- Email: [email protected]
Sense of mastery as mediator buffering psychological distress among people with diabetes
Sense of mastery as mediator buffering psychological distress among people with diabetes Karin Elisabeth Bennetter, Jocelyne Clench–Aas, Ruth Kjærsti Raanaas PII: DOI: Reference:
S1056-8727(16)30059-9 doi: 10.1016/j.jdiacomp.2016.03.022 JDC 6695
To appear in:
Journal of Diabetes and Its Complications
Received date: Revised date: Accepted date:
28 January 2016 16 March 2016 18 March 2016
Please cite this article as: Bennetter, K.E., Clench–Aas, J. & Raanaas, R.K., Sense of mastery as mediator buffering psychological distress among people with diabetes, Journal of Diabetes and Its Complications (2016), doi: 10.1016/j.jdiacomp.2016.03.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Sense of mastery as mediator buffering psychological distress
T
among people with diabetes
RI P
Karin Elisabeth Bennetter, MPH 1, Jocelyne Clench–Aas, Senior researcher 2 and Ruth
1
SC
Kjærsti Raanaas, Assoc. Prof. 1
Department of Landscape Architecture and Spatial Planning, Norwegian University of Life Sciences, Norway
MA NU
and 2 Division of mental health, Norwegian Institute of public health, Norway.
Corresponding Author: Ruth Kjærsti Raanaas, Section for Public Health Science, Department of Landscape Architecture and Spatial Planning, Norwegian University of Life Sciences, Post
AC
CE
PT
ED
Box 5003, N-1432 Ås, Tel: +47 67 23 12 68, Email address: [email protected]
ACCEPTED MANUSCRIPT Abstract Aims: The purpose of this study was to examine the association between diabetes with or
T
without other comorbid somatic diseases and depression and anxiety, and to explore the
RI P
mediating role of sense of mastery and social support. Methods: Data were obtained from a cross-sectional health survey conducted in Norway (n=6,827). People with diabetes alone or
SC
with simultaneous comorbid somatic diseases were compared to a group with no known
MA NU
somatic diseases. Results: Among people with diabetes alone, 16.3% reported having depression and anxiety. Having diabetes was associated with 3 times greater odds for anxiety compared to the control group, and 2 times greater odds for depression. Among individuals with diabetes and comorbid somatic diseases, 17.4% reported depression and 11.6% reported
ED
symptoms of anxiety. The odds for both were approximately 2 times greater than in the control group. Sense of mastery, but not social support, protected against depression in both
PT
groups and against anxiety in the diabetes with comorbidity group. Conclusions:
CE
Comorbidity between diabetes and other somatic diseases seems to be related to depression to a larger degree, whereas having diabetes alone relates more to anxiety. This can possibly be
mastery.
AC
explained by the overall burden in the comorbidity group and the related absence of sense of
Keywords: Mental health problems, psychosocial resources, chronic diseases, social support
2
ACCEPTED MANUSCRIPT 1.0 Introduction Diabetes mellitus, whether it is type 1 (T1DM) or type 2 (T2DM), are chronic and life
T
threatening diseases that requires regular monitoring of blood glucose and administration of
RI P
diet and medication. In Norway, approximately 165 000 people (3.2 % of the population) have blood glucose-lowering drugs dispensed, indicating that they are diagnosed with either
SC
T1DM or T2DM (Strom et al. 2014).
MA NU
Chronic somatic diseases, which in this case is defined as an impairment of normal physiological function affecting all or part of the organism, are frequently associated with psychological distress (Gili et al. 2010). Accordingly, people with diabetes are found to be at particular risk of developing mental health problems such as depression (Anderson et al.
ED
2001; Andreoulakis et al. 2012). In a meta-analysis, the prevalence of depression ranges from 8-18% for severe types to 15-35% for milder types among people with diabetes
PT
(Andreoulakis et al. 2012). In the same analysis, rates for depression are found for individuals
CE
with diabetes to be about 1.4-3 times as likely compared to non-diabetic controls. Comorbid depression among people with diabetes is associated with nonadherence to diabetes self-care
AC
(including dietary, exercise, medication etc.), disease control, symptom burden, and development of medical complications (Katon 2008). Diabetes is also associated with anxiety defined both as a generalized anxiety disorder and elevated symptoms of anxiety (Egede & Dismuke 2012; Smith et al. 2013). Generalized anxiety disorder are reported to be present in 14% of people living with diabetes, and a prevalence of up to 40% is reported looking at elevated symptoms of anxiety (Bener et al. 2011). In the meta-analysis by Smith and colleagues (2013) a pooled odds ratio of 1.25 for anxiety among people with diabetes was found. However, the association between diabetes and anxiety is suggested to be weaker than between diabetes and depression (Egede & 3
ACCEPTED MANUSCRIPT Dismuke 2012). The association with anxiety is linked to glycaemic control, age and lifetime severe hypoglycaemia (Labad et al. 2010; McDade-Montez & Watson 2011).
T
People with diabetes are found to be at particular risk of developing somatic
RI P
comorbidities (Fillenbaum et al. 2000; Topic et al. 2013) and the prevalence rate of psychological distress appears to increase according to the number of comorbid somatic
SC
diseases (Gili et al. 2010). An overall prevalence of serious psychological distress for individuals with diabetes and comorbidities have been reported to be 7,6% compared to 3,6%
MA NU
for people without diabetes (Egede & Dismuke 2012; Li et al. 2009). In addition, comorbid abdominal obesity and cardiovascular disease are identified as risk factors for depression among people with diabetes (Labad et al. 2010). Despite that many studies have found a
ED
relationship between diabetes and psychological distress, to our knowledge few studies have systematically differentiated between diabetes with and without somatic comorbidity when
PT
studying this relationship.
Psychosocial resources, such as sense of mastery (Raaijmakers et al. 2014) and social
CE
support (van Dam et al. 2005) are reported to protect against developing psychological
AC
distress. Sense of mastery is associated with self-management behaviors, preventative care and proper utilization of health care services, as well as responsiveness to interventions for health promotion, which may possibly account for its protective ability (Skaff et al. 2003). The concept is similar to self-efficacy (Bandura 1977), but whereas self-efficacy often is applied to the person’s feeling of competence in a specific task, sense of mastery is used more globally, referring to one’s overall sense of control in life (Skaff et al. 1996). The risk of diminished sense of mastery, is associated with an increase in both chronic conditions and functional disability (Jang et al. 2009). It has also been reported that poor sense of mastery significantly increases the risk of developing anxiety (Gordon et al. 2007).
4
ACCEPTED MANUSCRIPT Social support from family and friends are in some studies reported to facilitate better diabetes self-management (Gallant 2003; van Dam et al. 2005), although the literature is not
T
consistent (van Dam et al. 2005). Social support includes social behaviors such as seeking
RI P
advice, emotional support and instrumental support such as direct assistance with illness management activities. Self-care tasks, such as dietary behaviors and physical activity appear
SC
to be particularly susceptible to social support (Gallant 2003; Qiu et al. 2012). Since earlier studies examining the potential buffering effects of psychosocial
MA NU
resources on individuals with diabetes have not distinguished between people with diabetes alone and those with somatic comorbidity, the aim of the present study was to examine the association between diabetes and psychological distress, distinguishing between people with
ED
only diabetes and those with somatic comorbidity. Additionally, the aim was to explore the extent to which the possible association was mediated by sense of mastery and social support.
PT
The research questions were: Do relatively more people with diabetes, with or without other comorbid somatic diseases, suffer from anxiety and depression compared to people with no
CE
known somatic disease? Can sense of mastery or social support mediate the reduction of
AC
anxiety and/ or depression among people with diabetes, with or without other comorbid somatic diseases?
2.0 Materials and Methods 2.1 Design and participants The present study presents data from a cross-sectional health survey conducted in Norway in 2002 by Statistics Norway (SN) (Hougen & Gløboden 2002). A random sample of 10 000 non-institutionalized subjects over the age of 15 years was drawn to participate in the survey. The sample consisted of two subsamples, a main and a supplementary sample, each containing 5000 subjects. The main sample was drawn following SN’s standard sample plan 5
ACCEPTED MANUSCRIPT in which Norway is divided in 109 strata. The supplementary sample was drawn randomly from all of the municipalities in Norway. Participants from both samples were interviewed
T
(1/2 hour) by home visits or by phone (30%) (Hougen & Gløboden 2002). Some participants
RI P
were lost from the sample due to death, travel abroad, or institutionalization. A total of 6,827 (89.5%) participants completed the interview, with 3,410 males and 3,417 females. A written
SC
“helping card” which provided a list of 59 diseases, was used in the interview to simplify a sensitive and slightly difficult question (participants interviewed by phone were sent the cards
MA NU
in the mail). After the interview, all participants received a health questionnaire by post where data on mental health and psychosocial variables were obtained. Due to further attrition, a total of 5,343 participants responded to both interview and postal questionnaire.
and linked to the data set.
PT
2.2 Measures
ED
Information about education and household income were retrieved from national registries
Diabetes was measured with the following question: “Do you have, or have you had (i.e.)
CE
diabetes mellitus?” Score opportunities were 1= have, 2= have had, 3= have never had.
