In this large prospective cohort study of women, higher weekly physical activity was associated with a significantly lower risk of developing incident primary hyperparathyroidism.
Abstract
Context:
Primary hyperparathyroidism (P-HPTH) is relatively common and predominantly affects women. Prior studies have shown that physical activity (PA) can lower PTH levels.
Objective:
Our objective was to evaluate the hypothesis that lower PA is a risk factor for developing P-HPTH.
Design, Setting, and Participants:
This prospective cohort study included 69 621 female participants in the Nurses' Health Study I followed for 22 years.
Exposures:
PA and other dietary and demographic exposures were quantified via detailed, and validated, biennial questionnaires.
Outcomes:
Incident P-HPTH was confirmed by medical record review after initial assessment by questionnaire. Adjusted Cox proportional hazards models were used to evaluate whether PA was an independent risk factor for developing P-HPTH. We also evaluated the risk of developing P-HPTH when combining low PA (<16 metabolic equivalent hours/week) with a previously identified independent risk factor for developing P-HPTH: low calcium intake (<800 mg/day). The relation between PA and PTH levels was evaluated in 625 participants.
Results:
We confirmed 302 incident cases of P-HPTH during 1 474 993 person-years of follow-up. Participants in the highest quintile (Q) of PA had a 50% lower risk of developing P-HPTH: age-adjusted relative risks and 95% confidence intervals for incident P-HPTH by lowest to highest of PA were Q1 = 1.0 (reference); Q2 = 0.83 (0.60–1.15); Q3 = 0.84 (0.61–1.15); Q4 = 0.50 (0.34–0.74); Q5 = 0.50 (0.35–0.73); P for trend <.001. Extensive multivariable adjustments did not materially change these findings. The adjusted relative risk for developing P-HPTH among participants with the combination lower PA and lower calcium intake was 2.37-fold (1.60–3.51) higher than in participants with higher PA and higher calcium intake. PA was inversely correlated with serum PTH (ρ = −0.09, P = .03); the mean adjusted serum PTH in Q 2–5 of PA was lower than in Q 1 (36.3 vs 39.1 pg/mL, P = .02).
Conclusion:
Low physical activity may be a modifiable risk factor for developing P-HPTH in women.
Primary hyperparathyroidism (P-HPTH) is a relatively common disorder that has significant health implications (1, 2): constitutive and autonomous PTH secretion causes hypercalcemia, reduced bone density, and osteoporosis. Despite these important public health consequences, which predominantly impact postmenopausal women (2), our understanding of modifiable risk factors for developing P-HPTH remains limited.
A notable and frequently studied modifier of parathyroid function is physical activity (PA). A sedentary lifestyle has been implicated as a major contributor to many metabolic and bone diseases (3–5). Prior studies have shown that moderate-to-intense exercise activates a unique physiologic phenomenon whereby PTH secretion increases during episodes of exercise and subsequently decreases following exercise (6–9). Further, regular participation in a long-term exercise program lowers the setpoint of baseline circulating PTH levels (10). To date, no prospective studies have evaluated whether PA, or maintenance of a physically active lifestyle, can lower the development of incident P-HPTH.
We hypothesized that individuals who practice a more active lifestyle, incorporating higher levels of regular PA, would have a reduced risk of developing P-HPTH, whereas individuals who practice a more sedentary lifestyle could have a higher risk of developing P-HPTH. We evaluated this hypothesis by conducting a large prospective cohort study whereby we assessed whether the frequency and intensity of PA, over a 22-year span, was associated with the risk of developing incident P-HPTH.
Materials and Methods
Study population
The Nurses' Health Study I (NHS) is an ongoing, national, prospective cohort study which began in 1976, enrolling 121 700 female registered nurses between 30 and 55 years of age. The cohort is followed by questionnaires mailed every 2 years that ask about lifestyle practices and newly diagnosed diseases. The follow-up of participants exceeds 90% of the eligible person-time. Our analysis included 69 621 women who answered either the 2006 or 2008 questionnaires, which assessed lifetime history of P-HPTH, and had questionnaire assessment of PA. The study protocol was approved by the Brigham and Women's Hospital institutional review board.
Assessment of the main exposure: PA
PA in the NHS was assessed by the biennial questionnaires (3, 11). Participants were asked to report the average amount of time spent per week during the previous year in each of the following seven activities: 1) walking or hiking outdoors; 2) jogging (>10 minutes/mile); 3) running; 4) bicycling (including stationary machine); 5) racquet sports (tennis, squash, racquetball); 6) lap swimming; 7) and other aerobic activity (for example, aerobic dance, rowing machine). For each activity, women chose one of 11 duration categories that ranged from zero to 11 hours/week or more. Participants also reported their usual walking pace and number of flights of stairs climbed daily. Walking pace was reported as either easy (<2 mph), average (2–2.9 mph), brisk (3–3.9 mph), very brisk (≥4 mph), or unable to walk. Beginning in 1992, activity questionnaires also asked about other vigorous activities (for example, mowing the lawn) and lower intensity exercises (such as yoga or stretching).
