Relative contribution of HIV infection, demographics and body mass index to bone mineral density Aoife G. Cottera,b, Caroline A. Sabinc, Sibongile Simelanea, Alan Mackena, Eoin Kavanaghd, Jennifer J. Bradye, Geraldine McCarthyf, Juliet Compstong, Patrick W.G. Mallona,b, on behalf of the HIV UPBEAT Study Group Introduction: Low bone mineral density (BMD) is common in HIV-positive patients, although the role played by HIV infection versus sociodemographic and metabolic factors remains unclear. Methods: Understanding the Pathology of Bone Disease in HIV-infected individuals (HIV UPBEAT) is a prospective cohort study, enrolled HIV-positive and HIV-negative participants from similar demographic backgrounds. Dual X-ray absorptiometry at femoral neck, total hip and lumbar spine and blood tests were performed. Associations between BMD and factors of interest were assessed using multivariable linear regression. Results: A total of 474 participants were recruited. Two hundred and ten were HIVpositive, of whom, 59% were male, 40% African and median (interquartile range) age was 39 (33, 46) years. HIV acquisition risks were heterosexual sex (46.9%), homosexual sex (25.4%) and intravenous drug use (18.7%). Of the HIV-negative participants, 44% were male, 25% were African and median (interquartile range) age was 42 (34–49) years. HIV infection was independently associated with a 0.062 (P < 0.0001), 0.078 (P < 0.0001) and 0.060 g/cm2 (P ¼ 0.0002) lower BMD at femoral neck, total hip and lumbar spine, respectively, after adjustment for demographic/ lifestyle factors and BMI. After further adjustment for bone biomarkers, HIV remained independently associated with reduced BMD at each site, although effect sizes were reduced. The HIVpositive group had significantly higher bone turnover (all between-group P < 0.0001). Treatment variables and cumulative exposure to antiretroviral therapy were not associated with lower BMD at femoral neck or total hip, but acquisition of HIV infection via intravenous drug use and longer time since HIV diagnosis were independently associated with lower lumbar spine BMD. Discussion: HIV is independently associated with lower BMD, and its effect is likely mediated, in part, by alterations in bone metabolism. ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
AIDS 2014, 28:2051–2060 Keywords: BMI, bone disease, bone mineral density, bone turnover, HIV
a HIV Molecular Research Group, School of Medicine and Medical Science, University College Dublin, bDepartment of Infectious Diseases, Mater Misericordiae University Hospital, Dublin, Ireland, cResearch Department of Infection & Population Health, UCL, Royal Free Campus, London, UK, dDepartment of Radiology, eDepartment of Clinical Chemistry and Diagnostic Endocrinology, f Department of Rheumatology, Mater Misericordiae University Hospital, Dublin, Ireland, and gDepartment of Medicine, School of Clinical Medicine, Addenbrooke’s NHS Trust, University of Cambridge, UK. Correspondence to Aoife G. Cotter, MB, BCh, BAO, BMedSc, PhD, Clinical Research Centre, Mater Misericordiae University Hospital, 21 Nelson St, Dublin 7, Ireland. Tel: +353 1 716 4584; fax: +353 1 716 4539; e-mail: [email protected]; [email protected] Received: 10 April 2014; revised: 17 May 2014; accepted: 19 May 2014.
