| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Endocrinology (R.B., M.L.-Z., G.J.), Sahlgrenska University Hospital, SE-413 45 Gothenburg, Sweden; and Department of Mathematical Sciences (A.O.), Chalmers University of Technology, SE-412 96 Gothenburg, Sweden
Address all correspondence and requests for reprints to: Ragnhildur Bergthorsdottir, M.D., Department of Endocrinology, gr str 8, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden. E-mail: Ragnhildur.Bergthorsdottir{at}medic.gu.se.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Methods: A population-based, retrospective, observational study was performed, using the National Swedish Hospital and Cause of Death Registers, covering the period from 1987–2001. After a diagnosis of Addison’s disease, each patient was followed until the end of follow-up or death. Mortality was compared with that of the Swedish background population.
Findings: We identified 1675 patients (995 women and 680 men) diagnosed with primary adrenal insufficiency. The average follow-up from initial diagnosis was 6.5 yr. Five hundred seven patients died during the study period compared with an expected 199. The risk ratio for all-cause mortality was 2.19 (confidence interval 1.91–2.51) for men and 2.86 (confidence interval 2.54–3.20) for women. The excess mortality in both men and women was attributed to cardiovascular, malignant, and infectious diseases. Concomitant diabetes mellitus was observed in 12% of the patients, but only contributed to the increased mortality to a minor extent.
Interpretation: Compared with the background population, we observed that the risk ratio for death was more than 2-fold higher in patients with Addison’s disease. Cardiovascular, malignant, and infectious diseases were responsible for the higher mortality rate.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the 1950s, before the availability of glucocorticoids, the 1-yr survival rate in patients with Addison’s disease was 20% or less (12). However, Mason (8) studied survival after the introduction of glucocorticoid replacement therapy and estimated that mortality was similar to that of the background population, except in the case of those patients in whom the disease was undiagnosed at the time of death and in patients living under poor social conditions. Therefore, survival now is considered to be normal or comparable with that of the background population in patients receiving replacement therapy and adequate follow-up.
The dramatic improvement in survival observed after the availability of glucocorticoid replacement therapy, the low prevalence of the disease, and the reassuring data from Dunlop (12) and Mason (8) have not prompted additional survival studies in Addison’s patients. However, recent studies have demonstrated excess mortality in hypopituitary patients, a large proportion of whom are similarly dependent on life-long glucocorticoid replacement therapy (13, 14, 15). Whether these observations are the result of inadequate glucocorticoid replacement therapy (15), inadequate replacement with sex steroids (14), or untreated growth hormone deficiency (13) is not clear. To study the long-term outcome of patients with a life-long daily requirement for glucocorticoid replacement therapy, we investigated the standardized mortality rate of patients with Addison’s disease between 1987–2001 in Sweden. The existence of low population diversity within Sweden, a homogenous health care system, and the Swedish National Board of Health and Welfare (SNBW) Registers all offer unique conditions under which to study the long-term outcome of this rare disease.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The National Hospital and Cause of Death Registers at the SNBW were used to identify and track patients. We evaluated all patients hospitalized in Swedish hospitals for primary adrenal insufficiency between 1987–2001 using a technique similar to that described previously (16, 17). Patients with primary adrenal insufficiency who were coded 255.4 according to the International Classification of Diseases, ninth revision (ICD 9) and E27.1 or E27.2 (adrenal crisis) according to the International Classification of Diseases, tenth revision (ICD 10), were identified. Patients who were, at any time, diagnosed with Cushing’s syndrome or any pituitary disease were excluded from the analysis. We also looked for the presence of diabetes mellitus (DM) at the time of the initial diagnosis of primary adrenal insufficiency. By making use of unique, personal identification codes, each individual was counted once only. The patient’s county of residence and date of hospitalization were recorded. After the registered hospitalization during which the diagnosis of primary adrenal failure first occurred, each patient was tracked until the end of follow-up or death. Cause of death was identified using the Swedish Cause of Death Register. The mortality rate after hospitalization for primary adrenal failure was compared with that of the general Swedish population.
