Can J Diabetes xxx (2015) 1e7
Contents lists available at ScienceDirect
Canadian Journal of Diabetes journal homepage: www.canadianjournalofdiabetes.com
Original Research
Depression and Risk for Diabetes: A Meta-Analysis Min Yu MD a, Xingliang Zhang MD b, Feng Lu MD a, Le Fang PhD a, * a b
Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, Zhejiang, China Hangzhou Center for Disease Control and Prevention, Hangzhou, Zhejiang, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 April 2014 Received in revised form 28 November 2014 Accepted 28 November 2014 Available online xxx
Objective: Many studies have reported the relationship between depression and diabetes, but the results have been inconsistent. Our aim was to conduct a systematic review through meta-analysis to assess the association of depression with the risk for developing diabetes. Methods: We retrieved the studies concerning depression and the risk for diabetes. Meta-analysis was applied to calculate the combined effect values and their 95% confidence intervals. The risk for publication bias was assessed by the Egger regression asymmetry test. Results: As many as 33 articles were included in the meta-analysis, for a total of 2 411 641 participants. The pooled relative risk for diabetes was 1.41 (95% CI, 1.25e1.59) for depression, and the combined relative risk for type 2 diabetes mellitus was 1.32 (95% CI, 1.18e1.47). Conclusions: Depressed people have a 41% increased risk for developing diabetes mellitus and a 32% increased risk for developing type 2 diabetes. The mechanisms underlying this relationship are still unclear and need further research. Ó 2015 Canadian Diabetes Association
Keywords: depression diabetes meta-analysis risk factors
r é s u m é Mots clés : dépression diabète méta-analyse facteurs de risque
Objectif : De nombreuses études ont rapporté le lien entre la dépression et le diabète, mais les résultats sont apparus contradictoires. Notre objectif était de mener une revue systématique par le recours à la méta-analyse pour évaluer le lien entre la dépression et le risque de développement du diabète. Méthodes : Nous avons extrait les études concernant la dépression et le risque de diabète. Nous avons eu recours à la méta-analyse pour calculer les valeurs de l’effet combiné et leurs intervalles de confiance à 95 %. Le risque de biais de publication a été évalué à l’aide du test de régression d’Egger. Résultats : La méta-analyse comportait 33 articles, soit un total de 2 411 641 participants. Lors de dépression, le risque relatif global du diabète était de 1,41 (IC à 95%, 1,25e1,59) et le risque relatif combiné de diabète sucré de type 2 était de 1,32 (IC à 95 %, 1,18e1,47). Conclusions : Les personnes dépressives montrent une augmentation du risque de développement du diabète sucré de 41 % et une augmentation du risque de développement du diabète de type 2 de 32 %. Les mécanismes sous-jacents à ce lien ne sont pas encore élucidés. D’autres recherches sont nécessaires. Ó 2015 Canadian Diabetes Association
Introduction Diabetes can damage the heart, blood vessels, eyes, kidneys and nerves, and 50% of people with diabetes die of cardiovascular disease (1). Diabetic retinopathy is an important cause of blindness, and 1% of blindness worldwide can be attributed to diabetes (2). Diabetes is also among the leading causes of kidney
* Address for correspondence: Le Fang, MD, Zhejiang Provincial Center for Disease Control and Prevention, 3399 Binsheng Road, Hangzhou, Zhejiang 310051, China. E-mail address:
[email protected]
failure (3). The overall risk for dying among people with diabetes is at least double the risk of their peers without diabetes (4). There are currently about 347 million people with diabetes worldwide (5). In 2004, an estimated 3.4 million people died from consequences of fasting high blood sugar, and a similar number of deaths has been estimated for 2010 (6). More than 80% of diabetes deaths occur in low- and middle-income countries (7). The World Health Organization (WHO) projects that diabetes will be the seventh leading cause of death in 2030 (3). The causes of diabetes are complex but are in large part due to rapid increases in overweight, obesity, physical inactivity, sedentary lifestyles and certain dietary behaviours, such as high fat intake
1499-2671/$ e see front matter Ó 2015 Canadian Diabetes Association http://dx.doi.org/10.1016/j.jcjd.2014.11.006
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
2
M. Yu et al. / Can J Diabetes xxx (2015) 1e7
(8). In addition to these standard risk factors, it has been suggested that depression can increase the risk for diabetes (9,10). Depressive disorders are among the most common of the psychiatric disorders; a recent survey of 38 states in the United States reported the overall prevalence of current depressive symptoms to be 8.7% and found a 15.7% lifetime prevalence rate of diagnosis of a depressive disorder by a doctor or healthcare provider (11). The nature of depression is such that sufferers experience dysphoric mood, loss of interest or pleasure, appetite and sleep disturbances and changes in energy levels. Thus, decreases in selfcare behaviour, such as decreased medication adherence, poor nutrition and lack of exercise, are often associated with depression (12). A number of studies have investigated the relationship between depression and onset of diabetes longitudinally and have shown inconsistent findings. Some report that depression is associated with an increased risk for developing diabetes, whereas other studies do not find a significant association. The aim of this study was to examine the relationship between depression and the risk for onset of diabetes by conducting a metaanalysis of studies published on this subject in the peer-reviewed literature.
