Hip fractures in patients with primary aldosteronism – a Swedish nationwide study

Abstract

Summary

In this large population-based matched cohort study, patients with primary aldosteronism were at increased risk of hip fracture, particularly subgroups traditionally considered at higher risk of osteoporosis such as women, patients older than 56 years at diagnosis, patients with established cardiovascular disease at diagnosis, and patients treated with MRA.

Purpose

Previous studies suggest that primary aldosteronism (PA) is associated with dysregulated bone homeostasis. The aim of this study was to evaluate the incidence of hip fractures in patients with PA.

Methods

We studied a nationwide cohort of 2419 patients with PA (1997–2019) and 24 187 age and sex matched controls from the general population. Hip fractures were identified by ICD codes in the Swedish National Patient Register. We estimated hazard ratios (HRs) for incident hip fractures, adjusted for prior fractures, socioeconomic factors, diabetes, osteoporosis, hyperparathyroidism, and cardiovascular disease (CVD). Pairwise subgroup comparisons were performed by age (18–56 and > 56 years), sex, CVD at baseline, and treatment for PA.

Results

During a mean follow up of 8 ± 5 years, 64 (2.6%) patients had a hip fracture after being diagnosed with PA, compared to 401 (1.7%) controls. After adjustments, PA was associated with a 55% increased risk of hip fracture compared to controls (HR 1.55 [1.18–2.03]). HRs were increased in women (HR 1.76 [95% CI 1.24–2.52]), patients aged > 56 years (HR 1.62 [95% CI 1.21–2.17]), and patients with CVD at diagnosis (HR 2.15 [95% CI 1.37–3.37]). PA patients treated with adrenalectomy did not have higher risk than controls (HR 0.84 [95% CI 0.35–2.0]), while patients treated with mineralocorticoid receptor antagonists (MRA) retained a greater risk (HR 1.84 [95% CI 1.20–2.83]).

Conclusion

PA is associated with increased hip fracture risk, especially in women, patients diagnosed after the age of 56 years and patients with established CVD at diagnosis. Also, patients treated with MRA seem to have an increased risk of hip fractures, while adrenalectomy may be protective.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00198-024-07132-2.

Introduction

Primary aldosteronism (PA) is a common cause of hypertension with a prevalence of at least 6% in patients with high blood pressure [1]. PA is caused by renin-independent autonomous hypersecretion of aldosterone, which in one-third of the patients originates from a single adrenal (unilateral PA), and in two-thirds of patients from both adrenals (bilateral PA) [1]. Unilateral adrenalectomy is the treatment of choice in case of unilateral PA, while bilateral PA is treated with mineralocorticoid receptor antagonist (MRA).

PA is associated with impaired quality of life, 2–fourfold increased risk for cardiovascular morbidity, impaired renal function, and increased mortality, compared to controls [2–5]. The influence of PA on calcium and bone homeostasis is less well known. Nevertheless, patients with PA have higher urinary calcium excretion, higher concentrations of parathyroid hormone and lower calcium concentrations in serum, compared to controls with and without hypertension [6–9]. Furthermore, three studies have shown that the prevalence of radiographic vertebral fractures is higher in patients with PA (29–46%) compared to controls (12–23%) [8, 10, 11]. The number of patients in these studies (n = 11, 56 and 152, respectively) was however not sufficient to evaluate the prevalence of clinical fractures such as hip fractures.

One previous study has investigated the occurrence of clinical fractures in patients with PA. The study, a population-based cohort study including 2533 patients from Taiwan, found an increased risk for any fracture (HR; 1.8, 95% CI; 1.5–2.2) and for osteoporotic fractures (femoral neck, pelvic and vertebral; HR; 1.8, 95% CI; 1.4–2.4), in comparison to patients with essential hypertension [12]. The specific risk for hip fractures was not provided. Thus, the aim of this study was to determine the incidence of hip fractures in a large and unselected cohort of patients diagnosed with PA in Sweden. Additionally, we aimed to explore the influence of age, sex, other comorbidities at diagnosis, and treatment for PA, on the risk for hip fractures.