AC
Participants who indicated “having had” diabetes (13 participants with diabetes alone and 55 participants with diabetes and comorbidity) were excluded from statistical analyses since the current study focuses on chronic symptoms of diabetes. Due to lack of information in the dataset concerning type of diabetes (T1DM versus T2DM), the term diabetes will be used in this article without differentiating further. Other somatic diseases were measured with a similar question as above, but with reference to the particular disease of interest. Somatic diseases included in the current study were: epilepsy, osteoporosis, angina pectoris, coronary heart disease, stroke, cancer, allergy, high blood pressure, metabolism disease, ankylosing spondylitis (previously known as Bekhterev's disease), arthritis, chronic
6
ACCEPTED MANUSCRIPT bronchitis/emphysema/COPD, psoriasis, atopic eczema, urinary incontinence, fractures, removed organ, and ulcers. Four population subgroups were studied: 1) individuals with
T
diabetes alone; 2) individuals with diabetes and other comorbid somatic diseases; 3)
RI P
individuals who do not have diabetes but have other somatic diseases; and 4) the control group, individuals with no known somatic disease.
SC
Psychological distress in terms of depression and anxiety was assessed using the Hopkins Symptom Check List (HSCL-25), which measures symptoms over the previous 14
MA NU
days (Strand et al. 2003). It contains 25 items covering two subscales for depression and anxiety. Responses are given on a four-point scale (1=“not at all” to 4=“extremely”). In the present study, averages for each subscale are dichotomized into “low” and “high” scores with
ED
a cut-off point of 1.75 (Strand et al. 2003). A “case” is accepted if no more than two items are missing from the items measuring depression or anxiety, in which case mean values were
PT
substituted for the missing items. Cronbach’s alpha for the total HSCL-25 scale was .93; for depression it was .91 and for anxiety .84.
CE
Sense of mastery was measured using the five-item version of the seven-item scale
AC
developed by Pearlin and colleagues (1981). The five selected items ask about the experience of controlling and coping with life such as: “I have little control over the things that happen to me” and “I often feel helpless in dealing with the problems of life”. Responses were given on a five point scale (0=“agree” to 4=“do not agree”). For those with no more than two missing data points, mean values were substituted for the missing items. A sum score was calculated ranging from 0-20. Cronbach’s alpha was 0.86. Social support was measured using the Oslo 3 Support Scale (OSS-3 scale) (Dalgard et al. 2006). It is comprised of 3 questions about the number of close confidants they can count on if they have serious problems (scores were: 1=”no one”, 2=“1 or 2”, 3=“3-5” or
7
ACCEPTED MANUSCRIPT 4=“more than 5”), sense of concern or interest from other people (scores were: 1=“Great concern and interest” to 5=”no concern and interest”), and sense of instrumental support from
T
neighbors (scores were: 1=”very easy” to 5=“very difficult”). The corresponding scores in the
RI P
two last questions were recoded in reverse order. The questions were further merged into a social support index by adding up the scores for each item, ranging from 3-14. Data had to be
SC
complete (no missing) to count as a case.
Data on gender and age were obtained from the interview section. Age was
MA NU
categorized into 4 groups (1=”16-24”, 2=”25-44”, 3=”45-67” and 4=”>67 years”). Education and income were obtained from linked public registries. Education was categorized into 3 groups; Low Education (primary School and junior high school (year 1-10)), Middle
ED
Education (high school and first level of junior college without accomplished degree (year 11-14)), High Education (university or university college (above 14)). Household income
PT
defined as the total gross annual income of the family was divided into quartiles. Physical activity was measured by how many times per week the subject exercised (0=”never”,
CE
1=”less than 1/week” and 2=”1/week or more”). BMI was measured by kilograms/ height
1995).
AC
(m2) and classified into 6 categories after World Health Organizations guidelines (WHO
2.3 Statistical analysis All analyses were performed using the Statistical Package for the Social Sciences (SPSS for Windows, version 22.0). Hierarchical models of multiple logistic regression were conducted to test the associations between the predicting variables (diabetes with and without comorbidity, and other somatic diseases) and the dependent variables anxiety and depression (dichotomous versions; > 1.75). Odds ratio (OR) with a 95% confidence interval as well as beta (B) with SE were reported as the estimated outcomes. The data was weighted to account
8
ACCEPTED MANUSCRIPT for information collected from non-respondents in national registries by adjusting for gender, age, highest education level and family size, by way of adjusting the standard errors for the
T
modified sample size (Osborne 2011) using the complex sample module of SPSS.
RI P
Demographics, BMI and physical activity were all entered in the first step. Type of diagnosis was entered as a categorical factor in the next step. All the three groups of
SC
diagnosis were compared to the control group. Sense of mastery and social support (potential mediators) were entered separately at the final step and one at a time, testing mediating
MA NU
effects using the methods described by Kim and colleagues (2001). Moderation was tested using the PROCESS module for SPSS provided by Hayes and Little (2013) (model 1). Mediation (both individually and parallel with both mediators in simultaneously) was further
ED
tested using the bootstrap method and the Sobel test both provided in the PROCESS module for SPSS (model 4).
PT
Model fit was evaluated through R2, which were significant for both anxiety and depression. Missing data were excluded casewise. To test the inter-correlation among the
CE
predicting variables, multicollinearity tests were conducted and showed no violation of
AC
multicollinearity assumptions (Tolerance values were above 0.1 and VIF values under 10).
3.0 Results
Table 1 presents the weighted prevalence of responses to questions regarding depression, anxiety, sense of mastery, social support, demographics, BMI and physical activity for the four diagnostic groups. Among the 5,342 (non-weighted N) participants who responded to both interview and postal questionnaire, 3.6% reported being diagnosed with diabetes, of which 2.7% reported additional comorbidity, and 0.9% of participants reported being diagnosed with diabetes alone. In total, 54.5% of participants had somatic diseases other than
9
ACCEPTED MANUSCRIPT diabetes, and 41.9% constituted a control group without a known somatic disease. Nearly 10% of participants reported having depression.
T
Relatively more people with diabetes alone and with comorbid somatic diseases have
RI P
depression (17.8 %) compared to the control group (8.2 %). This is also the case for people with somatic diseases other than diabetes. Adjusting for demographics, BMI and physical
depression than the control group (Table 2).
SC
activity, people with diabetes and comorbid somatic diseases have twice the odds for
MA NU
Although not significant, there is a trend for higher levels of anxiety in people with diabetes alone than in the control group (18.8 % vs 5.6 %). Adjusting for demographics, BMI and physical activity, people with diabetes alone were significantly associated with more than
ED
3 times greater odds for anxiety (Table 3), whereas for people with diabetes and comorbid somatic diseases there was no significant relationship with anxiety. The association between
PT
anxiety and the disease category somatic diseases other than diabetes was significant and the odds were approximately 1.7.
CE
In total, 18.5% reported low levels of sense of mastery. More than 40 % of the
AC
population with diabetes with comorbidity, and a third of those with diabetes alone report low sense of mastery. No significant moderation effects were found for sense of mastery. Examining mediator effects of sense of mastery, in the hierarchal analysis of both depression and anxiety (Tables 2 and 3), the addition of sense of mastery resulted in a reduction by nearly half of all the betas. The odds were also considerably reduced. Sense of mastery in itself was significantly and negatively associated with the outcome variables of depression and anxiety. Using both the bootstrap method and the Sobel test, the indirect mediator pathway was significant for all disease categories.
10
ACCEPTED MANUSCRIPT Ten % of the population report low levels of social support. Levels are only slightly higher in those with other somatic disease alone and diabetes alone, whereas they are nearly
T
25% in those with diabetes with comorbidity. Also for social support the moderation
RI P
analyses, found no significant moderation effects. Examining mediator effects of social support, the addition of social support did not alter any of the betas for either depression or
SC
anxiety (Tables 2 and 3). The odds also remained unchanged. Despite that social support was in itself significantly and negatively associated with the outcome variables of depression and
MA NU
anxiety, using both the bootstrap method and the Sobel test, the indirect mediator path was non-significant for all disease categories. It should be noted that the relationship between the disease category and sense of mastery was highly significant, with estimates of beta ranging
ED
from 0.3 for somatic diseases to 1.96 for diabetes alone (data not shown). However, the similar relationship between disease category and social support were non-significant for all
PT
categories.
The results of the mediator analyses with simultaneous testing for social support and
CE
sense of mastery resulted in non-significance for social support both for anxiety and
AC
depression and for all disease categories, but significance for sense of mastery in all situations (data not shown).
4.0 Discussion
In the current study people with diabetes and comorbid somatic diseases were twice as likely to suffer from depression than people with no known somatic disease, when controlling for socio-demographics, BMI and physical activity. Elevated odds for depression were, however, not found among people with diabetes alone in the present study. These findings highlights the need to differentiate between people with diabetes with and without comorbidities when examining the association with psychological distress, and may be explained by the total 11
ACCEPTED MANUSCRIPT burden of having more chronic diseases (Gili et al. 2010), or the combination of diabetes with somatic diseases in the cardiovascular domain previously identified as risk factor for
T
depression (Labad et al. 2010). Previous studies, having found associations between diabetes
RI P
and depression (Anderson et al. 2001; Andreoulakis et al. 2012), have not clearly differentiated between people with diabetes with and without comorbidities, and it is thus
SC
difficult to compare findings.
People with diabetes alone where found in the current study to be three times more
MA NU
likely to experience anxiety compared to the healthy population. Elevated odds for anxiety among people with diabetes are found in earlier studies (Smith et al. 2013). However, the literature is somewhat inconsistent, and Clarke and Currie (2009) report that anxiety is not
ED
common among people with diabetes, but rather in other somatic diseases such as heart disease, stroke and cancer. These conflicting results again bring up the importance of
PT
distinguishing people with diabetes with or without somatic comorbidities. Since anxiety has been linked to glycemic control, age and lifetime severe hypoglycaemia, one may suggest a
CE
lack of ability to cope with the disease as the most vital factor (Labad et al. 2010; McDade-
AC
Montez & Watson 2011).
