Prevalence and Risk Factors for Loss of Bone Mineral Density in Male Japanese Patients With HIV

To the Editors: As the mortality of HIV-infected individuals was improved by combination antiretroviral therapy (cART), opportunistic infections were replaced by long-term complications including loss of bone mineral density (BMD).1–9 It was first reported in 1999 and has emerged in the last decade10,11 and has probably placed HIV-infected individuals at higher risk of bone fractures.12,13 The causes are considered to be multifactorial and include HIV infection itself, body mass index (BMI), estimated glomerular filtration rate (eGFR), testosterone level, and hepatitis coinfection.14–17 Some articles link the use of protease inhibitors and the tenofovir (TDF) with lower BMD.18–20 Most of the articles associated with BMD loss in HIV-infected patients have been reported from North America, Europe, and Oceania. Accordingly, the BMD research on HIV-infected Asians is underrepresented. In this work, we focused on male Japanese HIV-infected patients and studied the prevalence and severity of BMD loss among them and attempted to determine clinical factors related to BMD loss. Forty male Japanese HIV-infected individuals aged from 21 to 70 years who visited the Teikyo University Hospital were enrolled in this study. This study was approved by the ethical committee of the Teikyo University School of Medicine, and the patients gave written informed consent. The patients underwent BMD analyses from March 2010 to February 2012 with the same dual energy X-ray absorptiometry scan (Discovery SI, Hologic Inc, Bedford, MA). Thirty-nine patients were analyzed in both the lumbar spines and bilateral femoral necks, whereas 1 patient was limited to the bilateral femoral necks. The BMD data of the lumbar spines and the smaller BMD value of the femoral necks were employed to calculate T-scores that were calculated as comparison with young normal reference value expressed as SD units. Osteopenia was defined as a T-score of between −1 and −2.5SD, and osteoporosis was defined as that of ≥ −2.5 negative score following the World Health Organization classification.21 Clinical factors were collected including age, height, body weight, BMI, smoking status, use of corticosteroid, CD4 cell counts, durations of cART, medication of nucleoside/nucleotide reverse transcriptase inhibitors, serum creatinine, bone alkaline phosphatase (BAP), eGFR, serum, and urine N-terminal telopeptide (NTx). The eGFR for Japanese male patients were calculated by 194 × serum creatinine−1.094 × age−0.287 (mL/min per 1.73 m2).22 To determine the high-risk groups, statistical differences of T-scores were calculated by using Wilcoxon signed ranks or Kruskal–Wallis test with the post hoc comparisons by the Dunnett test. The correlations between T-scores and each variable were examined by a linear regression employing the method of least squares. A multiple regression analysis was conducted by employing these variables with a P value <0.1 to predict BMD loss. All the analyses were performed by JMP version 8.0.1 (SAS Institute Inc, United States). The median age of the patients was 39 years. Their median height and body weight were 169.3 cm and 65.0 kg, respectively. Twenty-three patients were ≤170 cm, and 14 were <60 kg. The median CD4 count was 396/μL. Twenty-seven were on cART, and median duration of cART was 2.9 years. Five have been continued on cART for >10 years. Twelve out of 27 took TDF/emtricitabine (FTC), and 13 took abacavir/lamivudine (3TC). The median and interquartile ranges of BMD and T-scores in the lumbar spines and the femoral necks are summarized in Table 1. The median T-scores in the lumbar spines and the femoral necks were −0.8 and −1.2, respectively. The T-scores of the femoral necks were significantly lower than those of the lumbar spines (P = 0.0461). By employing the T-scores in the lumbar spines, the prevalence of osteopenia and osteoporosis was found to be 43.6% and 5.1%, respectively, and these percentages rose to 47.5% and 7.5%, respectively, if the least T-scores in the femoral necks were used. Moreover, the number (and percentage) of patients with osteopenia and osteoporosis were 21 (52.5%) and 4 (10%), respectively, in at least 1 of the T-score measured in the lumbar spines or in the femoral necks.TABLE 1: The Prevalence of BMD LossSubgroup analyses of patients' baseline characteristics revealed that the group of patients aged 50 years or older was the only group that showed significantly lower T-scores of the lumbar spines. In contrast, high-risk groups that showed significantly lower T-scores of the femoral necks were patients aged 40 years or older, ≤170 cm in height, weighed ≤60 kg, those on cART, particularly those continuing cART for ≥10 years. There were no significant differences