DOI:10.1097/QAD.0000000000000353
ISSN 0269-9370 Q 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
2051
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
2052
AIDS
2014, Vol 28 No 14
Introduction With an ageing population of people living with HIV infection comes the potential for increased prevalence of age-related illnesses secondary to the effects of chronic HIV infection or its treatment with antiretroviral therapy (ART) [1]. Low bone mineral density (BMD) and reductions in BMD have been associated with HIV in cross-sectional studies and longitudinal cohorts of specific populations [2–10]. The relative contribution of HIV infection to the development of low BMD remains unclear because of a lack of adequately controlled prospective studies. Based on the data from a meta-analysis of cross-sectional studies, the prevalence of low BMD in the HIV-positive population is estimated to be 67%, with those on ART at increased risk [11]. A number of studies have demonstrated clinically relevant reductions in BMD occurring with initiation of both first-line and second-line ART, largely limited to the first 48 weeks of therapy, with greater potential for loss of BMD with ART containing the nucleotide reverse transcriptase inhibitor tenofovir disoproxil fumarate (TDF) or some protease inhibitors [12–15]. Low BMI has also been identified as a risk factor for low BMD with some studies determining that the increased risk in those with HIV infection is driven by lower body weight or BMI rather than HIV infection itself [16]. Bone undergoes continuous remodelling via the coupled processes of bone formation and resorption. Levels of circulating bone biomarkers reflect these processes and are specific for either resorption or formation. C-terminal cross-linking telopeptide of type 1 collagen (CTX-1) is a product of collagen breakdown and correlates with bone resorption. Osteocalcin is synthesized and secreted by osteoblasts and correlates with osteoblast function and bone formation. Procollagen type 1 N-terminal propeptide (P1NP) is a marker of bone formation. As collagen is synthesized, P1NP is cleaved from procollagen and released into circulation. If the balance between formation and resorption is disturbed, a net reduction in bone mass may result. In adults, high bone turnover, reflected in increased bone biomarkers, is usually associated with net bone loss, as bone formation occurs more slowly than bone resorption [17–19]. We aimed to determine the relative contribution of HIV infection to low BMD after adjustment for demographic, socioeconomic and laboratory variables. Furthermore, we aimed to describe markers of bone metabolism in the HIV-positive group compared with the HIV-negative group and to determine whether the effect of HIV on BMD is mediated through changes in bone turnover. Finally, we aimed to investigate the independent effects of HIV acquisition risk group and exposure to ART on BMD. Our hypotheses are that low BMD is more prevalent in HIV-positive patients after adjusting for
differences in age, sex, smoking, socioeconomic status and BMI and that treatment factors are associated with lower BMD.
Methods Understanding the pathology of bone disease in HIVinfected individual (HIV UPBEAT) is a prospective, observational cohort study that enrolled HIV-positive and HIV-negative participants from similar demographic backgrounds. HIV-positive participants were recruited from the Infectious Diseases clinics at the Mater Misericordiae University Hospital, Dublin, Ireland, whereas HIV-negative participants were recruited from the local area. Participants over 18 years of age were eligible for inclusion. There were no exclusion criteria. All participants gave written, informed consent and approval was granted by the local research ethics committee. Baseline study visits occurred between February 2011 and July 2012 and included collection of demographic, clinical and treatment information. Change in any component of ART regimen was considered a new regimen. Socioeconomic status (current employment, income compared to the average industrial wage), any history of fracture and falls history were collected. Fasting blood tests including full blood count, urea and electrolytes, liver function tests [bilirubin, alkaline phosphatase (ALP), alanine aminotransferase (ALT), gamma-glutamyltransferase (GGT), albumin], bone profile (corrected calcium, phosphate), lipid profile, glucose, thyroid function tests, were analysed by routine methodologies. Estimated glomerular filtration rate (eGFR) was calculated using the abbreviated Modification of Diet in Renal Disease equation [20]. HIV and hepatitis B and C serology were assessed in the HIV-negative group. Intact parathyroid hormone (PTH) was measured using sandwich immunoassay with chemiluminescent detection (Abbott, Abbott Park, Illinois, USA); serum and plasma samples were stored at 808C. Serum 25 hydroxy vitamin D (OH)D was measured by liquid chromatography mass spectrometry. After baseline visit completion, stored plasma samples were utilized for bone biomarker analyses. CTX-1, osteocalcin [N-mid osteocalcin (1–43)] and P1NP were measured using electrochemiluminescence immunoassays (Roche Diagnostics, Basel, Switzerland). Coefficients of variation for all assays were within acceptable limits. All participants were scanned on a Lunar Prodigy Dual-energy X-ray absorptiometry scanner (GE Medical Systems, Madison, Wisconsin, USA) by a single, trained operator (A.G.C.) as per manufacturer’s recommendations. Total body scan (total body densitometry, body composition), dual femur (femoral neck, total hip) and lumbar spine (lumbar spine) scans were obtained using a standardized protocol. Encore software version 12.10 (GE Healthcare, Madison, Wisconsin, USA) was used for acquisition and analysis.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
HIV infection, demographics and body mass index Cotter et al.
