ABSTRACT
Objective
Standard glucocorticoid (GC) replacement therapy in autoimmune Addison′s disease (AAD) fails to replicate natural cortisol rhythms. Despite adherence, patients report persistent fatigue, reduced vitality, and impaired wellbeing, ultimately lowering health‐related quality of life (HRQoL). Cortisol is essential for sleep regulation, yet the impact of cortisol imbalance on sleep and HRQoL in AAD remains poorly understood. This study investigates self‐reported sleep impairments and their associations with fatigue and HRQoL in young adults with AAD.
Patients and Methods
Sixty‐four patients with AAD and 128 healthy controls completed validated questionnaires assessing sleep (Karolinska Sleep Questionnaire), fatigue (Multidimensional Fatigue Inventory), and HRQoL (Short Form‐36).
Results
Patients reported significantly more non‐restorative sleep (p = 0.015) than controls. While overall rates of clinically relevant sleep impairments were similar, patients more frequently experienced awakening difficulties (p = 0.011) and struggling to stay awake (p = 0.036). Patients also reported poorer physical health (p < 0.001) and greater general fatigue (p = 0.007), with female patients experiencing more mental fatigue (p = 0.013). Poor sleep and fatigue were associated with reduced HRQoL across the cohort, with patients showing a more pronounced decline in physical health in relation to these factors. Mental health scores remained similar between groups.
Conclusion
Non‐restorative sleep emerged as a distinct feature of AAD. While the prevalence and severity of sleep impairments were similar to controls, the association with poorer physical health was stronger in patients. Mental health remained similar despite sleep disturbances. These findings highlight the importance of addressing even modest sleep disturbances, which may worsen fatigue and reduce physical wellbeing in AAD.
Keywords: Addison′s disease, cortisol, fatigue, mental health, physical health, quality of life, sleep
1. Introduction
Autoimmune Addison′s disease (AAD), the most common form of primary adrenal insufficiency (PAI), results from autoimmune destruction of the adrenal glands and necessitates lifelong glucocorticoid (GC) replacement therapy [1]. While lifesaving, current GC regimens fail to replicate natural cortisol rhythmicity, causing predictable periods of infra‐ and supra‐physiological cortisol levels, particularly overnight and before the morning dose [2, 3]. Despite treatment adherence, patients with AAD report persistent symptoms including poor sleep, fatigue, reduced vitality, depressive symptoms, and impaired physical and mental health, ultimately lowering health‐related quality of life (HRQoL) [4, 5, 6, 7, 8, 9].
Sleep is a complex biological process tightly synchronized with the hypothalamic‐pituitary‐adrenal (HPA) axis, and circadian HPA disruption has been associated with sleep disturbances, mood disorders, and fatigue [10]. Cortisol plays a vital role in maintaining sleep architecture by regulating the timing, transitions and duration of different sleep stages [11, 12]. These processes rely on precisely regulated cortisol levels, with natural oscillations being essential for healthy sleep [13, 14].
Although cortisol′s impact on complex brain functions is well established, its influence on sleep, neurobiology, and wellbeing in AAD remains unclear. A few studies have assessed sleep impairments in PAI using actigraphy, EEG, sleep diaries, and self‐questionnaires [9, 12, 15, 16, 17, 18, 19, 20, 21]. However, findings are inconsistent and often limited by small samples (3–60 patients), with reported outcomes varying from general sleep disruptions and reduced efficiency to no AAD‐specific abnormalities. Given the rarity of the disease, cohorts encompass heterogeneous PAI populations with varying aetiologies and comorbidities, complicating isolation of AAD‐specific effects. Larger studies (20–60 patients) typically focus on middle‐aged adults (mean age 45–50 years) [17, 18, 19, 20, 21], while research on younger patients with AAD is scarce. Despite the well‐established relationship between sleep, fatigue and HRQoL, few studies have directly examined their interplay in AAD [20, 21].
This study aimed to examine the relationship between sleep, fatigue and HRQoL in a cohort of young adults with AAD and few comorbidities. Guided by the theoretical framework where (a) cortisol imbalances disrupt sleep, and (b) sleep disturbances contribute to fatigue and impaired HRQoL [22], we hypothesized that poor sleep may serve as a biological mechanism partially mediating the negative impact of AAD on wellbeing. Using self‐reported data on sleep, fatigue and HRQoL, we explored these interrelated factors. As the largest study to date on this topic, we focused on a younger cohort managing disease‐related challenges alongside education, work, and family life – highlighting the need to address sleep disturbances to support HRQoL in this population.
2. Materials and Methods
Research Ethics
This study was approved by the Regional Ethics Committee of Karolinska Institute and by the Swedish Ethical Review Authority (Dnr: 99‐153, 2011/1764‐32, 20140917, 2017/1658‐32, 2018/1037‐32, 2020‐00564, 2023‐05127‐02). All participants gave written informed consent.
2.1. Participants
All 64 patients and 44 of 128 healthy controls were initially recruited from a previous study on PAI at Karolinska Institute, Stockholm [23], where participants completed questionnaires, neuropsychological testing, and brain fMRI. For this study, 84 additional controls completed questionnaires remotely. Controls were matched to patients by sex, age, and education (≥ 3 years of university or not) in a 1:2 ratio, and recruited via the Swedish Population Registry and online advertisements. Exclusion criteria for controls included any condition requiring pharmacological treatment (including psychotropic medications), autoimmune diseases, GC treatment within 3 months, substance abuse, or pregnancy, as determined via telephone interview before inclusion.
Patients with AAD, aged 18–45 years and diagnosed ≥ 2 years before testing, were recruited via the Swedish Addison Registry [24]. Diagnosis was based on clinical and biochemical criteria, with all patients positive for 21‐OH‐antibodies and well‐monitored at their home clinics. Exclusions included autoimmune polyendocrine syndrome type 1 (APS‐1), diabetes, epilepsy, severe psychiatric history, substance abuse, or pregnancy. Comorbid hypothyroidism was allowed due to its high prevalence in AAD, as were antidepressant use.
Sleep medication was permitted in both groups to avoid underestimating sleep impairments and preserve sample authenticity. However, no participants reported regular use.