Methods

Study design and data source

This was a nationwide, observational, matched, case-cohort study based on data from four Swedish nationwide registers and databases, all maintained by the Swedish National Board of Health and Welfare, and Statistics Sweden: (a) the National Patient Registry (NPR) [13], (b) the Total Population Register [14] (c) the Swedish Prescribed Drug Register [13, 15] and (d) the Longitudinal Database of Health Insurance and Labour Market Studies [14]. A personal identification number, which is unique for all citizens of Sweden, was used to cross-link data from these registers [16]. Registration of data by the relevant authorities and/or health care providers is mandated by law. This report follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

Identification of the study population

Patients with PA were identified by a search for individuals registered with the diagnostic code for PA (E26.0) between 1997 and 2019 in the NPR. The NPR is a population-based Swedish register that collects diagnostic codes from all inpatient hospital stays since 1987, and from all specialist outpatient visits since 2001. All diagnoses that patients receive at discharge are registered according to the International Classification of Diseases, with the 9th revision (ICD-9) between 1987 and 1996, and the 10th revision (ICD-10) since 1997. All surgical procedures, categorized according to the Nordic Classification of Surgical Procedures, are also registered in the NPR. The date of diagnosis (index date) was defined as the date when a patient was first assigned a diagnostic code for PA. Patients < 18 years at the time of diagnosis, as well as patients diagnosed with PA before 1996 (ICD-9 code 255.1), were excluded.

The Total Population Register in Sweden was used to identify 10 randomly selected controls for each patient, matched by age (± 1 year), sex, and county of residence [14]. The controls were assigned the same index date as their matched case. All subjects were followed from the index date until the date of hip fractur, death, emigration, or the end of the study (31 December 2019).

Information on treatment for PA

The NPR was used to collect information on surgical treatment for PA, i.e. adrenal resection, or adrenalectomy, that is registered according to the Nordic Classification of Surgical Procedures (BCA20, adrenal resection; BCA30, unilateral adrenalectomy; BCA31, laparoscopic unilateral adrenalectomy).

The use of MRA was assessed by using data from the Swedish Prescribed Drug Register [15] which contains information on all medicines that have been dispensed in Sweden since July 2005, and are categorised according to the Anatomical Therapeutic Chemical (ATC) Classification System. Information on the quantity of MRA dispensed for every patient was collected and presented as a mean defined daily dose (DDD) units according to the World Health Organization [spironolactone (ATC code C03DA01; 1 DDD = 75 mg) and eplerenone (ATC code C03DA04; 1 DDD = 50 mg)] [17]. We defined chronic MRA treatment as ≥ 2 dispensed prescriptions per year. In this study, low dose MRA treatment was defined as doses lower than the median DDD in the study population, and high dose MRA treatment as doses higher than the median DDD. Information about medical treatment before July 2005 when the Swedish Prescribed Drug Register was established was not available.

Information on comorbidities and socioeconomic factors

Information on the following comorbidities were collected from the NPR for patients and controls: (a) ischemic heart disease (I20–I25), (b) myocardial infarction (I21), (c) atrial fibrillation or flutter (I48), (d) heart failure (I50), and (e) stroke (I61–I64). To capture patients with diabetes we used data from both the NPR (E10–E14) and the Swedish Prescribed Drug Register on dispensed drugs used for diabetes (ATC code A10). Osteoporosis (M80 – M82), and hyperparathyroidism (E21), as risk factors for hip fracture, was based on ICD-10 codes assigned before the date of diagnosis.

The Longitudinal Database of Health Insurance and Labour Market Studies [14] was used to collect data on socioeconomic factors, i.e. (a) country of birth, (b) cohabiting/marital status, (c) income, and (d) educational level.

Fracture data

Information on hip fractures was obtained from the National Patient Register by a search for individuals registered with the diagnostic code for femoral neck fracture (S72.0), intertrochanteric fracture (S72.1), and subtrochanteric fracture (S72.2). Prior fracture was defined as any ever-recorded ICD-10 fracture diagnosis code before the index date (S02, S12, S22, S32, S42, S52, S62, S72, S82, S92).