Figure 1. Selection of studies for inclusion in meta-analysis.
Statistical analysis
Methods Literature searches Studies published in English and Chinese were comprehensively identified in this research. Studies in English were identified through PubMed, MEDLINE, Elsevier Science and Springer Link Cochrane databases from their earliest available dates to March 20, 2013. Chinese articles were screened through China National Knowledge Infrastructure, Database of Chinese Scientific and Technical Periodicals and China biology medical literature databases, which were searched in 1979, 1989, 1970, respectively, through March 21, 2013. The keywords diabetes mellitus or diabetes and depressive disorder or depression or dysthymic disorders or risk factors were used in combination to retrieve the relevant literatures in all these databases. Moreover, the references of all included studies were screened, as were reference lists from reviews and meta-analysis. This systematic review was planned, conducted and reported in adherence to Preferred Reporting Items for Systematic Reviews (PRISMA) guidelines for reporting metaanalyses (13). Eligibility criteria Studies were included in the meta-analysis if they met the following criteria: 1) the exposure of interest was depression; 2) the outcome of interest was diabetes; 3) it was a cross-sectional study, case-control study or cohort study; 4) relative risk (RR) or odds ratio (OR) estimates with 95% confidence intervals (CIs) (or data to calculate them) were reported. If data were duplicated in more than 1 study, we included the study with the largest number of cases. Data extraction The following data were extracted from each study: first author’s last name, publication year, country where the study was performed, study design, range of age, follow up time in years, method of depression assessment, method of diabetes assessment, diabetes type, relative risk and 95% CI (the one adjusted for the largest number of confounders), and adjustment for confounders. Data extraction was conducted independently by 2 authors (Zhang and Lu), with disagreements resolved by consensus.
Data were abstracted from all the studies that met our eligibility criteria. All statistical tests in this study were 2-tailed, and p values of 0.05 or less were considered significant. Statistical analysis was done using Stata, v. 9.2 (StataCorp, College Station, Texas, United States). Estimates of association with diabetes risk were evaluated by RRs and corresponding 95% CIs. Evaluation of meta-analysis results included a test of heterogeneity, sensitivity analyses and examination for bias. Heterogeneity among studies in meta-analysis was assessed by the Cochrane Q statistic, and p values less than 0.10 indicated significant heterogeneity (14). We also used the I2 statistic to quantify heterogeneity (15). Generally, I2 values less than 25% correspond to mild heterogeneity; values between 25% and 50% correspond to moderate heterogeneity; and values greater than 50% correspond to large heterogeneity among studies. If the data were heterogeneous, the random effect model was adopted (16), and if the data were homogeneous, the fixed effect model was applied. Sensitivity analyses were done to assess robustness and to examine the results of our meta-analyses for possible bias. Potential publication bias was assessed by using funnel plots of effect sizes vs. standard errors; the Egger regression asymmetry test was used to identify significant asymmetry (17). An analysis of influence was conducted; it describes how robust the pooled estimator is to the removal of individual studies. An individual study is suspected of excessive influence if the point estimate of its omitted analysis lies outside the 95% CI of the combined analysis. Results The detailed steps of our literature search are shown in Figure 1. The first approach yielded 6272 publications, which were screened by title and abstract or by a full-text review, if necessary, to identify 47 potentially relevant articles. Two articles were excluded because of a duplicate report based on the same study population, and 12 articles were excluded because of failure to meet the eligibility criteria (8). In the end, 33 publications met our eligibility criteria and were included in the meta-analysis. The 33 articles (10,18,49) were published between 1991 and 2012 (Table) and involved a total of 2 411 641 participants. Of those, 15 studies were conducted in North America (14 in the United States and 1 in Canada), 13 in Europe, 4 in Asia and 1 in Latin America. The 33 publications included 24 cohort studies, 2 nest case-control studies and 7 cross-sectional studies. Of the articles,
