Abstract
Objectives
This study aimed to investigate the impact of diabetes mellitus (DM) on the risk of mortality in patients with giant cell arteritis (GCA).
Methods
This Danish nationwide cohort study included patients diagnosed with both GCA and DM, who were matched 1:3 with patients with only GCA or DM, on sex and year of birth. The Pseudo-observation method and Cox regression were used to evaluate the cumulative incidence proportion (CIP), risk difference (RD), and relative risk of mortality after 6 months and 5 years. Also, time-dependent hazard ratios (HR) were made to assess the impact of predefined HbA1c categories on the risk of mortality.
Results
Among 3674 cases, the CIP of mortality was 8.1% (7.1-9.1) at 6 months and 34.6% (32.7-36.4) at 5 years. Compared with patients with GCA only (n = 10,311), the RD was 5.5% (4.5-6.5) at 6 months and 15.1% (13.3-17.0) at 5 years. Compared with patients with DM only (n = 11,020), the RD was 4.9% (3.9-5.9) and 6.5% (4.7-8.3). The adjusted HR for mortality was significantly higher in patients with elevated HbA1c, especially after 5 years of follow-up.
Conclusions
Having both GCA and DM was associated with a higher mortality risk compared with either condition alone, especially in the initial phase after diagnosis. High HbA1c increased the risk of 5-year mortality in both patients with GCA and DM and GCA alone, suggesting glycaemic control could be relevant in this population.
INTRODUCTION
Giant cell arteritis (GCA) is the most common form of systemic vasculitis, causing inflammation of the large- and medium-sized arteries originating from the aortic arch. GCA occurs in persons older than 50 years of age, with the highest prevalence in women and has a reported incidence of 18 to 29 per 100.000 per year in Scandinavia [[1], [2], [3]]. Glucocorticoids (GCs) are the first-line treatment. This is initiated to achieve remission and prevent disease-related complications. However, as up to 50% of patients experience relapse during the first year of tapering GCs [4], many require prolonged treatment, which increases the risk of adverse events [[2], [3], [4], [5]].
WHAT IS ALREADY KNOWN ON THIS TOPIC
-
•
Previous studies have reported conflicting results regarding mortality in patients with giant cell arteritis (GCA) compared with the general population.
-
•
It is well known that patients with GCA are at a greater risk of developing diabetes mellitus (DM) compared with the general population.
-
•
Although some findings suggest a potential association between DM and mortality in GCA, the precise effect of DM on mortality risk in these patients remains unclear.
WHAT THIS STUDY ADDS
-
•
Patients with GCA and DM had higher 6-month and 5-year risk of mortality compared with matched controls with only GCA or DM.
-
•
Mortality in patients with GCA and DM was particularly increased in the first 6 months after diagnosis.
-
•
Elevated HbA1c levels could be associated with a higher risk of 5-year mortality in patients with GCA and DM.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
-
•
his study underscores the importance of close clinical attention for patients with both GCA and DM, especially during the first months after diagnosis.
Alt-text: Unlabelled box dummy alt text
Conflicting results regarding mortality in patients with GCA compared with the general population have previously been reported [[5], [6], [7], [8], [9], [10], [11], [12], [13], [14]]. Studies that observe excess mortality in patients with GCA often attribute it to GC treatment. Treatment with GC in patients with inflammatory diseases is the leading cause of drug-induced hyperglycaemia and diabetes mellitus (DM), and it can worsen glycaemic control in individuals with pre-existing DM [2]. In patients with GCA, the risk of developing DM has been reported to be up to 7 times higher during the early stages of the disease [15]. A Danish study found an increased cause-specific mortality due to endocrine diseases in patients with GCA included between 1996 and 2018 [12]. However, that study did not compare mortality between patients with GCA with and without DM. Similarly, a study from 2018, using French national death certificate data, found that DM was associated with deaths in patients with GCA [16].
Although these findings suggest a potential link between DM and mortality in GCA, the specific impact of DM on the mortality risk in patients with GCA remains unclear. To date, direct comparisons between mortality outcomes in patients with GCA with and without DM are lacking. Given the increased risk of DM in patients with GCA and the potential implications for mortality, it is important to clarify how DM influences outcomes in this population. Also, the GC sparing agents tocilizumab, an interleukin-6 inhibitor, and upadacitinib, a Janus kinase inhibitor, were approved and introduced into the treatment regimen in 2018 [17] and 2025, respectively, highlighting the need for updated data.
Therefore, the aim of this study was to investigate the impact of DM on mortality among patients with GCA. As a secondary aim, we investigated the impact of HbA1c on mortality in patients with both GCA and DM, and patients with GCA alone.
METHODS
Data sources
This Danish nationwide register-based matched-cohort study included all patients with GCA or DM residing in Denmark at the start of follow-up (January 1, 1996) to December 31, 2022. We obtained information on diagnosis of GCA and DM, type of diagnosis, redeemed prescriptions, laboratory data, migration and death through the Danish National Patient Registry, the Danish National Prescription Registry, and the Clinical Laboratory Information System. All Danish residents have a civil registration number in the Danish Civil Registration System, which allows record linkage between registries [18]. The Data Protection Committee of Northern Jutland, Denmark (2020-032) has approved the study. In this study, sex refers to the binary biological categorisation of male and female, typically assigned at birth based on external anatomy.