The final cohort included 64 patients with AAD (38 females) and 128 controls (76 females). Among patients, 28 had comorbid hypothyroidism, six females used antidepressants, and seven females received dehydroepiandrosterone (DHEA) treatment. 50 patients were treated with immediate‐release hydrocortisone (IR‐HC), and 14 with dual‐release hydrocortisone (DR‐HC; Plenadren®), with doses converted to IR‐HC equivalents using the formula: DR‐HC dose (mg)×0.806 [25].
2.2. Measures
All participants completed a sociodemographic questionnaire, where patients also reported medication details and age at diagnosis.
The Karolinska Sleep Questionnaire (KSQ) assesses habitual subjective sleep. Its first dimension includes 18 items forming four mean indexes: sleep quality, non‐restorative sleep, sleep apnea, and sleepiness. Originally scored on a five‐point scale [26], it has been revised to a six‐point format (1=always, 6=never), where lower scores reflect poorer sleep. An inverted scoring (0‐5, higher = poorer sleep) also exists [27] but was not used. This study includes individual first‐dimension items and three mean indexes (sleep quality, non‐restorative sleep, sleepiness), along with a final item on overall sleep quality (rated “Very good” to “Very poor”). The sleep apnea index was excluded, as it primarily applies to middle‐aged men and is linked to obesity and lifestyle factors. While lacking formal cutoffs, scores ≤ 3 on any first‐dimension item and/or overall sleep quality ratings of “Fairly poor” or “Very poor” generally indicate clinically relevant sleep problems.
The Multidimensional Fatigue Inventory (MFI‐20) [28] measures dimensions of general, mental, and physical fatigue. Each dimension consists of four items rated on a five‐point scale, where higher scores reflect more fatigue.
The 36‐Item Short Form Health Survey (SF‐36) [29] evaluates HRQoL across eight dimensions: (a) physical functioning, (b) role limitations due to physical health, (c) pain, (d) general health, (e) emotional wellbeing, (f) role limitations due to emotional problems, (g) social functioning, and (h) energy/fatigue, typically summarized into physical (a‐d) and mental (e‐h) health summary scores. Scores range 0‐100, with lower scores reflecting poorer health.
2.3. Statistical Analysis
Measures were summarized using standard procedures [26, 28, 29]. Categorical variables were presented as frequency and percentage; continuous variables as means and standard deviations (SD). Analyses were performed in Stata Statistical Software, Release 18 (StataCorp LLC, College Station, TX, USA), with statistical significance set at < 0.05 unless otherwise specfied.
2.3.1. Main Group Comparisons
Continuous variables were compared using t‐tests or rank‐sum tests, and categorical variables with Chi‐square or Fisher′s exact tests. Associations between AAD and KSQ indexes, MFI‐20 and SF‐36 dimensions were analyzed by linear regression, while overall sleep quality was assessed with ordinal logistic regression. Beyond matching, AAD*sex interactions were tested for all outcomes due to known sex differences in androgen production and symptom profiles in AAD [5, 30, 31]. Clinically relevant sleep disturbances (individual KSQ first‐dimension items, score ≤ 3) were compared using Chi‐square or Fisher′s exact tests.
2.3.2. Predictor Analysis Within the AAD Subgroup
In patients, linear regression assessed associations between demographic/clinical factors and outcomes from group comparisons. Predictors included sex, age, body mass index (BMI), antidepressant use, age at diagnosis, disease duration, GC type (IR‐HC or DR‐HC) and dose (mg and mg/m²), number of daily GC doses, hypothyroidism, and DHEA treatment in females. Variables with p < 0.1 in univariable analysis were included in multivariable models.
2.3.3. Associations Between Sleep, Fatigue and HRQoL
Adjusted linear regression with interaction terms was used to examine associations between sleep, fatigue and HRQoL, based on KSQ indexes (sleep quality, non‐restorative sleep, sleepiness), MFI‐20 (general, physical, and mental fatigue), and SF‐36 (physical and mental health summary scores).
The analysis followed four steps: (1) linear regression to assess the association between sleep (independent variable) and fatigue (dependent variable) in the full cohort, adjusting for relevant covariates; (2) assessment of the relationships between sleep and/or fatigue (independent) with HRQoL (dependent), applying similar adjustments; (3) testing for AAD*sleep/fatigue interactions on physical and mental health summary scores; and (4) conducting post‐hoc, group‐stratified analyses to clarify the direction and strength of significant interactions.
3. Results
3.1. Demographics
Table 1 presents clinical characteristics of the participants. The cohort included 64 patients (38 females) and 128 controls (76 females). Groups differed only in coffee consumption (p = 0.033); other demographics were similar.
Table 1.
Demographic data of all participants.
| Variable | AAD (n = 64) | Controls (n = 128) | p value | |
|---|---|---|---|---|
| Sex ‐ n (%) | Female | 38 (59.4%) | 76 (59.4%) | 1.000 |
| Male | 26 (40.6%) | 52 (40.6%) | ||
| Age (years) | Mean (SD), range | 32.5 (6.6), 19–41.9 | 32.7 (6.6), 18.1–43.4 | 0.808 |
| ≥ 3 years of university studies ‐ n (%) | Yes | 28 (43.8%) | 56 (43.8%) | 1.000 |
| No | 36 (56.3%) | 72 (56.3%) | ||
| Years of study | Mean (SD), range | 14.40 (2.51), 11–21 | 14.40 (2.51), 9–22 | 0.865 |
| Contraceptive pills a ‐ n (%) | 6 (9.4%) | 7 (5.5%) | 0.310 | |
| Alcohol b ‐ n (%) | 49 (76.6%) | 93 (72.7%) | 0.561 | |
| Smoking b ‐ n (%) | 6 (9.4%) | 6 (4.7%) | 0.206 | |
| Coffee consumption b ‐ n (%) | 46 (71.9%) | 108 (84.4%) | 0.033 | |
| Previous psychological treatment b ‐ n (%) | 21 (32.8%) | 48 (37.5%) | 0.523 | |
| Ongoing psychological treatment b ‐ n (%) | 6 (9.4%) | 4 (3.1%) | 0.066 | |
| Age at diagnosis of AAD (years) | Mean (SD), range | 23.1 (6.6), 8–34 | ||
| Disease duration (years) | Mean (SD), range | 9.3 (4.8), 2.5–24.9 | ||
| Type of primary GC ‐ n (%) | IR‐HC | 50 (78.1%) | ||
| DR‐HC | 14 (21.8%) | |||
| Total dose IR‐HC (mg) | Mean (SD), range | 23.7 (6.9), 7.5–40 | ||
| Total dose IR‐HC (mg/m2) | Mean (SD), range | 12.9 (3.5), 4.6–21 | ||
| Number of daily GC doses | Mean (SD), range | 2.3 (0.8), 1–4 | ||
| BMI (kg/m2) | Mean (SD), range | 23.6 (4.2), 17.6–41.9 | ||
| Hypothyroism ‐ n (%) | 28 (43.7%) | |||
| DHEA‐treatment a ‐ n (%) | 7 (10.9%) | |||
| Antidepressants a ‐ n (%) | 6 (9.4%) |
Note: Bold text in the p‐value column indicate statistically significant differences (p < 0.05).
