Comorbid obstructive sleep apnea and insomnia and its associations with mood and diabetes-related distress in type 2 diabetes mellitus
ABSTRACT
Study Objectives:
Previous research suggests that obstructive sleep apnea (OSA) and insomnia frequently coexist and are prevalent in persons with type 2 diabetes mellitus. This study compared mood and diabetes-related distress among OSA, insomnia, and comorbid OSA and insomnia groups in persons with type 2 diabetes mellitus.
Methods:
A secondary analysis was conducted with baseline data from 2 independent randomized controlled trials evaluating the efficacy of OSA and insomnia treatment. The pooled sample (n = 224) included participants with OSA only (n = 68 [30.4%]), insomnia only (n = 107 [47.8%]), and OSA and insomnia (OSA+insomnia; n = 49 [21.9%]). OSA was defined as an apnea-hypopnea index ≥ 15 events/h; insomnia was defined as an Insomnia Severity Index score ≥ 15. Mood was measured by the Profile of Mood States total and subscale scores; diabetes-related distress was assessed by the Problem Areas in Diabetes questionnaire. One-way analysis of covariance and multivariate analysis of covariance were conducted, controlling for demographic characteristics and restless legs syndrome.
Results:
The insomnia group had on average significantly higher scores for total mood disturbance (insomnia vs OSA = 45.32 vs 32.15, P = .049), tension-anxiety (insomnia vs OSA = 12.64 vs 9.47, P = .008), and confusion-bewilderment (insomnia vs OSA = 9.45 vs 7.46, P = .036) than the OSA group. The OSA+insomnia group had on average significantly greater diabetes-related distress than the OSA group (OSA+insomnia vs OSA = 40.61 vs 30.97, P = .036).
Conclusions:
Insomnia may have greater impact on mood disturbance and diabetes-related distress than OSA in persons with type 2 diabetes mellitus. In particular, comorbid insomnia may contribute to greater diabetes-related distress in persons with type 2 diabetes mellitus and OSA.
Citation:
Jeon B, Luyster FS, Sereika SM, DiNardo MM, Callan JA, Chasens ER. Comorbid obstructive sleep apnea and insomnia and its associations with mood and diabetes-related distress in type 2 diabetes mellitus. J Clin Sleep Med. 2022;18(4):1103–1111.
BRIEF SUMMARY
Current Knowledge/Study Rationale: Both obstructive sleep apnea (OSA) and insomnia are highly prevalent in persons with type 2 diabetes mellitus. Although comorbid insomnia contributes to greater psychological symptoms in persons with OSA, the effect of comorbid OSA and insomnia on mood and diabetes-related distress in persons with type 2 diabetes mellitus is unknown.
Study Impact: This study showed that insomnia is frequently comorbid with OSA and suggests that comorbid OSA and insomnia may have a synergistic effect on mood disturbances and diabetes-related distress in persons with type 2 diabetes mellitus. These associations highlight the importance of evaluation and treatment of comorbid insomnia in promoting optimal psychological health in persons with type 2 diabetes mellitus, even after a diagnosis of OSA.
INTRODUCTION
Diabetes affects 34 million adults in the United States and is the seventh leading cause of death. Approximately 90%–95% of adults with diabetes have type 2 diabetes mellitus (T2DM).1 Insomnia and obstructive sleep apnea (OSA) are 2 of the most common sleep disorders in the United States.2,3 Importantly, 60% of persons with T2DM have OSA4 and almost 40% of persons with T2DM experience insomnia symptoms.5
OSA is characterized by repetitive upper airway obstruction or narrowing during sleep that causes cessation of breathing or decreased airflow that leads to hypoxia, sleep fragmentation, and daytime sleepiness.6 Insomnia is characterized by a state of hyperarousal with difficulty initiating or maintaining sleep and early morning awakenings with an inability to return to sleep, which results in significant daytime impairments of fatigue, tiredness, or sleepiness due to the inability to obtain refreshing sleep.7 OSA and insomnia have been associated with poor glycemic control and self-care behaviors in person with T2DM.5,8–10 However, the effect of OSA or insomnia on psychological symptoms in T2DM remains unclear.
Impaired mood such as depression and diabetes-related distress are widespread psychological symptoms in persons with T2DM.11,12 The prevalence of depression is twice as high in persons with T2DM compared to those without.13 Up to 44.5% of persons with T2DM experience diabetes-related distress.12,14 These psychological symptoms traditionally have been associated with poor diabetes self-management behaviors and poor glucose outcome in T2DM.15–18 Psychological care for comorbid conditions, such as diabetes-related distress, depression, and anxiety, has been recommended to optimize diabetes health outcomes in persons with T2DM.19
OSA and insomnia have been shown to be independently associated with increased psychological symptoms, most commonly depression.20–22 A recent review described that the co-occurrence of sleep apnea and insomnia is common and associated with higher impairment of daytime function and quality of life.23 A recent integrative review reported that persons with comorbid OSA and insomnia (OSA+I) had higher depressive symptoms than persons with OSA alone, and insomnia, not OSA, contributed to increased levels of depressive symptoms in persons with OSA+I.24 Given the synergistic effects of comorbid OSA and insomnia on psychological symptoms, the co-occurrence of OSA and insomnia may contribute to greater psychological symptoms in persons with T2DM. With T2DM, the effect of comorbid OSA and insomnia on mood and diabetes-related distress has not been examined. It is still unknown which sleep disorders have a stronger influence on mood and diabetes-related distress.
