Lower Parathyroid Hormone Levels are Associated With Reduced Fracture Risk in Japanese Patients on Hemodialysis

Abstract

Introduction: Secondary hyperparathyroidism (SHPT) affects bone metabolism and may lead to bone fragility. However, there is conflicting evidence as to whether parathyroid hormone (PTH) levels are associated with fracture risk and whether the relationship is linear or U-shaped.

Methods: We examined the association between PTH levels and the risk of any fracture and site-specific fractures in a nationwide cohort of 180,333 patients on hemodialysis. We also examined the association between the percent change in PTH levels during the preceding 1 year and subsequent fracture.

Results: At baseline, the median intact PTH level was 141 pg/ml (interquartile range, 78-226 pg/ml). During 1 year of follow-up, there were a total of 3762 fractures requiring hospitalization (1361 hip, 551 vertebral, and 1850 other). In an adjusted analysis, higher baseline PTH levels were associated with an incrementally increased risk of any fracture (odds ratio [OR] per doubling of intact PTH, 1.06; 95% confidence interval, 1.03-1.09). The association between PTH levels and fracture risk was more pronounced for hip fractures but not found for vertebral fractures. The absolute risk difference associated with higher PTH levels appeared to be more pronounced in older individuals, females, and those with lower body mass index (BMI). Change in PTH levels was also associated with fracture risk: the adjusted OR for fracture decreased linearly with decreasing PTH levels over 1 year, regardless of the preceding PTH levels.

Conclusion: Lower PTH levels are associated with a graded reduction in fracture risk. Further studies are needed to determine whether intensive PTH control reduces fracture risk.

Keywords: fracture; hemodialysis; parathyroid hormone; secondary hyperparathyroidism.

Graphical abstract

Figure 1

Schematic illustration of the study design. The primary analysis examined the association between PTH levels at the end of 2016 and fracture events during 2017, adjusted for covariates at the end of 2016. Secondary analyses examined the association between 1-year and 2-year change in PTH levels between 2015 and 2016 and between 2014 and 2016, respectively, and fracture events during 2017. For adjustment, we used covariates at the time of the first PTH measurement (e.g., in the 1-year change analysis, covariates at the end of 2015 were used for adjustment). PTH, parathyroid hormone.

Figure 2

Study profile. PTH, parathyroid hormone.

Figure 3

Restricted cubic splines of the adjusted odds ratio for any fracture (left), hip fracture (middle), and vertebral fracture (right) according to intact parathyroid hormone (PTH). Models adjusted for age, sex, dialysis duration, cause of kidney failure, dialysis modality (hemodialysis or hemodiafiltration), body mass index, Kt/V, normalized protein catabolic rate, history of cardiovascular disease (myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation), history of hip fracture, hemoglobin, albumin, creatinine, calcium, phosphorus, total cholesterol, and C-reactive protein. The black solid line represents the odds ratio, and the gray area represents the 95% confidence interval.

Figure 4

Subgroup analysis of fracture risk associated with intact parathyroid hormone (PTH). Models adjusted for age, sex, dialysis duration, cause of kidney failure, dialysis modality (hemodialysis or hemodiafiltration), body mass index, Kt/V, normalized protein catabolic rate, history of cardiovascular disease (myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation), history of hip fracture, hemoglobin, albumin, creatinine, calcium, phosphorus, total cholesterol, and C-reactive protein. Squares represent point estimates of the odds ratio, and horizontal lines indicate 95% confidence intervals (CIs). DM, diabetes mellitus.

Figure 5

Predicted probability of any fracture (left), hip fracture (middle), and vertebral fracture (right) across intact parathyroid hormone (PTH) levels, stratified by age (upper), sex (middle), and body mass index (BMI). Models adjusted for age, sex, dialysis duration, cause of kidney failure, dialysis modality (hemodialysis or hemodiafiltration), BMI, Kt/V, normalized protein catabolic rate, history of cardiovascular disease (myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation), history of hip fracture, hemoglobin, albumin, creatinine, calcium, phosphorus, total cholesterol, and C-reactive protein. The dark solid line represents the predicted probability of fracture, and the light-colored area represents the 95% confidence interval.

Figure 6

Restricted cubic splines of the adjusted odds ratio for any fracture (left), hip fracture (middle), and vertebral fracture (right) according to 1-year fold change in intact parathyroid hormone (PTH). Models adjusted for age, sex, dialysis duration, cause of kidney failure, dialysis modality (hemodialysis or hemodiafiltration), body mass index, Kt/V, normalized protein catabolic rate, history of cardiovascular disease (myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation), history of hip fracture, hemoglobin, albumin, creatinine, calcium, phosphorus, intact PTH, total cholesterol, and C-reactive protein. All covariates used for adjustment are at the time of the first intact PTH measurement. The black solid line represents the odds ratio, and the gray area represents the 95% confidence interval.