Abbreviations: n, number; SD, standard deviation; AAD, autoimmune Addison′s disease; GC, glucocorticoid; IR‐HC, immediate‐released hydrocortisone; DR‐HC, dual‐release hydrocortisone; BMI, body mass index; DHEA, dehydroepiandrosterone.
Applied to female patients only.
Categorized as ‘Yes′ or ‘No′.
Most patients (78.1%) were treated with IR‐HC with 2‐4 daily doses, while 14 patients (21.8%) received DR‐HC once daily. Nearly half (43.7%; 19 females) had hypothyroidism. A few female patients used antidepressants (9.4% overall, 15.8% of female patients) and/or DHEA treatment (18.4% of female patients).
3.2. Main‐Group Comparisons
Table 2 summarizes group comparisons of sleep disturbances (KSQ), fatigue (MFI‐20), and HRQoL (SF‐36), adjusted for coffee consumption and antidepressant use (the latter allowed only in patients).
Table 2.
Group comparisons of sleep (KSQ), fatigue (MFI‐20), and HRQoL (SF‐36) in patients and controls. Adjusted linear regression (controls as reference) was used for KSQ indexes, MFI‐20, and SF‐36; ordinal logistic regression (“Fairly good” as reference) for subjective sleep quality. Interaction analyses evaluated combined effects of AAD and sex.
| Mean (SD) | Adjusted linear regression (controls = reference) | Interaction analysis (AAD*sex) | |||
|---|---|---|---|---|---|
| AAD | Controls | β‐coefficient (95% CI) | p value | p value | |
| KSQ a | |||||
| Sleep quality index | 4.4 (1.05) | 4.65 (0.93) | −0.27 (−0.58, 0.04) | 0.082 | 0.356 |
| Non‐restorative sleep index | 3.9 (1.27) | 4.3 (1.03) | −0.44 (−0.80, −0.09) | 0.015 | 0.746 |
| Sleepiness index | 4.72 (0.74) | 4.92 (0.78) | −0.19 (−0.44, 0.05) | 0.122 | 0.619 |
| Overall sleep quality b ‐ n (%) | |||||
| Very good | 11 (17%) | 26 (20%) | 0.95 (0.41, 2.20) | 0.898 | |
| Fairly good | 32 (50%) | 63 (49%) | Reference | ||
| Neither good nor poor | 12 (19%) | 17 (13%) | 1.36 (0.55, 3.37) | 0.506 | |
| Fairly poor | 7 (11%) | 20 (16%) | 0.65 (0.23, 1.83) | 0.416 | |
| Very poor | 2 (3.1%) | 2 (1.6%) | 2.22 (0.29, 16.90) | 0.441 | |
| MFI‐20 c | |||||
| General fatigue | 12.7 (3.9) | 10.9 (3.8) | 1.70 (0.48, 2.93) | 0.007 | 0.97 |
| Physical fatigue | 9.5 (4.3) | 8.3 (3.7) | 1.16 (−0.08, 2.39) | 0.066 | 0.63 |
| Mental fatigue | 11.3 (3.5) | 10.5 (3.4) | 0.64 (−0.45, 1.74) | 0.247 | 0.021 |
| SF‐36 a | |||||
| Physical health summary score | 77.3 (19.3) | 89.2 (10.7) | −10.20 (−14.60, −5.81) | < 0.001 | 0.014 |
| Physical functioning | 93.9 (12.1) | 97.2 (6.2) | −2.79 (−5.53, −0.05) | 0.046 | 0.192 |
| Role limitations, physical | 75.0 (37.8) | 93.0 (20.1) | −14.44 (−22.84, −6.05) | 0.001 | 0.010 |
| Pain | 80.1 (23.4) | 88.3 (15.5) | −6.95 (−12.75, −1.14) | 0.019 | 0.143 |
| General health | 60.2 (22.4) | 78.5 (15.2) | −16.63 (−22.28, −10.99) | < 0.001 | 0.085 |
| Mental health summary score | 72.0 (19.1) | 77.8 (16.6) | −4.69 (−10.11, 0.74) | 0.090 | 0.576 |
| Emotional wellbeing | 73.9 (15.7) | 76.2 (15.0) | −1.25 (−5.91, 3.41) | 0.597 | 0.797 |
| Role limitations, emotional | 80.7 (31.9) | 83.9 (28.7) | −1.76 (−11.18, 7.65) | 0.712 | 0.494 |
| Social functioning | 79.9 (24.2) | 90.0 (16.4) | −9.06 (−15.13, −2.98) | 0.004 | 0.033 |
| Energy/fatigue | 53.6 (21.4) | 61.2 (19.4) | −6.68 (−12.97, −0.39) | 0.038 | 0.355 |
Note: Bold text in the p‐value column indicate statistically significant differences (p < 0.05).
Abbreviations: n, number; SD, standard deviation; CI, confidence interval; AAD, autoimmune Addison′s disease; KSQ, Karolinska Sleep Questionnaire; MFI‐20, Multidimensional Fatigue Inventory; SF‐36, 36‐Item Short Form Health Survey.
KSQ, SF‐36 = higher scores reflect better sleep/HRQoL.
Presented as numbers, percentage or relative risk ratio.
MFI‐20 = higher scores reflect more fatigue.