In this study, we compared mood states and diabetes-related distress among adults with T2DM and OSA, insomnia, and OSA+I. We hypothesized that (1) adults with OSA+I have greater mood disturbances and diabetes-related distress than those with OSA only, (2) adults with OSA+I have greater mood disturbance and diabetes-related distress than those with insomnia only among persons with T2DM, and (3) adults with insomnia only have greater mood disturbances and diabetes-related distress than those with OSA only.
METHODS
Study design, sample, and setting
This was a secondary analysis of pooled, cross-sectional baseline data obtained from 2 independent randomized controlled trials: Diabetes Sleep Treatment Trial (DSTT; R01-DK096023)25 and Diabetes Sleep Treatment Trial for Insomnia (DSTT-I; K24-NR016685). The purpose of the 2 parent studies was to examine the efficacy of the treatment of OSA (DSTT) and the treatment of insomnia (DSTT-I) on glucose control and diabetes self-management in persons with T2DM. The DSTT was a multisite trial that recruited 351 participants from the community and from endocrinology and sleep outpatient clinics using various strategies (eg, using focused mailing, flyers in the community, and social media advertisement). The DSTT-I was a single-site trial that recruited 55 participants using strategies similar to those used for DSTT and those excluded participants from DSTT who did not have an apnea-hypopnea index (AHI) of ≥ 10 events/h. The eligibility criteria for participants for the baseline assessment of the parent studies were self-reported T2DM, age 18 years or older, able to read and write English, AHI ≥ 10 events/h (DSTT) or moderate-severe insomnia (DSTT-I), and willingness to be randomized to the respective studies. The eligibility criteria for this secondary analysis were participants that (1) do not have missing data on variables for OSA severity, insomnia severity, mood states, and diabetes-related distress and (2) have either OSA or insomnia. The parent studies were approved by the Institutional Review Board at each study site. This study was independently approved by the University of Pittsburgh Institutional Review Board to combine the baseline data from the DSTT and DSTT-I for a secondary analysis.
Measures
OSA, insomnia, and comorbid OSA and insomnia definitions
OSA was defined by the level of AHI. Persons who had AHI ≥ 15 events/h were classified as having OSA. The ApneaLinkPlus (ResMed, San Diego, CA), an approved U.S. Food & Drug Administration level III device for unattended portable in-home sleep studies, was used to assess respiratory effort, nasal flow, pulse, and oxygen desaturations of 4% to derive the AHI.26 The home sleep apnea testing device, the ApneaLink, was found to have the 90.9% sensitivity and 94.6% specificity for screening of OSA with AHI ≥ 15 events/h compared with polysomnography.27
Insomnia was defined by the Insomnia Severity Index (ISI).28 The ISI, a 7-item self-report questionnaire, measures the presence, severity, and impact of insomnia symptoms. Each item was rated on a 5-point Likert scale from “0 = no problems” to “4 = very severe problem.” A higher overall score on the ISI indicates worse insomnia symptoms. For this study, persons who had an ISI ≥ 15 were classified as having insomnia. The sensitivity and specificity of the ISI for screening insomnia ≥ 15 are 78.1% and 100%, respectively, against the diagnosis derived from clinical interviews in a clinical sample.28
Persons who met both the OSA and insomnia criteria were classified as having OSA+I. OSA+I was defined as AHI ≥ 15 events/h and ISI ≥ 15.
Mood states
Mood states were measured by the Profile of Mood States (POMS).29 The POMS evaluates mood states with a set of 65 adjectives or phrases related to the 6 dimensions of mood: Tension-Anxiety, Depression-Dejection, Anger-Hostility, Vigor-Activity, Fatigue-Inertia, and Confusion-Bewilderment. Each item was rated on a 5-point Likert scale from “0 = not at all” to “4 = extremely.” Higher scores of each dimension of mood, except Vigor-Activity, indicate greater negative mood states. The POMS Total Mood Disturbance score is calculated by subtracting the score of Vigor-Activity (positive mood) from the sum of the 5 dimensions of mood (negative moods). Higher Total Mood Disturbance scores indicate greater mood disturbances.
Diabetes-related distress
Diabetes-related distress was measured by the Problem Areas in Diabetes (PAID).30 The PAID is a 20-item self-report questionnaire that evaluates diabetes-specific emotional distress, which is associated with the burden of diabetes management. Each item was rated on a 5-point Likert scale from “0 = not a problem” to “4 = serious problem.” The total score was calculated by summing all items and multiplying the total by 1.25 for a potential range from 0 to 100. Higher PAID scores indicate greater diabetes-related distress.
Sociodemographic and clinical information
Sociodemographic information comprised age (years), sex (male vs female), marital status (married/partnered vs single/divorced/widowed), race (non-White vs White), education level (less than 2-year degree of college or technical training vs ≥ 2-year degree of college or technical training), and financial hardship (no difficulty to meet one’s needs vs having somewhat to extreme difficulty to meet one’s needs). Diabetes-related clinical information included insulin use (yes vs no), duration of diagnosis of T2DM (years), HbA1c (%), and body mass index (kg/m2).