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
Table Characteristics of studies included in the meta-analysis Study
Country
Design
N
Follow up (years)
Palinkas 2004 [18] Almawi 2008 [19]
US Bahrain
Cohort Cross-sectional
971 8 50e89 275 Not applicable 49.9 10.9a 48.1 8.9b 1 026 625 Not applicable S18 65 381 10 50e75 58 056 5 S50
Cross-sectional Cohort Cohort
Eaton 1996 [23] Campayo 2010 [24] Kawakami 1999 [25] Arroyo 2004 [26] Saydah 2002 [27]
Cohort Cohort Cohort Cohort Cohort
1715 13 3521 5 2380 8 72 178 4 8870 9 11 615
US Spain Japan US US
S18 S55 18e53 45e72 32e86
Depression assessment
Diabetes assessment
Diabetes type
Relative risk (95% CI)
Adjustment for confoundersi
BDI DASS-21
Screening ADA, 1998 Screening WHO 1998
2 2
2.50 (1.29e4.87) 3.82 (1.43e10.25)
1, 2, 7, 8 None
ICD-9 MHI-5 ICD-9
Self-report Not specified 1.59 (1.54e1.65) National Diabetes Data Group criteria 2 1.17 (1.05e1.30) FPG 2 1.10 (1.02e1.20)
Interview: DIS GMS-AGECAT criteria Zung SDS MHI-5 CES-D
2 2.23 (0.90e5.55) Not specified 1.65 (1.02e2.66) 2 2.32 (1.06e5.08) Not specified 1.22 (1.00e1.50) 2 1.11 (0.79e1.56)
Not specified 1.46 (0.90e2.36) Not specified 2.29 (1.28e4.1) Not specified 0.9 (0.3e2.9)c 1.3 (0.4e3.6)d 2 1.23 (1.10e1.37)e 0.92 (0.84e1.00)f Not specified 0.73 (0.66e0.81) 2 0.98 (0.79e1.21) Not specified 1.63 (1.12e2.36) Not specified 1.53 (1.39e1.69) Not specified 3.52 (2.42e5.12)
6
48e67
VES
Self-report Doctor report Screening WHO 1981 National Diabetes Data group criteria Self-report, hospital diagnosis, Death certificates Screening ADA, 1998
2662 3 1000 10 1244 12
45e52 S65 45e65
CES-D PAS Questionnaires
Self-report, FPG Self-report Questionnaires
Depression medication use
Diabetes medication use
Self-report ICHPPC-2 CES-D Self-report Montogomery-Asberg Depression Rating Scale General Health Questionnaire
Self-report Diagnosis DCGP, 1999 Diabetes medication Self-report FPG
Golden 2004 [28]
US
Cohort
Everson-Rose 2004 [29] Atlantis 2010 [30] Mallon 2005 [31]
US Australia Sweden
Cohort Cohort Cohort
Brown 2005 [32]
Canada
Nest case control
Kessing 2004 [33] van den Akker 2004 [34] Carnethon 2007 [35] Chiena 2012 [36] Bhowmik 2012 [37]
Denmark Netherlands US Taiwan Bangladesh
Cohort Cohort Cohort Cross-sectional Cross-sectional
Kumari 2004 [38]
UK
Cohort
8320 11
35e55
Carnethon 2003 [10]
US
Cohort
6190 15.6
25e74
Stellato 2003 [39] Holt 2009 [40]
US UK
Cohort Cross-sectional
1156 9 40e70 2995 Not applicable 59e73
Engum 2007 [41]
Norway
Cohort
Palinkas 1991 [42] Kivimaki 2010 [43] Demakakos 2010 [44] Maty 2005 [45] Golden 2008 [46]
US Finland UK US US
Cross-sectional Nest case- control Cohort Cohort Cohort
Knol 2007 [47] Eriksson 2008 [48] Rajala 1997 [49]
Netherlands Cohort Sweden Cohort Finland Cross-sectional
92 677 Not applicable 20e95 92 597 68 004 4681 766 427 2293
20 Not reported 15 S20 8 S65 Not applicable S18 Not applicable S20
37 291 10
30e89
1585 Not applicable S50 5085 Not applicable 44.4 9.2 6111 4 S50 6148 34 17e94 5201 3 45e84 42 426 7 S18 5227 9 S35 734 Not applicable 55
2
1.31 (1.04e1.64)
1.03 (0.6e1.8)c 1.17 (0.8e1.7)d General Well-Being Depression Self-report, medical record diagnosis; Not specified 1.86 (1.27e2.71) subscale CES-D Self-report 2 3.09 (1.34e7.12) Hospital Anxiety and Self-report, OGTT Not specified 1.51 (0.47e4.84)c Depression Scale 3.89 (1.28e11.88)d Anxiety and Depression Index Self-report, FPG 1, 2 1.40 (1.16e1.69)g 0.80 (0.36e1.77)h BDI WHO criteria 1981 2 1.34 (0.78e2.31) Antidepressant used Diabetes medication 2 2.33 (1.74e3.12) CES-D Self-report, Diabetes medication 2 1.62 (1.15e2.29) HPL 18DI Self-report 2 1.08 (0.79e1.48) CES-D Self-report, FPG 2 1.21 (0.87e1.67) Antidepressant used questionnaire Zung SDS
Screening WHO1999
Diabetes medication OGTT OGTT
2