Patients with AAD were associated with, on average, a 0.44‐unit reduction in non‐restorative sleep relative to controls (β = −0.44, 95% CI [−0.80, −0.09], p = 0.015), with no other significant differences across remaining KSQ indexes. The proportion reporting clinically poor sleep quality (“Fairly poor” or “Very poor”) was similar between groups.
Patients reported higher general fatigue (p = 0.007) compared to controls. A significant AAD‐by‐sex interaction emerged for mental fatigue, with post‐hoc analysis showing higher scores in female patients than female controls (β = 1.80, 95% CI [0.47, 3.13], p = 0.008), but no difference in males.
Patients also reported lower physical health summary scores (p < 0.001) and impairments in physical functioning, physical limitations, pain, general health, and social functioning. AAD‐by‐sex interactions showed these differences were mainly driven by female patients, who reported greater physical limitations (β = −26.64, 95% CI [−38.25, −15.04], p = < 0.001), and worse social functioning (β = −15.30, 95% CI [−23.37, −7.22], p = 0.008) than female controls, with no male differences. Physical health summary scores were reduced in both female (β = −16.31, 95% CI [−22.28, −10.33], p < 0.001) and male patients (β = −5.58, 95% CI [−11.09, −0.07], p = 0.047), with a stronger effect in females.
Table 3 presents clinically relevant sleep disturbances based on KSQ first‐dimension items (score ≤ 3, disturbances occurring weekly or more often). Significant group differences were found only for awakening difficulties and struggling to staying awake at daytime.
Table 3.
Prevalence of clinically relevant sleep disturbances in patients and controls, defined as scores ≤ 3 on individual KSQ first‐dimension items (sleep problems occurring once a week or more often).
| AAD (n = 64) | Controls (n = 128) | p value | |
|---|---|---|---|
| Awakening difficulties | 22 (34.4%) | 23 (18.0%) | 0.011 |
| Not well‐rested at awakening | 28 (43.8%) | 45 (35.2%) | 0.248 |
| Disrupted sleep | 16 (25.0%) | 25 (19.5%) | 0.383 |
| Premature awakenings | 16 (25.0%) | 20 (15.6%) | 0.117 |
| Struggling to stay awake | 10 (15.6%) | 8 (6.3%) | 0.036 |
| Sleepy at work | 19 (29.7%) | 34 (26.6%) | 0.648 |
Note: Bold text in the p‐value column indicate statistically significant differences (p < 0.05).
Abbreviations: AAD, Autoimmune Addison′s disease; KSQ, Karolinska Sleep Questionnaire.
3.3. Predictor Analysis Within the AAD Subgroup
Among patients, older age was associated with better outcomes, including higher non‐restorative sleep index (β = 0.05, 95% CI [0.00, 0.09], p = 0.034) and less physical fatigue (β = −0.19, 95% CI [−0.35, −0.03], p = 0.019).
Female sex predicted poorer outcomes, including more mental fatigue (β = 2.46, 95% CI 0.77, 4.15], p = 0.005), lower physical health summary score (β = −10.43, 95% CI [−19.74, −1.12], p = 0.029), more physical limitations (β = −21.66, 95% CI [−39.61, −3.70], p = 0.019), and worse social dysfunction (β = −16.38, 95% CI [−27.24, −5.52], p = 0.004).
Higher BMI was linked to lower non‐restorative sleep index (β = −0.07, 95% CI [−0.14, −0.00], p = 0.047), lower mental health summary score (β = −1.29, 95% CI [−2.35, −0.22], p = 0.019), and worse social dysfunction (β = −2.22, 95% CI [−3.50, −0.95], p = 0.001).
Among disease‐specific variables, older age at diagnosis was linked to higher sleepiness index (β = 0.03, 95% CI [0.00, 0.06], p = 0.024). DR‐HC use predicted more physical (β = −24.09, 95% CI [−45.42, −2.75], p = 0.028) and emotional limitations (β = −21.05, 95% CI [−39.72, −2.38], p = 0.028). Higher total GC dose (mg/m2) was associated with lower physical functioning (β = −2.73, 95% CI [−4.55, −0.91], p = 0.004). Among female patients, DHEA treatment was associated with more emotional limitations (β = −29.4, 95% CI [−58.17, −0.51], p = 0.046). No significant associations were found for disease duration, GC dose in mg, number of daily GC doses, hypothyroidism, or antidepressant use.
3.4. Associations Between Sleep, Fatigue and HRQoL
Table 4 summarizes the results. After adjustment, higher non‐restorative sleep and sleepiness indexes were significantly associated with less general, physical, and mental fatigue across the full cohort. Higher sleep quality index was associated with better physical health summary scores. Physical fatigue was linked to poorer physical health, while general and mental fatigue were associated with lower mental health outcomes.
Table 4.
Associations between sleep, fatigue and HRQoL in patients and controls. Sleep variables (KSQ) modeled as independent variables; fatigue as both dependent and independent; and HRQoL (SF‐36) as dependent outcomes. Interaction terms (AAD × sleep/fatigue) tested if HRQoL associations differed by disease status.