Ethics

The study was conducted according to the Declaration of Helsinki and was approved by the Swedish Ethical Review Authority on 2017–02-20 (Dnr 029–17), and amendments on 2019–03-04 and 2020–06-04. Data generated for the study was based on pseudonymised individual data in accordance with the Swedish Public Access to Information and Secrecy Act (OSL, 2009:400). Thus, the requirement for informed consent was waived. The data used are available to researchers upon request to the Swedish National Board of Health and Welfare, and after ethical approval by the Swedish Ethical Review Authority.

Statistical analysis

We used SAS 9.4, SPSS, and R (version 4.2.2) for the statistical analyses. Normally distributed continuous variables are presented as mean ± standard deviation (SD), and non-normally distributed continuous variables are presented as median (interquartile range [IQR]). Categorical variables are presented as number (%).

Multivariate Cox regression models were used to compare hip fractures in patients with PA and matched controls. In the models, we calculated hazard ratio (HR) and 95% confidence interval (CI), after adjustments for sex, age at diagnosis of PA, socioeconomic factors at study entry (level of education, marital status, income, and country of birth), the presence of cardiovascular comorbidities (ischemic heart disease, atrial fibrillation or flutter, congestive heart failure, and stroke), diabetes mellitus, osteoporosis, hyperparathyroidism and a prior fracture at the time of diagnosis. The proportionality assumptions were tested with graphical methods. We conducted a stratified analysis by sex and median-split age groups.

Furthermore, to study the influence of treatment, we calculated HR separately for patients treated with adrenalectomy, patients treated with chronic MRA, and patients who had not received treatment with adrenalectomy or MRA. All reported HRs are based on pairwise case–control sets, i.e., we compared each treatment group with their matched controls. The association between MRA exposure and fractures was limited to patients diagnosed with PA since 2006 as the Swedish Prescribed Drug Register was established in July 2005.

Sensitivity analysis

We have previously shown that more than 94% of patients who receive the diagnostic code for PA in the National Patient Register on two or more occasions have PA [18]. We performed, therefore, a separate analysis where we restricted the cohort to patients fulfilling this criterion.

Results

Between 1997 and 2019, 2895 patients received the diagnostic code for PA. Of these patients, 476 were excluded due to following reasons: 465 patients had been diagnosed before 1997, eight patients were < 18 years at diagnosis, and three patients had an invalid personal identification number. Thus, 2419 patients [1344 (56%) men] and 24 187 controls matched for sex, age, and county of residence were included in the analysis (Table 1). The mean age at diagnosis was 56 ± 13 years and the median follow-up time 8.2 (3.9–13.5) years. Baseline characteristics as well as a summary of comorbidities are shown in Table 1.

Table 1.

Baseline characteristics of patients with primary aldosteronism (PA) and controls from the general population matched for age, sex, and county of residence

	Controls	Patients with PA	Patients with PA treated with adrenalectomy	Patients with PA not treated with adrenalectomy
N	24 187	2419	686	1733
Women	10 749 (44.4)	1075 (44.4)	316 (46.1)	759 (43.8)
Age at diagnosis, years	55.8 (12.8)	55.8 (12.8)	50.4 (11.2)	58.0 (12.7)
Year of diagnosis
1997–2007	7888	789	256	533
2008–2019	16 299	1630	430	1200
Socioeconomic factors#
Income, × 100SEK/year	2021 (1368–2938)	1924 (1334–2771)	2088 (1467–2971)	1843 (1298–2699)
Born in Sweden	19 935 (82.4)	1831 (75.7)	551 (80.3)	1280 (73.9)
Married/cohabiting	15 686 (64.9)	1627 (67.3)	475 (69.2)	1152 (66.5)
Level of education
≤ 9 years	5588 (23.1)	541 (22.4)	134 (19.5)	407 (23.5)
10–12 years	7076 (29.3)	691 (28.6)	200 (29.2)	491 (28.3)
≥ 13 years	11 193 (46.3)	1157 (47.8)	344 (50.1)	813 (46.9)
Comorbidities
Diabetes mellitus	952 (3.9)	352 (14.6)	76 (11.1)	276 (15.9)
Stroke	403 (1.7)	134 (5.5)	28 (4.1)	106 (6.1)
Ischaemic heart disease	1016 (4.2)	225 (9.3)	49 (7.1)	176 (10.2)
Myocardial infarction	453 (1.9)	79 (3.3)	11 (1.6)	68 (3.9)
Heart failure	330 (1.4)	104 (4.3)	15 (2.2)	89 (5.1)
Atrial fibrillation/flutter	621 (2.6)	204 (8.4)	34 (5.0)	170 (9.8)
Follow-up time (mean, SD)	8.2 (5.8)	8.1 (5.8)	9.3 (5.8)	7.6 (5.8)
Diagnostic code for PA on ≥ 2 occasions		2288 (94.6)	674 (98.3)	1614 (93.1)