Chronic conditions and functional disability can gradually decrease psychosocial resources such as sense of mastery and social support (Jang et al. 2009; Raaijmakers et al. 2014). In the present study, among people with diabetes with comorbid somatic diseases as well as for those with somatic diseases alone, sense of mastery played a mediator role both for depression and for anxiety. This indicates that not only may there be a direct association between having comorbid diabetes or somatic diseases and psychological distress, but these conditions may also be associated with depression via the indirect pathway of reducing sense of mastery. Thus the measured low levels of sense of mastery, especially among those with
12
ACCEPTED MANUSCRIPT diabetes as opposed to other somatic diseases, seemed to be associated with an increase in depression and anxiety. Low sense of mastery among people with diabetes may be associated
T
with more difficulties in adhering to treatment regimens (Skaff et al. 2003), or coping with
RI P
the disease. The reported higher prevalence of low sense of mastery among people with diabetes with comorbidity as compared to people with diabetes alone, can thus be explained
SC
by similar combination of challenges related to living with a chronic condition and functional disability (Jang et al. 2009). Increased coping skills and adherence to health promotion
MA NU
interventions may be one way of increasing the sense of mastery (Skaff et al. 2003; Thorpe et al. 2013), that can again serve as a psychological resource to possibly reduce both depression and anxiety among people with diabetes with and without comorbidity.
ED
Despite that low social support previously is reported to elevate the perceived psychological burden of a chronic disease (Renn et al. 2011), leading to increased depression,
PT
social support was not associated with reduce depression, or anxiety in this sample. Diabetes is a disease often associated with lifestyle choices that complicate the relationship with social
CE
support, and social support may change, pertaining to if or not, the individual adheres to
AC
recommended lifestyle changes (Gallant 2003; Qiu et al. 2012; van Dam et al. 2005). 4.1 Methodological limitations The study is based on a large and nationally representative sample, with a high response rate (70.4%). Despite the large sample, however, the numbers with diabetes alone were quite low. Further, the cross-sectional design restricts the ability to draw causal inferences concerning the direction of the association between diabetes and psychological distress. Since information on psychological distress and physical health was based on retrospective selfreport without independent verification by a medical professional, bias can occur.
13
ACCEPTED MANUSCRIPT 5.0 Conclusions Relatively more people with diabetes, with or without other comorbid somatic diseases,
T
suffer from psychological distress compared to people with no known somatic disease. The
RI P
present study indicates that people with diabetes and comorbid somatic diseases are to a larger degree suffering from depression, whereas those with diabetes alone are primarily
SC
associated with anxiety. To be able to target help to the different groups our study indicates
MA NU
that when examining comorbidity with psychological distress, people with the different somatic illnesses should be studied separately. Health care should include means for increasing sense of mastery in people with diabetes at an early stage to improve diabetes coping, such as self-management education and support groups.
ED
6.0 Acknowledgements
This study was performed in collaboration with the Norwegian Institute of Public Health,
PT
division of Mental Health. We would like to thank the institute and the division for providing
CE
us with access to data, workspace and support. In addition we would like to thank Statistics
AC
Norway (SSB) for collection and distribution of data.
The Author(s) declare(s) that there is no conflict of interest.
7.0 References
Anderson, R. J., Freedland, K. E., Clouse, R. E. & Lustman, P. J. (2001). The Prevalence of Comorbid Depression in Adults With Diabetes: A meta-analysis. Diabetes Care, 24 (6): 1069-1078. Andreoulakis, E., Hyphantis, T., Kandylis, D. & Iacovides, A. (2012). Depression in diabetes mellitus: a comprehensive review. Hippokratia, 16 (3): 205-214.
14
ACCEPTED MANUSCRIPT Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84 (2): 191-215.
T
Bener, A., Al-Hamaq, A. O. A. A. & Dafeeah, E. E. (2011). High Prevalence of Depression,
RI P
Anxiety and Stress Symptoms Among Diabetes Mellitus Patients. The Open Psychiatry
SC
Journal, 5: 5-12.
Clarke, D. M. & Currie, K. C. (2009). Depression, anxiety and their relationship with chronic
Australia, 190 (7 Suppl): S54-60.
MA NU
diseases: a review of the epidemiology, risk and treatment evidence. The Medical Journal of
Dalgard, O. S., Dowrick, C., Lehtinen, V., Vazquez-Barquero, J. L., Casey, P., Wilkinson,
ED
G., Ayuso-Mateos, J. L., Page, H. & Dunn, G. (2006). Negative life events, social support and gender difference in depression: a multinational community survey with data from the
PT
ODIN study. Social Psychiatry and Psychiatric Epidemiology, 41 (6): 444-51.
CE
Egede, L. E. & Dismuke, C. E. (2012). Serious psychological distress and diabetes: a review
AC
of the literature. Current Psychiatry Reports, 14 (1): 15-22. Fillenbaum, G. G., Pieper, C. F., Cohen, H. J., Cornoni-Huntley, J. C. & Guralnik, J. M. (2000). Comorbidity of Five Chronic Health Conditions in Elderly Community Residents: Determinants and Impact on Mortality. Journal of Gerontology: Medical Sciences, 55 (2): 84-89. Gallant, M. P. (2003). The Influence of Social Support on Chronic Illness Self-Management: A Review and Directions for Research. Health Education & Behavior, 30 (2): 170-195.
15
ACCEPTED MANUSCRIPT Gili, M., Comas, A., Garcia-Garcia, M., Monzon, S., Antoni, S. B. & Roca, M. (2010). Comorbidity between common mental disorders and chronic somatic diseases in primary care
T
patients. General Hospital Psychiatry, 32 (3): 240-245.
RI P
Gordon, A., Smith, R., Keramatian, K., Luus, B., Weinberg, A., Smallwood, J., Schooler, J. & Christoff, K. (2007). Mind-wandering, awareness, and task performance: An fMRI study.
SC
Canadian Journal of Experimental Psychology-Revue Canadienne De Psychologie
MA NU
Experimentale, 61 (4): 366-366.
Hayes, A. F. & Little, T. D. (2013). Introduction to mediation, moderation, and conditional process analysis: a regression-based approach. New York: Guilford Press. 507 pp.
ED
Hougen, H. C. & Gløboden, M. A. (2002). Samordnet levekårsundersøkelse 2002 -
PT
tverrsnittsundersøkelsen. Dokumentasjonsrapport. Oslo: Statistics Norway. Jang, Y., Chiriboga, D. A., Lee, J. & Cho, S. (2009). Determinants of a sense of mastery in
CE
Korean American elders: a longitudinal assessment. Aging & Mental Health, 13 (1): 99-105.
AC
Katon, W. J. (2008). The Comorbidity of Diabetes Mellitus and Depression. The American Journal of Medicine, 121 (11 Suppl 2): S8-15. Kim, J. S., Kaye, J. & Wright, L. K. (2001). Moderating and mediating effects in causal models. Issues Ment Health Nurs, 22 (1): 63-75. Labad, J., Price, J. F., Strachan, M. W. J., Fowkes, F. G. R., Ding, J., Deary, I. J., Lee, A. J., Frier, B. M., Seckl, J. R., Walker, B. R., et al. (2010). Symptoms of depression but not anxiety are associated with central obesity and cardiovascular disease in people with Type 2 diabetes: The Edinburgh Type 2 diabetes study. Diabetologia, 53 (3): 467-471.
16
ACCEPTED MANUSCRIPT Li, C., Ford, E., Zhao, G., Strine, T., Dhingra, S., Barker, L., Berry, J. & Mokdad, A. (2009). Association between diagnosed diabetes and serious psychological distress among U.S.
T
adults: the Behavioral Risk Factor Surveillance System, 2007. International Journal of
RI P
Public Health, 54 (1): 43-51.
McDade-Montez, E. A. & Watson, D. (2011). Examining the potential influence of diabetes
MA NU
Annals of Behavioral Medicine, 42 (3): 341-51.
SC
on depression and anxiety symptoms via multiple sample confirmatory factor analysis.
Osborne, J. W. (2011). Best Practices in using large, complex samples: The importance of using appropriate weights and design effect compensation. Practical Assessment, Research &
ED
Evaluation, 16 (12): 1-7.
Pearlin, L. I., Lieberman, M. A., Menaghan, E. G. & Mullan, J. T. (1981). The stress process.
PT
Journal of Health and Social Behavior. American Sociological Association, 22 (4): 337-56.
CE
Qiu, S. H., Sun, Z. L., Cai, X., Liu, L. & Yang, B. (2012). Improving patients' adherence to
5.
AC
physical activity in diabetes mellitus: a review. Diabetes and Metabolism Journal, 36 (1): 1-
Raaijmakers, L. G. M., Martens, M. K., Hesselink, A. E., de Weerdt, I., de Vries, N. K. & Kremers, S. P. J. (2014). Mastery and perceived autonomy support are correlates of Dutch diabetes patients’ self-management and quality of life. Patient Education and Counseling, 97 (1): 75-81. Renn, B. N., Feliciano, L. & Segal, D. L. (2011). The bidirectional relationship of depression and diabetes: a systematic review. Clinical Psychology Review, 31 (8): 1239-46.
17
ACCEPTED MANUSCRIPT Skaff, M. M., Pearlin, L. I. & Mullan, J. T. (1996). Transitions in the caregiving career: Effects on sense of mastery. Psychology and Aging, 11 (2): 247-257.
