A call for inclusion of work-related diabetes distress in the spectrum of diabetes management: Results from a cross-sectional survey among working people with type 1 diabetes

Accepted Manuscript A call for inclusion of work-related diabetes distress in the spectrum of diabetes management: Results from a cross-sectional survey among working people with type 1 diabetes Ulla M. Hansen, Kasper Olesen, Jessica L. Browne, Timothy C. Skinner, Ingrid Willaing PII: DOI: Reference:

S0168-8227(17)31912-5 https://doi.org/10.1016/j.diabres.2018.03.040 DIAB 7297

To appear in:

Diabetes Research and Clinical Practice

Received Date: Revised Date: Accepted Date:

4 December 2017 27 February 2018 23 March 2018

Please cite this article as: U.M. Hansen, K. Olesen, J.L. Browne, T.C. Skinner, I. Willaing, A call for inclusion of work-related diabetes distress in the spectrum of diabetes management: Results from a cross-sectional survey among working people with type 1 diabetes, Diabetes Research and Clinical Practice (2018), doi: https://doi.org/10.1016/ j.diabres.2018.03.040

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.

Work-related diabetes distress

1

A call for inclusion of work-related diabetes distress in the spectrum of diabetes management: Results from a cross-sectional survey among working people with type 1 diabetes

Ulla M. Hansen1,2, Kasper Olesen1, Jessica L. Browne3 4, Timothy C. Skinner5, Ingrid Willaing1 1

Steno Diabetes Center Copenhagen, Gentofte, Denmark

2

University of Southern Denmark, National Institute of Public Health, Copenhagen, Denmark

3

Deakin University, Geelong, School of Psychology, Victoria, Australia

4

The Australian Centre for Behavioural Research in Diabetes, Diabetes Victoria, Melbourne,

Victoria, Australia 5

University Copenhagen, Department of Psychology, Denmark

Corresponding author: Ulla Møller Hansen Steno Diabetes Center Copenhagen Diabetes Management Research Niels Steensens vej 6 DK-2820 Gentofte Denmark [email protected] +45 30913109 Word count: 4,224 including tables Tables: 5 Running head: Work-related diabetes distress

Work-related diabetes distress

2

ABSTRACT Aim Diabetes distress captures a range of emotional responses and reactions to life with diabetes and is considered a part of the experience of managing diabetes and its treatment. Given the importance of the social context of work life for people of working age we set out to explore whether work-related diabetes distress is a distinct and important dimension of diabetes-related emotional distress in working people with type 1 diabetes. Methods A questionnaire with self-reported measures of psychosocial health and well-being at work was completed by 1126 working people with type 1 diabetes from a specialist diabetes clinic in Denmark. Work-related diabetes distress was assessed with two questions about worry and exhaustion related to reconciling work life and diabetes. Diabetes-related emotional distress was assessed with the Problem Areas in Diabetes scale (PAID-5), a short form version of the full PAID scale. We performed inter-item correlation analyses, exploratory factor analysis, and hierarchical multiple regression analyses. Results Inter-item correlations and exploratory factor analysis indicated that work-related diabetes distress was distinct from diabetes-related emotional distress. Further, work-related diabetes distress was found to be a unique contributor to work ability, quality of life, intentional hyperglycaemia at work, and absenteeism, after adjusting for covariates and diabetes-related emotional distress. Conclusions The findings suggest that work-related diabetes distress captures an aspect of distress so far unaccounted for in workers with type 1 diabetes. Further studies are needed to strengthen the

Work-related diabetes distress

conceptual basis of work-related diabetes distress, explore its clinical usefulness and clarify its risk factors.

Keywords type 1 diabetes mellitus; work-related diabetes distress; well-being; diabetes distress

Highlights •

Work-related diabetes distress is a distinct problem area so far unaccounted for.

•

Work-related diabetes distress is a unique contributor to lower worker well-being.

•

Further validation of work-related diabetes distress is needed.