Study population
Patients with GCA included in this study were ≥50 years and had a first-time primary or secondary registration of a GCA diagnosis, according to the International Classification of Diseases 10th edition (ICD-10) codes used for GCA (M31.5 and M31.6), between January 1, 1996 and December 31, 2022. This case definition has a positive predictive value of 61% [19]. Patients with DM were included if they had a primary or secondary registration of DM according to the ICD-10 codes for DM (E10-E14, H36.0, and O24, excluding O24.4) or had redeemed 2 oral antidiabetic medications (Anatomical Therapeutic Chemical (ATC) codes A10B and A10X) or insulin prescriptions (ATC code A10A) within a 6-month period. The case definition for DM has previously been validated [20]. Data on HbA1c measurements was obtained with the following Nomenclature for Properties and Units codes from the laboratory information system: NPU03835 and NPU27300.
Exposure and comparator groups
In this study, 1 exposure group and 2 comparator groups were defined, resulting in 2 separate cohorts. The exposure group (cases) was composed of patients diagnosed with both GCA and DM no matter the sequence of diagnoses.
Comparator group 1 consisted of patients diagnosed with GCA who were free of DM at the time of GCA diagnosis. Comparator group 2 consisted of patients diagnosed with DM who were free of GCA at time of DM diagnosis.
Patients in the comparator groups were allowed to change to the exposure group during follow-up. If a patient in comparator group 1 developed DM, they would be censored and change their exposure group to cases. Conversely, if a patient in comparator group 2 developed GCA, the patient was censored and would change exposure group to cases. This ensures that comparator groups 1 and 2 were exclusively composed of patients with either GCA or DM, respectively. Cases were matched with up 3 controls on sex and year of birth in each comparator group, creating 2 different cohorts. Sensitivity analyses on the exposure group were performed, including (i) restricting the interval between GCA and DM diagnoses to 365 days and (ii) including only patients diagnosed with DM after GCA.
Follow-up
For both cohorts, a follow-up period of 6 months and 5 years was chosen to investigate both the short- and long-term impact of DM on the mortality in patients with GCA. For cases, the index date was defined as the date on which they were diagnosed with both GCA and DM. Cases were followed from the index date to the end of follow-up, emigration, or death, whichever came first. For comparator group 1, the index date was defined as the date of their respective matches, which could not be later than 1 year after diagnosis of GCA. Patients in comparator group 1 were followed from the index date to the end of follow-up, emigration, development of DM, or death, whichever came first. For comparator group 2, the index date was the date of matching, which could not be later than 1 year after DM diagnosis. Patients in comparator group 2 were followed from the index date, to the end of follow-up, emigration, development of GCA or death, whichever came first. In Figure 1, a timeline of follow-up, index dates, and censoring in these cohorts can be seen.
Figure 1.
Timeline over follow-up and the index date for patients with giant cell arteritis (GCA), diabetes mellitus (DM), and both.
Outcome and comorbidities
In this study, the primary outcome was mortality. Data on mortality among patients were retrieved from the Danish National Patient Registry. Comorbidities were identified based on records from the Danish National Patient Registry and/or redeemed prescriptions from the National Prescription Registry for relevant medications associated with each condition. These included chronic lung disease, cardiovascular disease (ischaemic heart disease, heart failure, hypertension, and stroke), chronic kidney disease, obesity, and cancer. For detailed definitions, see Supplementary Table S1.
Statistical analysis
In both cohorts, we used the pseudo-observation method to calculate the sex- and age-adjusted 6-month and 5-year cumulative incidence proportion (CIP) in %, risk difference (RD) in %, and relative risk (RR) of mortality. We chose not to adjust for chronic kidney disease and cardiovascular diseases, as we consider them intermediate variables rather than confounders. However, we did adjust for them in a sensitivity analysis, which can be seen in Supplementary Table S2. We also stratified this analysis on sex, age, calendar year, and DM onset. The pseudo-observation method was selected because it allows for the estimation of both RR and RD, providing more intuitive and interpretable results than hazard ratios (HR), which is why it was chosen as a tool for the primary analysis. It supports both absolute and relative comparisons and enables the estimation of cumulative incidence after 6 months and 5 years. Also, unlike Cox regression, it does not rely on the assumption of proportional hazards. To further explore the relationship between GCA and DM, we performed 3 supplementary analyses. In the first, patients with GCA were matched with individuals from the general population; in the second, patients with DM were matched with the general population; and in the third, patients with both GCA and DM were matched with the general population. These analyses allow us to explore the combined effect of GCA and DM.