Based on these responses, each activity on the questionnaire was assigned a metabolic equivalent (MET) score and each participant was assigned a weekly PA score expressed in MET hours per week (3, 11, 12). The scores for MET-hours per week for each activity were calculated from the reported hours per week engaged in that activity multiplied by the assigned MET score; the values from the individual activities were then summed for a total MET-hours per week score (13, 14). One MET is the energy expenditure for sitting quietly. MET scores for specific activities are defined as the ratio of the metabolic rate associated with that activity divided by the resting metabolic rate. Activities between 3 and 6 METs are considered moderate intensity (such as walking), whereas activities higher than 6 METs are considered vigorous intensity (such as jogging or running) (11, 15). This assessment of PA has been previously validated against PA diaries in a similar cohort (r = 0.79) (16).
Because intraindividual PA can vary over time, to obtain the best long-term measure of PA, total values were cumulatively averaged in our analyses: at the beginning of each 2-year follow-up period, the MET-hours per week was the mean of all MET-hours per week calculated from responses to the questionnaires up to that time. Complete assessments of PA were obtained in 1986, 1988, 1992, 1994, 1996, 1998, 2000, and 2004. Participants needed to have at least three complete assessments of PA over the course of the study to be included in the analysis. For the 1990, 2002, and 2006 questionnaires, in which partial assessments of PA were conducted, the prior biennial questionnaire PA data were carried forward.
Assessment of dietary exposures
We considered many dietary exposures that may be potential confounders of parathyroid function and neoplasia. To assess the participants' diet, we used semiquantitative food frequency questionnaires that asked about the average intake of more than 130 individual food items and 22 individual beverages during the previous year. The participants were asked to complete food frequency questionnaires in 1986 and every 4 years thereafter. Intake of specific dietary factors was computed from the reported frequency of consumption of each specified unit of food and from US Department of Agriculture data on the content of the relevant nutrient in specified portions. The food frequency questionnaire also asked about the use of calcium supplements, vitamin D supplements, and multivitamins. The food frequency questionnaire has been extensively validated (17, 18).
Assessment of other nondietary exposures
We also considered many nondietary exposures. Age, race, body mass index (BMI) (14), smoking status (never, past, current), history of diabetes, hypertension, osteoporosis, chronic kidney disease, heart failure, menopausal status (19), and postmenopausal hormone use were ascertained from the biennial questionnaires. The use of medications (including those used to treat hypertension and osteoporosis) was assessed at each biennial questionnaire.
Because participants who engage in greater PA may have greater exposure to sunlight and UV radiation, we adjusted for average annual UVB radiation, which has been demonstrated to be a reliable predictor of 25-hydroxyvitamin D (25[OH]D) levels in the NHS cohort (20). UVB flux is a composite measure of mean UVB radiation level reaching the Earth's surface that takes into account factors such as latitude, altitude, and cloud cover, based on state of residence (20).
Assessment of the outcome of interest: incident P-HPTH
Participants were asked about a diagnosis of hyperparathyroidism on the 2006 and 2008 questionnaires. To distinguish P-HPTH from nonprimary forms of hyperparathyroidism, we requested medical records of all participants who gave consent. We confirmed cases of P-HPTH using the conservative criteria of an elevated serum concentration of calcium (≥10.6 mg/dl) with a concomitantly high or insufficiently suppressed PTH (>50 pg/ml) and/or a pathology report indicating surgical resection of an adenomatous or hyperplastic parathyroid gland in the context of medical records indicating a diagnosis of P-HPTH. Of the medical records that were reviewed, we confirmed P-HPTH in 83% of participants. Self-reports were rejected after medical record review for a variety of reasons, most commonly identification of secondary hyperparathyroidism from vitamin D deficiency or renal insufficiency (9%), misreporting of a thyroid disorder as hyperparathyroidism (4%), or insufficient data to confirm the diagnosis of P-HPTH (3%). Participants with a history of P-HPTH at baseline were excluded.
Statistical analyses
The study design was prospective; information on PA was collected before the diagnosis of P-HPTH. For each participant, we counted person-time of follow-up from the date on which the first PA assessment questionnaire was returned to the date on which P-HPTH was diagnosed or death occurred, or May 31, 2008, whichever occurred first. We allocated person-time of follow-up according to exposure status at the start of each follow-up period.