Not specified 1.06 (0.89e1.26) 2 1.24 (0.88e1.8) 2 1.00 (0.30e2.90)
None 1, 7, 8, 9, 12, 13 1, 2, 7, 12, 14, 16, 18, 20, 26 1, 2, 3, 7 1, 2, 4, 23 1 1, 7, 8, 9, 12, 13, 15 1, 2, 3, 7, 8 1, 2, 3, 6, 7, 8, 11, 12, 14, 16, 18, 19, 20, 21 1, 3, 6, 8, 10, 23, 24, 1, 2, 3, 5, 6, 7, 8, 12, 13, 17 1, 5, 7, 12, 13 2 None 1, 2 4, 7 1, 2, 3, 20 None None 1, 3, 4, 7, 8, 9, 12, 14, 22, 25 1, 2, 3, 7, 8, 12, 13 7, 15, 26 None 1, 2, 5, 6, 7, 8, 11, 12, 16, 17 1, 2 None 1, 2, 5, 7, 26 None 1, 2, 3, 6, 7, 8, 12, 13, 14, 21, 24 1, 2, 17 None None
3
ADA, American Diabetes Association; BDI, Beck depression inventory; CES-D, Center for Epidemiological Studies Depression Scale; DASS, Depression Anxiety Stress Scales; DCGP, Dutch College of General Practitioners; DIS, diagnostic interview schedule; FPG, fasting plasma glucose; GHQ-D, general health questionnaire depression; GMS-AGECAT, Geriatric Mental State B3 diagnostic schedule with application of the Automated Geriatric Examination for ComputerAssisted Taxonomy algorithm; GWB-DS, general well-being depressive symptomsl; HPL 18DI, Human Population Laboratory 18-term Depression Index; ICD-9, International Classification of Diseases-9; ICHPPC, international classification of health problems in primary care; MHI, mental health index; OGTT, oral glucose tolerance test; OR, odds ratio; PAS, Psychogeriatric Assessment Scales; RR, relative risk; SDS, self-rating depression scale; US, United States; VES, vital exhaustion scale; WHO, World Health Organization. a Age of patients with diabetes. b Age of healthy people. c RR for females. d RR for males. e OR for age 20e50. f OR for age S50. g RR for type 2 diabetes. h RR for type 1 diabetes. i 1, age; 2, sex; 3, race; 4, socioeconomic status; 5, marital status; 6, education; 7, BMI; 8, physical activity; 9, family history of diabetes; 10, waist circumference; 11, waist-to-hip ratio; 12, smoking; 13, alcohol consumption; 14, systolic blood pressure; 15, history of hypertension; 16, HDL cholesterol; 17, chronic medical conditions; 18, triglycerides; 19, fasting insulin; 20, fasting glucose; 21, caloric intake; 22, ECG abnormalities; 23, use of medication for depression; 24, study site; 25, length of follow up; 26, cardiovascular disease.
M. Yu et al. / Can J Diabetes xxx (2015) 1e7
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
Disdier-Flores 2010 [20] Puerto Rico Pan 2010 [21] US Nichols 2011 [22] US
Age
M. Yu et al. / Can J Diabetes xxx (2015) 1e7
web 4C=FPO
4
Figure 2. Pooled relative risk for depression in patients with diabetes.
20 discussed the relationship between depression and type 2 diabetes, but the type of diabetes was not specified in the remaining 13 articles. The forest plot shows the relative risk and 95% CI of each study and the pooled relative risk (Figure 2). The heterogeneity among studies was heterogeneous (I2¼90.8%), so the random effect model was adopted, and the pooled relative risk (95% CI) was 1.41 (1.25e1.59). When we stratified the analysis by the type of diabetes, the combined RR of diabetes 2 for depression was 1.32 (1.18e1.47) (Figure 3). In addition, stratifying by the type of study design, the pooled RRs in the cohort study was 1.28 (1.14e1.44), and the pooled ORs for case-control studies and cross-sectional studies were 1.57 (0.73e3.39) and 1.77 (1.49e2.11), respectively. In addition, stratifying by geographic region, the RRs were 1.28 (95% CI, 1.16e1.42) for studies conducted in the United States, 1.31 (95% CI, 1.03e1.67) for studies in Europe and 2.49 (95% CI, 1.39e4.44) for studies in Asia. To explore the heterogeneity among studies, we performed sensitivity analyses. A sensitivity analysis omitting 1 study at a time and calculating the pooled RRs for the remainder of the studies showed that 5 studiesdthose by Disdier-Flores et al (20), Kessing et al (33), Chiena et al (36), Bhowmik et al (37) and Kivimaki et al (43)dsubstantially influenced the pooled RRs. After excluding the 5 studies, there was no study heterogeneity (I2¼47.7%), and the RR was 1.18 (95% CI, 1.13e1.24). The results of publication bias in the studies evaluated using the Egger test indicated that there was a very low possibility of publication bias (p¼0.684), and the funnel plot showed the lowest possible publication bias.