Sleep-related clinical information included daytime sleepiness, sleep quality, and history of restless legs syndrome (yes vs no). Daytime sleepiness was measured by the Epworth Sleepiness Scale.31 The Epworth Sleepiness Scale is an 8-item, self-report questionnaire used to rate the likelihood of falling sleep in common situations of daily living. Higher scores on the Epworth Sleepiness Scale indicate greater daytime sleepiness. Sleep quality was measured by the Pittsburgh Sleep Quality Index.32 The Pittsburgh Sleep Quality Index is a 19-item self-report questionnaire used to evaluate sleep quality. Higher scores indicate worse sleep quality.
Procedure
In both the parent DSTT and DSTT-I studies, brief telephone or in-person interviews were performed to identify participants who met the initial eligibility criteria for baseline assessment. At the in-person baseline assessments, written informed consent was obtained and anthropometric measurements (ie, height, weight) and venipuncture were performed for HbA1c. Participants were instructed to return the ApneaLinkPlus device and baseline questionnaires by prepaid mail packet (ie, the ISI, PAID, Epworth Sleepiness Scale, and Pittsburgh Sleep Quality Index) the next day and after 7 days, respectively. ApneaLinkPlus data were validated by trained polysomnography technicians. Data were stored in a secure server at the University of Pittsburgh. For this secondary analysis, the baseline datasets from the 2 parent studies (n = 406: 351 from the DSTT, 55 from the DSTT-I) were pooled and analyzed as 1 dataset based on shared measures and questionnaires. Among 406 participants recruited from the parent studies, 224 eligible participants for this secondary analysis were identified.
Statistical analysis
Data were analyzed by IBM SPSS Statistics Version 25.0 (IBM Corp., Armonk, NY). Baseline characteristics are reported as frequency count with percentage (for categorical variables) and mean with standard deviation (for continuous variables). The chi-square of independence (for categorical variables) and one-way analysis of variance (for continuous variables) were used to examine the differences in the characteristics among the OSA, insomnia, and OSA+I groups. Bonferroni or Games-Howell posthoc pairwise comparisons of the 3 types of sleep disorders were done for significant continuous variables found in one-way analysis of variance.
One-way analysis of covariance (ANCOVA) was performed to test whether mood states (eg, the POMS Total Mood Disturbance score) and diabetes-related distress (eg, total PAID score) were significantly different among the 3 types of sleep disorders. Pairwise comparisons with Bonferroni adjustment were conducted to determine where the mean differences existed among the 3 types of sleep disorders. Differences in the 6 dimensions of mood state (eg, the scores of Tension-Anxiety, Depression-Dejection, Anger-Hostility, Vigor-Activity, Fatigue-Inertia, and Confusion-Bewilderment) among the sleep disorder groups were analyzed by one-way multivariate analysis of variance (MANCOVA). One-way ANCOVA and pairwise comparisons with Bonferroni adjustment were performed to determine which dimensions of mood states were significantly different between the 3 types of sleep disorders. Covariates included in one-way ANCOVA and MANCOVA were age, sex, marital status, race, education level, financial hardship, and restless legs syndrome. The level of statistical significance was set at .05. The adjusted level of statistical significance for pairwise analyses was 0.016 or 0.0083, depending on the number of comparisons. Sample size was fixed at 224. To achieve 80% power with a significance level of .05, the given sample size for one-way ANCOVA detected differences among means of the 3 types of sleep disorders with small effect size (f = 0.20).33 For one-way MANCOVA, the given sample size detected differences of the 6 mood subscale scores among means of 3 types of sleep disorder with small effect size (f2 = 0.04).33
RESULTS
Sample characteristics
A total of 224 adults with T2DM were included in this study. The sample was primarily middle-aged (mean age 57.13 ± 10.30 years) with representation of non-Whites (38.4%) and with many expressing they had financial difficulty (42.2%). The sample was well-distributed by sex (51.3% female), marital status (43.3% married or partnered), and education level (54.5% with more than 2 years of college or technical training). Our sample of persons with T2DM was overweight or obese on average (mean body mass index = 35.06 ± 6.87 kg/m2), with an average duration of T2DM diagnosis of over 10 years (11.41 ± 9.06 years) and with almost half (49.1%) prescribed insulin. The sample had on average moderate to severe OSA,6 moderate insomnia,28 moderate levels of excessive daytime sleepiness,31 and poor sleep quality (Table 1).32