3

Work-related diabetes distress

4

1. INTRODUCTION Diabetes distress, defined as emotional burdens, stresses and worries associated with the constant behavioral demands of diabetes self-management is considered one of the most important psychosocial concerns in people with diabetes [1,2,3]. Individuals with type 1 diabetes carry the daily responsibility of managing the condition by keeping track of blood glucose levels and adjusting insulin, carbohydrate intake, and physical activity accordingly [4]. Due to the need for tight blood glucose control and the unpredictable nature of the disease, managing type 1 diabetes places significant physical, emotional, and social demands on the individuals affected [5]. According to Fisher et al., diabetes distress is context-specific and caused by specific stressors such as the emergence of a new complication [3]; however, one seminal context—work life—is missing from the most widely used measures of diabetes distress, namely the Problem Areas in Diabetes Scale [6] and the Diabetes Distress Scale [7] developed in samples in which the majority had type 1 diabetes and type 2 diabetes, respectively. A noteworthy exception is one item in the recently developed type 1-specific T1-DDS treating potential concerns that the diabetes makes the person less attractive to employers [8] but no other questions address the distress attributable to worries of reconciling diabetes management and work life. In this paper we explore whether diabetes distress related to work life should be considered a part of the “spectrum of diabetes management” [9] among working people with type 1 diabetes given the importance of the social context of work life for people of working age. Previous studies from Sweden have demonstrated that employment status and earnings are negatively affected by type 1 diabetes [10,11] and Danish cross-sectional evidence found that type 1 diabetes was associated with more sick leave and decreased health-related quality of life compared with the background population [12]. These quantitative results are corroborated by qualitative studies demonstrating how people with type 1 diabetes in the workforce may

Work-related diabetes distress

5

experience strong internal and external pressure to sideline diabetes management to avoid taking time away from work-related activities [13, 14]. These results indicate that reconciling diabetes and work life can be burdensome and ultimately lead to adverse outcomes for working people with diabetes, the workplace, and society at large. In a recent Finnish study [15], the first attempt to capture the psychosocial burden of diabetes in work life used the novel self-reported measure of work-related diabetes distress (WRDD). The measure consisted of two questions about worry and exhaustion related to balancing work with diabetes, which were combined to a sum score for WRDD [15]. Using structural equation modeling, the authors found that WRDD was independently associated with maintaining a high blood glucose level at work (hereafter referred to as intentional hyperglycaemia at work), which was in turn associated with a high haemoglobin A1c (HbA1c) level. The authors were the first to coin the term “work-related diabetes distress” in a population-level study; however, the measure construction did not allow direct comparison with the foundational concept of diabetes distress nor did the authors include a measure of diabetes distress in the study. To create new insights into the behavioral and psychosocial burden of type 1 diabetes in the context of work life, our aim was to test the hypothesis that work-related diabetes distress forms an independent and distinct dimension of diabetes-related emotional distress that predicts the wellbeing of workers. We used a broad concept of worker well-being as umbrella term for well-being and functioning in the context of work life [16] and diabetes-related emotional distress was assessed with a short form of the Problem Areas in Diabetes Scale, PAID-5 [17]. 2. MATERIALS AND METHODS 2.1 Participant recruitment and data collection The study was registered with The Danish Data Protection Agency (j.nr. SDC-2015033 DiA/I-Suite nr. 03813). A population of 3053 adults was identified through the electronic

Work-related diabetes distress

6

health records at a specialist diabetes clinic in Denmark using the following inclusion criteria: a diagnosis of type 1 diabetes, age 18-70 years, and currently receiving treatment at the clinic. Data collection occurred over a 12-week period in the last quarter of 2016 using an online survey. To minimize missing data, the online survey was designed so that participants were required to respond to all questions on one page before proceeding to the next. Reminders were sent to non-respondents after approximately 3 and 6 weeks. The last reminder included a paper version of the survey and a pre-paid return envelope. 2.2 Measures Diabetes-related emotional distress The PAID-5 is a psychometrically robust short-form measure of diabetes-related emotional distress [6][17]. Participants were asked to rate ”Which of the following issues are currently a problem for you?” 1) “Feeling scared when you think about living with diabetes?”, 2) “Feeling depressed when you think about living with diabetes?”, 3) “Worrying about the future and the possibility of serious complications?”, 4) “Feeling that diabetes is taking up too much of your mental and physical energy every day?”, 5) “Coping with complications of diabetes?”. Each of the five questions is scored on a 5-point Likert scale from “Not a problem” to “A serious problem” and can be summed to produce a total score of diabetes-related emotional distress. Higher scores indicate higher distress. Work-related diabetes distress We adapted the two questions from the previous study exploring WRDD [15] to the format of the widely used and validated PAID scale. WRDD was thus assessed by asking participants to rate ”Which of the following issues are currently a problem for you?” 1) “Worrying about your ability to do your job due to your diabetes” and 2) “ Often feeling exhausted by simultaneously reconciling work and diabetes.” Each item was scored on a 5-point Likert scale from “Not a