Discussion The findings of this meta-analysis indicate that depression is associated with an increased risk for diabetes mellitus. Our results also found that there is a link between depression and type 2 diabetes mellitus. Moreover, we did not find a relationship between the incidence of diabetes and the time since the depressive score; the RR estimates did not vary with duration of follow up. We used a wide search strategy and hand-searched references in original articles in English and Chinese. Although we cannot rule out publication bias, we made every effort to explore published and unpublished papers, and the statistical test for publication bias was not significant. In addition, there were a large number of participants in the included studies, and the quality of included studies was good. Moreover, our results revealed that there was heterogeneity among studies, but it is likely to be related to the wide range of scales used to measure depression, the differing methods of assessment of diabetes and the differing baseline characteristics (for example, mean age and sex distribution) of the studies’ participants. In the past few years, 4 similar meta-analyses have reported depression and its relationship to diabetes mellitus (50e53). Cosgrove et al (50) found that the RR for type 2 diabetes was 1.25 (95% CI, 1.02e1.48), and Knol et al (53) also believed that depression could increase risk for type 2 diabetes; the combined RR was 1.37 (1.14e1.63). The article conducted by Mezuk et al (51) revealed that the RR for incident diabetes associated with baseline depression was 1.60 (1.37e1.88), and another publication (52) found that the RR for diabetes was 1.38 (95% CI, 1.23e1.55). Of these articles, 3
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
5
web 4C=FPO
M. Yu et al. / Can J Diabetes xxx (2015) 1e7
Figure 3. Combined relative risk for depression in patients with diabetes.
were published before 2008 and detected only the baseline depression and risk for type 2 diabetes. Since 2008, more than 10 studies relating to the relationship between depression and diabetes have been published in Canadian Journal of Diabetes. Moreover, another article published in 2013 seems not to include all the relevant studies. This study is, to the best of our knowledge, the largest and most comprehensive assessment of literature concerning depression and risk for diabetes performed so far. It seems that there is a true association between depression and the subsequent development of diabetes mellitus. There are several possible explanations for this finding. First, depression may be a direct cause of diabetes. Depression may influence the activity of several neurotransmitters, including serotonin, norepinephrine, dopamine, acetylcholine and c-aminobutyric acid (54). Depressive symptoms are also associated with increased inflammation, and inflammatory markers are accepted as risk factors for diabetes (55). The hypothalamo-pituitary adrenal axis is abnormal during a major depressive episode due to increased hypothalamic levels of corticotropic-releasing hormone. As a consequence, there are raised circulating plasma corticosteroid hormone levels, and a blunted dexamethasone suppression test has been shown in subjects with major depressive disorder. Bjorntorp has suggested that the link between stress and diabetes is central adiposity and proposes that stress causes a more central distribution of fat which, in turn, causes diabetes (56). The Bjorntorp hypothesis (57) stated that the key to metabolic syndrome, insulin resistance and type 2 diabetes is hypothalamic arousaldthat is, excessive hypothalamopituitary adrenal axis activity and excessive sympathetic drive. As a consequence of the hypothalamic abnormality, there is increasing abdominal obesity, an increased free fatty acid concentration and increased heart rate, cardiac output and renin secretion, together with cortisol and androgen excess. Brain-derived neurotrophic factor (BDNF) is a neurotrophin, immunotrophin, epitheliotrophin
and metabotrophin (58). An association between BDNF and depression has been demonstrated in clinical and animal studies. Animal and human studies have revealed that BDNF may play an important role in the pathogenesis of depression and, therefore, might be involved in the therapeutic effects of antidepressants (59). Animal experiments have demonstrated that decreases in BDNF were associated with depressive states, and clinical studies have shown that administration of BDNF can reverse these states (60). Karge et al reported that serum BDNF concentrations were decreased in patients with depression (61), whereas another study found that serum BDNF concentrations were lower in depressed patients than in normal controls (62). BDNF is strongly associated with type 2 diabetes. BDNF has an important influence on the pathogenesis of obesity and type 2 diabetes because it modulates the secretion and activities of