We first assessed whether PA was an independent predictor for developing incident P-HPTH. PA was categorized into quintiles, with quintile 1 serving as the reference category. The measure of association used was the relative risk—the incidence rate of P-HPTH in each quintile with respect to the reference quintile 1. We used Cox proportional hazards regression to simultaneously adjust for known or potential confounders. The variables considered in these models were age; BMI (as a continuous variable); race (white or nonwhite); smoking status (past, current, never); menopausal status (pre or post), postmenopausal hormone use (yes or no); physical examination in prior 2 years; bisphosphonate use; dietary calcium intake (13), supplemental calcium intake (none, 1–500, >500 mg/d) (13), quintiles of intakes of total vitamin D and vitamin A intake, alcohol intake (none, 0.1–4.9, 5–14.9, ≥15 g/d), dietary magnesium intake, dietary protein intake; UVB radiation flux; and self-reported history of hypertension (1, 21), diabetes, chronic kidney disease, congestive heart failure, or osteoporosis. We repeated our analyses after including a minimum 2-year lag between the time of final cumulative PA assessment and the time of incident P-HPTH diagnosis. We calculated 95% confidence intervals for all relative risks. All P values were two-tailed.
Because we previously demonstrated that a lower total intake of calcium (13) and hypertension (1) are both independently associated with a higher risk of developing P-HPTH, we assessed whether lower PA when combined with these risk factors imparted an additive risk for developing P-HPTH. To conduct this analysis, we categorized each participant by their status for having any one these three risk factors. Lower PA was defined as cumulative PA less than 16 MET hours/week and lower calcium intake was defined as less than 800 mg per day of calcium from dietary and supplemental sources (13). Hypertension was described as self-reported history of hypertension (Y/N) (1). We conducted adjusted Cox proportional hazards regression to evaluate the independent risk for developing P-HPTH based on whether participants had the aforementioned risk factors: lower PA and/or lower calcium intake and/or hypertension.
Last, we explored whether serum PTH levels correlated with PA, as it has in prior observational and interventional studies (9, 10, 22). In previous nested case control studies of circulating factors and risk of incident coronary heart disease (CHD) in NHS (15), a total of 625 women (without CHD at time of blood draw) had PTH levels measured from stored aliquots of blood collected between 1989 and 1990 (15). None of these women had CHD at the time of blood draw. Intact PTH was measured by an electrochemiluminescence immunoassay on the 2010 Elecsys autoanalyzer (Roche Diagnostics). The coefficient of variation (CV) for this assay using quality control samples was 5.2%. Other relevant measurements included plasma 25(OH)D, measured by an enzyme immunoassay from Immunodiagnostic Systems Inc. (CV, 7.3%), serum calcium (colorimetric assay; CV, 2.9%), and serum phosphate (photometric assay; CV, 2.8%). We assessed the univariate associations between PTH levels and continuous demographic and biochemical factors using Spearman correlations. We further assessed PTH levels per quintile of PA using an analysis of covariance model that was adjusted for all previously mentioned variables used in our Cox proportional hazards regressions, as well as adjustments for available biochemical parameters (calcium, phosphate, creatinine, 25[OH]D, albumin), and for multiple testing.
Results
Study population
Baseline characteristics of the study population by quintile of PA are presented in Table 1. There was no appreciable difference in age, menopausal status, prevalence of osteoporosis, or bisphosphonate use among quintiles of PA. However, trends suggest that greater weekly PA associates with lower BMI, lower prevalence of hypertension and diabetes, and higher dietary and supplemental calcium intake, higher vitamin A and D intake, and higher magnesium intake (Table 1).
Table 1.
Study Population Characteristics in 1986 by Quintile of PAa
| Quintile of Physical Activity |
|||||
|---|---|---|---|---|---|
| Q1 (n = 11 683) | Q2 (n = 11 993) | Q3 (n = 12 356) | Q4 (n = 12 364) | Q5 (n = 12 392) | |
| Age, yb | 51.3 (6.8) | 51.7 (6.8) | 51.7 (7.0) | 51.8 (7.0) | 51.9 (7.0) |
| Physical activity, MET-h/wkb,c | <2.2 | 2.2–4.8 | 4.9–10.7 | 10.8–22.1 | >22.1 |
| BMI, kg/m2 | 26.2 (5.3) | 25.7 (4.8) | 25.2 (4.5) | 24.7 (4.2) | 24.1 (4.0) |
| White race, % | 93.6 | 94.4 | 95.0 | 94.8 | 94.4 |
| Smoking status | |||||
| Never, % | 45.0 | 47.6 | 47.7 | 46.4 | 44.1 |
| Past, % | 31.2 | 32.0 | 35.5 | 38.1 | 40.6 |
