Holl, R.W., Giani, G., Rosenbauer, J.,
2015. Associations between HbA1c and depressive symptoms in young adults
with early-onset type 1 diabetes. Psychoneuroendocrinology 55, 48–58.
Baucom, K.J., Queen, T.L., Wiebe, D.J., Turner, S.L., Wolfe, K.L., Godbey, E.I.,
Fortenberry, K.T., Mansfield, J.H., Berg, C.A., 2015. Depressive symptoms, daily
stress, and adherence in late adolescents with type 1 diabetes. Health Psychol.
34, 522–530.
Bot, M., Pouwer, F., de Jonge, P., Tack, C.J., Geelhoed-Duijvestijn, P.H., Snoek, F.J.,
2013. Differential associations between depressive symptoms and glycaemic
control in outpatients with diabetes. Diabet. Med. 30, e115–e122.
Braun, M., Tomasik, B., Wrona, E., Fendler, W., Jarosz-Chobot, P., Szadkowska, A.,
Zmysłowska, A., Wilson, J., Mlynarski, W., 2016. The stricter the better? The
relationship between targeted HbA1c values and metabolic control of pediatric
type 1 diabetes mellitus. J. Diabetes Res., http://dx.doi.org/10.1155/2016/
5490258.
Cameron, F.J., Northam, E.A., Ambler, G.R., Daneman, D., 2007. Routine
psychological screening in youth with type 1 diabetes and their parents: a
notion whose time has come? Diabetes Care 30, 2716–2724.
Clarke, D.M., Currie, K.C., 2009. Depression, anxiety and their relationship with
chronic diseases: a review of the epidemiology: risk and treatment evidence.
Med. J. Aust. 190, S54–S60.
De Wit, M., Snoek, F.J., 2011. Depressive symptoms and unmet psychological needs
of Dutch youth with type 1 diabetes: results of a web-survey. Pediatr. Diabetes
12, 172–176.
Delamater, A.M., de Wit, M., McDarby, V., Malik, J., Acerini, C.L., International
Society for Pediatric and Adolescent Diabetes, 2014. ISPAD Clinical practice
consensus Guidelines 2014. Psychological care of children and adolescents
with type 1 diabetes. Pediatr. Diabetes 15, 232–244.
Deutsche Diabetes Gesellschaft (DDG), 2009. Diagnostik, Therapie und
Verlaufskontrolle des Diabetes mellitus im Kindes- und Jugendalter (accessed
22.04.16.) http://www.deutsche-diabetes-gesellschaft.de/fileadmin/
Redakteur/Leitlinien/Evidenzbasierte Leitlinien/EBL Kindesalter 2010. pdf.
Deutsche Diabetes Gesellschaft (DDG), 2015. Diagnostik, Therapie und
Verlaufskontrolle des Diabetes mellitus im Kindes- und Jugendalter.
S3-Leitlinie der DDG und AGPD 2015 (accessed 15.02.16.) http://www.awmf.
org/uploads/tx szleitlinien/057-016l S3 Diabetes mellitus Kinder
Jugendliche 2016-02. pdf.
Diabetes.co.uk, 2016. Guide to HbA1c (accessed 12.03.16.) http://www.diabetes.co.
uk/what-is-hba1c.html.
Ducat, L., Rubenstein, A., Philipson, L.H., Anderson, B.J., 2015. A review of the
mental health issues of diabetes conference. Diabetes Care 38, 333–338.
Egede, L.E., Zheng, D., Simpson, K., 2002. Comorbid depression is associated with
increased health care use and expenditures in individuals with diabetes.
Diabetes Care 25, 464–470.
Eilander, M.M., de Wit, M., Rotteveel, J., Aanstoot, H.J., Waarde, W.M., Houdijk, E.C.,
Luman, M., Nuboer, R., Oosterlaan, J., Winterdijk, P., Snoek, F.J., 2015. Diabetes
IN develOpment (DINO): the bio-psychosocial, family functioning and parental
well-being of youth with type 1 diabetes: a longitudinal cohort study design.
BMC Pediatr. 15, 82, http://dx.doi.org/10.1186/s12887-015-0400-1.
