Cardiovascular and renal complications to postsurgical hypoparathyroidism: A Danish nationwide controlled historic follow-up study

Authors

  • Line Underbjerg,

    Corresponding author
    1. Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Tage Hansens Gade (THG), Aarhus, Denmark
    • Address correspondence to: Line Underbjerg, Aarhus University Hospital, Dept. of Endocrinology and Internal Medicine, THG, Tage-Hansens Gade 2, DK-8000 Aarhus C, Denmark. E-mail: linund@rm.dk

    Search for more papers by this author
  • Tanja Sikjaer,

    1. Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Tage Hansens Gade (THG), Aarhus, Denmark
    Search for more papers by this author
  • Leif Mosekilde,

    1. Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Tage Hansens Gade (THG), Aarhus, Denmark
    Search for more papers by this author
  • Lars Rejnmark

    1. Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Tage Hansens Gade (THG), Aarhus, Denmark
    Search for more papers by this author

ABSTRACT

We aimed to identify all patients with postsurgical hypoparathyroidism (HypoPT) and to evaluate their risks of renal complications and cardiovascular disease in relation to their disease and its treatment. We identified possible patients through the Danish National Patient Registry and a prescription database. Case status was adjudicated by review of individual patients' hospital records. For each patient with postsurgical HypoPT due to surgery for nonmalignant diseases between 1988 and 2012, three age-matched (± 2 years) and gender-matched controls were selected from the general background population. The prevalence of postsurgical HypoPT was 22 per 100,000 inhabitants. We identified 688 patients who had undergone neck surgery since 1988 with subsequent hypocalcaemia and inappropriate low parathyroid hormone (PTH) levels that necessitated treatment with calcium and/or vitamin D supplementation for more than 6 months. The average age at diagnosis was 49 years (range, 17–87 years), and 88% were women. Sixteen percent of all patients had had neck surgery prior to the operation causing HypoPT. Compared with controls, patients with HypoPT had an increased risk of renal complications (hazard ratio [HR], 3.67; 95% confidence interval [CI], 2.41–5.59) and hospitalization due to seizures (HR, 3.82; 95% CI, 2.15–6.79), whereas there was no increased risk of cardiac arrhythmias (HR, 1.11; 95% CI, 0.79–1.57) or cardiovascular disease or death (HR, 0.89; 95% CI, 0.73–1.09). In conclusion, although risk of seizures and renal complications is increased, mortality and risk of cardiovascular diseases or arrhythmias is not increased in patients with HypoPT. Further study should try to determine how to reduce the risk of seizures and renal complications in HypoPT. © 2013 American Society for Bone and Mineral Research.

Introduction

Hypoparathyroidism (HypoPT) is a relatively rare disease characterized by hypocalcemia with inappropriately low plasma levels of parathyroid hormone (PTH).[1] The most common cause is accidental removal of or damage to the parathyroid glands during neck surgery for goiter (toxic or nontoxic), cancer, or hyperparathyroidism.[2] The risk of HypoPT following different surgical neck procedures has been assessed in a number of studies.[3-5] In general, the risk is highest if bilateral neck exploration is performed during parathyroidectomy, if biopsies are obtained from the parathyroid glands, if radical thyroid surgery (total thyroidectomy) is performed, or if the patients previously have been operated on for thyroid or parathyroid diseases. In addition, the experience of the surgeon performing the operation is of importance to the outcome. Following total thyroidectomy as many as 2% to 5% of the patients may need subsequent medical treatment for HypoPT.[3-5] However, despite several studies on risk of HypoPT as a complication to specific surgical procedures, only a few data are available on the prevalence of HypoPT within a well-defined population. In a preliminary report from the Mayo Clinic (Rochester, New York, USA), a total of 54 patients (71% women) with HypoPT were identified, among whom 78% were classified as postsurgical, suggesting a prevalence of postsurgical HypoPT of approximately 29 in 100,000.[6]