Secondary analyses
We used Cox proportional hazard regression with time since index date as the underlying time scale to estimate crude and adjusted 6-month and 5-year HRs. This was first done in patients with GCA and DM compared with patients with GCA alone, using DM status as a time-varying exposure. Thereafter, we used HbA1c as a categorical (high [HbA1c > 48 mmol/mol] or low [HbA1c < 48 mmol/mol]) time-varying exposure. This means that patients may change the HbA1c group during follow-up, for example, a patient may have a low HbA1c at baseline but could shift to the high group after 3 months if a new HbA1c measurement is >48 mmol/mol. Lastly, in patients with both GCA and DM, we again used HbA1c as a categorical time-varying exposure, but HbA1c was stratified according to the following thresholds: <48 mmol/mol, 48-52 mmol/mol, 53-57 mmol/mol, 58-63 mmol/mol, 64-74 mmol/mol, 75-85 mmol/mol, and >85 mmol/mol [21]. All adjusted Cox regressions included covariates such as age, sex, social class, and chronic lung disease at baseline. Sex, social class, and chronic lung disease were included as categorical variables. Age was included as a continuous variable and adjusted for using restricted cubic splines with 3 degrees of freedom. Social class was derived from educational level and initially divided into 8 categories (1-8); however, due to low numbers in categories 1 and 8, these were combined with categories 2 and 7, respectively. Proportional hazard assumptions were checked for all included variables. All proportional hazard assumptions were satisfied.
Data management and all analyses were performed in R 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria). The level of statistical significance was set at 5%. All estimates were presented with 95% confidence intervals.
RESULTS
A total of 17,517 patients diagnosed with GCA between 1996 and 2022 were found, of which 3674 at some point were diagnosed with both GCA and DM. The 3674 cases were matched with 10,311 patients diagnosed with GCA only and 11,020 patients diagnosed with DM only. The baseline characteristics are summarised in Table 1. In this study, 11.5% of eligible patients with GCA had DM before GCA diagnosis and 9.5% developed DM after diagnosis of GCA. A total of 1708 and 1230 patients with both DM and GCA and only GCA, respectively, had available HbA1c measurements after the index date. Baseline characteristics of cases with and without HbA1c measurements at baseline can be seen in Supplementary Table S3. Patients with GCA and DM presented similar distributions of age and sex when compared with their matched cohorts. However, cases with both GCA and DM appeared to have a higher prevalence of comorbidities relative to both the GCA-only and DM-only cohorts. Among cases, 45.4% developed DM after the diagnosis of GCA. Of these cases, 16.1% developed DM within the first 3 months, whereas 27.2% experienced the onset of DM within the first 6 months of follow-up. Additional baseline characteristics stratified by sex can be found in Supplementary Table S4.
Table 1.
Baseline characteristics of patients with giant cell arteritis and diabetes mellitus, and their respective matched cohorts
| GCA and diabetes mellitus (n = 3674) |
Matched controls (GCA) (n = 10,311) |
Matched controls (Diabetes mellitus) (n = 11,020) |
|
|---|---|---|---|
| Female, n (%) | 2210 (60.2) | 6357 (61.7) | 6628 (60.1) |
| Age (SD) | 75.3 (8.8) | 75.1 (8.1) | 75.3 (8.8) |
| Cardiovascular disease, n (%) | 1572 (42.8) | 2726 (26.4) | 3667 (33.3) |
| Chronic kidney disease, n (%) | 359 (9.8) | 401 (4.9) | 532 (4.8) |
| Chronic lung disease, n (%) | 1253 (34.1) | 2720 (26.4) | 3101 (28.1) |
| Cancer history, n (%) | 449 (12.2) | 1106 (10.7) | 1375 (12.5) |
| Hospital diagnosis of obesity, n (%) | 342 (9.3) | 205 (2.0) | 779 (7.1) |
| Accumulated glucocorticoids 1 y after index date [IQR] | 3500 [2125-5750] | 2500 [1500-4000] | - |
| Accumulated glucocorticoids 1 y after GCA diagnosis date [IQR] | 5000 [3000-7125] | 5250 [3000-7000] | - |
| Diabetes mellitus before or at date of GCA, n (%) | 2006 (54.6) | - | - |
| New-onset diabetes mellitus after diagnosis of GCA, n (%) | 1668 (45.4) | - | - |
| - Within 3 mo, n (%) | 268 (16.1) | - | - |
| - Within 6 mo, n (%) | 453 (27.2) | - | - |
GCA, giant cell arteritis; N, number.