Work-related diabetes distress

7

problem” to “A serious problem”, and scores for both questions were summed to produce a total WRDD score. Higher scores indicate higher distress. Markers of worker well-being Worker well-being is a broad concept with no consensus definition [18, 19]. For the purposes of this study, we used four markers of worker well-being as dependent variables: work ability, global quality of life, intentional hyperglycaemia at work, and absenteeism. Using the Work Ability score of the Work Ability Index, respondents were asked to rate their current work ability on a scale from 0 (incapable of work) to 10 (lifetime best work ability) [20, 21]. This interval variable was treated as a continuous variable for further analysis. To assess global quality of life (QoL), we asked respondents to rate their quality of life on a 5-point Likert scale from “Very good” to “Very poor” [22]. A single item was also used to assess intentional hyperglycaemia; respondents were asked to rate the frequency with which they deliberately maintained a higher-than-recommended blood glucose level at work on a 5-point Likert scale from “Never” to ”Always” [15]. This diabetes-specific measure has previously been found to be associated with work-related diabetes distress and lower well-being [15]. We also asked respondents to report the number of days they had missed work due to their own illness or to attend healthcare appointments in the past 12 months. Response options included none, ≤ 9 days, 10-24 days, 25-99 days, and 100-365 days. The survey also included questions related to educational attainment. Information on age, gender, and diabetes duration, treatment, and complications was obtained from electronic health records. 2.3 Statistical analysis To explore the relationship between PAID-5 and WRDD, we performed pair-wise inter-item correlations of the five PAID-5 items and the two WRDD items. We then performed an exploratory

Work-related diabetes distress

8

factor analysis of the five PAID-5 items and the two WRDD items using maximum likelihood as the method of extraction. We used oblique rotation because we expected the factors to correlate. In the rotated pattern, we looked for a simple, differentiated structure to ease interpretation [23]. We used a cut-off of > 0.4 to identify meaningful factor loadings [24]. Finally, we constructed four multiple regression models with work ability, quality of life, intentional hyperglycaemia at work, and absenteeism, respectively. Variables were included stepwise in the following order: 1) covariates, 2) PAID-5, and 3) WRDD. Our aim was to see if WRDD was independently associated with each of the four dependent variables. Covariates were selected based on theory and existing evidence (gender, age, diabetes duration, treatment, complications, and education) and included in each model based on significant associations with the four dependent variables as assessed with Spearman’s correlation coefficients and Student's t-tests. Education was treated as a dummy variable in which each of five groups was compared against the baseline (primary school). To ease interpretation, work ability was reverse-scored so that higher scores for all outcome variables indicated more adverse outcomes. For the continuous work ability score, we performed hierarchical linear regression analyses with B estimates; for the three ordinal measures, we performed hierarchical ordinal regression analyses. For the latter, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated, with ORs > 1 indicating less favourable outcomes. We also analyzed the ordinal data with dichotomized outcomes at different thresholds and found that coefficients and confidence intervals for all independent variables were broadly consistent in direction and magnitude across all the splits in the data (data not shown). We report here the ordinal proportional odds models as a fair summary of the patterns in the data.

Work-related diabetes distress

9

Due to the exploratory nature of the study, we maintained a p-value of < 0.05 as indicating statistical significance although multiple analyses were performed. The analyses were carried out using SAS 9.4 (SAS Institute, Cary, NC, USA). 3. RESULTS 3.1 Participant characteristics A total of 1594 people with type 1 diabetes completed the survey, representing a response rate of 52%. Of these, 220 (14%) completed the paper version. Respondents were older (mean age was 49 years as compared to 45), had longer diabetes duration (mean duration was 26 years as compared to 23), more complications (based on ICD-10 codes) but more optimal glycemic control, as compared with non-respondents; no significant differences were found between respondents and nonrespondents in terms of gender, albumin-creatinine ratio, or treatment type (data not shown). For the purpose of this analysis, we used a subsample of 1126 people who were currently working. The characteristics of the subsample are presented in Table 1. 3.2 Inter-item correlations between PAID-5 and WRDD The internal consistency of the PAID-5 and WRDD scales was acceptable (Cronbach’s alpha 0.91 and 0.88, respectively). Pair-wise inter-item correlations between the PAID-5 items were 0.59-0.78, whereas the correlation between the two WRDD items was 0.66 (data not shown). The cross-construct correlations were approximately 0.5 with the notable exception of the correlation between the second WRDD item and the fourth PAID-5 item (r = 0.65). The proportion of respondents reporting each of the issues in PAID-5 and WRDD as a rather serious or a serious problem was 9.5%, 11.5%, 27.9%, 20.4%, and 24.4 %, respectively, for the PAID-5 items and 4.1% and 10.8 %, respectively, for the WRDD items. The item associated with the highest proportion at 27.9 % was worrying about the future and the possibility of serious complications (PAID-5 item 3).