| MFI‐20 | SF‐36 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| General fatigue | Physical fatigue | Mental fatigue | Physical health summary score | Mental health summary score | |||||||
| β‐coefficient (95% CI) | p value | β‐coefficient (95% CI) | p value | β‐coefficient (95% CI) | p value | β‐coefficient (95% CI) | p value | β‐coefficient (95% CI) | p value | ||
| Adjusted multivariable linear regression | |||||||||||
| KSQ | Sleep quality | −0.17 (−0.70, 0.35) | 0.521 | 0.43 (−0.24, 1.10) | 0.204 | −0.17 (−0.75, 0.41) | 0.559 | 2.32 (0.08, 4.57) | 0.042 | 1.83 (−0.72, 4.38) | 0.158 |
| Non‐restorative sleep | −1.07 (−1.56, −0.57) | < 0.001 | −1.24 (−1.87, −0.60) | < 0.001 | −0.74 (−1.28, −0.19) | 0.008 | −0.87 (−3.08, 1.34) | 0.436 | 0.24 (−2.27, 2.75) | 0.853 | |
| Sleepiness | −1.49 (−2.12, −0.85) | < 0.001 | −0.94 (−1.76, −0.13) | 0.024 | −0.89 (−1.59, −0.19) | 0.013 | −0.38 (−3.24, 2.48) | 0.795 | 2.01 (−1.24, 5.26) | 0.224 | |
| MFI‐20 | General fatigue | −0.59 (−1.37, 0.18) | 0.132 | −1.50 (−2.38, −0.62) | 0.001 | ||||||
| Physical fatigue | −1.92 (−2.60, −1.25) | < 0.001 | −0.48 (−1.24, 0.29) | 0.222 | |||||||
| Mental fatigue | −0.57 (−1.15, 0.01) | 0.053 | −0.68 (−1.34, −0.02) | 0.042 | |||||||
| Interactions | |||||||||||
| AAD*Sleep quality | 0.247 | 0.231 | 0.515 | 0.902 | 0.848 | ||||||
| AAD*Non‐restorative sleep | 0.970 | 0.901 | 0.595 | 0.014 | 0.648 | ||||||
| AAD*Sleepiness | 0.478 | 0.823 | 0.513 | 0.002 | 0.778 | ||||||
| AAD*General fatigue | < 0.001 | 0.249 | |||||||||
| AAD′Physical fatigue | 0.002 | 0.777 | |||||||||
| AAD*Mental fatigue | 0.070 | 0.068 | |||||||||
Note: Bold text in the p‐value column indicate statistically significant differences (p < 0.05).
Abbreviations: CI, confidence interval; AAD, autoimmune Addison′s disease; HRQoL, health‐related quality of life; KSQ, Karolinska Sleep Questionnaire; MFI‐20, Multidimensional Fatigue Inventory; SF‐36, 36‐Item Short Form Health Survey.
Interaction analyses revealed several significant AAD‐by‐symptom interactions for physical health. Post‐hoc analyses showed these associations existed in both patients and controls but were more pronounced in patients. Specifically, higher non‐restorative sleep index predicted better physical health in both patients (β = 6.98, 95% CI [3.49, 10.46], p < 0.001) and controls (β = 2.72, 95% CI [0.99, 4.46], p = 0.002), with a stronger effect in patients. Similar patterns appeared for sleepiness index (patients: β = 11.68, 95% CI [5.81, 17.54], p < 0.001; controls: β = 3.08, 95% CI [0.71, 5.44], p = 0.011). Both general and physical fatigue were negatively associated with physical health, but again more strongly in patients (general fatigue: β = −3.26, 95% CI [−4.18, −2.33], p < 0.001; physical fatigue: β = −2.89, 95% CI [−3.75, −2.03], p < 0.001) than controls (general fatigue: β = −1.27, 95% CI [−1.71, −0.83], p < 0.001; physical fatigue: β = −1.59, 95% CI [−2.02, −1.17], p < 0.001).
No significant interactions were found for mental health outcomes.
4. Discussion
This study highlights the clinical relevance and disproportionate burden of sleep impairments in AAD, particularly regarding fatigue and HRQoL. Focusing on a young, working‐age cohort, it emphasizes how sleep impairments may disrupt daily functioning in individuals balancing education, work, and family life. While overall sleep disturbances were similar to controls, non‐restorative sleep emerged as a distinguishing feature. When present, sleep impairments more negatively affected physical health in the AAD group, suggesting increased physical vulnerability, while mental health remained similar. Consistent with prior research, patients reported more general fatigue and poorer HRQoL. Female patients appeared particulary affected, also showing higher mental fatigue than female controls. These findings emphasize that even subtle sleep impairments may worsen disease burden and warrant increased clinical attention.
4.1. Sleep Impairments
AAD was associated with lower scores on the non‐restorative sleep index, reflecting difficulties with awakening, not feeling rested, and morning exhaustion. This aligns with the absence of a pre‐awakening cortisol peak in AAD, a physiological mechanism facilitating morning arousal [14, 15]. Compared to normative data, only the non‐restorative sleep index was elevated (mean: 3.9, reflecting difficulties several times per month to 1–2 times per week), though still within the 90th percentile ‐ a proposed threshold for clinically relevant problems [27].
Overall sleep quality did not differ significantly, but individual KSQ first‐dimension items revealed clinically relevant symptoms, particularly non‐restorative sleep and awakening difficulties. Compared to population data, patients with AAD reported more impairments in these areas, whereas other sleep issues were similar to the general population [32]. This suggests that while general sleep may be preserved, non‐restorative sleep and morning arousal remain specific concerns in AAD.
Our results contrast with Henry et al., who reported more widespread sleep impairments in AAD [21]. This may reflect cohort differences, particularly age, as our sample was considerably younger (mean 32.5 years vs. 50.6 years). In contrast, Lovås et al. found no major sleep disturbances but increased fatigue without excessive daytime sleepiness [20] – broadly consistent with our findings.
Consistent with findings in the general population [33], higher BMI was associated with poorer sleep in the patient group and may reflect overlapping mechanisms affecting both metabolic and neuroendocrine regulation in AAD. In contrast, increasing age and older age at diagnosis were linked to more favorable sleep outcomes, which differs from previous studies reporting more sleep difficulties with older age in AAD [20, 21]. This discrepancy may reflect our young cohort, in which age‐related sleep issues have not yet emerged. Additionally, earlier diagnosis in a young sample may indicate a more severe disease course, which could in turn contribute to sleep problems. However, whether age‐related factors drive these associations remain unclear, as our cross‐sectional design precludes disentangling these effects.
Given cortisol′s central role in sleep‐wake regulation, the non‐physiological profile of current GC therapy in AAD may contribute to sleep disturbances [10, 14]. Few studies have examined how different GC regimens affect sleep. Oksnes et al. found improved HRQoL with continuous subcutaneous hydrocortisone infusion (CSHI) compared to IR‐HC but no clear sleep benefits, noting that limited statistical power precluded firm conclusions on these effects [17]. Smans et al. showed GC over‐substitution impaired sleep, with improvements after dose adjustments guided by salivary cortisol [19]. Krekeler et al. reported fewer sleep disturbances with DR‐HC than IR‐HC, and better sleep with higher DR‐HC doses, though HRQoL differences were not significant [18]. These findings suggest non‐physiological cortisol rhythms may impair sleep in AAD, with individualized treatments offering potential benefits. We found no significant associations with GC type or dose, though interpretation is limited by few DR‐HC users (n = 14) and lack of cortisol measurements.