| Current, % | 23.5 | 20.2 | 16.6 | 15.2 | 15.0 |
| Missing, % | 0.2 | 0.2 | 0.2 | 0.3 | 0.3 |
| Postmenopausal status | |||||
| Pre, % | 35.5 | 35.8 | 36.0 | 36.4 | 35.6 |
| Post, % | 64.4 | 64.0 | 63.8 | 63.5 | 64.2 |
| Missing, % | 0.2 | 0.2 | 0.2 | 0.2 | 0.2 |
| Postmenopausal hormone use, % | 15.4 | 16.8 | 18.3 | 18.8 | 19.1 |
| History of hypertension, % | 24.3 | 23.6 | 22.8 | 21.7 | 19.9 |
| History of diabetes, % | 3.0 | 2.4 | 2.8 | 2.2 | 2.1 |
| Supplemental calcium use, Y/N, % | 48.9 | 53.2 | 57.4 | 61.1 | 63.6 |
| Supplemental calcium intake, mg/d | 296.4 (409.3) | 318.8 (407.6) | 350.5 (421.7) | 384.9 (431.1) | 418.3 (447.5) |
| Alcohol intake, g/d | 5.7 (11.0) | 5.5 (10.0) | 5.9 (9.9) | 6.2 (10.2) | 7.0 (10.7) |
| Dietary calcium intake, mg/d | 684.1 (253.9) | 705.5 (248.8) | 718.8 (243.5) | 735.8 (247.9) | 745.1 (252.4) |
| Vitamin D intake, IU/d | 304.0 (241.1) | 322.1 (242.0) | 335.8 (243.8) | 356.5 (252.5) | 373.2 (264.8) |
| Vitamin A intake, IU/d | 11 546.0 (7293.5) | 12 467.1 (7367.2) | 13 225.9 (7457.4) | 14 010.2 (7803.5) | 15 218.5 (8752.0) |
| Magnesium intake, mg/d | 281.6 (65.5) | 291.1 (64.9) | 299.4 (66.1) | 307.7 (68.6) | 316.4 (70.5) |
| Protein intake, g/d | 73.3 (13.0) | 74.5 (12.5) | 75.0 (12.6) | 75.9 (12.6) | 76.5 (13.1) |
| UVB radiation flux (Robertson-Berger count × 10−4) | 122.0 (24.6) | 120.8 (23.7) | 121.5 (24.2) | 121.6 (24.1) | 122.7 (24.7) |
| History of chronic kidney disease, 1990, %d | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| History of congestive heart failure, 1998, % | 1.5 | 1.1 | 1.1 | 1.2 | 1.0 |
| History of osteoporosis, % | 2.3 | 2.4 | 2.3 | 2.3 | 2.5 |
| Bisphosphonate use, 1998, % | 4.5 | 4.0 | 4.4 | 4.3 | 4.6 |
Abbreviations: N, no; Q, quintile; Y, yes.
Values are means(sd) or percentages and are standardized to the age distribution of the study population.
Values of polytomous variables may not sum to 100% because of rounding.
The number of participants at baseline in 1986 (60 788) is lower than the total number used in our analysis (69 621) because participants were allowed to enter follow-up over the course of the study as they became eligible.
Value is not age-adjusted.
Physical activity measures are from the 1986 baseline questionnaire assessment.
Chronic kidney disease values are all <0.05%.
PA and incident P-HPTH
We confirmed 302 incident cases of P-HPTH that developed during 1 474 993 person-years of follow-up (Table 2). When compared with participants in the lowest quintile of cumulative average PA, the age-adjusted relative risk for P-HPTH was significantly lower in quintiles 4 and 5 (50% lower), with a significant trend across quintiles (Table 2). These results were materially unchanged following additional comprehensive adjustments (Table 2). In additional lag-time analyses where a minimum of 2 years between the final cumulative average PA assessment and incident P-HPTH was required, the relative risks and confidence intervals in Table 2 did not materially change. We calculated the population attributable fraction for lower cumulative PA (<16 MET-hours per week) to be 27.4%.
Table 2.
Risk of Incident P-HPTH by Quintile of PA
| Q1 | Q2 | Q3 | Q4 | Q5 | P for Trend | |
|---|---|---|---|---|---|---|
| Cases | 76 | 68 | 71 | 43 | 44 | |
| Median PA (MET-h/wk)a | 3.6 | 8.2 | 13.4 | 20.7 | 36.9 | |
| Range of PA (MET-h/wk) | <5.8 | 5.8–10.5 | 10.6–16.5 | 16.6–26.4 | >26.4 | |
| Person-years | 276 760 | 291 841 | 298 912 | 302 212 | 305 268 | |
| Age-adjusted RR (95% CI) | Reference | 0.83 (0.60–1.15) | 0.84 (0.60–1.15) | 0.50 (0.34–0.73) | 0.50 (0.35–0.73) | <.001 |
| Multivariate-adjusted RRb | Reference | 0.84 (0.60–1.16) | 0.85 (0.61–1.19) | 0.52 (0.35–0.76) | 0.54 (0.37–0.80) | <.001 |
Abbreviations: CI, confidence interval; Q, quintile; RR, relative risk.
Median cumulative average physical activity levels are shown as of the 2006 questionnaire; however, cumulative PA values were updated throughout the course of the study. Participants were permitted to move in and out of the study population between 1986 and 2006 based on information on exposures and outcomes; therefore, the total number of participants in this table is 69 621, which is greater than the baseline of 60 788 participants in 1986.
Multivariable model includes age; BMI (continuous); race; smoking status (past, current, never); menopausal status (pre or post), postmenopausal hormone use (yes or no); physical examination in prior 2 years; bisphosphonate use; histories of hypertension, diabetes, chronic kidney disease, congestive heart failure, or osteoporosis; dietary calcium intake, supplemental calcium intake, total vitamin D and vitamin A intake, alcohol intake, dietary magnesium intake, dietary protein intake; and UVB radiation flux.