T
Skaff, M. M., Mullan, J. T., Fisher, L. & Chesla, C. A. (2003). A contextual model of control
RI P
beliefs, behavior, and health; Latino and European Americans with type 2 diabetes.
SC
Psychology and Health, 18 (3): 295-312.
Smith, K. J., Beland, M., Clyde, M., Gariepy, G., Page, V., Badawi, G., Rabasa-Lhoret, R. &
MA NU
Schmitz, N. (2013). Association of diabetes with anxiety: a systematic review and metaanalysis. Journal of Psychosomatic Research, 74 (2): 89-99. Strand, B. H., Dalgard, O. S., Tambs, K. & Rognerud, M. (2003). Measuring the mental
ED
health status of the Norwegian population: a comparison of the instruments SCL-25, SCL-10,
PT
SCL-5 and MHI-5 (SF-36). Nordic Journal of Psychiatry, 57 (2): 113-118. Strom, H., Selmer, R., Birkeland, K. I., Schirmer, H., Berg, T. J., Jenum, A. K., Midthjell, K.,
CE
Berg, C. & Stene, L. C. (2014). No increase in new users of blood glucose-lowering drugs in
AC
Norway 2006-2011: a nationwide prescription database study. BMC Public Health, 14: 520. Thorpe, C. T., Fahey, L. E., Johnson, H., Deshpande, M., Thorpe, J. M. & Fisher, E. B. (2013). Facilitating healthy coping in patients with diabetes: a systematic review. Diabetes Educator, 39 (1): 33-52. Topic, R., Milicic, D., Stimac, Z., Loncar, M., Velagic, V., Marcinko, D. & Jakovljevic, M. (2013). Somatic comorbidity, metabolic syndrome, cardiovascular risk, and CRP in patients with recurrent depressive disorders. Croatian medical journal, 54 (5): 453-459.
18
ACCEPTED MANUSCRIPT van Dam, H. A., van der Horst, F. G., Knoops, L., Ryckman, R. M., Crebolder, H. F. & van den Borne, B. H. (2005). Social support in diabetes: a systematic review of controlled
AC
CE
PT
ED
MA NU
SC
RI P
T
intervention studies. Patient Education and Counseling, 59 (1): 1-12.
19
ACCEPTED MANUSCRIPT
Table 1: Weighted prevalence W-N (%) of responses on the questions concerning depression, anxiety, sense of mastery, social support,
CR I
PT
demographics, BMI and physical activity, according to diagnostic category. Response alternatives
Control group (W-N=4075(41,9%))
No diabetes, but other somatic diseases (W-N=5307(54,5%))
Only diabetes (W-N=84(0,9%))
Diabetes with comorbidity (W-N=267(2,7%))
Depression
>1.75
280 (8.2)
629 (14.2)
12 (16.9)
38 (17.8)
Anxiety
>1.75
190 (5.6)
441 (9.9)
13 (18.8)
26 (11.9)
Sense of mastery
Mean (SE)
15.3 (.07)
14.0 (.07)
12.8 (.48)
12.6 (.33)
Social support
Mean (SE)
11.2 (.03)
11.0 (.03)
10.9 (.18)
10.2 (.15)
Gender
Male
2265 (55.6)
2472 (46.6)
54 (65.0)
134 (50.2)
Age
15-24 25-44 45-66 67
701 (17.2) 1807 (44.4) 1275 (31.3) 289 (7.1)
630 (11.9) 1732 (32.7) 1877 (35.4) 1061 (20.0)
11 (13.5) 31 (36.9) 17 (20.9) 24 (28.7)
5 (1.7) 34 (12.7) 119 (44.6) 110 (41.0)
Education
Low Middle High
616 (15.6) 2274 (57.7) 1057 (26.7)
1153 (22.3) 2900 (56.1) 1115 (21.6)
24 (28.8) 45 (55.0) 13 (16.3)
105 (34.0) 121 (46.3) 36 (13.7)
AC
CE P
TE
D
MA
NU S
Variables
20
773 (19.0) 726 (17.9) 1762 (43.4) 803 (19.8)
1334 (25.2) 1072 (20.2) 2069 (39.1) 823 (15.5)
No of days exercise/week
Never <1/week ≥1/week
953 (23.5) 476 (11.7) 2634 (64.8)
1420 (26.8) 545 (10.3) 3331 (62.9)
BMI categorized
Normal Underweight Overweight Obesity Obesity class 2 Obesity class 3
2391 (60.3) 83 (2.1) 1270 (32.0) 199 (5.0) 21 (0.5) 3 (0.1)
2677 (5246) 92 (1.8) 1849 (36.2) 411 (8.1) 65 (1.3) 14 (0.3)
CR I
Lowest quartile 2nd quartile 3rd quartile Highest quartile
21 (25.3) 24 (28.7) 29 (34.3) 10 (11.8)
97 (36.2) 67 (25.2) 77 (28.7) 27 (9.9)
25 (33.8) 5 (6.3) 50 (59.9)
108 (40.7) 16 (6.1) 142 (53.2)
48 (58.3) 0 (0.0) 28 (34.2) 3 (4.1) 3 (3.3) 0 (0.0)
75 (29.2) 6 (2.2) 113 (44.1) 48 (18.7) 8 (3.2) 6 (2.5)
AC
CE P
TE
D
MA
NU S
Household income (quartiles)
PT
ACCEPTED MANUSCRIPT
21
ACCEPTED MANUSCRIPT
Table 2: Hierarchal logistic regression of symptoms of depression, examining the association between people with only diabetes, diabetes with
W-N B(SE)
adjORa (95% CI)
No known somatic disease Only diabetes
3216
1 0.40 (0.45) 0.33 (0.34)
MA
2.22 (0.955.18) 202 0.71 (0.27)** 2.03 (1.203.43) 4195 0.53 1.70 (1.39(0.10)*** 2.08)
0.29 (0.12)**
D
0.80 (0.43)
-0.35 (0.02)***
B(SE)
adjORac (95% CI)
Tot/ Bts /Sobel
1
1.49 (0.62-3.58) S/S/S
0.82 (0.46)
1.39 (0.71-2.73) S/S/S
0.62 (0.29)*
1.33 (1.06-1.68) S/S/S
0.50 (0.11)***
2.27 (0.935.54) 1.86 (0.935.54) 1.65 (1.342.04)
S/NS/NS S/NS/NS S/NS/NS
0.71(0.69-0.73)
CE P
Social support
Tot/ Bts /Sobel
NU S
1
TE
Diabetes with comorbidity Other Somatic Diseases Sense of mastery
70
adjORab (95% CI)
B(SE)
CR I
Disease
PT
comorbidity, other somatic diseases compared to healthy adults (control) with no known somatic disease (W-N = 7618).
-0.25 (0.03)*** 0.78 (0.730.82)
AC
Definitions: W-N = weighted N; adjOR= adjusted OR after inclusion of mediator; Tot= total effect of X on Y; ts= Bootstrap; Sob= Sobel test, B= beta. Analyses done with complex samples. Bootstrapping and Sobel test done in the module PROCESS for SPSS. Sample sizes are weighted. Significance of individual regressions: * < .05; ** < .01; *** < .001. a
OR adjusted for socio-demographic variables, BMI and physical activity. b OR adjusted for sense of mastery. c OR adjusted for social support.
22
ACCEPTED MANUSCRIPT
Table 3: Hierarchal logistic regression of symptoms of anxiety, examining the association between people with only diabetes, diabetes with
Sense of mastery Social support
4204
1.26 (0.45) ** 0.54 (0.32) 0.51 (0.12) ***
3.52 (1.458.55)
0.99 (0.48)*
1.70 (0.913.18) 1.67 (1.312.13)
0.15 (0.36) 0.26 (0.13)
CR I
1
Tot/ B(SE) Bts /Sobel
adjORac (95% CI)
Tot/ Bts /Sobel
1
NU S
207
1
S/S/S
1.32 (0.46)**
3.74 (1.539.16)
S/NS/NS
1.1 (0.58-2.34)
S/S/S
0.46 (0.35)
S/NS/NS
1.29 (0.99-1.68)
S/S/S
0.55 (0.13)***
1.58 (0.803.12) 1.73 (1.352.22)
2.69 (1.04-6.95)
MA
Diabetes with comorbidity Other Somatic Diseases
adjORab (95% CI)
B(SE)
D
No known somatic 3224 disease 69 Only diabetes
B(SE) adjORa (95% CI)
TE
W-N
S/NS/NS
-0.29 (0.02) *** 0.75 (0.73-0.77)
CE P
Disease
PT
comorbidity, other somatic diseases compared to healthy adults (control) with no known somatic disease (W-N = 7705).
-0.23 (0.03)***
0.79 (0.740.84)
AC
Definitions: W-N = weighted N; adjOR= adjusted OR after inclusion of mediator; Tot= total effect of X on Y; ts= Bootstrap; Sob= Sobel test, B= beta. Analyses done with complex samples. Bootstrapping and Sobel test done in the module PROCESS in SPSS. Sample sizes are weighted. Significance of individual regressions: * < .05; ** < .01; *** < .001. a
OR adjusted for socio-demographic variables, BMI and physical activity. b OR adjusted sense of mastery. c OR adjusted for social support.
23
ACCEPTED MANUSCRIPT Highlights
CE
PT
ED
MA NU
SC
RI P
T
Having diabetes and comorbid somatic diseases is associated with depression. Having diabetes without comorbid somatic diseases is associated with anxiety. Sense of mastery is a mediator buffering psychological distress in both groups.