4.2. Fatigue and HRQoL
Despite adequate treatment, our AAD cohort reported greater general fatigue and reduced HRQoL than controls, consistent with prior research [4, 5, 6, 7, 8]. Risk factors included younger age, female sex, and higher BMI. Disease‐specific factors seemed less influential, though higher total GC dose was linked to poorer physical health. These findings align with studies associating female sex [4, 5, 6, 7, 8] and higher GC doses [4, 5] with lower HRQoL. In contrast, longer disease duration [6] and older age at diagnosis [7] were not identified as risk factors, likely due to our relatively young, narrow‐age cohort with limited variablity to detect these effects.
Female patients were more affected, reporting more mental fatigue than female controls. A key sex difference in AAD is women′s lower androgen levels due to absent adrenal DHEA, which may contribute to poorer outcomes. In our cohort, both DHEA supplementation in women and DR‐HC use were associated with HRQoL differences, though small sample sizes (n = 7 and n = 14) limit interpretation. Since neither treatment is standard in Sweden, recipients may possibly have had more severe symptoms beforehand, influencing both treatment decisions and outcomes. DHEA users reported more emotional limitations, and DR‐HC users reported both physical and emotional challenges. These findings echo earlier inconsistent results on DHEA and DR‐HC, underscoring the need for larger, controlled trials to clarify their clinical impact [30, 31, 34, 35].
4.3. Broader Implications
Healthy sleep is vital for wellbeing, supporting recovery, mental restoration, and cognitive‐emotional health, while disruptions in sleep or circadian rhythms are linked to metabolic imbalances, fatigue, and reduced HRQoL [10, 12]. Given cortisol′s key role in sleep‐wake regulation and the misalignment of GC therapy with natural rhythms in AAD, patients may be particularly vulnerable to sleep disturbances, which in turn may exacerbate fatigue and worsen HRQoL.
Our findings support prior research linking sleep disturbances to fatigue and poorer physical health in AAD [12, 21], but add nuance: despite similar sleep quality, sleep impairments corresponded to a greater decline in physical health in patients compared to controls. This suggests a heightened physical vulnerability, where even mild sleep issues may have amplified effects.
Notably, this increased vulnerability did not extend to mental health outcomes. Although sleep impairments were associated with poorer physical health, mental health scores remained similar to controls, even when sleep impairments were present. This may reflect psychological coping mechanisms that preserve mental wellbeing despite physical strain.
Importantly, the relationship between sleep and physical health is likely bidirectional. AAD, marked by fluctuating symptoms and reduced physical resilience, may itself disrupt sleep and increase fatigue. Rather than sleep being the sole driver of impaired HRQoL, disease‐related challenges may equally degrade sleep and recovery. This self‐reinforcing interplay highlights the importance of holistic management targeting both physiological and behavioral aspects of AAD.
4.4. Clinical and Research Implications
Our findings highlight the need for greater clinical attention to sleep in AAD. While sleep impairments may not be more frequent or severe than in the general population, non‐restorative sleep is a distinctive concern. Moreover, the negative effect of sleep disturbances on physical health appears disproportionately greater in AAD, suggesting that even modest disturbances warrant attention.
Future research should explore the mechanisms behind these associations and evaluate interventions optimizing GC regimens to improve sleep and HRQoL. Although evidence is limited, promising physiological approaches like CSHI and modified‐release preparations mimicking the early morning cortisol rise (e.g., Chronocort®), may offer benefits and call for further investigation [36]. Meanwhile, early identification and targeted strategies ‐ including treatment optimization, sleep hygiene, and behavioral or pharmacological support ‐ may help reduce fatigue and improve daily functioning in AAD.
4.5. Strengths and Limitations
This study features the largest AAD cohort to date with minimal comorbidities, offering valuable insight into sleep, fatigue, and HRQoL. However, as in most rare disease research, sample size limits statistical power, especially in subgroup analyses.
Selection bias also poses a concern. While antidepressant use was permitted in the AAD group, controls were required to be healthy and medication‐free. Although analyses were adjusted for this, controls may still represent a healthier‐than‐average population. For context, 6–9% of Swedish men and 13–17% of women aged 18–44 use antidepressants [37] compared to 15.8% of female patients in our study, which is consistent with national averages.
Another limitation is reliance on self‐reported data, which is subjective and prone to recall and interpretation biases. Some controls completed questionnaires online, potentially adding selection bias. Nevertheless, self‐administered tools enabled efficient data collection in a well‐matched, geographically dispersed cohort. While objective measures like actigraphy or polysomnography would add value, this study focused on subjective experiences. The use of validated instruments supports reliability and comparability of findings.
Despite limitations, our results offer important insights into how young adult patients with AAD experience sleep, fatigue and HRQoL while managing work, education, and family alongside disease‐related challenges.
5. Conclusion
This study highlights the interplay between sleep, fatigue, and HRQoL in young adults with AAD. While overall sleep quality was comparable to controls, non‐restorative sleep and awakening difficulties emerged as specific challenges. Consistent with prior research, patients reported lower HRQoL and greater fatigue, with female patients experiencing more mental fatigue than female controls. Across the cohort, poor sleep and fatigue were closely linked to reduced physical and mental health with a stronger negative effect on physical health in AAD, suggesting increased vulnerability. Mental health outcomes, however, remained similar between groups.
These findings underscore the need for greater clinical awareness of sleep issues in AAD, as even mild disturbances may negatively affect physical health. Future research should confirm and build on these results, as early detection and treatment of sleep disturbances may reduce fatigue and improve daily functioning. In time, therapies better mimicking natural hormone rhythms may improve sleep and HRQoL in AAD.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgements
This study was supported by the Regional Agreement on Medical Training and Clinical Research (ALF) between Stockholm County Council and Karolinska Institutet (to S.B.); the Marianne and Marcus Wallenberg Foundation; Stockholm County Council (ALF, to S.L.); the Swedish Research Council (DNR 2021‐02440); Region Stockholm (clinical research appointment, DNR RS 2019‐1140, to S.L.); the Lisa and Johan Grönberg Foundation; Stiftelsen Frimurare Barnhuset, Stockholm; Samariten; Jerringfonden; Sällskapet Barnavård; Wera Ekströms Stiftelse för Pediatrikforskning; and the Foundation for Research and Education in Pediatric Endocrinology.