Known risk factors and risk of incident P-HPTH
In addition to our current finding demonstrating lower PA as an independent risk factor for developing P-HPTH, our prior prospective studies in NHS indicated that lower dietary and supplemental calcium intake (13) as well as having hypertension (1) are also independent risk factors for developing incident P-HPTH. Although we accounted for both of these factors in our regression models (Table 2), we sought to examine whether the combination of these risk factors could have incremental effects on developing P-HPTH. We categorized participants based on how many established risk factors for P-HPTH they had, where risk factors were defined as: lower cumulative PA (PA <16 MET-hours per week), lower calcium intake (total dietary and supplemental calcium intake <800 mg/day), and having hypertension. In comparison to the reference of participants with zero risk factors, participants with both lower calcium intake and lower PA had a 2.37-fold (1.60–3.51) higher risk of developing P-HPTH (Table 3). When including all three risk factors in the analysis, participants with lower calcium intake, lower PA, and hypertension had a greater than 4-fold higher risk of developing P-HPTH when compared to those with zero risk factors (Supplemental Table 1); however, this result should be interpreted cautiously because there were only 31 cases of P-HPTH in the referent category of zero risk factors.
Table 3.
Risk of Incident P-HPTH According to Combinations of Risk Factors: Lower PA and Lower Total Calcium Intake
| No Risk Factors: | 1 Risk Factor: | 2 Risk Factors: | |
|---|---|---|---|
| PA ≥16 MET-h/wk | PA <16 MET-h/wk | PA <16 MET-h/wk | |
| AND | OR | AND | |
| Calcium Intake ≥800 mg/d | Calcium Intake <800 mg/d | Calcium Intake <800 mg/d | |
| Cases | 73 | 163 | 63 |
| Person-years | 436 598 | 708 306 | 273 275 |
| Age-adjusted RR (95% CI) | Reference | 1.58 (1.20–2.09) | 2.09 (1.48–2.94) |
| Multivariate-adjusted RRa | Reference | 1.56 (1.17–2.07) | 2.37 (1.60–3.51) |
Abbreviations: CI, confidence interval; RR, relative risk.
Established independent risk factors for P-HPTH are defined as total calcium intake <800 mg/d and cumulative PA levels <16 MET-h/wk. Three of the cases of P-HPTH were missing information regarding one of the risk factors.
Multivariable model includes age; BMI (continuous); race; smoking status (past, current, never); menopausal status (pre or post), postmenopausal hormone use (yes or no); physical examination in prior 2 years; bisphosphonate use; history of hypertension, diabetes, chronic kidney disease, congestive heart failure, or osteoporosis; total vitamin D and vitamin A intake, alcohol intake, dietary magnesium intake, dietary protein intake; and UVB radiation flux.
Exploratory analysis: PA and serum PTH levels
Among 625 participants with available PTH levels, higher PTH was associated with lower PA, older age, higher BMI, postmenopausal status, lower total calcium and vitamin A intake, higher prevalence of hypertension, lower prevalence of diabetes, and lower levels of 25(OH)D (Supplemental Table 2). There was a significant inverse correlation between PA and PTH levels (Supplemental Table 3). Adjusted regression analyses showed that PTH levels were highest in quintile 1 of PA, and were relatively lower and similar across quintiles 2–5 (Table 4). In age-adjusted regression analyses, participants in the fifth quintile of PA had significantly lower PTH levels when compared with those in quintile 1 (35.1 vs 39.9 pg/mL; P < .01). This difference in PTH values between quintiles 5 and 1 remained relatively stable, although without statistical significance, following multivariable adjustments for pertinent demographic and dietary confounders as well as biochemical predictors of PTH (Table 4). Because participants in PA quintiles 2–5 had similar mean PTH values, we also combined women in quintiles 2–5; the mean adjusted PTH level in quintiles 2–5 was significantly lower than quintile 1 (36.3 vs 39.1 pg/mL, P = .02) (Table 4).
Table 4.
Adjusted Levels of Intact PTH by PA Quintiles (n = 625)
| PA | Q1 | Q2 | Q3 | Q4 | Q5 |
|---|---|---|---|---|---|
| Median (MET-h/wk) | 1.2 | 4.6 | 9.7 | 17.7 | 36.1 |
| Age-adjusted PTH (pg/ml) | 39.9 | 36.1 | 35.9 | 37.3 | 35.1 |
| P valuea | Reference | 0.06 | 0.04 | 0.29 | <0.01 |
| MV-adjusted PTH (pg/ml)b | 39.1 | 36.2 | 36.0 | 37.9 | 35.3 |
| P valuea | Reference | 0.21 | 0.18 | 0.88 | 0.06 |
| MV-adjusted PTH (pg/ml)c | 39.1 | 36.0 | 36.1 | 37.6 | 35.5 |
| P valuea | Reference | 0.17 | 0.18 | 0.76 | 0.07 |
| Quintile 1 | Quintiles 2–5 | |
|---|---|---|
| Age-adjusted PTH (pg/ml) | 39.9 | 36.1 |
| P valuea | Reference | <0.01 |
| MV-adjusted PTH (pg/ml)b | 39.1 | 36.3 |
| P valuea | Reference | 0.02 |
| MV-adjusted PTH (pg/ml)c | 39.1 | 36.3 |
| P valuea | Reference | 0.02 |
Abbreviations: MV, multivariate; Q, quintile. Bold values indicate mean values.