Impact of DM on mortality in patients with GCA
Within the cohort of patients with GCA and DM, the CIP was 8.1% (7.1-9.1) at 6 months and 34.6% (32.7-36.4) at 5 years. Compared with patients with GCA only, an RD of 5.5% (4.5-6.5) at 6 months and 15.1% (13.3-17.0) at 5 years, along with an RR of 2.3 (1.9-2.6) at 6 months and 1.5 (1.4-1.6) at 5 years, were observed (Table 2). When compared with patients with only DM, the RD and RR were 4.9% (3.9-5.9) and 6.5% (4.7-8.3) at 6 months and 5 years, and 2.1 (1.8-2.4) and 1.1 (1.1-1.2) after 6 months and 5 years, respectively. As seen in Supplementary Table S2, adjustment for cardiovascular disease and chronic kidney disease caused a small decrease in overall mortality risk between both cohorts. This led to the 5-year mortality risk between cases and patients with only DM to turn nonsignificant. When stratified by sex, both the RD and RR remained significant for both sexes; however, excess mortality was more prominent in male patients, especially during the first 6 months (Table 2). No difference in 6-month and 5-year mortality was observed when stratifying for calendar time (before or after 2009) or if DM was before or after the diagnosis of GCA (Supplementary Table S5). Also, when allowing for a maximum of 365 days between diagnosis with GCA and DM for cases, the RD (6 months: 3.6% [1.8-5.3], 5 years: 12.0% [8.7-15.2]) and RR (6 months: 1.7 [1.3-2.4], 5 years: 1.4 [1.3-1.6]) remained more or less the same. In Supplementary Table S6, the mortality in patients with only GCA, only DM, and both compared with the general population can be observed. Overall, DM may contribute to the excess mortality, as patients with both GCA and DM and those with only DM had similar risks compared with their respective general populations. However, mortality was higher in patients with both GCA and DM compared with DM alone, which could indicate that the coexistence of GCA may further modify the impact of DM on mortality.
Table 2.
Risk of mortality in patients with giant cell arteritis and diabetes mellitus compared with matched controls
| Time since start of follow-up | GCA and diabetes mellitus (n = 3674) |
Matched controls (GCA) (n = 10,311) |
Matched controls (Diabetes mellitus) (n = 11,020) |
||
|---|---|---|---|---|---|
| Cumulative incidence proportion (95% CI) | Risk difference (95% CI) | Relative risk (95% CI) | Risk difference (95% CI) | Relative risk (95% CI) | |
| 6 mo | 8.1% (7.1-9.1) | 5.5% (4.5-6.5) | 2.3 (1.9-2.6) | 4.9% (3.9-5.9) | 2.1 (1.8-2.4) |
| 5 y | 34.6% (32.7-36.4) | 15.1% (13.3-17.0) | 1.5 (1.4-1.6) | 6.5% (4.7-8.3) | 1.1 (1.1-1.2) |
| Women (n = 2210) | Women (n = 6357) | Women (n = 6628) | |||
| 6 mo | 6.7% (5.5-7.9) | 4.5% (3.3-5.7) | 2.0 (1.6-2.5) | 3.9% (2.7-5.1) | 1.8 (1.5-2.2) |
| 5 y | 30.8% (28.5-33.1) | 15.1% (12.8-17.4) | 1.5 (1.4-1.6) | 6.1% (3.8-8.3) | 1.1 (1.1-1.2) |
| Men (n = 1464) | Men (n = 3954) | Men (n = 4392) | |||
| 6 mo | 9.9% (8.1-11.7) | 6.8% (5.1-8.6) | 2.5 (2.0-3.2) | 6.4% (4.7-8.1) | 2.5 (2.0-3.1) |
| 5 y | 39.7% (36.7-42.8) | 14.9% (11.9-17.9) | 1.5 (1.4-1.6) | 7.1% (4.2-10.0) | 1.2 (1.1-1.2) |
GCA, giant cell arteritis.
Age was included as a continuous variable and adjusted for using restricted cubic splines with 3 degrees of freedom.
Crude and adjusted HRs for mortality in patients with both GCA and DM, compared with patients with only GCA, are presented in Table 3. Both the crude and adjusted analyses revealed a significant HR for mortality after 6 months (Crude HR: 2.8 [2.4-3.3], P < .001, adjusted HR: 2.0 [1.9-2.2], P < .001) and 5 years (Crude HR: 2.3 [2.0-2.6], P < .001, adjusted HR: 1.8 [1.6-1.9], P < .001). In Figure 2, the crude and adjusted HR can be seen during the 5-year follow-up.
Table 3.