Continuous variables are presented as mean ± standard deviation or median (interquartile range); categorical variables are presented as number (%)

^# Information from National Registers

Hip fracture

Eighty (3.3%) out of 2419 patients with PA, and 560 (2.3%) out of 24 187 controls had a hip fracture (p = 0.002). At baseline, 16 (0.7%) patients and 159 (0.7%) controls had already had a hip fracture (p = 1.0). After PA was diagnosed, 64 (2.6%) patients and 401 (1.7%) controls suffered hip fracture (p = 0.0004). In a regression analysis with adjustments for age, sex, socioeconomic factors, prior fractures, osteoporosis, hyperparathyroidism, diabetes mellitus and cardiovascular disease, patients with PA had a higher risk of hip fracture compared to controls (HR 1.55 [95% CI 1.18–2.03]; p = 0.002; Fig. 1).

Fig. 1.

The total accumulated risk of experiencing hip fracture in patients with primary aldosteronism (PA; n = 2419) and their corresponding controls (n = 24,187) matched by sex, age at diagnosis and county of residence. Footnote: Hazard ratios (HR) and 95% confidence intervals (CI) were calculated by using Cox regression models adjusted for socioeconomic factors at study entry (level of education, marital status, income, and country of birth), prior fractures, osteoporosis, hyperparathyroidism, diabetes mellitus and cardiovascular disease

Influence of sex, age, and comorbidities on hip fracture

During follow-up, 38 hip fractures were recorded in women (3.5%) and in 26 (1.9%) men with PA (p = 0.015). The risk for hip fracture was increased in women compared to their controls (adjusted HR 1.76 [95% CI 1.24–2.52]; p = 0.002) but not in men (adjusted HR 1.38 [95% CI 0.90–2.11]; p = 0.14; Table 2).

Table 2.

The risk of hip fracture in patients with primary aldosteronism (PA) and controls, overall risk, and risk according to age (18–56 and ≥ 56 years), sex, established cardiovascular disease (CVD) at diagnosis and treatment