Work-related diabetes distress

10

3.3 Exploratory factor analysis An unforced two-factor structure emerged with no cross-loadings and all eigenvalues were > 1. The rotated pattern demonstrated simple structure; all items loaded highly onto one factor above the cut-off of 0.4. The five PAID-5 items loaded onto factor 1 (0.54-0.84), while the two WRDD items loaded onto factor 2 (0.58-1.02). However, the fourth PAID-5 item was just below the cut-off of cross-loading into factor 2 (0.39), which was consistent with the results from the inter-item correlation analysis. The visual inspection of the rotated factor pattern also indicated that the two WRDD items were clustered around factor 2 and all PAID-5 items clustered around factor 1, with the exception of the fourth PAID-5 item, which was only marginally closer to the PAID-5 cluster (data not shown). 3.4 Hierarchical regression analyses The adjusted B-estimates of lower work ability are displayed in Table 2 and adjusted ORs for lower QoL, more frequent intentional hyperglycaemia at work, and more frequent absenteeism are displayed in Tables 3-5. Work ability Adding PAID-5 to the linear model with covariates increased the explanatory power of the model from 7% to 17% (Table 2). When adding WRDD, the explanatory power of the model increased to 26%, suggesting that WRDD adds unique power for predicting work ability. Quality of life The odds of experiencing lower QoL decreased in the absence of complications and increased with more education and higher diabetes distress and WRDD. Adding PAID-5 to the ordinal model increased the explanatory power of the model from 8% to 28%. When WRDD was introduced, explanatory power further increased to 33%, suggesting an independent effect of WRDD on QoL (Table 3).

Work-related diabetes distress

11

Intentional hyperglycaemia at work The odds of deliberately maintaining high blood glucose levels while at work decreased with longer diabetes duration and increased with higher scores of PAID-5 and WRDD. Adding PAID-5 to the ordinal model increased its explanatory power from 3% to 13%; when WRDD was introduced, the model’s explanatory power increased to 16% (Table 4). This suggests that WRDD has an independent, albeit small, effect on intentional hyperglycaemia at work. Absenteeism The odds of frequent absenteeism decreased in the absence of diabetes complications and increased with female gender and higher scores of WRDD. Adding PAID-5 to the model with covariates increased its explanatory power from 4% to 7%; when WRDD was added, the explanatory power increased to 10% (Table 5). This suggests that WRDD has a small independent explanatory power for absenteeism. 4. DISCUSSION The results suggest that WRDD is a distinct facet of diabetes-related emotional distress although some conceptual overlap exists, especially between the second WRDD item and the fourth PAID-5 item as demonstrated by the inter-item correlations and exploratory factor analysis. However, this is not surprising because both questions explore the experience of exhaustion in daily life and work life. The percentage of variance in each dependent variable accounted for by the model improved significantly in each step, and WRDD made a significant independent contribution above and beyond the covariates and PAID-5 to predicting each of the four markers of worker well-being: work ability (9%), quality of life (5%), intentional hyperglycaemia at work (3%), and absenteeism (3%). Notably, WRDD added as much variance as PAID-5 for work ability and absenteeism. 4.1 Comparison with other research

Work-related diabetes distress

12

To the best of our knowledge, ours is the first study to examine whether WRDD represents a distinct dimension of diabetes-related emotional distress. However, other qualitative studies lend support to the notion of distress stemming from work life among people with type 1 diabetes [13, 14, 25]. In our sample, 4.1% and 10.8% of respondents respectively reported WRDD items 1 and 2 as rather serious or serious problems. In comparison, Hakkarainen et al. reported that 70% of the Finnish sample experienced WRDD; 20.6% were often stressed, and 49.4% were sometimes stressed [15]. This marked difference is likely to stem from the different measurement constructs; we assessed the magnitude of the problem instead of the frequency of worry. Fisher et al. note that worry may be frequent and even expected among people with a serious chronic condition such as type 1 diabetes [3], but worry does not necessarily equate to the magnitude of problems experienced by patients. 4.2 Strengths and limitations The cross-sectional nature of the data limits our ability to draw conclusions about causality, and self-reported data of behavioural constructs may be limited by response bias, including social desirability. We used the PAID-5 scale, a reliable, short form measure of the 20-item PAID scale, to accommodate space for other key measures of interest and avoid survey fatigue. To check the robustness of the results, including residual confounding, further work is required to explore whether this preliminary result is also seen when using the full PAID scale or the T1-DDS covering a range of potentially distressing issues pertinent to individuals with type 1 diabetes [8]. We used a novel measure of work-related diabetes distress adjusted to the format of the PAID scale and validated on its face by the target group and experts. To strengthen the conceptual basis of the construct we recommend further validation as part of a future research program. In this process the possibilities to encompass the work-specific item from the T1-DDS or other items should be considered.