Crude and adjusted hazard ratio of mortality in patients with both giant cell arteritis and diabetes mellitus and giant cell arteritis only, including stratification on HbA1c levels
| Overall | Crude HR (95% CI) |
Adjusted HRa (95% CI) |
||
|---|---|---|---|---|
| 6 mo | 5 y | 6 mo | 5 y | |
| GCA, n = 10,311 | Reference | Reference | Reference | Reference |
| GCA and Diabetes mellitusb, n = 3674 | 2.8 (2.4-3.3) | 2.0 (1.9-2.2) | 2.3 (2.0-2.6) | 1.8 (1.6-1.9) |
| HbA1c category | ||||
| GCA, n = 1230 | ||||
| HbA1c < 48 | Reference | Reference | Reference | Reference |
| HbA1c > 48 | 1.5 (0.7-2.8) | 2.7 (1.9-4.1) | 1.2 (0.6-2.4) | 2.1 (1.4-3.1) |
| GCA and Diabetes mellitus, n = 1708 | ||||
| HbA1c < 48 | Reference | Reference | Reference | Reference |
| HbA1c > 48 | 1.2 (0.9-1.6) | 1.4 (1.2-1.7) | 1.1 (0.8-1.5) | 1.5 (1.3-1.8) |
| GCA and Diabetes mellitus, n = 1708 | ||||
| HbA1c < 48 | Reference | Reference | Reference | Reference |
| 48 ≤ HbA1c < 53 | 1.0 (0.7-1.5) | 1.2 (0.9-1.5) | 0.9 (0.6-1.4) | 1.2 (1.0-1.6) |
| 53 ≤ HbA1c < 58 | 1.7 (1.1-2.5) | 1.5 (1.2-2.0) | 1.5 (1.0-2.3) | 1.6 (1.3-2.1) |
| 58 ≤ HbA1c < 64 | 1.0 (0.6-1.7) | 1.3 (1.0-1.7) | 1.0 (0.6-1.7) | 1.3 (1.0-1.8) |
| 64 ≤ HbA1c < 75 | 1.0 (0.6-1.7) | 1.6 (1.2-2.1) | 0.9 (0.5-1.5) | 1.7 (1.3-2.2) |
| 75 ≤ HbA1c < 85 | 1.7 (1.0-2.8) | 2.1 (1.5-3.0) | 1.6 (1.0-2.8) | 2.3 (1.6-3.3) |
| 85 ≤ HbA1c | 0.7 (0.3-1.6) | 1.4 (0.9-2.3) | 0.8 (0.4-1.8) | 1.6 (1.0-2.5) |
HbA1c, haemoglobin-A1c; HR, hazard ratio; GCA, giant cell arteritis; N, Number.
HbA1c in mmol/mol.
Adjusted for sex, age, chronic lung disease, and social class.
Diabetes mellitus was analysed as a time-varying exposure.
Figure 2.
Crude and adjusted hazard ratio of mortality in patients with giant cell arteritis and diabetes mellitus compared with patients with only giant cell arteritis. Adjusted hazard ratio is adjusted for sex, age, chronic lung disease, and social class.
Impact of HbA1c on mortality in patients with GCA and DM, and GCA alone
In addition, crude and adjusted HR for mortality in both patients with GCA and DM, and patients with only GCA, stratified by predetermined HbA1c groups, are also shown in Table 3. When patients were grouped by high or low HbA1c levels, no significant differences were observed at 6 months, neither among those with both GCA and DM nor among those with GCA alone. At 5 years, however, both groups showed significantly increased risks. When stratifying on multiple HbA1c levels, a similar pattern emerged, which indicated no increased risk after 6 months, but after 5 years, the increased risk was most pronounced among those with the highest HbA1c levels, and remained more or less unchanged after adjustment. Although HbA1c measurements were not available for all cases, Supplementary Table S3 indicates that the 2 patient groups are largely comparable with respect to demographics, which enhances the validity of our findings.
DISCUSSION
To our knowledge, this study is among the first to investigate both the absolute and RRs of all-cause mortality in patients with both GCA and DM compared with patients with only GCA or only DM. We demonstrated that individuals with both GCA and DM had a higher risk of mortality than age- and sex-matched patients with either GCA or DM alone. The RR was highest in the initial months following diagnosis, whereas the RD increased over time, suggesting that the excess number of deaths primarily occurred during the early disease course. This pattern aligns with previous studies reporting elevated risk of mortality in the initial phase of the disease, often attributed to the disease activity and intensive treatment regimen [5,9,10,12,13]. These findings emphasise the vulnerability of patients shortly after diagnosis, especially when additional comorbidities such as DM are present.
Although the increased risk of mortality in DM compared with the general population is well established [22], this study suggests that patients with both GCA and DM may have higher mortality than those with DM alone. This finding suggests that the excess mortality observed among patients with both GCA and DM may not be entirely explained by the presence of DM, but could reflect the combined impact of both diseases. This pattern is also observed in our sensitivity analyses. First, we compared patients with DM with the general population, and subsequently, patients with both GCA and DM with the general population. Both patients with DM alone and patients with GCA and DM demonstrated higher mortality than their respective matched general population (Supplementary Table S6). However, the higher mortality observed in patients with both GCA and DM compared with those with DM alone may suggest that the coexistence of the 2 conditions could be associated with an additional increase in mortality.
We also examined whether the timing of DM in relation to GCA diagnosis influenced mortality risk. In this study, 11.5% of eligible patients with GCA had DM before GCA diagnosis and 9.5% developed DM after diagnosis of GCA. These numbers are somewhat similar with what has been presented elsewhere [15,23]. Stratified analyses revealed no significant difference in RR or RD based on whether DM was diagnosed before or after GCA (Supplementary Table S5). This may suggest that the coexistence of both conditions, rather than the order, could be an important determinant of risk. This could be explained by the clinical interplay between disease activity and treatment: in patients with GCA developing DM, hyperglycaemia is often GC induced, leading to rapid onset or worsening of undiagnosed DM. Whereas in patients with pre-existing DM, glycaemic control may initially be stable, but is often disrupted by GC treatment following GCA diagnosis. Thus, regardless of diagnostic sequence and timing, the 2 diseases typically become concurrently active, potentially explaining the similar mortality risks across subgroups.