| Baseline Characteristic | Total | OSA | Insomnia | OSA+I | P |
|---|---|---|---|---|---|
| n = 224 | n = 68 | n = 107 | n = 49 | ||
| Sociodemographic information | |||||
| Age (y), mean ± SD | 57.13 ± 10.30 | 60.26 ± 9.66 | 54.93 ± 10.66a | 57.55 ± 9.33 | .003 |
| Sex, n (%) | |||||
| Female | 115 (51.3) | 22 (32.4) | 69 (64.5) | 24 (49.0) | < .001 |
| Male | 109 (48.7) | 46 (67.6) | 38 (35.5) | 25 (51.0) | |
| Marital status, n (%) | |||||
| Never married/separated/divorced/widowed | 127 (56.7) | 31 (45.6) | 73 (68.2) | 23 (46.9) | .004 |
| Married/partnered | 97 (43.3) | 37 (54.4)) | 34 (31.8) | 26 (53.1) | |
| Race, n (%) | |||||
| Non-White | 86 (38.4) | 15 (22.1) | 50 (46.7) | 21 (42.9) | .004 |
| White | 138 (61.6) | 53 (77.9) | 57 (53.3) | 28 (57.1) | |
| Education level, n (%) | |||||
| < 2 y degree of college, or technical training | 102 (45.5) | 27 (39.7) | 44 (41.1) | 31 (63.3) | .018 |
| ≥ 2 y degree of college, or technical training | 122 (54.5) | 41 (60.3) | 63 (58.9) | 18 (36.7) | |
| Financial hardship, n (%) | |||||
| Somewhat/extremely difficult | 94 (42.2) | 17 (25.0) | 56 (52.8) | 21 (42.9) | < .001 |
| Not at all difficult | 129 (57.8) | 51 (75.0) | 50 (47.2) | 28 (57.1) | |
| Sleep-related clinical information | |||||
| OSA severity | |||||
| AHI, events/h (mean ± SD) | 18.46 ± 17.72 | 32.14 ± 18.38 | 5.72 ± 4.13b | 27.31 ± 15.58c | < .001 |
| Insomnia severity | |||||
| ISI, mean ± SD | 16.41 ± 5.73 | 9.49 ± 3.41 | 19.58 ± 3.58a | 19.10 ± 3.22a | < .001 |
| Daytime sleepiness | |||||
| ESS, mean ± SD | 10.57 ± 4.87 | 8.63 ± 4.49 | 10.89 ± 4.72a | 12.57 ± 4.81a | < .001 |
| Sleep quality | |||||
| PSQI, mean ± SD | 11.19 ± 3.60 | 7.53 ± 3.07 | 13.00 ± 3.14a | 12.35 ± 3.39a | < .001 |
| Diabetes-related clinical information | |||||
| Insulin use status, n (%) | |||||
| Yes | 122 (47.8) | 47 (47.5) | 38 (44.7) | 37 (52.1) | .651 |
| No | 133 (52.2) | 52 (52.5) | 47 (55.3) | 34 (47.9) | |
| Duration of T2DM (y); mean ± SD | 11.41 ± 10.0 | 11.90 ± 10.43 | 10.49 ± 7.13 | 11.83 ± 10.54 | .560 |
| HbA1C (%), mean ± SD | 7.90 ± 1.63 | 7.55 ± 1.30 | 8.11 ± 1.96 | 8.15 ± 1.56b | .022 |
| BMI (kg/m2), mean ± SD | 35.29 ± 7.03 | 36.97 ± 7.12 | 32.56 ± 5.63b | 36.19 ± 7.53c | < .001 |
The characteristics of the participants in the 3 types of sleep disorders are summarized in Table 1. Among 224 participants, there were 68 (30.4%) with OSA only, 107 (47.8%) with insomnia only, and 49 (21.9%) with OSA+I. On average, persons with OSA were older than those with insomnia (P = .003). Persons with OSA+I had an even sex distribution, whereas those with OSA and those with insomnia were predominantly male and female, respectively (P < .001). Persons with OSA were more likely to be White (P = .004), have at least 2 years of college or technical training (P = .018), and reported not having financial difficulty (P < .001) compared with the other 2 groups. Persons with insomnia were less likely to be married or partnered (P = .004). Insomnia severity was not significantly different between persons with insomnia and those with OSA+I. OSA severity was not significantly different between persons with OSA and those with OSA+I. Daytime sleepiness and sleep quality were higher in those with insomnia and OSA+I than those with OSA (P < .001). Body mass index was higher in persons with OSA and those with comorbid OSA+I than in those with insomnia (P < .001). The level of HbA1c was higher in persons with comorbid OSA+I than in those with OSA (P = .022).
Effect of type of sleep disorder on mood states in adults with T2DM
After controlling for identified covariates, the POMS Total Mood Disturbance score was significantly different on average among the 3 types of sleep disorders, F(2,210) = 3.595, P = .029, partial η2 = .033. The adjusted group means for the POMS Total Mood Disturbance score were 32.15 ± 4.71 for persons with OSA, 45.32 ± 3.57 for those with insomnia, and 33.85 ± 4.82 for those with OSA+I (Table 2). Based on Bonferroni adjusted pairwise comparisons, the POMS Total Mood Disturbance Score for persons with OSA+I was not significantly higher compared to those with OSA (Mdiff = 1.70, 95% confidence interval [CI] [−13.19 to 16.58], P = 1.000). Persons with insomnia had a significantly higher POMS Total Mood Disturbance score than those with OSA (Mdiff = 13.17, 95% CI [0.04 to 26.29], P = .049). On average the POMS Total Mood Disturbance Score between persons with OSA+I and those with insomnia was not significantly different (Mdiff = −11.47, 95% CI [−25.27 to 2.33], P = .138) (Table 2).