insulin, leptin, ghrelin, neurotransmitters/neuropeptides and pro-inflammatory cytokines associated with energy homeostasis (63). Animal studies have also shown that BDNF has important effects on the regulation of eating behaviour, suggesting that BDNF may play a pathogenic role in the development of obesity and type 2 diabetes in humans (64). So BDNF may represent a link between depression and type 2 diabetes. A second explanation is that diabetes is not caused by depression; rather, there is a common underlying etiologic mechanism involved in both conditions. For example, none of the studies controlled for birth weight, and being small for the date of birth is known to increase the risk for developing both depression and diabetes. A potential mechanism is a defect in glucocorticoid sensitivity. A further explanation is that diabetes is preceded by symptoms, such as tiredness and malaise, that may be confused with depression when using a screening questionnaire. Finally, the development of diabetes in people with depression may be related to behaviours associated with depression (inactivity,
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
6
M. Yu et al. / Can J Diabetes xxx (2015) 1e7
poor diet and obesity) or to the drugs used to treat it rather than to the depression itself. Physical inactivity, poor diet and other behavioural factors associated with depression can lead to the development of diabetes (41). The diagnosis of diabetes could also be a result of experiencing depressive symptoms because patients with depressive symptoms are more likely to visit physicians, and they may identify the presence of type 2 diabetes also (65). Our study has a few limitations. First, the study included several case-control studies and cross-sectional studies in which the time sequence of depressive symptoms and incidence of diabetes was not clear. Second, the measurement of depression and diabetes varied among studies, and some of the depression scales are not in widespread use in clinical or research practice; heterogeneity may be introduced because of methodologic differences among studies, including differing ranges of exposure. Third, a depression score may vary with time, but the majority of the studies did not repeat recording of the depression scores. Moreover, a meta-analysis is not able to solve problems involved with confounding factors that could be inherent in the included studies. Inadequate control for confounders may bias the results in either direction, toward exaggeration or underestimation of risk estimates. Residual or unknown confounding cannot be excluded as a potential explanation for the observed findings. Further, meta-analyses can suffer from publication bias and can be limited by the data available from published studies done using different protocols because small studies with null results tend not to be published. In this metaanalysis, we found no evidence of publication bias. Nevertheless, meta-analyses are an important method of evaluating the consistency of study findings and can provide clues and guidelines for future studies. The findings of these observational studies need to be confirmed in large, randomized clinical trials. Acknowledgements This study was supported by Zhejiang Provincial Major Special Project of Science and Technology (2011C13032-1). Author Contributions Min Yu, Xingliang Zhang and Le Fang formulated the idea for conducting a systematic review; Xingliang Zhang and Feng Lu performed literature searches, study selection, data extraction, risk of bias assessment and meta-analysis and wrote the initial draft; Min Yu and Le Fang solved disagreements regarding study selection and risk of bias assessment and critically revised the article. References 1. Morrish NJ, Wang SL, Stevens LK, et al. Mortality and causes of death in the WHO Multinational Study of Vascular Disease in Diabetes. Diabetologia 2001; 44(Suppl 2):S14e21. 2. World Health Organization. Diabetes. 2012. 3. World Health Organization. Diabetes. 2011. 4. Roglic G, Unwin N, Bennett PH, et al. The burden of mortality attributable to diabetes: Realistic estimates for the year 2000. Diabetes Care 2005;28:2130e5. 5. Danaei G, Finucane MM, Lu Y, et al. National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: Systematic analysis of health examination surveys and epidemiological studies with 370 countryyears and 2.7 million participants. Lancet 2011;378:31e40. 6. World Health Organization. Diabetes. 2009. 7. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med 2006;3:e442. 8. Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med 2001;345:790e7. 9. Talbot F, Nouwen A. A review of the relationship between depression and diabetes in adults: Is there a link? Diabetes Care 2000;23:1556e62. 10. Carnethon MR, Kinder LS, Fair JM, et al. Symptoms of depression as a risk factor for incident diabetes: Findings from the National Health and Nutrition Examination Epidemiologic Follow-up Study, 1971-1992. Am J Epidemiol 2003;158: 416e23.