Work-related diabetes distress

13

We used four markers of worker well-being because, to the best of our knowledge, no consensus definition of worker well-being exists. By design, the self-reported questions on quality of life and work ability are broad measures that may be sensitive to underlying health conditions overlooked by more objective and specific measures of health [26]. Although none of these measures are diabetes-specific, the potential impact of diabetes is implicit when assessing both overall quality of life and self-rated work ability. 4.3 Implications The distinction between WRDD and diabetes-related emotional distress has three major implications for clinical care, workplace policies, and future research. First, the brief, two-item measure of WRDD may serve as a screening tool in clinical practice to identify people susceptible to work-related diabetes distress. This is likely to be most relevant at critical moments during the course of type 1 diabetes, e.g., at the onset of complications, or at important points in work life such as entering the labour market or starting a new job. In addition, understanding the nature of distress among working people with type 1 diabetes may make it easier to appropriately target support and counselling. Second, by mapping WRDD, we are opening an avenue of enquiry that can ultimately help to resource and equip workplaces to develop appropriate policies and procedures related to people with diabetes. As shown by Fisher et al. a low-cost distress-reduction trial can lower levels of distress and improve disease management in a type 2 population [27]. This may serve as inspiration for development of a type 1-specific intervention tailored to the specific needs of this population. Finally, adding the two-item measure of WRDD to the PAID scale will allow for more tailored research into worker well-being and quality of work life for people living and working with type 1 diabetes. To date, this work-specific dimension of the spectrum of diabetes management has been unaccounted for by the many existing measures of diabetes distress.

Work-related diabetes distress

14

4.4 Conclusion Our findings suggest that WRDD is a distinct dimension of diabetes-related emotional distress. The brief two-item measure used in the present study may be valuable in assessing the psychosocial burden of reconciling diabetes management and work life. However, as research into work-related diabetes distress is still in its infancy further studies are needed to strengthen the conceptual basis of the concept, explore its clinical usefulness and clarify risk factors for workrelated diabetes distress.

Contributors UMH was responsible for the conception and design of the study with advice from IW, KO and TCS. UMH conducted data analyses, which were reviewed by KO, JLB and TCS. UMH prepared the first draft of the manuscript and all authors provided feedback and revisions on the first and subsequent drafts. All authors approved the final submitted manuscript.

Funding This work was supported by an unrestricted grant from Innovation Fund Denmark. Acknowledgements We acknowledge Jennifer Green from Caduceus Strategies for proof reading the manuscript.

Work-related diabetes distress

15

Table 1. Participant characteristics Female gender, n (%) Age in years, mean ± SD (range) Cohabitation status, n (%) Alone, no children living at home Alone with children living at home Living with a partner, no children at home Living with a partner and children at home Living with other adults Other Education, n (%) Primary school High school Vocational training Short higher education Medium higher education Long higher education Diabetes duration in years, mean (range) Haemoglobin A1c, mean ± SD mmol/mol % Albumin-creatinine ratio, mg/g, mean ± SD) Complications, n (%) None ≥1 Treatment, n (%) Insulin pump Insulin pen Work-related diabetes distress score, mean ± SD (range) Diabetes-related emotional distress score, mean ± SD (range) Work ability score, mean ± SD (range) Intentional hyperglycaemia at work, n (%) Never A few times Sometimes Often Always Quality of life, n (%) Very good Good Neither good nor bad Bad Very bad Days absent from work due to illness in past 12 months, n (%) None ≤9 10-24 25-99 100-365 Missing data were excluded.

N = 1,126 532 (47) 46.5 ± 12.1 (18-70)

173 (15.4) 67 (6.0) 403 (35.9) 405 (36.0) 49 (4.4) 27 (2.4) 59 (5.3) 89 (7.9) 167 (14.9) 172 (15.3) 315 (28.1) 321 (28.6) 23.6 ± 13.6 (0-63) 61.0 ± 11.3 7.7 ± 3.2 36.5 ± 203.9 594 (52.8) 532 (47.3) 209 (18.6) 917 (81.4) 1.6 ± 1.8 (0-8) 7.1 ± 4.9 (0-20) 8.6 ± 1.6 (0-10) 400 (35.6) 354 (31.5) 243 (21.6) 110 (9.8) 18 (1.6) 302 (26.8) 579 (51.4) 201 (17.9) 43 (3.8) 1 (0.1) 359 (31.9) 554 (49.3) 149 (13.3) 50 (4.5) 12 (1.1)