AC
S1056-8727(16)30059-9 doi: 10.1016/j.jdiacomp.2016.03.022 JDC 6695
To appear in:
Journal of Diabetes and Its Complications
Received date: Revised date: Accepted date:
28 January 2016 16 March 2016 18 March 2016
Please cite this article as: Bennetter, K.E., Clench–Aas, J. & Raanaas, R.K., Sense of mastery as mediator buffering psychological distress among people with diabetes, Journal of Diabetes and Its Complications (2016), doi: 10.1016/j.jdiacomp.2016.03.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Sense of mastery as mediator buffering psychological distress
T
among people with diabetes
RI P
Karin Elisabeth Bennetter, MPH 1, Jocelyne Clench–Aas, Senior researcher 2 and Ruth
1
SC
Kjærsti Raanaas, Assoc. Prof. 1
Department of Landscape Architecture and Spatial Planning, Norwegian University of Life Sciences, Norway
MA NU
and 2 Division of mental health, Norwegian Institute of public health, Norway.
Corresponding Author: Ruth Kjærsti Raanaas, Section for Public Health Science, Department of Landscape Architecture and Spatial Planning, Norwegian University of Life Sciences, Post
AC
CE
PT
ED
Box 5003, N-1432 Ås, Tel: +47 67 23 12 68, Email address: [email protected]
ACCEPTED MANUSCRIPT Abstract Aims: The purpose of this study was to examine the association between diabetes with or
T
without other comorbid somatic diseases and depression and anxiety, and to explore the
RI P
mediating role of sense of mastery and social support. Methods: Data were obtained from a cross-sectional health survey conducted in Norway (n=6,827). People with diabetes alone or
SC
with simultaneous comorbid somatic diseases were compared to a group with no known
MA NU
somatic diseases. Results: Among people with diabetes alone, 16.3% reported having depression and anxiety. Having diabetes was associated with 3 times greater odds for anxiety compared to the control group, and 2 times greater odds for depression. Among individuals with diabetes and comorbid somatic diseases, 17.4% reported depression and 11.6% reported
ED
symptoms of anxiety. The odds for both were approximately 2 times greater than in the control group. Sense of mastery, but not social support, protected against depression in both
PT
groups and against anxiety in the diabetes with comorbidity group. Conclusions:
CE
Comorbidity between diabetes and other somatic diseases seems to be related to depression to a larger degree, whereas having diabetes alone relates more to anxiety. This can possibly be
mastery.
AC
explained by the overall burden in the comorbidity group and the related absence of sense of
Keywords: Mental health problems, psychosocial resources, chronic diseases, social support
2
ACCEPTED MANUSCRIPT 1.0 Introduction Diabetes mellitus, whether it is type 1 (T1DM) or type 2 (T2DM), are chronic and life
T
threatening diseases that requires regular monitoring of blood glucose and administration of
RI P
diet and medication. In Norway, approximately 165 000 people (3.2 % of the population) have blood glucose-lowering drugs dispensed, indicating that they are diagnosed with either
SC
T1DM or T2DM (Strom et al. 2014).
MA NU
Chronic somatic diseases, which in this case is defined as an impairment of normal physiological function affecting all or part of the organism, are frequently associated with psychological distress (Gili et al. 2010). Accordingly, people with diabetes are found to be at particular risk of developing mental health problems such as depression (Anderson et al.
ED
2001; Andreoulakis et al. 2012). In a meta-analysis, the prevalence of depression ranges from 8-18% for severe types to 15-35% for milder types among people with diabetes
PT
(Andreoulakis et al. 2012). In the same analysis, rates for depression are found for individuals
CE
with diabetes to be about 1.4-3 times as likely compared to non-diabetic controls. Comorbid depression among people with diabetes is associated with nonadherence to diabetes self-care
AC
(including dietary, exercise, medication etc.), disease control, symptom burden, and development of medical complications (Katon 2008). Diabetes is also associated with anxiety defined both as a generalized anxiety disorder and elevated symptoms of anxiety (Egede & Dismuke 2012; Smith et al. 2013). Generalized anxiety disorder are reported to be present in 14% of people living with diabetes, and a prevalence of up to 40% is reported looking at elevated symptoms of anxiety (Bener et al. 2011). In the meta-analysis by Smith and colleagues (2013) a pooled odds ratio of 1.25 for anxiety among people with diabetes was found. However, the association between diabetes and anxiety is suggested to be weaker than between diabetes and depression (Egede & 3
ACCEPTED MANUSCRIPT Dismuke 2012). The association with anxiety is linked to glycaemic control, age and lifetime severe hypoglycaemia (Labad et al. 2010; McDade-Montez & Watson 2011).
T
People with diabetes are found to be at particular risk of developing somatic
RI P
comorbidities (Fillenbaum et al. 2000; Topic et al. 2013) and the prevalence rate of psychological distress appears to increase according to the number of comorbid somatic
SC
diseases (Gili et al. 2010). An overall prevalence of serious psychological distress for individuals with diabetes and comorbidities have been reported to be 7,6% compared to 3,6%
MA NU
for people without diabetes (Egede & Dismuke 2012; Li et al. 2009). In addition, comorbid abdominal obesity and cardiovascular disease are identified as risk factors for depression among people with diabetes (Labad et al. 2010). Despite that many studies have found a
ED
relationship between diabetes and psychological distress, to our knowledge few studies have systematically differentiated between diabetes with and without somatic comorbidity when
PT
studying this relationship.
Psychosocial resources, such as sense of mastery (Raaijmakers et al. 2014) and social
CE
support (van Dam et al. 2005) are reported to protect against developing psychological
AC
distress. Sense of mastery is associated with self-management behaviors, preventative care and proper utilization of health care services, as well as responsiveness to interventions for health promotion, which may possibly account for its protective ability (Skaff et al. 2003). The concept is similar to self-efficacy (Bandura 1977), but whereas self-efficacy often is applied to the person’s feeling of competence in a specific task, sense of mastery is used more globally, referring to one’s overall sense of control in life (Skaff et al. 1996). The risk of diminished sense of mastery, is associated with an increase in both chronic conditions and functional disability (Jang et al. 2009). It has also been reported that poor sense of mastery significantly increases the risk of developing anxiety (Gordon et al. 2007).
4
ACCEPTED MANUSCRIPT Social support from family and friends are in some studies reported to facilitate better diabetes self-management (Gallant 2003; van Dam et al. 2005), although the literature is not
T
consistent (van Dam et al. 2005). Social support includes social behaviors such as seeking
RI P
advice, emotional support and instrumental support such as direct assistance with illness management activities. Self-care tasks, such as dietary behaviors and physical activity appear
SC
to be particularly susceptible to social support (Gallant 2003; Qiu et al. 2012). Since earlier studies examining the potential buffering effects of psychosocial
MA NU
resources on individuals with diabetes have not distinguished between people with diabetes alone and those with somatic comorbidity, the aim of the present study was to examine the association between diabetes and psychological distress, distinguishing between people with
ED
only diabetes and those with somatic comorbidity. Additionally, the aim was to explore the extent to which the possible association was mediated by sense of mastery and social support.
PT
The research questions were: Do relatively more people with diabetes, with or without other comorbid somatic diseases, suffer from anxiety and depression compared to people with no
CE
known somatic disease? Can sense of mastery or social support mediate the reduction of
AC
anxiety and/ or depression among people with diabetes, with or without other comorbid somatic diseases?
2.0 Materials and Methods 2.1 Design and participants The present study presents data from a cross-sectional health survey conducted in Norway in 2002 by Statistics Norway (SN) (Hougen & Gløboden 2002). A random sample of 10 000 non-institutionalized subjects over the age of 15 years was drawn to participate in the survey. The sample consisted of two subsamples, a main and a supplementary sample, each containing 5000 subjects. The main sample was drawn following SN’s standard sample plan 5
ACCEPTED MANUSCRIPT in which Norway is divided in 109 strata. The supplementary sample was drawn randomly from all of the municipalities in Norway. Participants from both samples were interviewed
T
(1/2 hour) by home visits or by phone (30%) (Hougen & Gløboden 2002). Some participants
RI P
were lost from the sample due to death, travel abroad, or institutionalization. A total of 6,827 (89.5%) participants completed the interview, with 3,410 males and 3,417 females. A written
SC
“helping card” which provided a list of 59 diseases, was used in the interview to simplify a sensitive and slightly difficult question (participants interviewed by phone were sent the cards
MA NU
in the mail). After the interview, all participants received a health questionnaire by post where data on mental health and psychosocial variables were obtained. Due to further attrition, a total of 5,343 participants responded to both interview and postal questionnaire.
and linked to the data set.
PT
2.2 Measures
ED
Information about education and household income were retrieved from national registries
Diabetes was measured with the following question: “Do you have, or have you had (i.e.)
CE
diabetes mellitus?” Score opportunities were 1= have, 2= have had, 3= have never had.