11. Strine TW, Mokdad AH, Balluz LS, et al. Depression and anxiety in the United States: Findings from the 2006 Behavioral Risk Factor Surveillance System. Psychiatr Serv 2008;59:1383e90. 12. Lin EH, Katon W, Von Korff M, et al. Relationship of depression and diabetes self-care, medication adherence, and preventive care. Diabetes Care 2004;27: 2154e60. 13. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Int J Surg 2010;8:336e41. 14. Lau J, Ioannidis JP, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med 1997;127:820e6. 15. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539e58. 16. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177e88. 17. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629e34. 18. Palinkas LA, Lee PP, Barrett-Connor E. A prospective study of type 2 diabetes and depressive symptoms in the elderly: The Rancho Bernardo Study. Diabet Med 2004;21:1185e91. 19. Almawi W, Tamim H, Al-Sayed N, et al. Association of comorbid depression, anxiety, and stress disorders with type 2 diabetes in Bahrain, a country with a very high prevalence of type 2 diabetes. J Endocrinol Invest 2008;11: 1020e4. 20. Disdier-Flores OM. Association of major depression and diabetes in medically indigent Puerto Rican adults. P R Health Sci J 2010;29:30e5. 21. Pan A, Lucas M, Sun Q, et al. Bidirectional association between depression and type 2 diabetes mellitus in women. Arch Intern Med 2010;170:1884e91. 22. Nichols GA, Moler EJ. Cardiovascular disease, heart failure, chronic kidney disease and depression independently increase the risk of incident diabetes. Diabetologia 2011;54:523e6. 23. Eaton WW, Armenian H, Gallo J, et al. Depression and risk for onset of type II diabetes: A prospective population-based study. Diabetes Care 1996;19:1097e102. 24. Campayo A, de Jonge P, Roy JF, et al. Depressive disorder and incident diabetes mellitus: The effect of characteristics of depression. Am J Psychiatry 2010;167: 580e8. 25. Kawakami N, Takatsuka N, Shimizu H, et al. Depressive symptoms and occurrence of type 2 diabetes among Japanese men. Diabetes Care 1999;22: 1071e6. 26. Arroyo C, Hu FB, Ryan LM, et al. Depressive symptoms and risk of type 2 diabetes in women. Diabetes Care 2004;27:129e33. 27. Saydah SH, Brancati FL, Golden SH, et al. Depressive symptoms and the risk of type 2 diabetes mellitus in a US sample. Diabetes Metab Res Rev 2003;19: 202e8. 28. Golden SH, Williams JE, Ford DE, et al. Depressive symptoms and the risk of type 2 diabetes: The Atherosclerosis Risk in Communities study. Diabetes Care 2004;27:429e35. 29. Everson-Rose SA, Meyer PM, Powell LH, et al. Depressive symptoms, insulin resistance, and risk of diabetes in women at midlife. Diabetes Care 2004;27: 2856e62. 30. Atlantis E, Browning C, Sims J, et al. Diabetes incidence associated with depression and antidepressants in the Melbourne Longitudinal Studies on Healthy Ageing (MELSHA). Int J Geriatr Psychiatry 2010;25:688e96. 31. Mallon L, Broman JE, Hetta J. High incidence of diabetes in men with sleep complaints or short sleep duration: A 12-year follow-up study of a middleaged population. Diabetes Care 2005;28:2762e7. 32. Brown LC, Majumdar SR, Newman SC, et al. History of depression increases risk of type 2 diabetes in younger adults. Diabetes Care 2005;28:1063e7. 33. Kessing LV, Nilsson FM, Siersma V, et al. Increased risk of developing diabetes in depressive and bipolar disorders? J Psychiatr Res 2004;38:395e402. 34. van den Akker M, Schuurman A, Metsemakers J, et al. Is depression related to subsequent diabetes mellitus? Acta Psychiatr Scand 2004;110:178e83. 35. Carnethon MR, Biggs ML, Barzilay JI, et al. Longitudinal association between depressive symptoms and incident type 2 diabetes mellitus in older adults: The cardiovascular health study. Arch Intern Med 2007;167:802e7. 36. Chien IC, Wu EL, Lin CH, et al. Prevalence of diabetes in patients with major depressive disorder: A population-based study. Compr Psychiatry 2012;53: 569e75. 37. Bhowmik B, Binte Munir S, Ara Hossain I, et al. Prevalence of type 2 diabetes and impaired glucose regulation with associated cardiometabolic risk factors and depression in an urbanizing rural community in bangladesh: A populationbased cross-sectional study. Diabetes Metab J 2012;36:422e32. 38. Kumari M, Head J, Marmot M. Prospective study of social and other risk factors for incidence of type 2 diabetes in the Whitehall II study. Arch Intern Med 2004;164:1873e80. 39. Stellato RK, Feldman HA, Hamdy O, et al. Testosterone, sex hormone-binding globulin, and the development of type 2 diabetes in middle-aged men: Prospective results from the Massachusetts male aging study. Diabetes Care 2000; 23:490e4. 40. Holt RI, Phillips DI, Jameson KA, et al. The relationship between depression and diabetes mellitus: Findings from the Hertfordshire Cohort Study. Diabet Med 2009;26:641e8. 41. Engum A. The role of depression and anxiety in onset of diabetes in a large population-based study. J Psychosom Res 2007;62:31e8. 42. Palinkas LA, Barrett-Connor E, Wingard DL. Type 2 diabetes and depressive symptoms in older adults: A population-based study. Diabet Med 1991;8: 532e9.