P value compared to lowest quintile, adjusted for multiple comparisons.
Adjusted for age; race; BMI; smoking (current, past, or never); CHD case (yes or no); history of osteoporosis (yes or no), hypertension (yes or no), and diabetes (yes or no); total intakes of calcium, magnesium, protein, vitamin A, and alcohol; UVB radiation flux; and menopausal status and postmenopausal hormone use.
Adjusted for age; race; BMI; smoking (current, past, or never); CHD case (yes or no); history of osteoporosis (yes or no), hypertension (yes or no), and diabetes (yes or no); menopausal status and postmenopausal hormone use; and plasma levels of 25(O)D, calcium, phosphate, creatinine, and albumin.
Discussion
Primary hyperparathyroidism is an increasingly prevalent condition that predominantly affects postmenopausal women (2, 23–26). Despite this important public health context, our understanding of modifiable risk factors for developing P-HPTH, and methods by which to prevent or mitigate its severity, remain limited. Here, we conducted a large prospective cohort study spanning 22 years and more than 1.4 million person-years of follow-up and found that participation in greater weekly PA (such as running more than once a week or walking briskly multiple times a week) was independently associated with a decreased risk of developing P-HPTH. Our results are unique in that they represent the first investigation of PA as a predictor of incident P-HPTH, and further, identify in PA a modifiable risk factor for P-HPTH that has also previously been shown to modify PTH levels.
The connection between PA and parathyroid function has been the subject of many prior interventional and observational investigations (9, 10, 22) that have suggested that a lifestyle that incorporates frequent vigorous PA may result in a cumulatively decreased parathyroid gland stimulation, or PTH secretion, over time. Intervention studies have shown that PTH levels increase acutely during a bout of moderate-to-intense exercise (running or bicycling) (6–9); however, this rise in PTH is not explained by traditional physiologic principles, because serum calcium (and phosphate) is observed to simultaneously increase with exercise in all of these studies. In this respect, this independently repeated outcome suggests a unique calcium-parathyroid physiology of exercise that has not been explained to date. Notably, this unique physiology continues beyond the active exercise period: PTH levels not only rise during moderate-to-intense exercise, but decline to a level that is lower than the circadian baseline of a comparative nonexercising control group for several hours following the exercise (9). Hence, our current knowledge of PTH exercise physiology suggests that intense PA acutely raises PTH during the exercise, but results in a subsequent prolonged decrease of the PTH set point. Whether this resultant exercise-induced decrease in PTH secretion reduces the risk of subsequent parathyroid overactivity or neoplastic transformation has never been directly studied; however, participants who were randomized to a high-impact exercise program for 12 months exhibited a 22% decrease in circulating baseline PTH levels when compared with a control group that was not prescribed an exercise program (10). Similarly, in one longitudinal prospective observational study, women who participated in regular PA were observed to have a dose-dependent decline in PTH levels over a 10-year period when compared to women not participating in regular exercise (22). Interestingly, in another large prospective cohort study, participation in greater PA was associated with a lower risk of hip fracture (3); although parathyroid function or PTH levels were not directly investigated in this study, this observation raises the question of whether exercise-induced reductions in PTH secretion could represent a mechanism to improve skeletal integrity.
Our current study significantly extends and builds upon prior knowledge. Herein, we demonstrate that in comparison with a sedentary lifestyle of minimal PA (<5.8 MET-hours per week), participation in higher weekly PA (≥16.6 MET-hours per week, the approximate equivalent of brisk walking >5 hours per week or running >2.5 hours per week) may decrease this risk by nearly 50%. Our analyses suggest that the population-attributable fraction for lower PA (the fraction of P-HPTH cases that theoretically could be prevented by eliminating exposure to lower PA assuming that lower PA is a causal and independent predictor of P-HPTH) is 27.4%. This relationship was robust and independent of numerous potential dietary and disease-related confounders, with a trend suggesting a dose-dependent association between PA and incident P-HPTH. In the context of intervention studies showing a reduced PTH set point following an acute and intense episode of exercise (9), our findings suggest that with regular participation in more PA, this exercise-induced parathyroid “hypoactivity” may be one potential explanation for a lower future risk of developing P-HPTH. Indeed, our subset analyses focused on the correlation between PTH levels and PA suggested that greater PA may be associated with lower PTH levels, as previously reported (22). Our findings are also better appreciated in the context of our own prior prospective studies that have identified other independent risk factors (lower calcium intake and hypertension) for developing P-HPTH. We previously showed that lower calcium intake (<800 mg/day) was associated with an approximately 30–60% higher risk of developing P-HPTH when compared with higher calcium intake (>1000 mg/day) (13), and having hypertension was associated with an approximately 50% higher risk of developing P-HPTH when compared with not having hypertension (1). When woven together, our current analyses suggest that individuals with all three of these risk factors may have more than a 4-fold higher risk of developing P-HPTH when compared with individuals with none of these risk factors. These mentioned studies, in conjunction with the current one, are the only prospective studies to date to evaluate nongenetic nonmolecular determinants of P-HPTH and postulate that maintaining a combination of a physically active lifestyle, in addition to intake of sufficient calcium and normal blood pressure, may substantially reduce the risk of developing P-HPTH. Prospective intervention studies are needed to confirm this hypothesis, and whether modification of any of these risk factors could subsequently alter the future risk of developing P-HPTH.