Work-related diabetes distress

16

Table 2. Hierarchical multiple linear regression model predicting lower work ability from gender, diabetes complications, education, diabetes distress, and work-related diabetes distress R2 Step 1

B

P value

0.07

Female gender

12.72** 0.27

0.005

-0.38

< .001

High school

1.27

< .001

Vocational training

0.55

0.02

Short higher education

1.02

< .001

Medium higher education

1.14

< .001

Long higher education

1.40

< .001

No complications vs. ≥ 1 complication

Model F

Education vs. primary school

Step 2

0.17

Female gender

29.16** 0.07

0.458

-0.38

< .001

High school

1.31

< .001

Vocational training

0.64

0.005

Short higher education

0.91

< .001

Medium higher education

0.98

< .001

Long higher education

1.29

< .001

0.11

< .001

No complications vs. ≥ 1 complication Education vs. primary school

PAID-5 score Step 3

0.26

Female gender

42.42** 0.03

0.701

-0.35

< .001

High school vs. primary school

1.20

< .001

Vocational training vs. primary school

0.63

0.004

Short higher education vs. primary school

0.91

< .001

Medium higher education vs. primary school

0.89

< .001

Long higher education vs. primary school

1.22

< .001

PAID-5 score

0.03

0.019

WRDD score

0.34

< .001

No complications Education

*p <0.05; **p <0.0001. Missing data were excluded.

Work-related diabetes distress

17

Table 3. Hierarchical multiple ordinal regression model predicting lower quality of life from gender, diabetes duration, complications, treatment, education, diabetes distress, and work-related diabetes distress R2

OR

95% CI

Female gender

1.39

1.10-1.75

Diabetes duration

1.00

0.99-1.01

Step 1

0.08

No diabetes complications vs. ≥ 1 complication

0.63

0.48-0.84

Insulin pen vs. pump

0.68

0.50-0.90

1.50

0.79-2.81

Education vs. primary school High school Vocational training

0.71

0.40-1.25

Short higher education

0.97

0.55-1.71

Medium higher education

2.00

1.17-3.42

Long higher education

2.37

1.39-4.04

Female gender

0.93

0.73-1.19

Diabetes duration

1.01

1.00-1.02

No diabetes complications vs. ≥ 1 complication

0.71

0.53-0.95

Insulin pen vs. pump

0.77

0.57-1.03

Step 2

0.28

Education vs. primary school High school

1.43

0.74-2.77

Vocational training

0.72

0.40-1.31

Short higher education

1.36

0.75-2.47

Medium higher education

1.39

0.79-2.44

Long higher education

1.91

1.09-2.35

1.25

1.22-1.29

Female gender

0.88

0.69-1.13

Diabetes duration

1.01

1.00-1.02

No diabetes complications vs. ≥ 1 complication

0.70

0.52-0.94

Insulin pen vs. pump

0.80

0.59-1.01

PAID-5 score Step 3

0.33

Education vs. primary school High school

1.32

0.68-2.57

Vocational training

0.71

0.39-1.30

Short higher education

1.35

0.74-2.45

Medium higher education

1.32

0.75-2.33

Long higher education

1.84

1.05-3.24

PAID-5 score

1.16

1.12-1.20

WRDD score

1.45

1.33-1.59

Missing data were excluded.

Work-related diabetes distress

18

Table 4. Hierarchical multiple ordinal regression modeling to predict intentional hyperglycaemia at work from age, diabetes duration, diabetes complications, education, diabetes distress, and work-related diabetes distress R2 Step 1

OR

95% CI

0.03

Age

1.00

0.99-1.01

Diabetes duration

0.98

0.97-0.99

No diabetes complications vs. ≥ 1 complication

0.90

0.69-1.18

High school

0.90

0.49-1.66

Vocational training

0.93

0.54-1.59

Short higher education

0.78

0.46-1.34

Medium higher education

1.55

0.93-2.57

Long higher education

1.29

0.78-2.14

Age

1.00

0.99-1.01

Diabetes duration

0.98

0.97-0.99

No diabetes complications vs. ≥ 1 complication

0.99

0.76-1.30

High school

1.00

0.54-1.88

Vocational training

1.07

0.62-1.86

Short higher education

0.82

0.47-1.42

Medium higher education

1.36

0.81-2.30

Long higher education

1.18

0.70-1.98

1.14

1.11-1.17

Age

1.00

0.99-1.02

Diabetes duration

0.98

0.97-0.99

No diabetes complications vs. ≥ 1 complication

0.99

0.75-1.30

High school

0.97

0.52-1.81

Vocational training

1.10

0.63-1.91

Short higher education

0.80

0.46-1.39

Medium higher education

1.32

0.78-2.23

Long higher education

1.15

0.68-1.94

PAID-5 score

1.08

1.05-1.12

WRDD score

1.26

1.16-1.36

Education vs. primary school

Step 2

0.13

Education vs. primary school

PAID-5 score Step 3

0.16

Education vs. primary school

Missing data were excluded.