Standard treatment for HypoPT consists of oral calcium and vitamin D analog supplements. In most cases, calcium supplements with vitamin D (ergocalciferol or cholecalciferol) are combined with active vitamin D analogs in terms of alfacalcidol or calcitriol.[7, 8] Although meta-analyses have shown a decreased mortality in elderly men and women treated with calcium and vitamin D,[9, 10] recent studies have suggested an increased risk of cardiovascular diseases (CVDs) and renal complications, including nephrocalcinosis, nephrolithiasis, and renal insufficiency, in users of calcium supplements.[11, 12] If such risks also apply to patients with HypoPT, they may be at a noticeable risk, because high oral doses of calcium often are used in these patients. Risk of renal stones may be further increased in HypoPT due to the lack of a stimulatory effect of PTH on the reabsorption of calcium in the renal tubules.[1] Moreover, phosphate levels are often increased, causing an increased calcium-phosphate product, which may enhance the risk of extraskeletal calcifications.[13] In a number of case reports, extraskeletal calcifications and renal complications have been reported in individual patients or small groups of patients, but only few data have so far been reported on risk of different disease outcomes in larger groups of patients. In a recent clinical workup on 120 patients with HypoPT (66% postsurgical), Mitchell and colleagues[14] found a twofold to 17-fold increased risk of chronic kidney disease stage 3 with renal calcifications being present in 17 of the 54 (31%) patients who had a renal imaging study performed. Similarly, in cross-sectional studies including a variable number of patients with HypoPT due to different etiologies, nephrolithiasis has been reported to be present in 8% to 41% of patients supporting the notion of an increased risk in patients with HypoPT.[15-17]

In order to provide further knowledge on the occurrence of HypoPT and potential complications to the disease and its treatment, we aimed to identify all patients diagnosed with postsurgical HypoPT in Denmark. Within the group of patients who had acquired HypoPT as a result of neck surgery for nonmalignant causes within the last 24 years (cases), we assessed mortality and incident CVD and renal diseases. By comparing risks in patients with HypoPT with a group of age-matched and gender-matched population-based controls, we determined possible renal and cardiovascular adverse health outcomes in patients with postsurgical HypoPT.

Patients and Methods

In Denmark, the National Health Service provides tax-supported healthcare for all inhabitants, allowing free access to general practitioners and hospitals. All Danish citizens are assigned a personal Civil Registration Number at the time of birth; since 1968, the Danish Civil Registration System has kept electronic records on all changes in vital status, including change of address, date of emigration, and date of death for the entire Danish population. Moreover, all contacts (on an inpatient or outpatient basis) to hospitals are filed in the National Hospital Patient Registry (NPR), which is a database run by the National Danish Board of Health. The register was founded in 1977, and outpatient records were completely incorporated from 1995. The files of the register include information on the civil registry number of the patient, date of discharge, and discharge diagnoses, assigned exclusively by the physician at discharge according to the Danish version of the International Classification of Diseases, 8th revision (ICD-8) until the end of 1993, and after January 1, 1994, according to the Danish version of the International Classification of Diseases, 10th revision (ICD-10). The register has nationwide coverage of public and private hospitals with an almost 100% completeness of recordings and a high precision of diagnoses.[18] Using these databases, we identified patients with postsurgical HypoPT and assessed mortality as well as renal and cardiovascular complications to the disease. The study was approved by the Central Denmark Region Committees on Health Research Ethics (M-20110074), the Danish Data Protection Agency (2011–41–5955), and the Central Denmark Region (1–16–02–101–11).

Identification of patients with postsurgical HypoPT (cases)

In July 2011, we retrieved information from the NPR on hospital discharge codes covering a diagnosis of postsurgical HypoPT or a diagnosis compatible with possible postsurgical HypoPT (ICD 8: 193.99, 194.19, 239.19, 252, 252.10, 252.11, 252.18, and 252.19; ICD 10: C73.9, C75.0, D44.0, D44.2, D44.8, E20.8, E20.9, E58, E89.2a, E89.2b, E89.8, and E89.9) assigned after January 1, 1988.

In Denmark, reimbursement by the state is provided for a variable proportion of the costs of prescribed drugs. Patients who had redeemed a prescription for a drug are recorded in a regional prescription database. In order to improve the completeness of identification of all patients with HypoPT in Denmark, we also retrieved data on patients who had filed at least two prescriptions for an active vitamin D analog (ATC group A11CC), because these drugs are used almost exclusively by patients with HypoPT, chronic kidney diseases, or osteoporosis. Due to technical difficulties, however, data on drug prescriptions were only available from approximately two-thirds of the Danish population.

For all patients identified by either of the two methods, case status was adjudicated by review of the individual hospital patient charts with extraction of information on the disease necessitating surgery causing HypoPT and date of surgery. Patients with chronic renal failure, defined as plasma creatinine above 150 µmol/L, equal to an estimated glomerular filtration rate (eGFR) above 30 mL/min (corresponding to stage 3 chronic kidney disease) prior to parathyroid surgery were excluded from the present study.