Fletcher‐Sandersjöö S., Westeinde A. V., Spelman T., Hirvikosi T., Lajic S., and Bensing S., “The Relationship Between Sleep, Fatigue and Quality of Life in Young Adults With Autoimmune Addison′s Disease,” Clinical Endocrinology 103 (2025): 795‐804. 10.1111/cen.70028.
Svetlana Lajic and Sophie Bensing should be considered joint senior author.
Data Availability Statement
Data will be made available from the corresponding author upon reasonable request.
References
- 1. Martin‐Grace J., Dineen R., Sherlock M., and Thompson C. J., “Adrenal Insufficiency: Physiology, Clinical Presentation and Diagnostic Challenges,” Clinica Chimica Acta 505 (2020): 78–91. [DOI] [PubMed] [Google Scholar]
- 2. Choudhury S., Lightman S., and Meeran K., “Improving Glucocorticoid Replacement Profiles in Adrenal Insufficiency,” Clinical Endocrinology 91, no. 3 (2019): 367–371. [DOI] [PubMed] [Google Scholar]
- 3. Lightman S. and Terry J. R., “The Importance of Dynamic Signalling for Endocrine Regulation and Drug Development: Relevance for Glucocorticoid Hormones,” Lancet Diabetes & Endocrinology 2, no. 7 (2014): 593–599. [DOI] [PubMed] [Google Scholar]
- 4. Hahner S., Loeffler M., Fassnacht M., et al., “Impaired Subjective Health Status in 256 Patients With Adrenal Insufficiency on Standard Therapy Based on Cross‐Sectional Analysis,” Journal of Clinical Endocrinology & Metabolism 92, no. 10 (2007): 3912–3922. [DOI] [PubMed] [Google Scholar]
- 5. Tiemensma J., Andela C. D., Kaptein A. A., et al., “Psychological Morbidity and Impaired Quality of Life in Patients With Stable Treatment for Primary Adrenal Insufficiency: Cross‐Sectional Study and Review of the Literature,” European Journal of Endocrinology 171, no. 2 (2014): 171–182. [DOI] [PubMed] [Google Scholar]
- 6. Kluger N., Matikainen N., Sintonen H., Ranki A., Roine R. P., and Schalin‐Jäntti C., “Impaired Health‐Related Quality of Life in Addison′s Disease—Impact of Replacement Therapy, Comorbidities and Socio‐Economic Factors,” Clinical Endocrinology 81, no. 4 (2014): 511–518. [DOI] [PubMed] [Google Scholar]
- 7. Meyer G., Hackemann A., Penna‐Martinez M., and Badenhoop K., “What Affects the Quality of Life in Autoimmune Addison′s Disease?,” Hormone and Metabolic Research 45, no. 2 (2013): 92–95. [DOI] [PubMed] [Google Scholar]
- 8. van der Valk E. S., Smans L. C. C. J., Hofstetter H., et al., “Decreased Physical Activity, Reduced QoL and Presence of Debilitating Fatigue in Patients With Addison′s Disease,” Clinical Endocrinology 85, no. 3 (2016): 354–360. [DOI] [PubMed] [Google Scholar]
- 9. Henry M., Ross I. L., Wolf P. S. A., and Thomas K. G. F., “Impaired Quality and Efficiency of Sleep Impairs Cognitive Functioning in Addison′s Disease,” Psychoneuroendocrinology 78 (2017): 237–245. [DOI] [PubMed] [Google Scholar]
- 10. Buckley T. M. and Schatzberg A. F., “On the Interactions of the Hypothalamic‐Pituitary‐Adrenal (HPA) Axis and Sleep: Normal HPA Axis Activity and Circadian Rhythm, Exemplary Sleep Disorders,” Journal of Clinical Endocrinology & Metabolism 90, no. 5 (2005): 3106–3114. [DOI] [PubMed] [Google Scholar]
- 11. Follenius M., Brandenberger G., Bandesapt J. J., Libert J. P., and Ehrhart J., “Nocturnal Cortisol Release in Relation to Sleep Structure,” Sleep 15, no. 1 (1992): 21–27. [DOI] [PubMed] [Google Scholar]
- 12. Henry M., Thomas K. G. F., and Ross I. L. S., “Sleep, Cognition and Cortisol in Addison′s Disease: A Mechanistic Relationship,” Front Endocrinol 12 (2021): 694046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Kalafatakis K., Russell G. M., Harmer C. J., et al., “Ultradian Rhythmicity of Plasma Cortisol Is Necessary for Normal Emotional and Cognitive Responses in Man,” Proceedings of the National Academy of Sciences 115, no. 17 (2018): E4091–E4100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Oster H., Challet E., Ott V., et al., “The Functional and Clinical Significance of the 24‐hour Rhythm of Circulating Glucocorticoids,” Endocrine Reviews 38, no. 1 (2017): 3–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Henry M., Ross I. L., and Thomas K. G. F., “Reduced Slow‐Wave Sleep and Altered Diurnal Cortisol Rhythms in Patients With Addison′s Disease,” European Journal of Endocrinology 179, no. 5 (2018): 319–330. [DOI] [PubMed] [Google Scholar]
- 16. Garcia‐Borreguero D., “Glucocorticoid Replacement Is Permissive for Rapid Eye Movement Sleep and Sleep Consolidation in Patients With Adrenal Insufficiency,” Journal of Clinical Endocrinology & Metabolism 85, no. 11 (2000): 4201–4206. [DOI] [PubMed] [Google Scholar]
- 17. Øksnes M., Björnsdottir S., Isaksson M., et al., “Continuous Subcutaneous Hydrocortisone Infusion Versus Oral Hydrocortisone Replacement for Treatment of Addison′s Disease: A Randomized Clinical Trial,” Journal of Clinical Endocrinology & Metabolism 99, no. 5 (2014): 1665–1674. [DOI] [PubMed] [Google Scholar]
- 18. Krekeler C., Kropp P., Blacha A. K., Rahvar A. H., and Harbeck B., “Dual‐Release Hydrocortisone and Its Benefits on Cognitive Function and Quality of Sleep,” Endocrine 72 (2021): 223–233. [DOI] [PubMed] [Google Scholar]