Work-related diabetes distress

19

Table 5. Hierarchical multiple ordinal regression modeling to predict absenteeism at work from age, diabetes duration, complication status, education, diabetes distress and work-related diabetes distress R2

OR

95% CI

Female gender

1.69

1.34-2.13

No diabetes complications vs. ≥ 1 complication

0.75

0.60-0.94

High school

1.60

0.85-3.00

Vocational training

0.56

0.32-0.99

Short higher education

1.01

0.58-1.77

Medium higher education

1.10

0.65-1.86

Long higher education

1.25

0.74-2.12

Female gender

1.52

1.21-1.93

No diabetes complications vs. ≥ 1 complication

0.74

0.59-0.93

High school

1.65

0.87-3.13

Vocational training

0.61

0.34-1.08

Short higher education

1.07

0.60-1.90

Medium higher education

1.01

0.59-1.74

Long higher education

1.19

0.69-2.03

1.07

1.04-1.09

Female gender

1.48

1.17-1.88

No diabetes complications vs. ≥ 1 complication

0.74

0.59-0.93

High school

1.60

0.84-3.04

Vocational training

0.62

0.35-1.11

Short higher education

1.01

0.57-1.79

Medium higher education

1.01

0.59-1.75

Long higher education

1.19

0.69-2.04

PAID-5 score

1.00

0.97-1.03

WRDD score

1.34

1.23-1.45

Step 1

0.04

Education vs. primary school

Step 2

0.07

Education vs. primary school

PAID-5 score Step 3

0.10

Education vs. primary school

Missing data were excluded.

Work-related diabetes distress

20

REFERENCES 1.

Young-Hyman D, de Groot M, Hill-Briggs F, Gonzalez JS, Hood K, and Peyrot M.

Psychosocial care for people with diabetes: a position statement of the American Diabetes Association. Diabetes Care 2016; 39:2126-2140. https://doi.org/10.2337/dc16-2053 2.

Hessler DM, Fisher L, Polonsky WH, Masharani U, Strycker LA, Peters AL, Blumer

I, Bowyer V.Diabetes distress is linked with worsening diabetes management over time in adults with Type 1 diabetes. Diabetic Med 2017;34, 1228–1234. https://doi.org/10.1111/dme.13381 3.

Fisher L, Gonzalez JS, and Polonsky WH. The confusing tale of depression and

distress in patients with diabetes: a call for greater clarity and precision. Diabetic Med 2014; 31:764-772. https://doi.org/10.1111/dme.12428 4.

American Diabetes Association, 4. Lifestyle management: Standards of medical care

in diabetes—2018. Diabetes Care, 2018. 41(Supplement 1): p. S38-S50. 5.

Reddy, M., et al., Type 1 diabetes in adults: Supporting self management. British

Medical Journal, 2016. 352: p. 1998. 6.

Polonsky WH, Anderson BJ, Lohrer PA, Welch G, Jacobson AM, Aponte JE, and

Schwartz CE. Assessment of diabetes-related distress. Diabetes Care 1995; 18:754-760. 7.

Polonsky WH, Fisher L, Earles J, Dudl RJ, Lees J, Mullan J, and Jackson RA.

Assessing Psychosocial Distress in Diabetes. Development of the Diabetes Distress Scale. Diabetes Care 2005 Mar; 28(3): 626-631. https://doi.org/10.2337/diacare.28.3.626 8.

Fisher, L., et al., Understanding the sources of diabetes distress in adults with type 1

diabetes. Journal of Diabetes and Its Complications 2015. 29(4): p. 572-577. 9.

Gonzalez JS, Fisher L, and Polonsky WH. Depression in diabetes: have we been

missing something important? Diabetes Care 2011; 34:236-239. https://doi.org/10.2337/dc10-1970 10.

Persson S, Gerdtham UG, Steen Carlsson K, Swedish Childhood Diabetes Study

Work-related diabetes distress

21

Group. Labor market consequences of childhood onset type 1 diabetes. Economics and Human Biology 2016. 23:180-192. 11.