More than 99% of deaths are reported in the Cause of Death Register and the frequency of miscoding is between 1.2–6.3%, as has been shown by repeated quality control tests (18). The National Hospital Register covers more than 99% of all hospital admissions, and the rate of miscoding reported during quality control checks is less than 8.3% (19).
Local study
To validate the selection criteria used in the national survey, a similar search was performed using a local hospital register. Patients with primary adrenal insufficiency, adrenal crises, drug-induced adrenal insufficiency and other unspecified adrenal insufficiencies (ICD 10 codes E27.1–4), and receiving inpatient or outpatient medical care between 1999 and 2003 were identified and their medical records reviewed. A broader search strategy was performed in the local register to estimate the proportion of patients with Addison’s disease who were not identified in the National Survey, as well as the proportion classified with adrenal crises or primary adrenal insufficiency who did not have Addison’s disease.
Statistical analysis
The expected numbers of deaths, if the risk coincided with that of the general population, were calculated for men and women separately, taking age and calendar time into account. Comparisons between observed and expected numbers were performed using Poisson distributions, which were also applied to the calculation of 95% confidence intervals (CI) of risk ratios (RR). A Poisson model (16, 20) was used to estimate the risk of Addison’s disease with respect to latitude. Poisson regression was also used to estimate the hazard function of death as a continuous function of time from the initial Addison’s diagnosis and depending on the presence of DM at the time of diagnosis. We have used the term "hazard ratio" for the quotient between two hazard functions and RR for the quotient between observed and expected number of deaths. The two terms essentially reflect the same function. The 
2 test was used to assess the difference between regions with respect to the risk of Addison’s. All tests were two-tailed.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The search in the National Hospital Register identified 1675 patients with Addison’s disease between 1987–2001 (Table 1
). The mean age at initial identification was 52.8 yr (SD 22.0), and the average follow-up period was 6.5 yr; 3.4 yr for those patients who died and were found in the National Cause of Death Register and 7.9 yr for those who survived the period of observation.
|
Mortality
Within the time period of 1987–2001, 507 deaths were noted compared with the expected 199 deaths (Fig. 1). The overall RR was 2.19 (CI 1.91–2.51) for men and 2.86 (CI 2.54–3.20) for women. The greatest number of deaths occurred from cardiovascular (n = 239), malignant (n = 73), endocrine (n = 64), respiratory (n = 45), and infectious diseases (n = 12). The relative death rate from cardiovascular and malignant diseases was increased in both men and women (Fig. 2). Ischemic heart disease was the most common cardiovascular cause of death (n = 133) followed by cerebrovascular disease (n = 46). No clustering of cancer type or cancer from a specific organ system was observed (Table 2).
|
|
|
). If patients who died from infectious disease from all disease categories were counted together, a total of 35 patients died from infection, the majority (66%) from pneumonia. One death was related to tuberculosis.
|
Impact of concurrent DM
At the time of the initial identification in the register, 12.6% of the men and 11.4% of the women had a concomitant diagnosis of DM (Table 1
). The RR for death was 1.82 (CI 1.29–2.06) for men and 1.52 (CI 1.11–2.07) for women with Addison’s disease and DM compared with those patients with Addison’s disease without DM. Therefore, DM had a significant influence on total mortality in both men and women.
However, the impact of DM on the excess mortality of the whole group of Addison’s patients was limited. The RR of death for patients with Addison’s disease without DM was 2.04 (CI 1.74–2.37) for men and 2.68 (CI 2.36–3.04) for women as compared with the background population, i.e. 7% less for men and women than the ratios obtained for the whole cohort. The excess mortality among patients with Addison’s disease with and without concomitant DM was most pronounced in the time period close to the initial detection (Fig. 4).
|
Using wider search criteria than in the national study, we identified 122 patients; of these, 105 were included in the analysis. The remaining 17 patients were excluded due to secondary adrenal insufficiency (two), tertiary adrenal insufficiency (three), Waterhouse-Friedrichsen syndrome (one), adrenalectomy due to Cushing’s disease (four) or due to malignant disease (two), miscoded diagnosis (two), and insufficient medical records (three). Therefore, depending on whether using ICD 9 or ICD 10 coding, seven (6%) or six (5%) patients, respectively, who did not have Addison’s disease could have been included in our national cohort.