| Mood States | OSA | Insomnia | OSA+I | F (2,210) | Difference in Adjusted Means (95% CI) | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| n = 68 | n = 103 | n = 49 | OSA+I vs OSA | Insomnia vs OSA | OSA+I vs Insomnia | |||||
| M (SD) | Madj (SE) | M (SD) | Madj (SE) | M (SD) | Madj (SE) | |||||
| Total Mood Disturbance | 19.87 (27.79) | 32.15 (4.71) | 44.35 (40.11) | 45.32 (3.57) | 28.18 (30.18) | 33.85 (4.82) | 3.60a | 1.70 (−13.19 to 16.58) | 13.17 (0.04 to 26.29)c | −11.47 (−25.27 to 2.33) |
| Tension-Anxiety | 7.40 (5.14) | 9.47 (0.90) | 12.50 (7.81) | 12.64 (0.68) | 9.78 (5.37) | 10.60 (0.92) | 4.93b | 1.14 (−1.71 to 0.40) | 3.18 (1.29 to 6.19)c | −2.04 (−4.20 to 0.79) |
| Depression-Dejection | 6.56 (7.62) | 10.11 (1.36) | 12.96 (11.69) | 13.62 (1.03) | 7.73 (7.87) | 9.66 (1.39) | 3.85 | — | — | — |
| Anger-Hostility | 7.06 (7.07) | 9.94 (1.14) | 9.06 (9.17) | 9.85 (0.86) | 6.76 (7.43) | 8.27 (1.16) | 0.81 | — | — | — |
| Vigor-Activity | 16.41 (6.28) | 15.86 (0.88) | 13.23 (6.01) | 13.92 (0.67) | 14.69 (6.46) | 14.63 (0.90) | 1.81 | — | — | — |
| Fatigue-Inertia | 9.43 (5.68) | 11.04 (0.90) | 13.81 (6.80) | 13.69 (0.68) | 12.14 (6.49) | 12.82 (0.92) | 3.30 | — | — | — |
| Confusion-Bewilderment | 5.84 (3.97) | 7.46 (0.68) | 9.25 (5.69) | 9.45 (0.52) | 6.47 (4.02) | 7.13 (0.69) | 5.18b | −0.33 (−2.48 to 1.81) | 1.99 (0.09 to 3.88)c | −2.32 (−4.31 to −0.33)c |
As the POMS Total Mood Disturbance score consists of 6 dimensions of mood states, we examined the association of the type of sleep disorder with the 6 mood subscale scores, Tension-Anxiety, Depression-Dejection, Anger-Hostility, Vigor-Activity, Fatigue-Inertia, and Confusion-Bewilderment. The type of sleep disorder was significantly associated with the 6 mood states after controlling for covariates, F(12,410) = 2.640, Wilks λ = .862, partial η2 = .072. Exploring each of the POMS subscale scores individually via one-way ANCOVA revealed that there were significant mean differences in Tension-Anxiety (F[2,210] = 4.972, P = .008, partial η2 = .045) and Confusion-Bewilderment (F[2,210] = 5.183, P = .006, partial η2 = .047) subscale scores among the 3 types of sleep disorders (P < .0083 after a Bonferroni correction) (Table 2). The adjusted mean score of Tension-Anxiety for persons with insomnia was significantly higher compared to those with OSA (Mdiff = 3.18, 95% CI [0.67 to 5.69], P = .008). The adjusted mean score of Confusion-Bewilderment in persons with insomnia was significantly higher compared to those with OSA (Mdiff = 1.99, 95% CI [0.94 to 3.88], P = .036) and those with OSA+I (Mdiff = 2.32, 95% CI [0.33 to 4.31], P = .016). There was no significant difference in the mean Tension-Anxiety score of persons with OSA+I and those with insomnia (Mdiff = −2.04, 95% CI [−4.68 to 0.60], P = .191) (Table 2).
Effect of type of sleep disorder on diabetes-related distress in adults with T2DM
On average, the PAID score was significantly different among the 3 types of sleep disorders after controlling for covariates, F(2,210) = 3.404, P = .035, partial η2 = .031. The adjusted group means for the PAID score were 30.97 ± 2.82 for persons with OSA, 35.42 ± 2.14 for those with insomnia, and 40.61 ± 2.89 for those with OSA+I (Table 3). The mean PAID score for persons with OSA+I was significantly higher compared to those with OSA (Mdiff = 9.65, 95% CI [0.73 to 18.57], P = .029). The mean difference for persons with OSA was not significant compared to those with insomnia (Mdiff = 4.46, 95% CI [−3.41 to 12.32], P = .519). There was no significant difference between the mean PAID score between persons with OSA+I and those with insomnia (Mdiff = 5.19, 95% CI [−3.08 to 13.46], P = .394) (Table 3).
| OSA | Insomnia | OSA+I | F (2, 210) | Difference in Adjusted Means (95% CI) | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| n = 68 | n = 103 | n = 49 | OSA+I vs OSA | Insomnia vs OSA | OSA+I vs Insomnia | |||||
| M (SD) | Madj (SE) | M (SD) | Madj (SE) | M (SD) | Madj (SE) | |||||
| Diabetes-related distress | 24.21 (17.31) | 30.97 (2.82) | 34.86 (21.62) | 35.42 (2.14) | 38.75 (22.15) | 40.61 (2.89) | 3.40a | 9.65 (0.73 to 18.57)b | 4.46 (−3.41 to 12.32) | 5.19 (−3.08 to 13.46) |
DISCUSSION
This study investigated the effect of OSA and insomnia individually and jointly on mood states and diabetes-related distress in persons with T2DM. Insomnia was associated with greater mood disturbance than OSA, and OSA+I was associated with greater diabetes-related distress than OSA. These findings suggest that insomnia leads to additive impairments on mood, and insomnia increases the level of diabetes-related distress in persons with OSA+I. It is possible that insomnia, not OSA, might be a more potent contributing factor to increased severity of mood disturbances and diabetes-related distress in persons with T2DM.