In this study, we found that high HbA1c levels (>48 mmol/mol) were associated with an increased risk of 5-year mortality, but not 6 months mortality, in patients with both GCA and DM, as well as in those with GCA alone. However, as some patients with only GCA had registered HbA1c values above 48 mmol/mol, this could indicate that they had undiagnosed DM or significant hyperglycaemia, which may have introduced some bias into our analysis. Further stratification of HbA1c showed that HbA1c levels above 64 mmol/mol, except values above 85 mmol/mol, were linked to a higher 5-year mortality risk. This could indicate that high HbA1c primarily impacts the long-term mortality risk. This does somewhat support previous evidence linking hyperglycaemia to increased risk of mortality in patients with DM and underlines the importance of achieving good glycaemic control in patients with both GCA and DM [22]. It is important to note that multiple factors could contribute to the increased mortality observed in patients with GCA, and this cannot be solely attributed to DM. Patients with GCA experience systemic inflammation, particularly during the early phase of the disease, which likely contributes to the peak in mortality observed shortly after diagnosis. In addition, treatment involves high-dose immunosuppressive therapy, which further increases the risk of infections [24] and cardiovascular diseases [25]. Cardiovascular diseases are frequently reported as one of the leading causes of mortality in this population [11,12,16,26].
Some limitations in this study need mentioning. In this study, we defined patients with GCA solely based on registration with ICD-10 codes, and not the widely used algorithm requiring at least 3 redeemed prescriptions for GC within 6 months of a GCA diagnosis [19]. Although requiring at least 3 redeemed prescriptions enhances diagnostic specificity, it introduces immortal time bias by excluding patients who may experience an event before redeeming their third prescription. This is particularly problematic in patients with GCA, as the highest doses of GCs are typically prescribed early in the treatment course. Moreover, for a patient to be classified as having developed DM, they must have survived long enough after GCA diagnosis to receive that diagnosis. This creates 2 potential sources of immortal time bias: first, patients must survive long enough to redeem at least 3 prescriptions of GC to be included in the exposure group; second, they must remain alive until DM is diagnosed. Also, the validation study indicated that most of the misclassified patients likely have polymyalgia rheumatica [19]. If anything, this misclassification would underestimate the risk of mortality in patients with GCA and DM, as patients receive higher doses of GCs in the initial phase than patients with polymyalgia rheumatic [19]. During matching, we also restricted controls to be diagnosed with DM or GCA for no longer than 12 months before being matched. Because we also matched on year of birth, and all patients with GCA were over 50 years old, most matched DM controls were also above 50 at the time of DM onset. This likely resulted in a predominance of type 2 DM among controls, as type 2 DM is more common in older adults [27]. This could help explain the observed differences in hospital-diagnosed obesity between cases and controls. However, due to a lack of data on lifestyle factors such as smoking and exercise, this was not something we could account for in the analysis, though we did include chronic obstructive lung disease as a proxy for smoking. It is therefore plausible that the DM controls may have been less healthy overall, and the higher observed prevalence of obesity in this group could support this interpretation, which could underestimate the differences found. However, residual cofounding cannot be excluded from this analysis. Finally, as this study is based on a Danish nationwide cohort, the generalisability of the findings to other populations may be limited. This should be considered when interpretating the results of this study. This study also has some notable strengths. One key strength is the inclusion of a large, nationwide cohort and the use of Danish national registries, which enabled individual-level linkage between registers and ensured nearly complete follow-up. In addition, access to laboratory data allowed for a more comprehensive analysis of the cohort.
The presence of both GCA and DM was associated with an increased risk of mortality compared with having either condition alone. This increased risk was most pronounced in the early phase following disease onset and remained significant after adjusting for relevant covariates, which could also reflect high inflammation and GC burden. Elevated HbA1c levels appeared to be linked to an increased 5-year mortality risk in both patients with GCA and DM, and GCA only, suggesting that glycaemic control could be relevant in this high-risk population. Overall, these findings emphasise the need for clinical attention to patients with both GCA and DM simultaneously.
CRediT authorship contribution statement
Mads Engell Refstrup Sørensen: Writing – original draft, Methodology, Data curation. Kirsten Duch: Writing – review & editing, Methodology, Data curation. Line Uhrenholt: Writing – review & editing, Methodology. Lene Wohlfahrt Dreyer: Writing – review & editing, Methodology. Johan Røikjer: Writing – review & editing, Methodology. Niels Ejskjaer: Writing – review & editing, Methodology. Bolette Gylden Soussi: Writing – review & editing, Methodology. Salome Kristensen: Writing – review & editing, Supervision, Methodology.
Acknowledgments
Funding
The study was funded by Aalborg University Hospital, Denmark, and Danish National Research Foundation (grant no. DNRF148). The sponsor was not involved in the study’s design or execution; the collection, management, analysis, or interpretation of data; the preparation, review, or approval of the manuscript; or the decision to submit it for publication.