Only one patient was diagnosed with certainty in the outpatient clinic. Eighty-two percent of the patients (86 of 105) were hospitalized when the diagnosis of Addison’s disease was confirmed. In 18 cases, information on whether patients were diagnosed in an outpatient unit or during hospitalization was unavailable.
The mean age at diagnosis was 37.3 yr (SD 15.5); 42.1 (14.7) in women and 31.1 (14.3) in men. Fifty-two percent (55 of 105) had other endocrine disorders. Forty-one percent (43 of 105) had primary hypothyroidism and 11% (12 of 105) had DM (four type 1, five type 2, and three unknown type). One hundred and two patients received cortisone acetate and three patients received hydrocortisone as replacement therapy. The median daily cortisone acetate dose was 50 mg (range, 18.8–75 mg) administered twice daily in 85% (89 of 105) of the patients. Ninety-four percent (99 of 105) of the patients also received fludrocortisone with a median daily dose of 0.1 mg (range, 0.025–0.2 mg).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A possible explanation for an increased mortality rate in Addison’s patients is inappropriate glucocorticoid replacement therapy in the case of both excess and inadequate glucocorticoid exposure in response to stress and concurrent illnesses. Excess glucocorticoid exposure can induce hypertension, obesity with abdominal fat distribution, and DM; all strong independent risk factors for cardiovascular disease (21, 22). In addition, an attenuated diurnal variation in the serum cortisol profile has been associated with abdominal obesity and the metabolic syndrome supporting the importance of the diurnal profile (23). Finally, a thorough reevaluation of patients undergoing glucocorticoid replacement therapy revealed that doses were too high and could be reduced in the case of more than half of the patients (24). Therefore, it is likely that many patients in the national cohort were receiving overly high doses of glucocorticoids delivered in a nonphysiological pattern, as indicated by the local study.
The increased frequency of other endocrine disorders in patients with autoimmune primary adrenal insufficiency such as mineralocorticoid insufficiency, primary hypothyroidism, and type 1 DM may also compromise survival in these patients. Overly high doses of fludrocortisone, which result in an increase in the activity of the mineralocorticoid receptor, may have deleterious effects on the heart and vascular system (25). Also, untreated hypothyroidism is associated with an increased risk of atherosclerotic disease (26), which may have contributed to increased cardiovascular mortality in our patient cohort if the thyroid status was inadequately monitored and treated. Untreated dehydroepiandrosterone deficiency could also be a contributing factor to the poor outcome in patients with Addison’s disease as indicated in recent dehydroepiandrosterone replacement trials (27, 28, 29).
In the national cohort, the frequency of DM among patients with primary adrenal failure was 12.0%, which is similar to that found in a recent survey (3). This is three to four times higher than the expected prevalence of 2.7–4.3% reported in the Swedish population (30, 31). We were not able to determine the proportion of type 1 and 2 DM in the national cohort, but others have demonstrated an increased frequency of type 1 DM among patients with Addison’s disease (3). The relative risk of death for patients with DM is 3.8 in Sweden (31). Having DM and primary adrenal insufficiency significantly increased the risk of death when compared with having adrenal insufficiency alone. However, the overall impact of concomitant DM on the total mortality was minor. By excluding patients with DM, the overall mortality rate reduced only by 7% in men and women, demonstrating the independent influence of Addison’s disease on the increased relative risk of death.
The relative risk of death from cancer was increased in both men and women. The increased cancer mortality may be an effect of obesity associated with an inadequate replacement therapy or it may be due to the underlying autoimmune disorder. However, the absence of clustering of cancer types such as colon, rectum, breast, endometrial, and kidney cancer, which are associated with obesity (32), or non-Hodgkin’s lymphoma, which is associated with autoimmunity (33), does not support such a notion.