between the 2 groups divided based on the medication of nucleotide reverse transcriptase inhibitors or use of corticosteroid. Correlation between BMD loss and clinical factors was analyzed by univariate analyses. Age, height, body weight, and duration of cART showed correlation with T-scores of both the lumbar spines and the femoral necks. In addition, serum creatinine, BAP, and eGFR were related to BMD loss measured in the femoral necks. Urinary phosphate excretion showed a tendency of correlation to T-scores in the lumbar spines (P = 0.0511). Serum NTx, 25-hydroxyvitamin D, cystatin C, free testosterone, urinary NTx, and smoking status were not correlated to BMD loss. Those variables that showed correlation or tendency of correlation were employed in a multiple regression analysis. Age, BAP, and eGFR were determined as independent clinical factors of T-scores of the femoral necks, whereas urinary phosphate excretion was the only independent variable to the T-score of the lumbar spines (Table 2).TABLE 2: Independent Risk Factors of BMD LossOur study revealed that male Japanese HIV-infected patients are highly concomitant with low BMD. More than 60% of individuals were diagnosed with either osteopenia or osteoporosis. This is the first report from Asia, and the prevalence of low BMD in male Japanese HIV-infected patients is as high as that of previous reports from Europe, North America, or Oceania.11 Increased age, short stature, and low body weight indicated from our study are well-known risk factors for BMD loss.23 However, the age of these patients suffering from BMD loss is much younger, and the prevalence of BMD loss is significantly higher than that in the general population. It is reported previously that BMD in both the lumbar spines and the femoral necks decreases almost simultaneously in the general population of Japanese males.24 In our study, the prevalence of osteopenia and osteoporosis diagnosed by the BMD in the femoral necks is higher than that in the lumbar spines. A meta-analysis by Bolland et al25 showed consistent evidence of BMD loss especially among antiretroviral drug naive HIV-infected patients and that larger BMD loss were observed in the femoral necks than the lumbar spines. Osteoblast and osteoclast functions are influenced by a number of factors modulated during HIV infection and HIV itself or with other concomitant factors stimulate the progression of BMD loss.26,27 It is possible that progression of BMD loss accelerated by HIV infection proceeds primarily targeting on cortical bone tissues such as the femoral necks. A multiple regression analysis revealed that age, BAP, and eGFR are indicated as independent risk factors for BMD loss in the femoral necks. Decreased renal function is one of the risk factor for BMD loss among the general population. However, serum creatinine and eGFR remain within the normal range in many of our patients with BMD loss. We employed serum and urine NTx as markers for bone resorption and BAP as a marker for bone formation. Madeddu et al28 reported that Italian HIV-infected patients with BMD loss in the lumbar spines had a higher mean BAP level, although it was not statistically significant in men. BAP might be a predictor of BMD loss in patients with HIV, implying an accelerated bone turnover. Interestingly, urinary phosphate excretion was determined as an independent variable of T-scores of the lumbar spines by a multiple regression analysis (P = 0.0110). Urinary phosphate excretion is mainly regulated by NaPi cotransporters that express in the brush border membrane of proximal tubular epithelia.29 Although the mechanisms are unclear, it is hypothesized that HIV might have some effects on these transporters to stimulate excretion of phosphate. It has been investigated by both cross-sectional and longitudinal studies whether or not cART adds to the burden of metabolic bone disease in HIV infection with mixed conclusions.20 In our study, no significant differences were observed in BMD loss between TDF/FTC group and abacavir/3TC group. In conclusion, this is the first report on BMD loss in Asian HIV-infected patients. More than 60% of the patients are diagnosed as having osteopenia or osteoporosis. A multivariate regression analysis indicated that age, BAP, and eGFR are independent factors of BMD loss in the femoral necks, whereas urinary phosphate excretion is that of BMD loss of the lumbar spines. DXA scan examination not only in the lumbar spines but also in the femoral necks are recommended in patient more than 40 years old or those with increased BAP, decreased eGFR, or decreased urinary phosphate excretion. ACKNOWLEDGMENT The authors thank Dr. Greg J. German, University of Ottawa, for his time and efforts in proofreading this article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.