Competing interests
LD has received research grant/research support from BMS, and speaker’s bureau from Eli Lilly and Galderma (unrelated to this manuscript). LU has received travel reimbursement from AstraZeneca (unrelated to this manuscript). SK has received travel reimbursement from Johnson & Johnson (unrelated to his manuscript). All other authors declare that they have no competing interests.
Patient consent for publication
Not applicable.
Ethics approval
According to Danish legislation, no ethics approval is needed for register-based studies. The Data Protection Committee of Northern Jutland, Denmark (2020-032) has approved the study.
Provenance and peer review
Not commissioned; externally peer reviewed.
Data availability statement
Enquiries regarding data sharing will be managed in accordance with the European General Data Protection Regulation (GDPR). In compliance with regulations from the Danish Data Protection Agency and Statistics Denmark, data that could identify individuals cannot be extracted or shared.
Declaration of generative AI and AI-assisted technologies in the writing process
During the preparation of this work, the authors used ChatGPT (GPT-5 mini, OpenAI) in order to assist with refining selected sentences for clarity and readability. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.
Footnotes
Handling editor Gerd Burmester.
Supplementary material associated with this article can be found in the online version at doi:10.1016/j.ero.2026.100198.
Appendix. Supplementary materials
References
- 1.Hunder G.G., Bloch D.A., Michel B.A., Stevens M.B., Arend W.P., Calabrese L.H., et al. The American College of Rheumatology 1990 criteria for the classification of giant cell arteritis. Arthritis Rheum. 1990;33:1122–1128. doi: 10.1002/art.1780330810. [DOI] [PubMed] [Google Scholar]
- 2.Fathallah N., Slim R., Larif S., Hmouda H., Ben Salem C. Drug-induced hyperglycaemia and diabetes. Drug Saf. 2015;38:1153–1168. doi: 10.1007/s40264-015-0339-z. [DOI] [PubMed] [Google Scholar]
- 3.Salvarani C., Pipitone N., Versari A., Hunder G.G. Clinical features of polymyalgia rheumatica and giant cell arteritis. Nat Rev Rheumatol. 2012;8:509–521. doi: 10.1038/nrrheum.2012.97. [DOI] [PubMed] [Google Scholar]
- 4.Maibourg S., Addario A., Samson M., Puéchal X., Francois M., Durupt S., et al. Prevalence of giant cell arteritis relapse in patients treated with glucocorticoids: a meta-analysis. Arthritis Care Res (Hoboken). 2020;725:838–849. doi: 10.1002/acr.23901. [DOI] [PubMed] [Google Scholar]
- 5.Uddhammar A., Eriksson A-L, Nyström L., Stenling R., Rantapää-Dahlqvist S. Increased mortality due to cardiovascular disease in patients with giant cell arteritis in northern Sweden. J Rheumatol. 2002;29:737–742. [PubMed] [Google Scholar]
- 6.Gran J.T., Myklebust G., Wilsgaard T., Jacobsen B.K. Survival in polymyalgia rheumatica and temporal arteritis: a study of 398 cases and matched population controls. Rheumatology (Oxford) 2001;40:1238–1242. doi: 10.1093/rheumatology/40.11.1238. [DOI] [PubMed] [Google Scholar]
- 7.Ninan j, Nguyen A., Cole A., Rischmueller M., Dodd T., Roberts-Thomson P., et al. Mortality in patients with biopsy-proven giant cell arteritis: a South Australian population-based study. J Rheumatol. 2011;38:2215–2217. doi: 10.3899/jrheum.101254. [DOI] [PubMed] [Google Scholar]
- 8.Andersen J., Mykleburst G., Haugeberg G., Pripp A., Diamantopoulos A. Incidence trends and mortality of giant cell arteritis in Southern Norway. Arthritis Care Res (Hoboken). 2021;73:409–414. doi: 10.1002/acr.24133. [DOI] [PubMed] [Google Scholar]
- 9.Mohammad A., Nilsson J., Jacobsson L., Merkel P., Turesson C. Incidence and mortality rates of biopsy-proven giant cell arteritis in southern Sweden. Ann Rheum Dis. 2015;74:993–997. doi: 10.1136/annrheumdis-2013-204652. [DOI] [PubMed] [Google Scholar]
- 10.Li l, Neogi T., Jick S. Mortality in patients with giant cell arteritis: a cohort study in UK primary care. Arthritis Care Res (Hoboken). 2018;70:1251–1256. doi: 10.1002/acr.23538. [DOI] [PubMed] [Google Scholar]
- 11.Brekke L., Fevang B., Diamantopoulos A., Assmus J., Esperø E., Gjesdal G. Survival and death causes of patients with giant cell arteritis in Western Norway 1972-2012: a retrospective cohort study. Arthritis Res Ther. 2019;21:154. doi: 10.1186/s13075-019-1945-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Therkildsen P., Nielsen B., Thurah A., Hansen I., Nørgaard M., Hauge E. All-cause and cause-specific mortality in patients with giant cell arteritis: a nationwide population-based cohort study. Rheumatology (Oxford). 2022;61:1195–1203. doi: 10.1093/rheumatology/keab507. [DOI] [PubMed] [Google Scholar]