Limitations
Our study has limitations. First, our study population was female and almost entirely white and therefore our findings are not necessarily generalizable to men or other races; however, a recent large study indicated that the demographic of our study population represents most at-risk individuals for P-HPTH (2). Second, our research design could not elucidate the molecular mechanism underlying our findings; however, our findings do represent an extension of prior research suggesting decreased parathyroid stimulation following exercise (6–10, 22). Third, we were unable to exclude selection bias because we only included cases confirmed by medical record review and could not obtain medical records for all women who self-reported P-HPTH. We also used very conservative criteria to define P-HPTH. Milder cases of P-HPTH would not have been classified as cases using our strict definition; however, this could have attenuated our results by biasing toward the null hypothesis. Fourth, many cases of P-HPTH may be asymptomatic and detected by routine bloodwork, which may be more frequent in women who pursue regular preventive care; however, we adjusted our analyses for regular physical examinations to account for this. Fifth, we did not have biochemical measurements on the entire cohort of 69 621 participants. Specifically, we did not have measurements of 25(OH)D levels on every participant, but we did adjust our models for factors that have been shown to predict 25(OH)D levels in NHS participants, including dietary and supplemental vitamin D intake based on validated food frequency questionnaires (17, 18, 20), age, race, BMI, alcohol intake, and UVB exposure (20). We were also able to conduct a subset analysis in those with available biochemical parameters that supported the general implications of our findings.
Conclusions
In this large prospective cohort study of women, we observed that participation in higher weekly PA was associated with a significantly lower risk of developing P-HPTH when compared with a more sedentary lifestyle. Our findings extend prior studies demonstrating a reduction in PTH secretion following acute and chronic exercise (9, 22). This association between sedentary behavior and development of P-HPTH was independent of other potential risk factors for P-HPTH and may be further amplified in individuals who also consume less calcium and who have hypertension. Because low PA is a potentially modifiable risk factor, future studies are needed to evaluate whether increasing PA can prevent the incidence of, or even mitigate the severity of, primary hyperparathyroidism.
Acknowledgments
We thank Dr. Henry Kronenberg from the Endocrine Unit at Massachusetts General Hospital for playing an important role in designing our study and his review of this manuscript.
Funding was provided by National Institutes of Health (NIH) Grants UM1-CA186107, CA087969, and DK099739; the National Heart, Lung, And Blood Institute of the NIH under Award K23HL111771, the National Institutes of Diabetes and Digestive and Kidney Disease (NIDDK) of the NIH under Award R01 DK107407, and by Grant 2015085 from the Doris Duke Charitable Foundation (to A.V.); and by the NIDDK of the NIH Award K24DK091417 (to. G.C.) and Award K23DK100447 (to J.M.P.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Disclosure Summary: The authors have nothing to disclose.
Footnotes
- BMI
- body mass index
- CHD
- coronary heart disease
- CV
- coefficient of variation
- MET
- metabolic equivalent
- NHS
- Nurses' Health Study
- 25(OH)D
- 25-hydroxyvitamin D
- P-HPTH
- primary hyperparathyroidism
- PA
- physical activity.