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
M. Yu et al. / Can J Diabetes xxx (2015) 1e7 43. Kivimaki M, Hamer M, Batty GD, et al. Antidepressant medication use, weight gain, and risk of type 2 diabetes: a population-based study. Diabetes Care 2010; 33:2611e6. 44. Demakakos P, Pierce MB, Hardy R. Depressive symptoms and risk of type 2 diabetes in a national sample of middle-aged and older adults: The English longitudinal study of aging. Diabetes Care 2010;33:792e7. 45. Maty SC, Everson-Rose SA, Haan MN, et al. Education, income, occupation, and the 34-year incidence (1965-99) of type 2 diabetes in the Alameda County Study. Int J Epidemiol 2005;34:1274e81. 46. Golden SH, Lazo M, Carnethon M, et al. Examining a bidirectional association between depressive symptoms and diabetes. JAMA 2008;299:2751e9. 47. Knol MJ, Geerlings MI, Egberts AC, et al. No increased incidence of diabetes in antidepressant users. Int Clin Psychopharmacol 2007;22:382e6. 48. Eriksson AK, Ekbom A, Granath F, et al. Psychological distress and risk of prediabetes and type 2 diabetes in a prospective study of Swedish middle-aged men and women. Diabet Med 2008;25:834e42. 49. Rajala U, Keinanen-Kiukaanniemi S, Kivela SL. Non-insulin-dependent diabetes mellitus and depression in a middle-aged Finnish population. Soc Psychiatry Psychiatr Epidemiol 1997;32:363e7. 50. Cosgrove MP, Sargeant LA, Griffin SJ. Does depression increase the risk of developing type 2 diabetes? Occup Med (Lond) 2008;58:7e14. 51. Mezuk B, Eaton WW, Albrecht S, et al. Depression and type 2 diabetes over the lifespan: A meta-analysis. Diabetes Care 2008;31:2383e90. 52. Rotella F, Mannucci E. Depression as a risk factor for diabetes: A meta-analysis of longitudinal studies. J Clin Psychiatry 2013;74:31e7. 53. Knol MJ, Twisk JW, Beekman AT, et al. Depression as a risk factor for the onset of type 2 diabetes mellitus: A meta-analysis. Diabetologia 2006;49:837e45. 54. Harris MD. Psychosocial aspects of diabetes with an emphasis on depression. Curr Diab Rep 2003;3:49e55.
7
55. Duncan BB, Schmidt MI, Pankow JS, et al. Low-grade systemic inflammation and the development of type 2 diabetes: The atherosclerosis risk in communities study. Diabetes 2003;52:1799e805. 56. Bjorntorp P. Visceral fat accumulation: the missing link between psychosocial factors and cardiovascular disease? J Intern Med 1991;230:195e201. 57. Bjorntorp P. Abdominal obesity and the development of noninsulin-dependent diabetes mellitus. Diabetes Metab Rev 1988;4:615e22. 58. Chaldakov GN, Fiore M, Tonchev AB, et al. Homo obesus: A metabotrophindeficient species. Pharmacol Nutr Insight Curr Pharm Des 2007;13:2176e9. 59. Hou SJ, Yen FC, Tsai SJ. Is dysfunction of the tissue plasminogen activator (tPA)plasmin pathway a link between major depression and cardiovascular disease? Med Hypotheses 2009;72:166e8. 60. Berton O, McClung CA, Dileone RJ, et al. Essential role of BDNF in the mesolimbic dopamine pathway in social defeat stress. Science 2006;311:864e8. 61. Karege F, Bondolfi G, Gervasoni N, et al. Low brain-derived neurotrophic factor (BDNF) levels in serum of depressed patients probably results from lowered platelet BDNF release unrelated to platelet reactivity. Biol Psychiatry 2005;57: 1068e72. 62. Gervasoni N, Aubry JM, Bondolfi G, et al. Partial normalization of serum brainderived neurotrophic factor in remitted patients after a major depressive episode. Neuropsychobiology 2005;51:234e8. 63. Rao AA, Sridhar GR, Srinivas B, et al. Bioinformatics analysis of functional protein sequences reveals a role for brain-derived neurotrophic factor in obesity and type 2 diabetes mellitus. Med Hypotheses 2008;70:424e9. 64. Fujinami A, Ohta K, Obayashi H, et al. Serum brain-derived neurotrophic factor in patients with type 2 diabetes mellitus: Relationship to glucose metabolism and biomarkers of insulin resistance. Clin Biochem 2008;41:812e7. 65. Knol MJ, Heerdink ER, Egberts AC, et al. Depressive symptoms in subjects with diagnosed and undiagnosed type 2 diabetes. Psychosom Med 2007;69:300e5.
FLA 5.2.0 DTD JCJD523_proof 6 March 2015 5:11 pm ce