- 13.Baslund B., Helleberg M., Faurschou M., Obel N. Mortality in patients with giant cell arteritis. Rheumatology. 2015;54:139–143. doi: 10.1093/rheumatology/keu303. [DOI] [PubMed] [Google Scholar]
- 14.Catanose M., Macchionu P., Boiardi L., Muratore F., Restuccia G., Cavazza A., et al. Incidence, prevalence and survival of biopsy-proven giant cell arteritis in Northern Italy during a 26-year period. Arthritis Care Res (Hoboken). 2017;69:430–438. doi: 10.1002/acr.22942. [DOI] [PubMed] [Google Scholar]
- 15.Faurschou M., Ahlström M.G., Lindhardsen J., Obel N., Baslund B. Risk of Diabetes Mellitus among patients diagnosed with giant cell arteritis or granulomatosis with polyangiitis: comparison with the general population. J Rheumatol. 2017;44:78–83. doi: 10.3899/jrheum.160797. [DOI] [PubMed] [Google Scholar]
- 16.Aouba A., Chiappe S., Eb M., Delmas C., Boyssin H., Bienvenu B., et al. Mortality causes and trends associated with giant cell arteritis: analysis of the French national death certificate databse (1980-2011) Rheumatology (Oxford). 2018;57:1047–1055. doi: 10.1093/rheumatology/key028. [DOI] [PubMed] [Google Scholar]
- 17.Helmich b, Agueda A., Monti S., Buttgereit F., de Boysson H., Brouwer E., et al. 2018 update of the EULAR recommendations for the management of large vessel vasculitis. Ann Rheum Dis. 2020;79:19–30. doi: 10.1136/annrheumdis-2019-215672. [DOI] [PubMed] [Google Scholar]
- 18.Schmidt M., Pedersen L., Sørensen H.T. The Danish Civil Registration System as a tool in epidemiology. Eur J Epidemiol. 2014;29:541–549. doi: 10.1007/s10654-014-9930-3. [DOI] [PubMed] [Google Scholar]
- 19.Hjort P.E., Therkildsen P., Nielsen B.D., Hansen I.T., Nørgaard M., Thurah A., et al. Positive predictive value of the giant cell arteritis diagnosis in the Danish National Patient Registry: a Validation Study. Clin Epidemiol. 2020;12:731–736. doi: 10.2147/CLEP.S258219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Carstensen B., Kristensen J.K., Marcussen M.M., Borch-Johnsen K. The National Diabetes Register. Scand J Public Health. 2011;39:58–61. doi: 10.1177/1403494811404278. [DOI] [PubMed] [Google Scholar]
- 21.Knudsen J.S., Knudsen S.S., Hulman A., Witte D.R., Gregg E.W., Lauritzen T., et al. Changes in type 2 diabetes incidence and mortality associated with introduction of HbA1c as diagnostic option: a Danish 24-year population-based study. Lancet Reg Health Eur. 2022;14 doi: 10.1016/j.lanepe.2021.100291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rao Kondapally Seshasai S., Kaptoge S., Thompson A., Di Angelantonio E., Gao P., Sarwar N., et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med. 2011;364:829–841. doi: 10.1056/NEJMoa1008862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mahr A., Hachulla E., de Boysson H., Guerroui N., Héron E., Vinzio S., et al. Presentation and Real-World Management of Giant Cell Arteritis (Artemis Study) Front Med (Lausanne). 2021;8 doi: 10.3389/fmed.2021.732934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Therkildsen P., de Thurah A., Nielsne B.D., Faurschou M., Baslund B., Hansen I., et al. The one-year infection risk among patients diagnosed with giant cell arteritis: use of antibiotics and hospitalisations. Rheumatology (Oxford). 2025;64:740–746. doi: 10.1093/rheumatology/keae107. [DOI] [PubMed] [Google Scholar]
- 25.Amiri N., De Vera M., Choi H., Sayre E., Avina-Zubieta J. Increased risk of cardiovascular disease in giant cell arteritis: a general population-based study. Rheumatology (Oxford). 2016;55:33–40. doi: 10.1093/rheumatology/kev262. [DOI] [PubMed] [Google Scholar]
- 26.Raghavan S., Vassy J., Ho Y., Song R., Gagnon D., Cho K., et al. Diabetes mellitus–related all-cause and cardiovascular mortality in a National Cohort of Adults. J AM Heart Assoc. 2019;8 doi: 10.1161/JAHA.118.011295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bradley D., Hsueh W. Type 2 diabetes in the elderly: challenges in a unique patient population. J Geriatr Med Gerontol. 2016;2:14. doi: 10.23937/2469-5858/1510014. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Enquiries regarding data sharing will be managed in accordance with the European General Data Protection Regulation (GDPR). In compliance with regulations from the Danish Data Protection Agency and Statistics Denmark, data that could identify individuals cannot be extracted or shared.