Cohort		Subjects at risk N	Hip fractures N (%)	Age and sex adjusted HR (95% CI)	Pa	Multivariable adjusted HRb (95% CI)	P
Overall	PA	2419	64 (2.6)	1.74 (1.34–2.28)	< 0.001	1.55 (1.18–2.03)	0.002
	Controls	24 187	401 (1.7)	1 (ref.)		1 (ref.)
Age at diagnosis
< 56 years	PA	1253	8 (0.6)	1.37 (0.66–2.87)	0.403	1.17 (0.54–2.55)	0.685
	Controls	12 529	58 (0.5)	1 (ref.)		1 (ref.)
≥ 56 years	PA	1166	56 (4.8)	1.82 (1.37–2.41)	< 0.001	1.62 (1.21–2.17)	0.001
	Controls	11 658	343 (2.9)	1 (ref.)		1 (ref.)
Sex
Female	PA	1075	38 (3.5)	1.98 (1.40–2.79)	0.001	1.76 (1.24–2.52)	0.002
	Controls	10 749	220 (2.0)	1 (ref.)		1 (ref.)
Male	PA	1344	26 (1.9)	1.50 (0.99–2.26)	0.06	1.38 (0.90–2.11)	0.135
	Controls	13 438	181 (1.3)	1 (ref.)		1 (ref.)
History of CVDc	PA	519	25 (4.8)	2.39 (1.55–3.67)	< 0.001	2.15 (1.37–3.37)	< 0.001
	Controls	5188	127 (2.4)	1 (ref.)		1 (ref.)
No History of CVD	PA	1 900	39 (2.1)	1.48 (1.06–2.07)	0.023	1.37 (0.97–1.93)	0.077
	Controls	18 999	274 (1.4)	1 (ref.)		1 (ref.)
Treatment
Adrenalectomy	PA	686	6 (0.9)	0.96 (0.42–2.23)	0.930	0.84 (0.35–2.00)	0.690
	Controls	6860	62 (0.9)	1 (ref.)		1 (ref.)
MRAd	PA	1152	27 (2.3)	2.20 (1.46–3.33)	< 0.001	1.84 (1.20–2.83)	0.005
	Controls	11 518	126 (1.1)	1 (ref.)		1 (ref.)
MRA high dosee	PA	575	10 (1.7)	2.28 (1.15–4.51)	0.018	2.0 (0.99–4.03)	0.054
	Controls	5749	47 (0.8)	1 (ref.)		1 (ref.)
MRA low dosee	PA	577	17 (2.9)	2.14 (1.26–3.61)	0.005	1.74 (1.01–3.01)	0.047
	Controls	5769	79 (1.4)	1 (ref.)		1 (ref.)
No treatmentf	PA	215	4 (1.9)	2.36 (0.81–6.83)	0.115	1.41 (0.41–4.88)	0.592
	Controls	2150	24 (1.1)	1 (ref.)		1 (ref.)

CI confidence interval, CVD cardiovascular disease, HR hazard ratio, MRA mineralocorticoid receptor antagonist

^a p values were calculated using the cox proportional hazard model

^b Multivariate analyses were adjusted for age, sex, education, country of birth, income, marital status, osteoporosis diagnosis (M80–M82), previous history of fracture (S02, S12, S22, S32, S42, S52, S62, S72, S82, S92), primary hyperparathyroidism (E21), cardiovascular comorbidity, and diabetes mellitus (E10–E14) at baseline

^c CVD was defined as any record of ischemic heart disease (I20–I25), myocardial infarction (I21), atrial fibrillation or flutter (I48), (heart failure (I50) or stroke (I61–I64) before start of follow up

^d Medically treated patients with at least two dispensed prescriptions per year of eplerenone or spironolactone. Data on medical treatment have been available since 2006

^e Low and high MRA use was defined as lower and higher than the median defined daily dose (DDD) of MRA dispensed per year, respectively

^f Non-surgically treated patients, and not treated with MRA

The risk of hip fracture was greater in patients > 56 years compared to their controls (adjusted HR 1.62 [95% CI 1.21–2.17]; p = 0.001). In patients aged 18–56 years, the risk was not increased (adjusted HR 1.17 [95% CI 0.54–2.55]; p = 0.7). In women > 56 years the adjusted HR was 1.69 [95% CI 1.14–2.50]; p = 0.009; (Fig. 2).

Fig. 2.

The accumulated risk of experiencing hip fracture in women > 56 years with primary aldosteronism (PA; n = 466) and their corresponding controls (n = 4,660) matched by sex, age at diagnosis and county of residence. Footnote: Hazard ratios (HR) and 95% confidence intervals (CI) were calculated by using Cox regression models adjusted for socioeconomic factors at study entry (level of education, marital status, income, and country of birth), prior fractures, osteoporosis, hyperparathyroidism, diabetes mellitus and cardiovascular disease

Patients with established cardiovascular disease (ischemic heart disease, atrial fibrillation or flutter, congestive heart failure, and/or stroke) when PA was diagnosed had a twofold increased risk of hip fracture compared to their controls (adjusted HR 2.15 [95% CI 1.37–3.37]; p < 0.001). The risk of hip fractures in patients without cardiovascular comorbidities was not significantly increased (adjusted HR 1.37 [95% CI 0.97–1.93]; p = 0.08).