Garrison, C.Z., Addy, C.L., Jackson, K.L., McKeown, R.E., Waller, J.L., 1992. Major
depressive disorder and dysthymia in young adolescents. Am. J. Epidemiol.
135, 792–802.
Goldston, D.B., Kelley, A.E., Reboussin, D.M., Daniel, S.S., Smith, J.A., Schwartz, R.P.,
Lorentz, W., Hill, C., 1997. Suicidal ideation and behavior and noncompliance
with the medical regimen among diabetic adolescents. J. Am. Acad. Child.
Adolesc. Psychiatr. 36, 1528–1536.
Grey, M., Whittemore, R., Tamborlane, W., 2002. Depression in type 1 diabetes in
children: natural history and correlates. J. Psychosom. Res. 53, 907–911.
Guo, J., Whittemore, R., Grey, M., Wang, J., Zhou, Z.G., He, G.P., 2013. Diabetes
self-management, depressive symptoms, quality of life and metabolic control
in youth with type 1 diabetes in China. J. Clin. Nurs. 22, 69–79.
Herzer, M., Hood, K.K., 2010. Anxiety symptoms in adolescents with type 1
diabetes: association with blood glucose monitoring and glycemic control. J.
Pediatr. Psychol. 35, 415–425.
Herzer, M., Vesco, A., Ingerski, L.M., Dolan, L.M., Hood, K.K., 2011. Explaining the
family conflict-glycemic control link through psychological variables in
adolescents with type 1 diabetes. J. Behav. Med. 34, 268–274.
Higgins, J.P.T., Green, S., 2011. Cochrane Handbook for Systematic Reviews of
Interventions 5.1.0. Oxford (updated March 2011).
Hilliard, M.E., Herzer, M., Dolan, L.M., Hood, K.K., 2011. Psychological screening in
adolescents with type 1 diabetes predicts outcomes one year later. Diabetes
Res. Clin. Pract. 94, 39–44.
Hood, K.K., Huestis, S., Maher, A., Butler, D., Volkening, L., Laffel, L.M., 2006.
Depressive symptoms in children and adolescents with type 1 diabetes:
association with diabetes-specific characteristics. Diabetes Care 29,
1389–1391.
Hood, K.K., Rausch, J.R., Dolan, L.M., 2011. Depressive symptoms predict change in
glycemic control in adolescents with type 1 diabetes: rates, magnitude, and
moderators change. Pediatr. Diabetes 12, 718–723.
Hood, K.K., Lawrence, J.M., Anderson, A., Bell, R., Dabelea, D., Daniels, S., Rodriguez,
B., Dolan, L.M., SEARCH for Diabetes in Youth Study Group, 2012. Metabolic
and inflammatory links to depression in youth with diabetes. Diabetes Care 35,
2443–2446.
Ingerski, L.M., Laffel, L., Drotar, D., Repaske, D., Hood, K.K., 2010. Correlates of
glycemic control and quality of life outcomes in adolescents with type 1
diabetes. Pediatr. Diabetes 11, 563–571.
Johnson, B., Eiser, C., Young, V., Brierley, S., Heller, S., 2013. Prevalence of
depression among young people with Type 1 diabetes: a systematic review.
Diabet. Med. 30, 199–208.
Khan, W.I., Rabbani, M.W., Afzal, E., Adnan, M., 2013. Psychological screening in
children with diabetes mellitus type-I at the children’s hospital and the
Institute of Child Health, Multan. J. Pak. Med. Assoc. 63, 1520–1522.
84 B. Buchberger et al. / Psychoneuroendocrinology 70 (2016) 70–84
Kirisci, L., Clark, D.B., Moss, H.B., 1996. Reliability and validity of the State-Trait
Anxiety Inventory for Children in an adolescent sample: confirmatory factor
analysis and item response theory. J. Child Adolesc. Subst. Abuse 5,
57–69.
Kongkaew, C., Jampachaisri, K., Chaturongkul, C.A., Scholfield, C.N., 2014.
Depression and adherence to treatment in diabetic children and adolescents: a
systematic review and meta-analysis of observational studies. Eur. J. Pediatr.
173, 203–212.