AC
Participants who indicated “having had” diabetes (13 participants with diabetes alone and 55 participants with diabetes and comorbidity) were excluded from statistical analyses since the current study focuses on chronic symptoms of diabetes. Due to lack of information in the dataset concerning type of diabetes (T1DM versus T2DM), the term diabetes will be used in this article without differentiating further. Other somatic diseases were measured with a similar question as above, but with reference to the particular disease of interest. Somatic diseases included in the current study were: epilepsy, osteoporosis, angina pectoris, coronary heart disease, stroke, cancer, allergy, high blood pressure, metabolism disease, ankylosing spondylitis (previously known as Bekhterev's disease), arthritis, chronic
6
ACCEPTED MANUSCRIPT bronchitis/emphysema/COPD, psoriasis, atopic eczema, urinary incontinence, fractures, removed organ, and ulcers. Four population subgroups were studied: 1) individuals with
T
diabetes alone; 2) individuals with diabetes and other comorbid somatic diseases; 3)
RI P
individuals who do not have diabetes but have other somatic diseases; and 4) the control group, individuals with no known somatic disease.
SC
Psychological distress in terms of depression and anxiety was assessed using the Hopkins Symptom Check List (HSCL-25), which measures symptoms over the previous 14
MA NU
days (Strand et al. 2003). It contains 25 items covering two subscales for depression and anxiety. Responses are given on a four-point scale (1=“not at all” to 4=“extremely”). In the present study, averages for each subscale are dichotomized into “low” and “high” scores with
ED
a cut-off point of 1.75 (Strand et al. 2003). A “case” is accepted if no more than two items are missing from the items measuring depression or anxiety, in which case mean values were
PT
substituted for the missing items. Cronbach’s alpha for the total HSCL-25 scale was .93; for depression it was .91 and for anxiety .84.
CE
Sense of mastery was measured using the five-item version of the seven-item scale
AC
developed by Pearlin and colleagues (1981). The five selected items ask about the experience of controlling and coping with life such as: “I have little control over the things that happen to me” and “I often feel helpless in dealing with the problems of life”. Responses were given on a five point scale (0=“agree” to 4=“do not agree”). For those with no more than two missing data points, mean values were substituted for the missing items. A sum score was calculated ranging from 0-20. Cronbach’s alpha was 0.86. Social support was measured using the Oslo 3 Support Scale (OSS-3 scale) (Dalgard et al. 2006). It is comprised of 3 questions about the number of close confidants they can count on if they have serious problems (scores were: 1=”no one”, 2=“1 or 2”, 3=“3-5” or
7
ACCEPTED MANUSCRIPT 4=“more than 5”), sense of concern or interest from other people (scores were: 1=“Great concern and interest” to 5=”no concern and interest”), and sense of instrumental support from
T
neighbors (scores were: 1=”very easy” to 5=“very difficult”). The corresponding scores in the
RI P
two last questions were recoded in reverse order. The questions were further merged into a social support index by adding up the scores for each item, ranging from 3-14. Data had to be
SC
complete (no missing) to count as a case.
Data on gender and age were obtained from the interview section. Age was
MA NU
categorized into 4 groups (1=”16-24”, 2=”25-44”, 3=”45-67” and 4=”>67 years”). Education and income were obtained from linked public registries. Education was categorized into 3 groups; Low Education (primary School and junior high school (year 1-10)), Middle
ED
Education (high school and first level of junior college without accomplished degree (year 11-14)), High Education (university or university college (above 14)). Household income
PT
defined as the total gross annual income of the family was divided into quartiles. Physical activity was measured by how many times per week the subject exercised (0=”never”,
CE
1=”less than 1/week” and 2=”1/week or more”). BMI was measured by kilograms/ height
1995).
AC
(m2) and classified into 6 categories after World Health Organizations guidelines (WHO
2.3 Statistical analysis All analyses were performed using the Statistical Package for the Social Sciences (SPSS for Windows, version 22.0). Hierarchical models of multiple logistic regression were conducted to test the associations between the predicting variables (diabetes with and without comorbidity, and other somatic diseases) and the dependent variables anxiety and depression (dichotomous versions; > 1.75). Odds ratio (OR) with a 95% confidence interval as well as beta (B) with SE were reported as the estimated outcomes. The data was weighted to account
8
ACCEPTED MANUSCRIPT for information collected from non-respondents in national registries by adjusting for gender, age, highest education level and family size, by way of adjusting the standard errors for the
T
modified sample size (Osborne 2011) using the complex sample module of SPSS.
RI P
Demographics, BMI and physical activity were all entered in the first step. Type of diagnosis was entered as a categorical factor in the next step. All the three groups of
SC
diagnosis were compared to the control group. Sense of mastery and social support (potential mediators) were entered separately at the final step and one at a time, testing mediating
MA NU
effects using the methods described by Kim and colleagues (2001). Moderation was tested using the PROCESS module for SPSS provided by Hayes and Little (2013) (model 1). Mediation (both individually and parallel with both mediators in simultaneously) was further
ED
tested using the bootstrap method and the Sobel test both provided in the PROCESS module for SPSS (model 4).
PT
Model fit was evaluated through R2, which were significant for both anxiety and depression. Missing data were excluded casewise. To test the inter-correlation among the
CE
predicting variables, multicollinearity tests were conducted and showed no violation of
AC
multicollinearity assumptions (Tolerance values were above 0.1 and VIF values under 10).
3.0 Results
Table 1 presents the weighted prevalence of responses to questions regarding depression, anxiety, sense of mastery, social support, demographics, BMI and physical activity for the four diagnostic groups. Among the 5,342 (non-weighted N) participants who responded to both interview and postal questionnaire, 3.6% reported being diagnosed with diabetes, of which 2.7% reported additional comorbidity, and 0.9% of participants reported being diagnosed with diabetes alone. In total, 54.5% of participants had somatic diseases other than
9
ACCEPTED MANUSCRIPT diabetes, and 41.9% constituted a control group without a known somatic disease. Nearly 10% of participants reported having depression.
T
Relatively more people with diabetes alone and with comorbid somatic diseases have
RI P
depression (17.8 %) compared to the control group (8.2 %). This is also the case for people with somatic diseases other than diabetes. Adjusting for demographics, BMI and physical
depression than the control group (Table 2).
SC
activity, people with diabetes and comorbid somatic diseases have twice the odds for
MA NU
Although not significant, there is a trend for higher levels of anxiety in people with diabetes alone than in the control group (18.8 % vs 5.6 %). Adjusting for demographics, BMI and physical activity, people with diabetes alone were significantly associated with more than
ED
3 times greater odds for anxiety (Table 3), whereas for people with diabetes and comorbid somatic diseases there was no significant relationship with anxiety. The association between
PT
anxiety and the disease category somatic diseases other than diabetes was significant and the odds were approximately 1.7.
CE
In total, 18.5% reported low levels of sense of mastery. More than 40 % of the
AC
population with diabetes with comorbidity, and a third of those with diabetes alone report low sense of mastery. No significant moderation effects were found for sense of mastery. Examining mediator effects of sense of mastery, in the hierarchal analysis of both depression and anxiety (Tables 2 and 3), the addition of sense of mastery resulted in a reduction by nearly half of all the betas. The odds were also considerably reduced. Sense of mastery in itself was significantly and negatively associated with the outcome variables of depression and anxiety. Using both the bootstrap method and the Sobel test, the indirect mediator pathway was significant for all disease categories.
10
ACCEPTED MANUSCRIPT Ten % of the population report low levels of social support. Levels are only slightly higher in those with other somatic disease alone and diabetes alone, whereas they are nearly
T
25% in those with diabetes with comorbidity. Also for social support the moderation
RI P
analyses, found no significant moderation effects. Examining mediator effects of social support, the addition of social support did not alter any of the betas for either depression or
SC
anxiety (Tables 2 and 3). The odds also remained unchanged. Despite that social support was in itself significantly and negatively associated with the outcome variables of depression and
MA NU
anxiety, using both the bootstrap method and the Sobel test, the indirect mediator path was non-significant for all disease categories. It should be noted that the relationship between the disease category and sense of mastery was highly significant, with estimates of beta ranging
ED
from 0.3 for somatic diseases to 1.96 for diabetes alone (data not shown). However, the similar relationship between disease category and social support were non-significant for all
PT
categories.
The results of the mediator analyses with simultaneous testing for social support and
CE
sense of mastery resulted in non-significance for social support both for anxiety and
AC
depression and for all disease categories, but significance for sense of mastery in all situations (data not shown).
4.0 Discussion
In the current study people with diabetes and comorbid somatic diseases were twice as likely to suffer from depression than people with no known somatic disease, when controlling for socio-demographics, BMI and physical activity. Elevated odds for depression were, however, not found among people with diabetes alone in the present study. These findings highlights the need to differentiate between people with diabetes with and without comorbidities when examining the association with psychological distress, and may be explained by the total 11
ACCEPTED MANUSCRIPT burden of having more chronic diseases (Gili et al. 2010), or the combination of diabetes with somatic diseases in the cardiovascular domain previously identified as risk factor for
T
depression (Labad et al. 2010). Previous studies, having found associations between diabetes
RI P
and depression (Anderson et al. 2001; Andreoulakis et al. 2012), have not clearly differentiated between people with diabetes with and without comorbidities, and it is thus
SC
difficult to compare findings.
People with diabetes alone where found in the current study to be three times more
MA NU
likely to experience anxiety compared to the healthy population. Elevated odds for anxiety among people with diabetes are found in earlier studies (Smith et al. 2013). However, the literature is somewhat inconsistent, and Clarke and Currie (2009) report that anxiety is not
ED
common among people with diabetes, but rather in other somatic diseases such as heart disease, stroke and cancer. These conflicting results again bring up the importance of
PT
distinguishing people with diabetes with or without somatic comorbidities. Since anxiety has been linked to glycemic control, age and lifetime severe hypoglycaemia, one may suggest a
CE
lack of ability to cope with the disease as the most vital factor (Labad et al. 2010; McDade-
AC
Montez & Watson 2011).