We defined chronic postsurgical HypoPT as hypocalcaemia (plasma calcium below the lowest reference level) with inappropriately low PTH levels following neck surgery that necessitated treatment with calcium and/or vitamin D analog supplementation for more than 6 months. In cases of doubt, consensus was reached among two to three independent investigators. In addition, we collected data on coexistence of postsurgical hypothyroidism, and treatment with calcium and magnesium supplements, vitamin D analogs, and thyroid hormones.

Analytical approach

In order to assess mortality and complications of HypoPT in terms of renal impairment and CVD, we identified a group of controls using the Danish Civil Registration System. For each patient with HypoPT (case), we randomly selected three controls, matched by gender and year of birth (± 2 years). Controls were selected using the incidence-density sampling technique[19]; ie, the controls had to be alive at the time the corresponding case had surgery causing HypoPT (index date). Each control subject was assigned an index date (start of follow-up) identical to the date of surgery causing HypoPT in the case subject. A renal or cardiovascular disease causing hospitalization following the index date was considered as an incident episode in patients and controls, whereas participants assigned a diagnosis prior to the index date were considered as having a history of the disease.

For the entire group of patients with HypoPT and matched controls, we retrieved data on May 10, 2012, on vital status from the Danish Civil Registration System and discharge codes from the NPR covering hospitalizations due to CVDs or renal diseases.

An acute myocardial infarction (MI) was considered in the participant's assigned ICD-10 codes I21 to I23 or sudden cardiac arrest (I46) and the corresponding ICD-8 codes (410.09, 410.99, and 427.27). An ischemic heart disease (IHD) was considered in participants with an acute MI (as defined) or hospitalization due to IHD or congestive heart failure (ICD-10 codes I11, I20, I24, I25, or I50; ICD-8 codes 411.00–414.99 and 427.09–427.19, 427.27, 428.99, 429.08, 429,09 or 782.49). Strokes (nontraumatic cerebrovascular accidents) were considered in terms of cerebral infarctions (ICD-10 code I63, I69, or G46; ICD-8 codes 432.00–434.99), intracerebral hemorrhage including nontraumatic subarachnoid hemorrhage (ICD-10 codes I61–I62; ICD-8 codes 430.00–430.09 or 431.00–431.99), or a stroke for which it was not specified whether it occurred due to ischemia or hemorrhage (ICD-8 436.01, 436.90, or 437.00–438.99; ICD-10 I64).

Arrhythmia was considered if there was an ICD-8 code of 427.20 to 427.99 or ICD-10 codes I44 to I44.7 (atrioventricular block), I45 to I45.9 (other types of block), I47 to I47.9 (tachycardia), I48 to I48.9 (atrial heart flutter), or I49 to I49.9 (other types of heart rhythm disturbances). ICD-8 code 42.727 and the ICD-10 codes I46 to I46.1 were used to identify arrhythmia-induced deaths.

In the following, the term CVD is used to describe the presence of either IHD or strokes. Renal insufficiency was defined as ICD-8 code 792.99 (uremia) or ICD-10 code N18.0 to N18.9 (chronic renal insufficiency) or renal insufficiency without specification (N19–N19.9). Stones in the kidney or ureter were considered in patients assigned an ICD-8 code of 591.00 to 592.09, 594.00 to 594.09 (nephrolithiasis and ureterolithiasis) or an ICD-10 code (N20–N20.9) for stones in the kidney and ureter or stones in the lower urinary tract (N21–N23.9).

Patients with diabetes were identified as those assigned an ICD-8 code of 249.00 to 250.09 or a corresponding ICD-10 code (E10–E14).

The occurrence of seizures assessed by identifying patients assigned an ICD-8 code of 252.11 (tetania post therapium), 345.09 to 345.99 (epilepsy), 780.31 to 780.39 (convulsions), 788.59 (tetania), 998.99 (tetania postoperativa), or ICD-10 codes of epilepsy (G40–G40.9), cramps (R25.2), or tetany (R29.0–R290A).

In our analyses on mortality and comorbidity, we excluded patients with HypoPT due to cancer surgery (including multiple endocrine neoplasia), because risk of death and comorbidity supposedly is increased in these patients.