Our findings support the conclusions of our integrative review that OSA+I has a synergistic effect on depressive symptoms compared to OSA and that comorbid insomnia contributes to increased severity of depressive symptoms.24 Understanding the mechanisms underlying the effect of OSA+I on mood and diabetes-related distress in T2DM remains limited. Data from previous studies found that OSA and insomnia were independently related to dysregulation of the hypothalamic-pituitary-adrenal axis (eg, increased level of cortisol)34,35 and alteration in monoamine neurotransmitters release (eg, decreased serotonin, dopamine),34,36,37 which were also found in persons with depression.38–40 It can be hypothesized that this physiological dysregulation may be more severe when OSA and insomnia coexist, thus exacerbating mood disturbances and diabetes-related distress in individuals with OSA+I rather than in those with OSA alone. A recent functional neuroimaging study also suggests that increased activity in the emotion-regulating system and decreased activity in the cognitive-executive system in the brain are commonly found in OSA, insomnia, and major depressive disorder.41–44 Specifically, insomnia is mainly associated with the hyperactivation of the emotion-regulating system and OSA attenuates the activity of the cognitive-executive system.44 It is plausible, then, that insomnia has a closer association with the dysregulation of moods than OSA.
This is the first study to demonstrate that persons with T2DM and OSA+I showed greater diabetes-related distress than those with OSA alone and that comorbid insomnia contributed to increased diabetes-related distress. Diabetes-related distress is a unique emotional burden experienced by persons with T2DM and is associated with the responsibility of maintaining the daily demands of diabetes management. Diabetes-related distress is known as a key influential psychological symptom in persons with T2DM, along with depression and anxiety.45 Current literature has focused on the negative impact of diabetes-related distress on health outcomes in T2DM; this includes diabetes self-care behaviors and glucose control.15–18 The evidence regarding the contributing factors to diabetes-related distress is limited. Findings of this study provide evidence that insomnia, especially comorbid with OSA, is a significant modifiable factor associated with diabetes-related distress. Our findings also indicate that persons with OSA+I had an average HbA1c value 0.57% higher than those with OSA alone. Previous research has indicated a difference in HbA1c of 0.5% as clinically significant.46 The association between insomnia and diabetes-related distress may negatively affect glycemic outcome in T2DM.
In our findings, persons with insomnia and OSA+I had significantly higher daytime sleepiness than those with OSA. This suggests that comorbid insomnia may exacerbate daytime sleepiness in persons with OSA. In terms of T2DM, excessive daytime sleepiness is associated with worse glucose outcomes.47,48 Insomnia not only negatively affects mood and diabetes-related distress but also increases the severity of excessive daytime sleepiness more than in persons with OSA. Exacerbated excessive daytime sleepiness, greater mood impairment, and diabetes-related distress from comorbid insomnia may hinder effective glucose management in persons with OSA.
Although daytime sleepiness is a frequent symptom of OSA, there is a distinct phenotype of nonsleepy persons with OSA.49 Data from a large study (n = 16,583) of community-dwelling adults found that the presence of excessive daytime sleepiness was significantly (P < .05) associated with depression and diabetes but not with the presence of OSA.50 However, this study did not evaluate if participants had coexisting insomnia, a known sleep disorder associated with impaired mood. Although persons with sleepy OSA may experience greater mood disturbances compared to persons with nonsleepy OSA, our findings suggest that comorbid insomnia may be a factor in sleepy persons with OSA having worse mood disturbances compared to nonsleepy persons with OSA. In a recent study for persons in a community sleep apnea cohort, excessive daytime sleepiness was associated with higher prevalence of insomnia symptoms, poorer mental health, and disease-specific functional status.51
A recent review article described potential mechanisms that may be involved in the relationship between insomnia and OSA.23 In persons with insomnia, consecutive sleep loss may reduce upper airway muscle tone, or hyperarousal during sleep, a cardinal symptom of insomnia, may decrease the respiratory arousal threshold, which contributes to increase apnea and hypopnea during sleep. In persons with OSA, the frequent awakenings after apnea and hypopnea events may exacerbate insomnia complaints and perpetuate insomnia symptoms.23 Indeed, insomnia and OSA may contribute to the development or exacerbation of severity in each other.
Despite the bidirectional relationship between insomnia and OSA, insomnia is often overlooked in persons with OSA because it has been assumed to be a secondary symptom and is believed to improve with successful treatment of OSA.52 The stereotypical characteristics of insomnia (eg, common in middle-aged and older women) are contradictory to those of OSA (eg, common in middle-aged and older men).33,53 In particular, several international clinical guidelines for T2DM have focused mainly on understanding the adverse impact of OSA on diabetes care and rarely mention insomnia.54 Therefore, comorbid insomnia is likely to be ignored in persons with T2DM and OSA. Mood disturbances and diabetes-related distress are likely to remain unresolved in T2DM due to comorbid insomnia. Addressing insomnia may provide an effective strategy to support psychological care in T2DM.