References
- 1. Vaidya A, Paik JM, Taylor EN, Curhan GC. Hypertension, anti-hypertensive medication use, and the risk for incident primary hyperparathyroidism. J Clin Endocrinol Metab. 2015;100:2396–2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Yeh MW, Ituarte PH, Zhou HC, et al. Incidence and prevalence of primary hyperparathyroidism in a racially mixed population. J Clin Endocrinol Metab. 2013;98:1122–1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Feskanich D, Willett W, Colditz G. Walking and leisure-time activity and risk of hip fracture in postmenopausal women. JAMA. 2002;288:2300–2306. [DOI] [PubMed] [Google Scholar]
- 4. Reddy KS, Hunter DJ. Noncommunicable diseases. N Engl J Med. 2013;369:2563. [DOI] [PubMed] [Google Scholar]
- 5. Borer KT. Physical activity in the prevention and amelioration of osteoporosis in women: interaction of mechanical, hormonal and dietary factors. Sports Med. 2005;35:779–830. [DOI] [PubMed] [Google Scholar]
- 6. Barry DW, Kohrt WM. Acute effects of 2 hours of moderate-intensity cycling on serum parathyroid hormone and calcium. Calcif Tissue Int. 2007;80:359–365. [DOI] [PubMed] [Google Scholar]
- 7. Scott JP, Sale C, Greeves JP, Casey A, Dutton J, Fraser WD. The effect of training status on the metabolic response of bone to an acute bout of exhaustive treadmill running. J Clin Endocrinol Metab. 2010;95:3918–3925. [DOI] [PubMed] [Google Scholar]
- 8. Scott JP, Sale C, Greeves JP, Casey A, Dutton J, Fraser WD. The role of exercise intensity in the bone metabolic response to an acute bout of weight-bearing exercise. J Appl Physiol (1985). 2011;110:423–432. [DOI] [PubMed] [Google Scholar]
- 9. Scott JP, Sale C, Greeves JP, Casey A, Dutton J, Fraser WD. Treadmill running reduces parathyroid hormone concentrations during recovery compared with a nonexercising control group. J Clin Endocrinol Metab. 2014;99:1774–1782. [DOI] [PubMed] [Google Scholar]
- 10. Vainionpaa A, Korpelainen R, Vaananen HK, Haapalahti J, Jamsa T, Leppaluoto J. Effect of impact exercise on bone metabolism. Osteoporos Int. 2009;20:1725–1733. [DOI] [PubMed] [Google Scholar]
- 11. Wolin KY, Lee IM, Colditz GA, Glynn RJ, Fuchs C, Giovannucci E. Leisure-time physical activity patterns and risk of colon cancer in women. Int J Cancer. 2007;121:2776–2781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ainsworth BE, Haskell WL, Leon AS, et al. Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc. 1993;25:71–80. [DOI] [PubMed] [Google Scholar]
- 13. Paik JM, Curhan GC, Taylor EN. Calcium intake and risk of primary hyperparathyroidism in women: prospective cohort study. BMJ. 2012;345:e6390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Rimm EB, Stampfer MJ, Colditz GA, Chute CG, Litin LB, Willett WC. Validity of self-reported waist and hip circumferences in men and women. Epidemiology. 1990;1:466–473. [DOI] [PubMed] [Google Scholar]
- 15. Shai I, Stampfer MJ, Ma J, et al. Homocysteine as a risk factor for coronary heart diseases and its association with inflammatory biomarkers, lipids and dietary factors. Atherosclerosis. 2004;177:375–381. [DOI] [PubMed] [Google Scholar]
- 16. Wolf AM, Hunter DJ, Colditz GA, et al. Reproducibility and validity of a self-administered physical activity questionnaire. Int J Epidemiol. 1994;23:991–999. [DOI] [PubMed] [Google Scholar]
- 17. Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;18:858–867. [DOI] [PubMed] [Google Scholar]
- 18. Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122:51–65. [DOI] [PubMed] [Google Scholar]
- 19. Colditz GA, Stampfer MJ, Willett WC, et al. Reproducibility and validity of self-reported menopausal status in a prospective cohort study. Am J Epidemiol. 1987;126:319–325. [DOI] [PubMed] [Google Scholar]
- 20. Bertrand KA, Giovannucci E, Liu Y, et al. Determinants of plasma 25-hydroxyvitamin D and development of prediction models in three US cohorts. Br J Nutr. 2012;108:1889–1896. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Brown JM, de Boer IH, Robinson-Cohen C, et al. Aldosterone, parathyroid hormone, and the use of renin-angiotensin-aldosterone system inhibitors: the multi-ethnic study of atherosclerosis. J Clin Endocrinol Metab. 2015;100:490–499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Berger C, Greene-Finestone LS, Langsetmo L, et al. Temporal trends and determinants of longitudinal change in 25-hydroxyvitamin D and parathyroid hormone levels. J Bone Miner Res. 2012;27:1381–1389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Bilezikian JP. Primary hyperparathyroidism. In: De Groot LJ, Beck-Peccoz P, Chrousos G, et al., eds. Endotext. South Dartmouth, MA: MDText.com, Inc., 2012. [Google Scholar]
- 24. Bilezikian JP, Brandi ML, Eastell R, et al. Guidelines for the management of asymptomatic primary hyperparathyroidism: summary statement from the Fourth International Workshop. J Clin Endocrinol Metab. 2014;99:3561–3569. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Bilezikian JP, Silverberg SJ. Clinical practice. Asymptomatic primary hyperparathyroidism. N Engl J Med. 2004;350:1746–1751. [DOI] [PubMed] [Google Scholar]
- 26. Fraser WD. Hyperparathyroidism. Lancet. 2009;374:145–158. [DOI] [PubMed] [Google Scholar]