Introduction.
Whilst antiretroviral therapy (ART) has transformed the prognosis of HIV-1 infection, metabolic consequences such as osteoporosis, of both infection and its therapy, have become more common. A meta-analysis found that the prevalence of osteoporosis in HIV-infected subjects was over three times higher than in HIV-uninfected persons. [1] The exact mechanisms are unclear, but a combination of factors are thought to contribute to decreased bone mineral density in HIV-infected people including lifestyle (e.g. diet, exercise, smoking and alcohol), genetic influences, as well as direct effects of HIV infection [2] and administration of antiretrovirals, such as Tenofovir disoproxil fumarate (TDF) [1], [3], [4], protease inhibitors (PIs) [1] and zidovudine. [5]

Administration of TDF has also been associated with renal phosphate wasting, [6] osteomalacia [7], [8] and more recently, increased parathyroid hormone (PTH) levels [9]–[11]. The mechanism behind the potential alteration of calcium and phosphate homeostasis by TDF is not well understood. Additionally, it remains unclear whether renal phosphate wasting is related to either reduced bone mineral density and/or hyperparathyroidism in patients receiving TDF.

The aim of this study therefore was to examine the effect of tenofovir on PTH, calcium and phosphate homeostasis, vitamin D metabolism and bone mineral density by comparing patients receiving TDF with patients not receiving TDF. Furthermore, we aimed to investigate whether these effects are modified by ethnicity, age and sex.

Abstract.
Background
Tenofovir has been associated with renal phosphate wasting, reduced bone mineral density, and higher parathyroid hormone levels. The aim of this study was to carry out a detailed comparison of the effects of tenofovir versus non-tenofovir use on calcium, phosphate and, vitamin D, parathyroid hormone (PTH), and bone mineral density.

Methods
A cohort study of 56 HIV-1 infected adults at a single centre in the UK on stable antiretroviral regimes comparing biochemical and bone mineral density parameters between patients receiving either tenofovir or another nucleoside reverse transcriptase inhibitor.

Principal Findings
In the unadjusted analysis, there was no significant difference between the two groups in PTH levels (tenofovir mean 5.9 pmol/L, 95% confidence intervals 5.0 to 6.8, versus non-tenofovir; 5.9, 4.9 to 6.9; p = 0.98). Patients on tenofovir had significantly reduced urinary calcium excretion (median 3.01 mmol/24 hours) compared to non-tenofovir users (4.56; p<0.0001). Stratification of the analysis by age and ethnicity revealed that non-white men but not women, on tenofovir had higher PTH levels than non-white men not on tenofovir (mean difference 3.1 pmol/L, 95% CI 5.3 to 0.9; p = 0.007). Those patients with optimal 25-hydroxyvitamin D (>75 nmol/L) on tenofovir had higher 1,25-dihydroxyvitamin D [1,25(OH)2D] (median 48 pg/mL versus 31; p = 0.012), fractional excretion of phosphate (median 26.1%, versus 14.6; p = 0.025) and lower serum phosphate (median 0.79 mmol/L versus 1.02; p = 0.040) than those not taking tenofovir.

Conclusions
The effects of tenofovir on PTH levels were modified by sex and ethnicity in this cohort. Vitamin D status also modified the effects of tenofovir on serum concentrations of 1,25(OH)2D and phosphate.