- 19. Smans L., Lentjes E., Hermus A., and Zelissen P., “Salivary Cortisol Day Curves in Assessing Glucocorticoid Replacement Therapy In Addison′s Disease,” Hormones 12, no. 1 (2013): 93–100. [DOI] [PubMed] [Google Scholar]
- 20. Lovas K., Husebye E., Holsten F., and Bjorvatn B., “Sleep Disturbances in Patients With Addison′s Disease,” European Journal of Endocrinology 148, no. 4 (2003): 449–456. [DOI] [PubMed] [Google Scholar]
- 21. Henry M., Wolf P. S. A., Ross I. L., and Thomas K. G. F., “Poor Quality of Life, Depressed Mood, and Memory Impairment May Be Mediated by Sleep Disruption In Patients With Addison′s Disease,” Physiology & Behavior 151 (2015): 379–385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Matsui K., Yoshiike T., Nagao K., et al., “Association of Subjective Quality and Quantity of Sleep With Quality of Life Among a General Population,” International Journal of Environmental Research and Public Health 18, no. 23 (2021): 12835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. van't Westeinde A., Ström S., Hirvikoski T., et al., “Young Adult Swedish Patients With Autoimmune Addison′s Disease Report Difficulties With Executive Functions in Daily Life Despite Overall Good Cognitive Performance,” Psychoneuroendocrinology 140 (2022): 105714. [DOI] [PubMed] [Google Scholar]
- 24. Dalin F., Nordling Eriksson G., Dahlqvist P., et al., “Clinical and Immunological Characteristics of Autoimmune Addison Disease: A Nationwide Swedish Multicenter Study,” Journal of Clinical Endocrinology and Metabolism 102, no. 2 (2017): 379–389. [DOI] [PubMed] [Google Scholar]
- 25. Johannsson G., Nilsson A. G., Bergthorsdottir R., et al., “Improved Cortisol Exposure‐Time Profile and Outcome In Patients With Adrenal Insufficiency: A Prospective Randomized Trial of a Novel Hydrocortisone Dual‐Release Formulation,” Journal of Clinical Endocrinology & Metabolism 97, no. 2 (2012): 473–481. [DOI] [PubMed] [Google Scholar]
- 26. Westerlund A., Lagerros Y. T., Kecklund G., Axelsson J., and Åkerstedt T., “Relationships Between Questionnaire Ratings of Sleep Quality and Polysomnography in Healthy Adults,” Behavioral Sleep Medicine 14, no. 2 (2016): 185–199. [DOI] [PubMed] [Google Scholar]
- 27. Nordin M., Åkerstedt T., and Nordin S., “Psychometric Evaluation and Normative Data for the Karolinska Sleep Questionnaire,” Sleep and Biological Rhythms 11, no. 4 (2013): 216–226. [Google Scholar]
- 28. Lundh Hagelin C., Wengström Y., Runesdotter S., and Johan Fürst C., “The Psychometric Properties of the Swedish Multidimensional Fatigue Inventory MFI‐20 in Four Different Populations,” Acta Oncologica 46, no. 1 (2007): 97–104. [DOI] [PubMed] [Google Scholar]
- 29. Ware J. E. and Sherbourne C. D., “The MOS 36‐ltem Short‐Form Health Survey (SF‐36): I. Conceptual Framework and Item Selection,” Medical Care 30, no. 6 (1992): 473–483. [PubMed] [Google Scholar]
- 30. Alkatib A. A., Cosma M., Elamin M. B., et al., “A Systematic Review and Meta‐Analysis of Randomized Placebo‐Controlled Trials of DHEA Treatment Effects on Quality of Life in Women With Adrenal Insufficiency,” Journal of Clinical Endocrinology & Metabolism 94, no. 10 (2009): 3676–3681. [DOI] [PubMed] [Google Scholar]
- 31. Løvås K., Gebre‐Medhin G., Trovik T. S., et al., “Replacement of Dehydroepiandrosterone in Adrenal Failure: No Benefit for Subjective Health Status and Sexuality in a 9‐month, Randomized, Parallel Group Clinical Trial,” Journal of Clinical Endocrinology & Metabolism 88, no. 3 (2003): 1112–1118. [DOI] [PubMed] [Google Scholar]
- 32. Ohayon M. M. and Bader G., “Prevalence and Correlates of Insomnia in the Swedish Population Aged 19‐75 Years,” Sleep Medicine 11, no. 10 (2010): 980–986. [DOI] [PubMed] [Google Scholar]
- 33. Krističević T., Štefan L., and Sporiš G., “The Associations Between Sleep Duration and Sleep Quality With Body‐Mass Index in a Large Sample of Young Adults,” International Journal of Environmental Research and Public Health 15, no. 4 (2018): pii:E758, 10.3390/ijerph15040758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Delle Cese F., Corsello A., Cintoni M., et al., “Switching From Immediate‐Release to Fractionated Dual‐Release Hydrocortisone May Improve Metabolic Control and QoL In Selected Primary Adrenal Insufficiency Patients,” Front Endocrinol (Lausanne) 11 (2020): 610904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Dineen R., Martin‐Grace J., Ahmed K. M. S., et al., “Cardiometabolic and Psychological Effects of Dual‐Release Hydrocortisone: A Cross‐Over Study,” European Journal of Endocrinology 184, no. 2 (2021): 253–265. [DOI] [PubMed] [Google Scholar]
- 36. Peel A., Rushworth R. L., and Torpy D. J., “Novel Agents to Treat Adrenal Insufficiency: Findings of Preclinical and Early Clinical Trials,” Expert Opinion on Investigational Drugs 33, no. 2 (2024): 115–126. [DOI] [PubMed] [Google Scholar]
- 37. Public Health Agency of Sweden . Statistics on Mental Health in Sweden 2023. [Internet], https://www.folkhalsomyndigheten.se/livsvillkor-levnadsvanor/psykisk-halsa-och-suicidprevention/statistik-psykisk-halsa/. Published 2023, accessed May 28, 2025.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data will be made available from the corresponding author upon reasonable request.