Statistics

Denmark's population in year 2000 (5,336,394 inhabitants), was used to calculate incidence rates. We assessed differences between groups using Fisher's exact test for categorical variables and two-sample t test or Mann-Whitney U test for continuous variables, as appropriate, after testing for normal distributions. We used Kaplan-Meier methods to construct cumulative time-to-event curves for the two groups and calculated relative risk in terms of hazard ratios (HR) with 95% confidence intervals (95% CI) using Cox proportional-hazards models. Within each analysis (eg, risk of MI, risk of stroke, etc.), we only included the first-incident episode in analyses. We also performed analyses adjusted for a history of CVD, renal diseases, and diabetes. For subjects with a CVD or renal outcome, a history of diabetes was considered if the diagnosis of diabetes was assigned prior to or within 180 days following the CVD outcome. We performed all calculations using the IBM Statistical Package for Social Sciences (SPSS 20.0) (IBM, New York, USA) for Windows.

Results

Included patients

We identified a total of 1849 patients with postsurgical HypoPT, among whom 1127 were alive at the date of follow-up, equal to a prevalence of 21.0 per 100,000 inhabitants (Fig. 1). However, medical charts were not available for review in 15% of the patients. Accordingly, we were able to adjudicate the diagnosis in 1568 patients, of whom 1101 were alive at follow-up (prevalence 20.6/100,000). Approximately 15% of the patients were not coded in the NPR as having HypoPT; ie, they were identified through the prescription database covering two-thirds of the Danish population. Assuming that a similar proportion of patients were not coded as having HypoPT within the one-third of the population not covered by data from the prescription database, the adjusted prevalence of HypoPT in Denmark is 22 per 100,000 inhabitants. During the last 24 years (from 1988 and onward), a total of 980 patients were diagnosed with postsurgical HypoPT, equal to a yearly incidence of 0.77 per 100,000 inhabitants; after adjustment for the prescription registry we found a yearly incidence of 0.80 per 100,000 inhabitants.

Figure 1.

Identification of patients with postsurgical hypoparathyroidism.

Among the group of patients diagnosed since 1988, 292 (30%) had acquired HypoPT due to cancer surgery, whereas 688 (70%) patients had developed HypoPT as a complication to surgery for nonmalignant diseases.

Within the group of patients with postsurgical HypoPT since 1988 due to nonmalignant causes, 88% of the patients were women, with an average age at time of diagnosis (index date) of 49 years (Table 1). One-half of the cases had HypoPT due to surgery for nontoxic goiter (Table 1). A substantial part of the patients (n = 109; 16%) had had neck surgery for thyroid or parathyroid diseases prior to the operation causing HypoPT. Most of the patients were on treatment with calcium supplements (93%) and/or active vitamin D analogs (93%). Alfacalcidol was used in all cases, except two who were on treatment with calcitriol. Twenty-five (4%) patients received magnesium supplements and 519 (75%) received levothyroxine substitution therapy.

Table 1. Characteristics of Patients With Postsurgical Hypoparathyroidism as a Result of Nonmalignant Causes Diagnosed After 1987 and Their Matched Controls
 All (n = 2752)Patients (n = 688)Controls (n = 2064)p
  1. Number of subjects (n) with percentages within group (%), median with range (minimum–maximum), or median with interquartile (25% ; 75% percentile) range.
  2. NA = not applicable.
Gender
Male, n (%)333 (12)85 (12)248 (12)0.81
Female, n (%)2419 (88)603 (88)1816 (88) 
Age, years49 (16–88)49 (17–87)49 (16–88)0.97
Etiology, n (%)
Surgery for goiter without specification 21 (3)  
Surgery for atoxic goiter 343 (50)  
Surgery for toxic goiter 244 (35)  
Surgery for primary hyperparathyroidism 80 (12)  
Duration of disease, years 8 (4 ; 12)  
Treatment
Calcium, n (%) 641 (93)NA 
Dose, mg/d 1000 (800; 1600)NA 
Alfacalcidol, n (%) 639 (93)NA 
Dose, µg/d 1.0 (0.5; 2.0)NA 
Magnesium, n (%) 25 (4)NA 
Dose, mg/d 0 (0–1080)NA 
Levothyroxine, n (%) 519 (75)NA 

Controlled historical follow-up study

As shown in Table 1, the patients with postsurgical HypoPT due to nonmalignant causes (cases) were well-matched with controls on age and gender. The mean duration of follow-up (time from index date until end of follow-up) was 8.4 years in cases and 7.9 years in controls (p = 0.26).