We found that among the 6 dimensions of mood states, insomnia was associated with higher scores on Tension-Anxiety and Confusion-Bewilderment than OSA. The American Diabetes Association emphasized that treatment of psychosocial difficulties and reduction of diabetes-related distress are important because they are associated with fear of hyperglycemia, failure to meet blood glucose targets, insulin injection induced anxiety, and worse glycemic control.19
There are several limitations in our study. This study was a secondary analysis using a cross-sectional design of baseline data, so we cannot infer causal relationships. Regarding measurements, in-laboratory polysomnography was not used to diagnosis OSA in the parent studies. Thus, the use of home sleep apnea testing, the ApneaLinkPlus, may have resulted in an underestimated proportion of persons with OSA. The ISI was designed as a screening tool for insomnia rather than a diagnosis tool and it has lower accuracy compared to diagnosis derived from a clinical interview. Furthermore, although we found that insomnia is a contributing factor in increased mood disturbances and diabetes-related distress in persons with T2DM, this study cannot explain how the severity of insomnia interacts with the severity of OSA in affecting mood and diabetes-related distress. This suggests that future studies evaluating the effects of OSA+I need to examine the interaction effects of OSA and insomnia severity on outcomes. Future intervention studies should examine whether the reduction in insomnia severity or OSA severity is associated with better outcomes in persons with T2DM.
CONCLUSIONS
In our study, insomnia played a significant role in increasing mood disturbances and diabetes-related distress in persons with OSA+I. These results suggest that insomnia, not OSA, may have a more potent influence on mood disturbances and diabetes-related distress in persons with T2DM. Insomnia was associated with increased HbA1c, which may have implications for the management of persons with T2DM and comorbid insomnia. Future research is needed to see if treatment of insomnia improves impaired mood and diabetes-related distress in persons with T2DM and is associated with improved glycemic control. However, at this time, this association highlights the need for appropriate clinical screening for insomnia symptoms even after a diagnosis of OSA in persons with T2DM.
ABBREVIATIONS
| ANCOVA | analysis of covariance |
| AHI | apnea-hypopnea index |
| CI | confidence interval |
| DSTT | Diabetes Sleep Treatment Trial |
| DSTT-I | Diabetes Sleep Treatment Trial: Insomnia |
| ISI | Insomnia Severity Index |
| OSA | obstructive sleep apnea |
| OSA+I | comorbid obstructive sleep apnea and insomnia |
| PAID | Problem Areas in Diabetes |
| POMS | Profile of Mood States |
| T2DM | type 2 diabetes mellitus |
DISCLOSURE STATEMENT
All authors have seen and approved this manuscript. Work for this study was performed at University of Pittsburgh School of Nursing, Pittsburgh, PA. This study was funded by the National Institute of Diabetes and Digestive and Kidney Diseases (E.R. Chasens; R01-DK096028) and the National Institute of Nursing Research (E.R. Chasens; K24-NR016685). This research was also funded by CTSI grants UL1-RR024153 and UL1-TR000005. The authors report no conflicts of interest.
REFERENCES
1. Centers for Disease Control and Prevention. National Diabetes Statistic Report, 2020. https://www.cdc.gov/diabetes/library/features/diabetes-stat-report.html. Accessed May 1, 2021.
2. . Sleep epidemiology—a rapidly growing field. Int J Epidemiol. 2011;40(6):1431–1437.
3. . Prevalence and impact of sleep disorders and sleep habits in the United States. Sleep Breath. 2010;14(1): 63–70.
4. . The association between sleep and diabetes outcomes—a systematic review. Diabetes Res Clin Pract. 2020;161:108035.
5. . Prevalence of insomnia (symptoms) in T2D and association with metabolic parameters and glycemic control: meta-analysis. J Clin Endocrinol Metab. 2020;105(3):614–643.
6. . Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine. Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events. J Clin Sleep Med. 2012;8(5):597–619.
7. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Washington, DC: American Psychiatric Association; 2013.
8. . A comparison of diabetes self-care behavior in people with type 2 diabetes with and without insomnia symptoms. Acta Diabetol. 2020;57(6):651–659.
9. . The effects of cognitive behavioral therapy for insomnia in people with type 2 diabetes mellitus, pilot RCT part II: diabetes health outcomes. BMC Endocr Disord. 2020;20(1):136.
10. . Obstructive sleep apnea and diabetes: a state of the art review. Chest. 2017;152(5):1070–1086.
11. . The prevalence of comorbid depression in patients with type 2 diabetes: an updated systematic review and meta-analysis on huge number of observational studies. Acta Diabetol. 2019; 56(6):631–650.
12. . The prevalence of diabetes-specific emotional distress in people with type 2 diabetes: a systematic review and meta-analysis. Diabet Med. 2017;34(11):1508–1520.
13. . The prevalence of comorbid depression in adults with diabetes: a meta-analysis. Diabetes Care. 2001;24(6):1069–1078.
14. . DAWN2 Study Group. Diabetes Attitudes, Wishes and Needs second study (DAWN2™): cross-national benchmarking of diabetes-related psychosocial outcomes for people with diabetes. Diabet Med. 2013;30(7):767–777.
15. . The relationship between diabetes distress and clinical depression with glycemic control among patients with type 2 diabetes. Diabetes Care. 2010;33(5):1034–1036.