Persson S, Dahlquist G, Gerdtham UG, Steen Carlsson K, Swedish Childhood

Diabetes Study Group. Why childhood-onset type 1 diabetes impacts labour market outcomes: a mediation analysis. Diabetologia 2018. 61: p. 342-353. 12.

Nielsen HB, Ovesen LL, Mortensen LH, Lau CJ, and Joensen LE. Type 1 diabetes,

quality of life, occupational status and education level - A comparative population-based study. Diabetes Res Clin Pract 2016; 121:62-68. https://doi.org/10.1016/j.diabres.2016.08.021 13.

Hansen UM, Cleal, B., Tjørnhøj-Thomsen, T., Willaing, I. Type 1 diabetes in work

life: a matter of containment? In: Järvelin-Pasanen S, ed. Proceedings of the 48th Annual Conference of Nordic Ergonomics and Human Factors Society. Kuopio, Finland: University of Eastern Finland 2016: 222-224. 14.

Balfe M BR, Smith D, Sreenan S, Doyle F, Conroy R. . Why do young adults with

type 1 diabetes find it difficult to manage diabetes in the workplace? Health & Place 2014; 26:7. 15.

Hakkarainen P, Moilanen L, Hanninen V, Heikkinen J, and Rasanen K. Work-related

diabetes distress among Finnish workers with type 1 diabetes: a national cross-sectional survey. Journal of occupational medicine and toxicology (London, England) 2016; 11:11. https://doi.org/10.1186/s12995-016-0099-4 16.

Centers for Disease Control and Prevention. Measuring Healthy Days. Atlanta,

Georgia: CDC, November 2000. 17.

McGuire BE, Morrison TG, Hermanns N, Skovlund S, Eldrup E, Gagliardino J,

Kokoszka A, Matthews D, Pibernik-Okanovic M, Rodriguez-Saldana J, de Wit M, and Snoek FJ. Short-form measures of diabetes-related emotional distress: the Problem Areas in Diabetes Scale (PAID)-5 and PAID-1. Diabetologia 2010; 53:66-69. https://doi.org/10.1007/s00125-009-1559-5

Work-related diabetes distress

18.

22

Sirgy MJ. Employee well-being: an integrative perspective. In: Reilly NP, Sirgy MJ,

Gorman CA (eds). Work and Quality of Life Ethical Practices in Organizations. New York: Springer 2012, 35-63. 19.

Schulte P and Vainio H. Well-being at work--overview and perspective. Scandinavian

journal of work, environment & health 2010; 36:422-429. 20.

El Fassi M, Bocquet V, Majery N, Lair ML, Couffignal S, and Mairiaux P. Work

ability assessment in a worker population: comparison and determinants of Work Ability Index and Work Ability score. BMC Public Health 2013; 13:305. https://doi.org/10.1186/1471-2458-13-305 21.

Tuomi K, Ilmarinen J, Jahkola A, Katajarinne L, and Tulkki A. Work Ability Index

Helsinki: Finnish Institute of Occupational Health, 1994. 22.

Development of the World Health Organization WHOQOL-BREF quality of life

assessment. The WHOQOL Group. Psychological medicine 1998; 28:551-558. 23.

Thurstone L. Multiple-factor analysis. Chicago: University of Chicago Press, 1947.

24.

Brown TA. Confirmatory Factor Analysis for Applied Research, 2nd ed. New York:

The Guilford Press, 2015. 25.

Hinder S and Greenhalgh T. "This does my head in". Ethnographic study of self-

management by people with diabetes. BMC health services research 2012; 12:83. https://doi.org/10.1186/1472-6963-12-83 26.

Golenbock S, Kazman JB, Krauss S, and Deuster PA. General health status in army

personnel: relations with health behaviors and psychosocial variables. Quality of life research : an international journal of quality of life aspects of treatment, care and rehabilitation 2017; 26:18391851. https://doi.org/10.1007/s11136-017-1523-7

Work-related diabetes distress

27.

23

Fisher L, Hessler D, Glasgow RE, Arean PA, Masharani U, Naranjo D, and Strycker

LA. REDEEM: a pragmatic trial to reduce diabetes distress. Diabetes Care 2013; 36:2551-2558. https://doi.org/10.2337/dc12-2493

Highlights •

Work-related diabetes distress is a distinct problem area so far unaccounted for.

•

Work-related diabetes distress is a unique contributor to lower worker well-being.

•

Further validation of work-related diabetes distress is needed.