Mortality

A similar number of cases and controls died during follow-up (Table 2). Accordingly, survival analyses showed no effect of postsurgical HypoPT on mortality (Fig. 2).

Table 2. Number of Subjects Who Died or Were Diagnosed With a Cardiovascular or Renal Disease During Follow-Up
 Cases (n = 688)Controls (n = 2064)p
  1. CVD = cardiovascular disease.
Death, n (%)79 (11.5)235 (11.4)0.95
Acute myocardial infarction, n (%)18 (2.6)52 (2.5)0.89
Ischemic heart disease, n (%)77 (11.2)162 (7.8)0.01
Stroke, n (%)35 (5.1)80 (3.9)0.19
Arrhythmia, n (%)50 (7.2)110 (5.3)0.07
Arrhythmia-induced deaths, n (%)2 (0.3)8 (0.4)1.00
Cardiovascular diseases, n (%)101 (14.7)221 (10.7)<0.01
Death or CVD, n (%)138 (20.1)374 (18.1)0.26
Seizures, n (%)26 (3.8)21 (1.0)<0.001
Renal stones, n (%)13 (1.9)8 (0.4)<0.001
Renal insufficiency, n (%)35 (5.1)21 (1.0)<0.001
Renal diseases, n (%)54 (7.8)38 (1.8)<0.01
Dialysis, n (%)2 (0.3)2(0.1)0.26
Figure 2.

Risk of death in patients with postsurgical hypoparathyroidism due to nonmalignant causes and matched controls. Hazard ratio (HR) with 95% confidence interval (95% CI).

Cardiovascular diseases

Prior to the date of surgery causing HypoPT, a diagnosis of a CVD was significantly (p < 0.01) more prevalent among cases (n = 63; 9.2%) than controls (n = 102; 4.9%). As shown in Table 2, significantly more cases than controls were diagnosed with an IHD or a CVD during follow-up. However, after adjusting for a prevalent CVD prior to the index date, risk of hospitalization due to an IHD or CVD was no longer increased. Further adjustments for a diagnosis of diabetes did not change the results (Table 3). As shown in Fig. 3A, risk of death or hospitalization due to a CVD did not differ between cases or controls.

Table 3. Risk of Cardiovascular and Renal Diseases in Patients With Postsurgical Hypoparathyroidism
 CrudeAdjustment model 1Adjustment model 2
  • Values are HR (95% CI).
  • HR = hazard ratio; CI = confidence interval; CVD = cardiovascular disease.
  • aAdjusted for prior CVD.
  • bAdjusted for prior renal diseases.
  • cAdjusted for prior CVD and diabetes.
  • dAdjusted for prior diabetes and renal diseases.
  • eAdjusted for prior seizures.
Myocardial infarction, acute1.00 (0.59–1.72)0.86 (0.50–1.49)a0.77 (0.44–1.34)c
Ischemic heart disease1.40 (1.07–1.84)1.16 (0.88–1.52)a1.09 (0.83–1.45)c
Stroke1.30 (0.87–1.93)1.20 (0.81–1.80)a1.09 (0.73–1.64)c
Arrhythmia1.32 (0.95–1.86)1.14 (0.83–1.60)a1.11 (0.79–1.57)d
Cardiac arrest0.72 (0.15–3.41)0.75 (0.16–3.54)a0.68 (0.15–3.23)c
Renal stones4.82 (2.00–11.64)4.22 (1.73–10.30)b4.02 (1.64–9.90)d
Renal insufficiency4.95 (2.88–8.50)4.54 (2.63–7.84)b3.10 (1.73–5.55)d
Renal disease4.31 (2.84–6.52)3.97 (2.61–6.03)b3.67 (2.41–5.59)d
Seizures3.77 (2.12–6.70)3.82 (2.15–6.79)e 
Figure 3.

Risk of death or hospitalization for a cardiovascular disease (A) or hospitalization due to renal (nephrolithiasis or renal insufficiency) diseases (B) in patients with postsurgical hypoparathyroidism due to non-malignant causes and matched controls. Hazard ratio (HR) with 95% confidence interval (95% CI) adjusted for a history of diabetes and a prior cardiovascular disease (A) or a prior renal disease (B).