Influence of treatment for PA on hip fracture

During the study period, 686 patients (370 men and 316 women) were treated with adrenalectomy. These patients were younger compared to non-surgically treated patients. Six (0.9%) patients treated with adrenalectomy had a hip fracture after the diagnosis of PA, and 62 (0.9%) of their controls (n = 6860). Thus, the risk for hip fracture in patients treated with adrenalectomy did not differ from controls (adjusted HR 0.84 [95% CI 0.35–2.00]).

Among 1360 non-surgically treated patients diagnosed with PA since 2006, 1152 (85%) were receiving chronic treatment with MRA. Of these 1152 patients, 643 (56%) received spironolactone, 232 (20%) eplerenone, and 277 (24%) had received both spironolactone and eplerenone. The median (interquartile range) DDD of MRA was 0.6 (0.3–1.1) units, 0.5 (0.3–0.8) DDD in women and 0.7 (0.3–1.2) DDD in men.

Twenty-seven (2.3%) MRA users had hip fracture after being diagnosed with PA, compared to 126 of 11,518 (1.1%) controls. Of the 27 MRA users, 18 were on spironolactone (3% of spironolactone users), one on eplerenone (0.4% of eplerenone users), and 8 (3%) had received both spironolactone and eplerenone. Of 27 MRA users with PA and hip fracture, 10 were treated with high dose MRA and 17 with a low dose.

MRA treatment was associated with an increased risk of hip fracture (adjusted HR 1.84 [95% CI 1.20–2.83]; p = 0.005). When the analysis was stratified by the median dispensed MRA dose, differences in the hip fracture risk between MRA users and controls reached significance in patients treated with a low dose MRA (adjusted HR 1.74 [95% CI 1.01–3.01]; p = 0.047), but not patients treated with a high dose (adjusted HR 2.0 [95% CI 0.99–4.03]; p = 0.054).

Sensitivity analysis

A sensitivity analysis restricted to patients who had received a diagnostic code for PA on two or more occasions did not modify the results (Supplementary Table 1).

Discussion

In this longitudinal nationwide study on 2419 patients with PA, we found higher hip fracture risk in patients compared to age and sex matched controls from the general population. The risk was especially high in women with PA, patients older than 56 years, patients with cardiovascular comorbidity at the time of diagnosis, and in patients treated with MRA.

The available data on the prevalence of clinical fractures in PA is scarce. Three previous studies have shown higher prevalence of radiographic vertebral fractures in patients with PA compared to controls [8, 10, 11], and one previous study has shown an increased risk of osteoporotic fractures, either hip or vertebral, compared with hypertensive controls [12]. Radiographic vertebral fractures are mainly associated with impaired bone quality, but radiological findings can be difficult to relate to clinical symptoms. In contrast, the current study reports on hip fractures, a robust outcome, which is associated with high economic costs, impaired quality of life and reduced life expectancy [19, 20]. Of equal importance, as almost all patients with hip fractures receive inpatient orthopaedic specialist care, the coverage in the NPR is excellent [21].

The reason for the increased fracture risk in patients with PA is still not fully understood. The mineralocorticoid receptor, the target receptor for aldosterone, is expressed in both osteoblasts and osteocytes [22]. It is therefore plausible that aldosterone excess has a direct effect on bone metabolism. In mice with glucocorticoid induced osteoporosis, blockage of the mineralocorticoid receptor with eplerenone stimulates bone formation and increases bone mass [22]. Furthermore, it has previously been shown that patients with PA have higher concentrations of PTH compared to controls [23]. Patients in our cohort had more often been diagnosed with primary hyperparathyroidism compared to the controls, i.e., a disease caused by overproduction of parathyroid hormone (PTH). PTH increases bone resorption and patients with primary hyperparathyroidism are at increased risk for fractures, irrespective of bone mineral density [24]. Thus, it is possible that disrupted PTH secretion may also contribute to the increased fracture risk in patients with PA. The association between PA and hip fractures may also be indirect. We observed that older patients and patients with cardiovascular comorbidity had a greater risk of hip fracture, regardless of prior fractures, diabetes, and socioeconomic factors. Thus, our results suggest that frailty at the time of the diagnosis may play an important role for the higher prevalence of hip fractures among patients with PA.