Chronic conditions and functional disability can gradually decrease psychosocial resources such as sense of mastery and social support (Jang et al. 2009; Raaijmakers et al. 2014). In the present study, among people with diabetes with comorbid somatic diseases as well as for those with somatic diseases alone, sense of mastery played a mediator role both for depression and for anxiety. This indicates that not only may there be a direct association between having comorbid diabetes or somatic diseases and psychological distress, but these conditions may also be associated with depression via the indirect pathway of reducing sense of mastery. Thus the measured low levels of sense of mastery, especially among those with
12
ACCEPTED MANUSCRIPT diabetes as opposed to other somatic diseases, seemed to be associated with an increase in depression and anxiety. Low sense of mastery among people with diabetes may be associated
T
with more difficulties in adhering to treatment regimens (Skaff et al. 2003), or coping with
RI P
the disease. The reported higher prevalence of low sense of mastery among people with diabetes with comorbidity as compared to people with diabetes alone, can thus be explained
SC
by similar combination of challenges related to living with a chronic condition and functional disability (Jang et al. 2009). Increased coping skills and adherence to health promotion
MA NU
interventions may be one way of increasing the sense of mastery (Skaff et al. 2003; Thorpe et al. 2013), that can again serve as a psychological resource to possibly reduce both depression and anxiety among people with diabetes with and without comorbidity.
ED
Despite that low social support previously is reported to elevate the perceived psychological burden of a chronic disease (Renn et al. 2011), leading to increased depression,
PT
social support was not associated with reduce depression, or anxiety in this sample. Diabetes is a disease often associated with lifestyle choices that complicate the relationship with social
CE
support, and social support may change, pertaining to if or not, the individual adheres to
AC
recommended lifestyle changes (Gallant 2003; Qiu et al. 2012; van Dam et al. 2005). 4.1 Methodological limitations The study is based on a large and nationally representative sample, with a high response rate (70.4%). Despite the large sample, however, the numbers with diabetes alone were quite low. Further, the cross-sectional design restricts the ability to draw causal inferences concerning the direction of the association between diabetes and psychological distress. Since information on psychological distress and physical health was based on retrospective selfreport without independent verification by a medical professional, bias can occur.
13
ACCEPTED MANUSCRIPT 5.0 Conclusions Relatively more people with diabetes, with or without other comorbid somatic diseases,
T
suffer from psychological distress compared to people with no known somatic disease. The
RI P
present study indicates that people with diabetes and comorbid somatic diseases are to a larger degree suffering from depression, whereas those with diabetes alone are primarily
SC
associated with anxiety. To be able to target help to the different groups our study indicates
MA NU
that when examining comorbidity with psychological distress, people with the different somatic illnesses should be studied separately. Health care should include means for increasing sense of mastery in people with diabetes at an early stage to improve diabetes coping, such as self-management education and support groups.
ED
6.0 Acknowledgements
This study was performed in collaboration with the Norwegian Institute of Public Health,
PT
division of Mental Health. We would like to thank the institute and the division for providing
CE
us with access to data, workspace and support. In addition we would like to thank Statistics
AC
Norway (SSB) for collection and distribution of data.
The Author(s) declare(s) that there is no conflict of interest.
7.0 References
Anderson, R. J., Freedland, K. E., Clouse, R. E. & Lustman, P. J. (2001). The Prevalence of Comorbid Depression in Adults With Diabetes: A meta-analysis. Diabetes Care, 24 (6): 1069-1078. Andreoulakis, E., Hyphantis, T., Kandylis, D. & Iacovides, A. (2012). Depression in diabetes mellitus: a comprehensive review. Hippokratia, 16 (3): 205-214.
14
ACCEPTED MANUSCRIPT Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84 (2): 191-215.
T
Bener, A., Al-Hamaq, A. O. A. A. & Dafeeah, E. E. (2011). High Prevalence of Depression,
RI P
Anxiety and Stress Symptoms Among Diabetes Mellitus Patients. The Open Psychiatry
SC
Journal, 5: 5-12.
Clarke, D. M. & Currie, K. C. (2009). Depression, anxiety and their relationship with chronic
Australia, 190 (7 Suppl): S54-60.
MA NU
diseases: a review of the epidemiology, risk and treatment evidence. The Medical Journal of
Dalgard, O. S., Dowrick, C., Lehtinen, V., Vazquez-Barquero, J. L., Casey, P., Wilkinson,
ED
G., Ayuso-Mateos, J. L., Page, H. & Dunn, G. (2006). Negative life events, social support and gender difference in depression: a multinational community survey with data from the
PT
ODIN study. Social Psychiatry and Psychiatric Epidemiology, 41 (6): 444-51.
CE
Egede, L. E. & Dismuke, C. E. (2012). Serious psychological distress and diabetes: a review
AC
of the literature. Current Psychiatry Reports, 14 (1): 15-22. Fillenbaum, G. G., Pieper, C. F., Cohen, H. J., Cornoni-Huntley, J. C. & Guralnik, J. M. (2000). Comorbidity of Five Chronic Health Conditions in Elderly Community Residents: Determinants and Impact on Mortality. Journal of Gerontology: Medical Sciences, 55 (2): 84-89. Gallant, M. P. (2003). The Influence of Social Support on Chronic Illness Self-Management: A Review and Directions for Research. Health Education & Behavior, 30 (2): 170-195.
15
ACCEPTED MANUSCRIPT Gili, M., Comas, A., Garcia-Garcia, M., Monzon, S., Antoni, S. B. & Roca, M. (2010). Comorbidity between common mental disorders and chronic somatic diseases in primary care
T
patients. General Hospital Psychiatry, 32 (3): 240-245.
RI P
Gordon, A., Smith, R., Keramatian, K., Luus, B., Weinberg, A., Smallwood, J., Schooler, J. & Christoff, K. (2007). Mind-wandering, awareness, and task performance: An fMRI study.
SC
Canadian Journal of Experimental Psychology-Revue Canadienne De Psychologie
MA NU
Experimentale, 61 (4): 366-366.
Hayes, A. F. & Little, T. D. (2013). Introduction to mediation, moderation, and conditional process analysis: a regression-based approach. New York: Guilford Press. 507 pp.
ED
Hougen, H. C. & Gløboden, M. A. (2002). Samordnet levekårsundersøkelse 2002 -
PT
tverrsnittsundersøkelsen. Dokumentasjonsrapport. Oslo: Statistics Norway. Jang, Y., Chiriboga, D. A., Lee, J. & Cho, S. (2009). Determinants of a sense of mastery in
CE
Korean American elders: a longitudinal assessment. Aging & Mental Health, 13 (1): 99-105.
AC
Katon, W. J. (2008). The Comorbidity of Diabetes Mellitus and Depression. The American Journal of Medicine, 121 (11 Suppl 2): S8-15. Kim, J. S., Kaye, J. & Wright, L. K. (2001). Moderating and mediating effects in causal models. Issues Ment Health Nurs, 22 (1): 63-75. Labad, J., Price, J. F., Strachan, M. W. J., Fowkes, F. G. R., Ding, J., Deary, I. J., Lee, A. J., Frier, B. M., Seckl, J. R., Walker, B. R., et al. (2010). Symptoms of depression but not anxiety are associated with central obesity and cardiovascular disease in people with Type 2 diabetes: The Edinburgh Type 2 diabetes study. Diabetologia, 53 (3): 467-471.
16
ACCEPTED MANUSCRIPT Li, C., Ford, E., Zhao, G., Strine, T., Dhingra, S., Barker, L., Berry, J. & Mokdad, A. (2009). Association between diagnosed diabetes and serious psychological distress among U.S.
T
adults: the Behavioral Risk Factor Surveillance System, 2007. International Journal of
RI P
Public Health, 54 (1): 43-51.
McDade-Montez, E. A. & Watson, D. (2011). Examining the potential influence of diabetes
MA NU
Annals of Behavioral Medicine, 42 (3): 341-51.
SC
on depression and anxiety symptoms via multiple sample confirmatory factor analysis.
Osborne, J. W. (2011). Best Practices in using large, complex samples: The importance of using appropriate weights and design effect compensation. Practical Assessment, Research &
ED
Evaluation, 16 (12): 1-7.
Pearlin, L. I., Lieberman, M. A., Menaghan, E. G. & Mullan, J. T. (1981). The stress process.
PT
Journal of Health and Social Behavior. American Sociological Association, 22 (4): 337-56.
CE
Qiu, S. H., Sun, Z. L., Cai, X., Liu, L. & Yang, B. (2012). Improving patients' adherence to
5.
AC
physical activity in diabetes mellitus: a review. Diabetes and Metabolism Journal, 36 (1): 1-
Raaijmakers, L. G. M., Martens, M. K., Hesselink, A. E., de Weerdt, I., de Vries, N. K. & Kremers, S. P. J. (2014). Mastery and perceived autonomy support are correlates of Dutch diabetes patients’ self-management and quality of life. Patient Education and Counseling, 97 (1): 75-81. Renn, B. N., Feliciano, L. & Segal, D. L. (2011). The bidirectional relationship of depression and diabetes: a systematic review. Clinical Psychology Review, 31 (8): 1239-46.
17
ACCEPTED MANUSCRIPT Skaff, M. M., Pearlin, L. I. & Mullan, J. T. (1996). Transitions in the caregiving career: Effects on sense of mastery. Psychology and Aging, 11 (2): 247-257.