Cardiac arrhythmias was borderline significantly more prevalent in patients with postsurgical HypoPT compared with the controls (Table 2). However, after adjusting for prior CVD and diabetes, the associations were no longer significant (Table 3).

Risk of renal complications

Compared with the controls, patients with postsurgical HypoPT had a significantly increased risk of renal diseases in terms of calcifications or renal insufficiency (HR, 4.31; 95% CI, 2.84–6.52). As shown in Fig. 3B, risk was only slightly reduced by adjustment for a history of renal diseases or diabetes prior to the index date (HR, 3.67; 95% CI, 2.41–5.59). Compared with the controls, patients had an almost five times increased risk of renal stones (HR, 4.82; 95% CI, 2.00–11.64) and renal insufficiency (HR, 4.95; 95% CI, 2.88–8.50). Adjustment for prior renal disease and diabetes did not change results to any major degree (Table 3). Moreover, although not statistical significant, more patients than controls received treatment with dialyses (Table 2).

Seizures

Patients with postsurgical HypoPT were more often hospitalized due to seizures than the controls (Table 2). As shown in Table 3, adjustment for a history of seizures prior to the operation causing HypoPT did not change results (HR, 3.82; 95% CI, 2.15–6.79).

Discussion

Taking advantage of the extensive number of national databases on health outcomes in Denmark, we were able to identify patients with postsurgical HypoPT and compare risk of death, CVD, seizures, and renal complications in HypoPT with a group of age-matched and gender-matched population-based controls. To the best of our knowledge, the present analyses represent one of the first large-scale assessments of the epidemiology of HypoPT and possible complications to the disease.

In previous studies, risk of HypoPT has been assessed in terms of complication rates to specific surgical procedures, showing that risk increases with the extent of thyroid dissection and if biopsies are obtained from the parathyroid glands during parathyroid surgery.[20] Importantly, we did not aim to assess risk of HypoPT as a complication to specific diseases treated by neck surgery. However, our data adds to previous findings by showing that a substantial number of patients with HypoPT have acquired the disease as either a complication to cancer surgery (30%) or following repeated neck surgery due to nonmalignant causes (16%).[7]

PTH normally stimulates the renal tubular reabsorption of calcium.[1] The calcium loads in the nephron is therefore increased if PTH levels are low and lack of PTH may therefore by itself increase risk of renal stones. In addition to adverse renal effects of calcium supplements per se, calcium-phosphate product is typically increased in HypoPT, which may further increase risk of renal stones.[7]

Our findings of an increased risk of kidney diseases in terms of renal calculi or insufficiency agree with previous findings of an increased risk with use of calcium supplements. In the Women's Health Initiative (WHI) randomized trial, risk of renal stones was significantly increased by 17% in participants randomized to receive 1000 mg of elemental calcium with 400 IU of vitamin D3 compared with placebo.[12] In addition, treatment with vitamin D and calcium increases the risk of vitamin D intoxication. Intoxication will lead to hypercalcemia and hyperphosphatemia, which may cause nephrocalcinosis with a progressive loss of kidney functions. In contrast to the adverse renal effects of calcium supplements, several observational studies have suggested that a high intake of calcium through the diet may actually decrease risk of kidney stones.[21-23] In an analysis from the Nurses' Health Study II, Curhan and colleagues[21] found a trend toward a reduced risk of kidney stones with an increased dietary calcium intake. Similarly, in the Nurses' Health Study I, those with a dietary calcium intake in the highest quintile had a significantly reduced risk of renal stones compared with those in the lowest quintile (relative risk [RR], 0.65; 95% CI, 0.5–0.83).[22] Further studies should aim to determine whether an increased intake of calcium from diet instead of use of calcium supplements or use of PTH replacement therapy may reduce the risk of renal complications in patients with HypoPT.[24]