Women with PA, but not men, had an increased risk of hip fracture in our study. It is well known that postmenopausal women, in general, have an increased risk of fractures due to bone loss [25]. Moreover, after the age of 50, the risk for hip fracture in women increases gradually, while in men the risk remains the same despite increasing age [26]. Thus, our results are in line with this evidence, and indicate that postmenopausal women with PA are especially vulnerable for developing osteoporotic fractures. Interestingly, among MRA users in our cohort, the proportion of women receiving treatment with low doses was high (65%), indicating undertreatment which may have contributed to worse bone health compared to men.

Few studies have addressed the effect of treatment for PA on calcium homeostasis [6, 10, 12, 27]. Both adrenalectomy and spironolactone seem to increase calcium in plasma and bone mineral density, whereas they decrease PTH concentrations and urinary calcium excretion, [6, 10, 27]. The effect of treatment of PA on fractures is less well known. The fracture risk in one previous large study was not increased in patients treated with adrenalectomy [12], results that agree with our study. However, the effect of treatment with MRA on hip fractures has not been studied before. We found that patients treated with MRA had increased hip fracture risk as compared to their controls. Despite a slight statistical significance for low dose users, HRs were numerically similar between low and high dose users. The reason for worse outcome in MRA users is not clear and needs to be studied further. However, from a mechanistical point of view, MRA counteracts the aldosterone effect at the mineralocorticoid receptor, while adrenalectomy effectively reduces aldosterone excess. In this context, a previous nationwide Swedish study showed a complete biochemical cure rate of 92% following adrenalectomy [28]. If there is indeed a qualitative difference between adrenalectomy and MRA, or if the difference is simply due undertreatment with MRA, especially in women, remains to be explained.

Important strengths of this study are the large number of patients, the long follow-up time, the quality of the data used, and the pairwise comparisons with age and sex matched controls from the general population. Moreover, we consider the cohort to be highly representative for patients with PA in Sweden, as our previous study demonstrated high coverage (97%) of patients with PA in the NPR, and a high rate of accurate diagnosis [18]. Similarly, we restricted our analysis to hip fractures, which is highly robust outcome with an excellent coverage and high positive predictive value in the NPR [21]. Nevertheless, the study has also limitations, mainly related to the register-based design, especially the lack of important clinical data such as information on blood pressure, renin and potassium concentrations, BMI, degree of physical activity and tobacco use.

In conclusion, this large nationwide study demonstrated an increased risk of hip fracture in patients with PA, particularly in women, patients diagnosed after the age of 56 years and patients with established CVD at diagnosis. Patients treated with adrenalectomy did not have higher risk compared to controls. In contrast, medically treated patients had increased risk, possibly due to undertreatment with MRA, given the generally low doses used in the cohort.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Concept and design: EG, AR, GJ, OR. Data requisition: OR. Data validation: EG. Acquisition of the financial support: EG, OR, AR. Statistical analysis: TZS. Interpretation of data: All authors. Drafting of the manuscript: EG, OR. Critical revision of the manuscript for important intellectual content and approval of final version: All authors. All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. OR is responsible for the overall content as the guarantor of this work. The corresponding author attests that all listed authors meet authorship criteria.

Funding

Open access funding provided by University of Gothenburg. The study was financed by grants from the Swedish state under the agreement between the Swedish government and the county councils, the ALF-agreement (Grant Number: 971027), the Gothenburg Society of Medicine, the Swedish Heart and Lung Foundation [2021–0345], and the Swedish Research Council [2019–00193]. The funders of the study had no role in study design. The interpretations of the data which have been issued are those of the authors and not of the funders.

Declarations

Conflicts of interest

GJ has served as a consultant for Astra Zeneca and Novo Nordisk and has received lecture fees from Novo Nordisk, Pfizer and Takeda/Shire. EG, TZS, AR, PT, AM, and OR declare that they have no conflict of interest.