16. . Diabetes distress but not clinical depression or depressive symptoms is associated with glycemic control in both cross-sectional and longitudinal analyses. Diabetes Care. 2010;33(1):23–28.
17. . Differentiating symptoms of depression from diabetes-specific distress: relationships with self-care in type 2 diabetes. Diabetologia. 2008;51(10):1822–1825.
18. . Depression and poor glycemic control: a meta-analytic review of the literature. Diabetes Care. 2000;23(7):934–942.
19. . Psychosocial care for people with diabetes: a position statement of the American Diabetes Association. Diabetes Care. 2016;39(12):2126–2140.
20. . Insomnia as a predictor of depression: a meta-analytic evaluation of longitudinal epidemiological studies. J Affect Disord. 2011;135(1-3):10–19.
21. . Insomnia and the risk of depression: a meta-analysis of prospective cohort studies. BMC Psychiatry. 2016;16(1):375.
22. . Obstructive sleep apnea and depression: a systematic review and meta-analysis. Maturitas. 2020;142:45–54.
23. . Co-morbid insomnia and sleep apnea (COMISA): prevalence, consequences, methodological considerations, and recent randomized controlled trials. Brain Sci. 2019;9(12):371.
24. . Depressive symptoms in comorbid obstructive sleep apnea and insomnia: an integrative review. West J Nurs Res. 2021; 43(11):1061–1072.
25. . Diabetes sleep treatment trial: premise, design, and methodology. Contemp Clin Trials. 2019;76:104–111.
26. . Portable Monitoring Task Force of the American Academy of Sleep Medicine. Clinical guidelines for the use of unattended portable monitors in the diagnosis of obstructive sleep apnea in adult patients. J Clin Sleep Med. 2007;3(7): 737–747.
27. . Validation of the ApneaLink for the screening of sleep apnea: a novel and simple single-channel recording device. J Clin Sleep Med. 2007;3(4):387–392.
28. . The Insomnia Severity Index: psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep. 2011;34(5):601–608.
29. . Profile of Mood States: POMS: Technical Update. Multi-Health Systems; 2007.
30. . Assessment of diabetes-related distress. Diabetes Care. 1995;18(6):754–760.
31. . A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep. 1991;14(6):540–545.
32. . The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193–213.
33. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Lawrence Erlbaum Associates; 1988.
34. . Insomnia disorder. Nat Rev Dis Primers. 2015;1(1):15026.
35. . Comorbidity between sleep apnea and insomnia. Sleep Med Rev. 2009;13(4):287–293.
36. . Sleep, insomnia, and depression. Neuropsychopharmacology. 2020;45(1):74–89.
37. . Obstructive sleep apnea and depression. Sleep Med Rev. 2009;13(6):437–444.
38. . Insomnia and depression: a multifaceted interplay. Curr Psychiatry Rep. 2009;11(6):437–442.
39. . Depression. Lancet. 2018;392(10161):2299–2312.
40. . Sleep disturbances in depression. Sleep Med Clin. 2015;10(1):17–23.
41. . Multimodal abnormalities of brain structure and function in major depressive disorder: a meta-analysis of neuroimaging studies. Am J Psychiatry. 2020;177(5):422–434.
42. . Structural and functional brain alterations in obstructive sleep apnea: a multimodal meta-analysis. Sleep Med. 2019;54:195–204.
43. . Explore structural and functional brain changes in insomnia disorder: a PRISMA-compliant whole brain ALE meta-analysis for multimodal MRI. Medicine (Baltimore). 2020;99(14):e19151.
44. . Neuroimaging insights into the link between depression and insomnia: a systematic review. J Affect Disord. 2019;258:133–143.
45. . A systematic review of the relationships among psychosocial factors and coping in adults with type 2 diabetes mellitus. Int J Nurs Sci. 2019;6(4):468–477.
46. ; National Glycohemoglobin Standardization Program (NGSP) Steering Committee. Status of hemoglobin A1c measurement and goals for improvement: from chaos to order for improving diabetes care. Clin Chem. 2011;57(2):205–214.
47. . Effects of sleep disorders on hemoglobin A1c levels in type 2 diabetic patients. Chin Med J (Engl). 2015;128(24): 3292–3297.
48. . The relationship between excessive daytime sleepiness and metabolic syndrome in severe obstructive sleep apnea syndrome. Clin Respir J. 2016;10(6):714–721.
49. . Non-sleepy obstructive sleep apnoea: to treat or not to treat? Eur Respir Rev. 2019;28(154):190031.
50. . Excessive daytime sleepiness in a general population sample: the role of sleep apnea, age, obesity, diabetes, and depression. J Clin Endocrinol Metab. 2005;90(8):4510–4515.
51. . Prevalence of sleepiness and associations with quality of life in patients with sleep apnea in an online cohort [published online 2021, Dec 1]. J Clin Sleep Med. 2021. doi:
52. . Secondary insomnia: a myth dismissed. Sleep Med Rev. 2006;10(1):3–5.
53. . Prevalence of obstructive sleep apnea in the general population: a systematic review. Sleep Med Rev. 2017;34:70–81.
54. . Systematic review of clinical practice guidelines to identify recommendations for sleep in type 2 diabetes mellitus management. Diabetes Res Clin Pract. 2020;170:108532.