Discrepant results have been reported on the effects of calcium intake on cardiovascular health.[25] In epidemiological studies, risk of CVD has consistently been reported to be inversely associated with dietary calcium intake.[26-28] Studies on use of calcium supplements have, however, shown conflicting results, with either an increased risk[11, 29, 30] or no effect on CVD.[31, 32] Our findings of no excess adverse cardiovascular outcomes in patients with HypoPT do not support harmful effects of calcium supplements within this group of patients. It has been suggested that the apparent discrepant results between observational and randomized studies may be explained by selection bias; ie, that patients who choose to use calcium supplements are more healthy than nonusers, which may blunt potential harmful effects. It should be noted that in our group of patients, calcium supplements were administered as part of their medical treatment; therefore, a self-selection (healthy user) bias probably does not apply to our results. However, our findings in patients with HypoPT may not be applicable to the general population. An adverse effect of calcium supplements on cardiovascular health has been suggested to be a result of transient hypercalcemia following intake of calcium supplements.[33] In HypoPT, plasma calcium levels are preferentially kept in the low-normal range[7] and intake of calcium supplements may therefore not cause hypercalcemia to the same extent as in patients with an intact parathyroid function. Our findings of no excess cardiovascular outcomes in patients with HypoPT do not support adverse effects of calcium supplements on cardiovascular health within this group of patients.

Our findings of an almost four times increased risk of seizures in patients with HypoPT adds further concerns to the question on whether conventional treatment with calcium supplements and active vitamin D analogs may result in severe adverse health outcomes. Although the actual number of patients hospitalized due to seizures was relatively low, an episode of a seizure may have debilitating consequences.

Strengths and limitations to our results

A major strength of our study is that we restricted our analyses on adverse health outcomes to patients with postsurgical HypoPT, rather than all causes of HypoPT, such as idiopathic or autoimmune HypoPT. This approach allowed us to identify a specific starting point and to adjust for potential confounders such as history of the disease studied. Another strength of our study is the careful adjudication of case status. However, by identifying patients using ICD codes, our analysis does not exclude that we may have missed some episodes of renal stones or nephrocalcinosis, for example, that did not result in hospitalizations. Moreover, in addition to HypoPT, most of our patients also suffered from postsurgical hypothyroidism, which we cannot exclude may have had an influence on studied outcomes. Furthermore, we cannot exclude that we may have missed patients who acquired HypoPT prior to the establishment of the Danish NPR in 1977, if the patients had not been hospitalized and assigned a diagnosis of HypoPT since then. Nevertheless, identification of patients using a combination of different types of registers strengthens our case-finding strategy because a portion of our patients were identified through the prescription database. This is due, most likely, to the fact that these patients have only been coded by their disease necessitating surgery and have not been assigned an additional code corresponding to HypoPT. Although most patients started treatment shortly after the surgery causing HypoPT, some patients may have changed therapy during the course of their disease, which we have not been able to address precisely in the current analyses.

In our study, the prevalence of HypoPT was slightly lower than reported in a preliminary report by Clarke and colleagues,[6] which may be due to several reasons, including differences in the methodological approach as well as possible differences in indications for surgery and surgical techniques. Accordingly, the prevalence of postsurgical HypoPT reported in this work is not necessarily applicable to other populations, as it may depend on the diagnostic strategy and preferred surgical procedures for thyroid and parathyroid diseases and the prescribed drug for treatment of HypoPT. In North America, calcitriol is used more often than alfacalcidol. Nevertheless, because the hospitals in Denmark refer to international standards and participate in international surveillance programs for thyroid and parathyroid surgery, we believe that our results may somehow be representative for most westernized countries.[3, 34]

Conclusion

Postsurgical HypoPT is a rare disease with an estimated prevalence of 22 per 100,000 inhabitants in Denmark. Approximately one-third of the cases result from surgery for cancers at the neck. Survival is not impaired in HypoPT resulting from nonmalignant causes and neither do patients have an increased risk of CVDs. Our data are reassuring in terms of the cardiovascular safety of calcium and vitamin D analog supplementation to patients with postsurgical HypoPT. However, risk of seizures and renal complications in terms of kidney stones and renal impairment is markedly increased. Our study did not suggest an increased overall mortality; however, our mean follow-up period of 8 years might be too short to show an increased mortality. Further studies should focus on how to optimize treatment in order to avoid such complications.

Disclosures

LM is primary investigator on the REPLACE-study initiated by NPS Pharmaceuticals. All other authors state that they have no conflicts of interest.

Acknowledgments

This work was supported the Augustinus Foundation and the Th. Maigaard's Eftf. Fru Lily Benthine Lunds Foundation of 1.6.1978.

Authors' roles: Line Underbjerg: Drafting of the manuscript. Tanja Sikjaer and Leif Mosekilde: Critical revision of the manuscript and approval of the manuscript. Lars Rejnmark: Contributing to the concept and design as well as critical revision of the manuscript and approval of the article.

Ancillary