T
Skaff, M. M., Mullan, J. T., Fisher, L. & Chesla, C. A. (2003). A contextual model of control
RI P
beliefs, behavior, and health; Latino and European Americans with type 2 diabetes.
SC
Psychology and Health, 18 (3): 295-312.
Smith, K. J., Beland, M., Clyde, M., Gariepy, G., Page, V., Badawi, G., Rabasa-Lhoret, R. &
MA NU
Schmitz, N. (2013). Association of diabetes with anxiety: a systematic review and metaanalysis. Journal of Psychosomatic Research, 74 (2): 89-99. Strand, B. H., Dalgard, O. S., Tambs, K. & Rognerud, M. (2003). Measuring the mental
ED
health status of the Norwegian population: a comparison of the instruments SCL-25, SCL-10,
PT
SCL-5 and MHI-5 (SF-36). Nordic Journal of Psychiatry, 57 (2): 113-118. Strom, H., Selmer, R., Birkeland, K. I., Schirmer, H., Berg, T. J., Jenum, A. K., Midthjell, K.,
CE
Berg, C. & Stene, L. C. (2014). No increase in new users of blood glucose-lowering drugs in
AC
Norway 2006-2011: a nationwide prescription database study. BMC Public Health, 14: 520. Thorpe, C. T., Fahey, L. E., Johnson, H., Deshpande, M., Thorpe, J. M. & Fisher, E. B. (2013). Facilitating healthy coping in patients with diabetes: a systematic review. Diabetes Educator, 39 (1): 33-52. Topic, R., Milicic, D., Stimac, Z., Loncar, M., Velagic, V., Marcinko, D. & Jakovljevic, M. (2013). Somatic comorbidity, metabolic syndrome, cardiovascular risk, and CRP in patients with recurrent depressive disorders. Croatian medical journal, 54 (5): 453-459.
18
ACCEPTED MANUSCRIPT van Dam, H. A., van der Horst, F. G., Knoops, L., Ryckman, R. M., Crebolder, H. F. & van den Borne, B. H. (2005). Social support in diabetes: a systematic review of controlled
AC
CE
PT
ED
MA NU
SC
RI P
T
intervention studies. Patient Education and Counseling, 59 (1): 1-12.
19
ACCEPTED MANUSCRIPT
Table 1: Weighted prevalence W-N (%) of responses on the questions concerning depression, anxiety, sense of mastery, social support,
CR I
PT
demographics, BMI and physical activity, according to diagnostic category. Response alternatives
Control group (W-N=4075(41,9%))
No diabetes, but other somatic diseases (W-N=5307(54,5%))
Only diabetes (W-N=84(0,9%))
Diabetes with comorbidity (W-N=267(2,7%))
Depression
>1.75
280 (8.2)
629 (14.2)
12 (16.9)
38 (17.8)
Anxiety
>1.75
190 (5.6)
441 (9.9)
13 (18.8)
26 (11.9)
Sense of mastery
Mean (SE)
15.3 (.07)
14.0 (.07)
12.8 (.48)
12.6 (.33)
Social support
Mean (SE)
11.2 (.03)
11.0 (.03)
10.9 (.18)
10.2 (.15)
Gender
Male
2265 (55.6)
2472 (46.6)
54 (65.0)
134 (50.2)
Age
15-24 25-44 45-66 67
701 (17.2) 1807 (44.4) 1275 (31.3) 289 (7.1)
630 (11.9) 1732 (32.7) 1877 (35.4) 1061 (20.0)
11 (13.5) 31 (36.9) 17 (20.9) 24 (28.7)
5 (1.7) 34 (12.7) 119 (44.6) 110 (41.0)
Education
Low Middle High
616 (15.6) 2274 (57.7) 1057 (26.7)
1153 (22.3) 2900 (56.1) 1115 (21.6)
24 (28.8) 45 (55.0) 13 (16.3)
105 (34.0) 121 (46.3) 36 (13.7)
AC
CE P
TE
D
MA
NU S
Variables
20
773 (19.0) 726 (17.9) 1762 (43.4) 803 (19.8)
1334 (25.2) 1072 (20.2) 2069 (39.1) 823 (15.5)
No of days exercise/week
Never <1/week ≥1/week
953 (23.5) 476 (11.7) 2634 (64.8)
1420 (26.8) 545 (10.3) 3331 (62.9)
BMI categorized
Normal Underweight Overweight Obesity Obesity class 2 Obesity class 3
2391 (60.3) 83 (2.1) 1270 (32.0) 199 (5.0) 21 (0.5) 3 (0.1)
2677 (5246) 92 (1.8) 1849 (36.2) 411 (8.1) 65 (1.3) 14 (0.3)
CR I
Lowest quartile 2nd quartile 3rd quartile Highest quartile
21 (25.3) 24 (28.7) 29 (34.3) 10 (11.8)
97 (36.2) 67 (25.2) 77 (28.7) 27 (9.9)
25 (33.8) 5 (6.3) 50 (59.9)
108 (40.7) 16 (6.1) 142 (53.2)
48 (58.3) 0 (0.0) 28 (34.2) 3 (4.1) 3 (3.3) 0 (0.0)
75 (29.2) 6 (2.2) 113 (44.1) 48 (18.7) 8 (3.2) 6 (2.5)
AC
CE P
TE
D
MA
NU S
Household income (quartiles)
PT
ACCEPTED MANUSCRIPT
21
ACCEPTED MANUSCRIPT
Table 2: Hierarchal logistic regression of symptoms of depression, examining the association between people with only diabetes, diabetes with
W-N B(SE)
adjORa (95% CI)
No known somatic disease Only diabetes
3216
1 0.40 (0.45) 0.33 (0.34)
MA
2.22 (0.955.18) 202 0.71 (0.27)** 2.03 (1.203.43) 4195 0.53 1.70 (1.39(0.10)*** 2.08)
0.29 (0.12)**
D
0.80 (0.43)
-0.35 (0.02)***
B(SE)
adjORac (95% CI)
Tot/ Bts /Sobel
1
1.49 (0.62-3.58) S/S/S
0.82 (0.46)
1.39 (0.71-2.73) S/S/S
0.62 (0.29)*
1.33 (1.06-1.68) S/S/S
0.50 (0.11)***
2.27 (0.935.54) 1.86 (0.935.54) 1.65 (1.342.04)
S/NS/NS S/NS/NS S/NS/NS
0.71(0.69-0.73)
CE P
Social support
Tot/ Bts /Sobel
NU S
1
TE
Diabetes with comorbidity Other Somatic Diseases Sense of mastery
70
adjORab (95% CI)
B(SE)
CR I
Disease
PT
comorbidity, other somatic diseases compared to healthy adults (control) with no known somatic disease (W-N = 7618).
-0.25 (0.03)*** 0.78 (0.730.82)
AC
Definitions: W-N = weighted N; adjOR= adjusted OR after inclusion of mediator; Tot= total effect of X on Y; ts= Bootstrap; Sob= Sobel test, B= beta. Analyses done with complex samples. Bootstrapping and Sobel test done in the module PROCESS for SPSS. Sample sizes are weighted. Significance of individual regressions: * < .05; ** < .01; *** < .001. a
OR adjusted for socio-demographic variables, BMI and physical activity. b OR adjusted for sense of mastery. c OR adjusted for social support.
22
ACCEPTED MANUSCRIPT
Table 3: Hierarchal logistic regression of symptoms of anxiety, examining the association between people with only diabetes, diabetes with
Sense of mastery Social support
4204
1.26 (0.45) ** 0.54 (0.32) 0.51 (0.12) ***
3.52 (1.458.55)
0.99 (0.48)*
1.70 (0.913.18) 1.67 (1.312.13)
0.15 (0.36) 0.26 (0.13)
CR I
1
Tot/ B(SE) Bts /Sobel
adjORac (95% CI)
Tot/ Bts /Sobel
1
NU S
207
1
S/S/S
1.32 (0.46)**
3.74 (1.539.16)
S/NS/NS
1.1 (0.58-2.34)
S/S/S
0.46 (0.35)
S/NS/NS
1.29 (0.99-1.68)
S/S/S
0.55 (0.13)***
1.58 (0.803.12) 1.73 (1.352.22)
2.69 (1.04-6.95)
MA
Diabetes with comorbidity Other Somatic Diseases
adjORab (95% CI)
B(SE)
D
No known somatic 3224 disease 69 Only diabetes
B(SE) adjORa (95% CI)
TE
W-N
S/NS/NS
-0.29 (0.02) *** 0.75 (0.73-0.77)
CE P
Disease
PT
comorbidity, other somatic diseases compared to healthy adults (control) with no known somatic disease (W-N = 7705).
-0.23 (0.03)***
0.79 (0.740.84)
AC
Definitions: W-N = weighted N; adjOR= adjusted OR after inclusion of mediator; Tot= total effect of X on Y; ts= Bootstrap; Sob= Sobel test, B= beta. Analyses done with complex samples. Bootstrapping and Sobel test done in the module PROCESS in SPSS. Sample sizes are weighted. Significance of individual regressions: * < .05; ** < .01; *** < .001. a
OR adjusted for socio-demographic variables, BMI and physical activity. b OR adjusted sense of mastery. c OR adjusted for social support.
23
ACCEPTED MANUSCRIPT Highlights
CE
PT
ED
MA NU
SC
RI P
T
Having diabetes and comorbid somatic diseases is associated with depression. Having diabetes without comorbid somatic diseases is associated with anxiety. Sense of mastery is a mediator buffering psychological distress in both groups.
AC