The mortality rate resulting from infectious diseases was more than five times that expected. This is an uncommon cause of death in the age-adjusted background population, and, therefore, the numbers are small. The goal should be to prevent these cases through continuous education of patients and a constant awareness of this risk in the overall care of patients with adrenal insufficiency(15). Tuberculosis did not contribute to the increased mortality rate because only one case was identified.
Published mortality data in patients with hypopituitarism where the majority have secondary adrenal insufficiency have demonstrated increased relative risk of death of approximately 1.8 (13). This increased risk has not been associated statistically with the presence of secondary adrenal insufficiency (14). However, as the excess mortality is of similar magnitude as demonstrated in this study, it might therefore be speculated that adrenal insufficiency is a contributing factor to the increased mortality rate also in hypopituitarism.
Although the study is based on official registers that have a very high yield, there are some limitations to be considered. Patients enter the study when being hospitalized, and this may occur when the diagnosis is first made; however, a patient in whom Addison’s disease has previously been diagnosed can also enter the study because of a concurrent illness. Selection of subjects by this means may explain the higher average age in the national cohort as compared with the mean age at the time of diagnosis in the local survey and in previous cross-sectional studies (3). Very rare causes of primary adrenal failure are likely to be represented within the national cohort, but the contribution to overall mortality is probably small. Furthermore, the study does not allow the detection of patients dying from adrenal insufficiency before diagnosis, which would increase the relative risk of death further. In addition, an unavoidable aspect of the study is that a small proportion of cases may have been incorrectly coded, which is more likely to have occurred in ICD 9 than in ICD 10.
The strength of this study is its size: it is the largest cohort of primary adrenal insufficiency reported in the literature. Moreover, the analysis included a longitudinal follow-up from initial detection to either death or to the end of the study period. Finally, because the majority of new cases were diagnosed during hospitalization in the local study, detection of cases in the national study was evenly distributed within the country and throughout the study period and suggests a high yield of all newly diagnosed cases during the 14-yr study period.
This study has demonstrated that the relative risk of death is more than 2-fold greater in patients with primary adrenal failure. The excess mortality was highest in the period immediately after detection and hospitalization. The excess mortality from infectious diseases may be prevented by adequately educating patients and health care workers and through other safety measures. An unexpected finding was the excess mortality observed as a result of cardiovascular disease and cancer. This could be due to overly high doses of glucocorticoids and a nonphysiological serum cortisol profile associated with the currently available replacement therapy. Concomitant endocrine disorders may have a significant impact on the outcome of this patient group, although our analysis of the impact of DM only explained the increased death rate to a minor extent. The results of this study should pave the way for improvements in the overall care of patients with adrenal insufficiency.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure statement: The authors have nothing to disclose.
First Published Online September 12, 2006
Abbreviations: CI, Confidence interval; DM, diabetes mellitus; ICD 9, International Classification of Diseases, ninth revision; ICD 10, International Classification of Diseases, tenth revision; RR, risk ratio.
Received January 13, 2006.
Accepted September 5, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. R Walker Glucocorticoids and Cardiovascular Disease Eur. J. Endocrinol., November 1, 2007; 157(5): 545 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Hahner, M. Loeffler, M. Fassnacht, D. Weismann, A.-C. Koschker, M. Quinkler, O. Decker, W. Arlt, and B. Allolio Impaired Subjective Health Status in 256 Patients with Adrenal Insufficiency on Standard Therapy Based on Cross-Sectional Analysis J. Clin. Endocrinol. Metab., October 1, 2007; 92(10): 3912 - 3922. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Elfstrom, S. M. Montgomery, O. Kampe, A. Ekbom, and J. F. Ludvigsson Risk of Primary Adrenal Insufficiency in Patients with Celiac Disease J. Clin. Endocrinol. Metab., September 1, 2007; 92(9): 3595 - 3598. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