Kovacs, M., Ho, V., Pollock, M.H., 1995. Criterion and predictive validity of the
diagnosis of adjustment disorder: a prospective study of youths with
new-onset insulin-dependent diabetes mellitus. Am. J. Psychiatry 152,
523–528.
Kovacs, M., Goldston, D., Obrosky, D.S., Bonar, L.K., 1997. Psychiatric disorders in
youth with IDDM: Rates and risk factors. Diabetes Care 20, 36–44.
Kovacs, M., 1980. Rating scales used to assess depression in school-aged children.
Acta Paediopsychiatr. 46, 305–315.
Kovacs, M., 2003. The Children’s Depression Inventory (CDI). Psychopharmacol.
Bull. 21, 995–998.
Kristensen, L.J., Birkebaek, N.H., Mose, A.H., Hohwü, L., Thastum, M., 2014.
Symptoms of emotional, behavioral, and social difficulties in the danish
population of children and adolescents with type 1 diabetes–results of a
national survey. PLoS One 9, e97543.
Lernmark, B., Persson, B., Fisher, L., Rydelius, P.A., 1999. Symptoms of depression
are important to psychological adaptation and metabolic control in children
with diabetes mellitus. Diabet. Med. 16, 14–22.
McGrady, M.E., Hood, K.K., 2010. Depressive symptoms in adolescents with type 1
diabetes: associations with longitudinal outcomes. Diabetes Res. Clin. Pract.
88, e35–e37.
McGrady, M.E., Laffel, L., Drotar, D., Repaske, D., Hood, K.K., 2009. Depressive
symptoms and glycemic control in adolescents with type 1 diabetes:
meditational role of blood glucose monitoring. Diabetes Care 32, 804–806.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., The PRISMA Group, 2009. Preferred
reporting items for systematic reviews and meta- analyses: the PRISMA
statement. PLoS Med. 6, e1000097, http://dx.doi.org/10.1371/journal.
pmed1000097.
Nardi, L., Zucchini, S., D’Alberton, F., Salardi, S., Maltoni, G., Bisacchi, N., Elleri, D.,
Cicognani, A., 2008. Quality of life, psychological adjustment and metabolic
control in youths with type 1 diabetes: a study with self- and parent-report
questionnaires. Pediatr. Diabetes 9, 496–503.
Overman, L.C., Bossema, E.R., van Middendorp, H., Wijngaards-de Meij, L.,
Verstappen, S.M., Bulder, M., Jacobs, J.W., Bijlsma, J.W., Geenen, R., 2012. The
prospective association between psychological distress and disease activity in
rheumatoid arthritis: a multilevel regression analysis. Ann. Rheum. Dis. 71,
192–197.
Raymond, J., 2015. Updates in behavioural and psychosocial literature in
adolescents with type 1 diabetes. Curr. Opin. Endocrinol. Diabetes Obes. 22,
265–269.
Rewers, A., Chase, H.P., Mackenzie, T., Walravens, P., Roback, M., Rewers, M.,
Hamman, R.F., Klingensmith, G., 2002. Predictors of acute complications in
children with type 1 diabetes. JAMA 287, 2511–2518.
Reynolds, K.A., Helgeson, V.S., 2011. Children with diabetes compared to peers:
depressed? Distressed? A meta-analytic review. Ann. Behav. Med. 42,
29–41.Roy, T., Lloyd, C.E., 2012. Epidemiology of depression and diabetes: a
systematic review. J. Affect. Disord. 142, S8–S21.
Stewart, S.M., Rao, U., Emslie, G.J., Klein, D., White, P.C., 2005. Depressive
symptoms predict hospitalization for adolescents with type 1 diabetes
mellitus. Pediatrics 115, 1315–1319.
Stroup, D.F., Berlin, J.A., Morton, S.C., Olkin, I., Williamson, G.D., Rennie, D., Moher,
D., Becker, B.J., Sipe, T.A., Thacker, S.B., 2000. Meta-analysis of observational
studies in epidemiology: a proposal for reporting. JAMA 283, 2008–2012.
Thomas, B.H., Ciliska, D., Dobbins, M., Micucci, S., 2004. A process for
systematically reviewing the literature: providing the research evidence for
public health nursing interventions. Worldviews Evid. Based. Nurs. 1, 176–184.