Low BMD was defined as T-score less than 1.0 in those older than 40 years or Z less than 2.0 in those younger than 40 years, respectively. Differences between the HIV-positive and HIV-negative groups were assessed for significance using Wilcoxon Mann-Whitney (continuous variables) and chi-squared or Fisher’s exact tests (categorical variables), whereas associations between BMD and the clinical, demographic and laboratory parameters were assessed using Wilcoxon Mann-Whitney tests, analysis of variance, chi-squared tests and Spearman correlation where appropriate. Bone biomarkers were correlated using Spearman correlation. Univariate followed by multivariable linear regression was performed to identify variables independently associated with BMD at each site. Models were constructed to include relevant clinical or demographic variables in this order: HIV status (positive; negative), age (per 5-year increment), sex (male; female), ethnicity (African; non-African), smoking status (current smoker; noncurrent smoker) and education status (third level; primary/secondary level completed) (model 1). The additional effect of BMI was determined by adding BMI at this point (model 2) and finally laboratory variables that were significantly associated with BMD on univariate testing (P < 0.05) were added (model 3). A HIV-specific subanalysis was performed to identify independent associations between HIV-specific variables [current CD4þ T-cell count, nadir CD4þ T-cell count, HIV RNA
AIDS 2014, 28:2051–2060 Keywords: BMI, bone disease, bone mineral density, bone turnover, HIV
a HIV Molecular Research Group, School of Medicine and Medical Science, University College Dublin, bDepartment of Infectious Diseases, Mater Misericordiae University Hospital, Dublin, Ireland, cResearch Department of Infection & Population Health, UCL, Royal Free Campus, London, UK, dDepartment of Radiology, eDepartment of Clinical Chemistry and Diagnostic Endocrinology, f Department of Rheumatology, Mater Misericordiae University Hospital, Dublin, Ireland, and gDepartment of Medicine, School of Clinical Medicine, Addenbrooke’s NHS Trust, University of Cambridge, UK. Correspondence to Aoife G. Cotter, MB, BCh, BAO, BMedSc, PhD, Clinical Research Centre, Mater Misericordiae University Hospital, 21 Nelson St, Dublin 7, Ireland. Tel: +353 1 716 4584; fax: +353 1 716 4539; e-mail: [email protected]; [email protected] Received: 10 April 2014; revised: 17 May 2014; accepted: 19 May 2014.
DOI:10.1097/QAD.0000000000000353
ISSN 0269-9370 Q 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
2051
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
2052
AIDS
2014, Vol 28 No 14
Introduction With an ageing population of people living with HIV infection comes the potential for increased prevalence of age-related illnesses secondary to the effects of chronic HIV infection or its treatment with antiretroviral therapy (ART) [1]. Low bone mineral density (BMD) and reductions in BMD have been associated with HIV in cross-sectional studies and longitudinal cohorts of specific populations [2–10]. The relative contribution of HIV infection to the development of low BMD remains unclear because of a lack of adequately controlled prospective studies. Based on the data from a meta-analysis of cross-sectional studies, the prevalence of low BMD in the HIV-positive population is estimated to be 67%, with those on ART at increased risk [11]. A number of studies have demonstrated clinically relevant reductions in BMD occurring with initiation of both first-line and second-line ART, largely limited to the first 48 weeks of therapy, with greater potential for loss of BMD with ART containing the nucleotide reverse transcriptase inhibitor tenofovir disoproxil fumarate (TDF) or some protease inhibitors [12–15]. Low BMI has also been identified as a risk factor for low BMD with some studies determining that the increased risk in those with HIV infection is driven by lower body weight or BMI rather than HIV infection itself [16]. Bone undergoes continuous remodelling via the coupled processes of bone formation and resorption. Levels of circulating bone biomarkers reflect these processes and are specific for either resorption or formation. C-terminal cross-linking telopeptide of type 1 collagen (CTX-1) is a product of collagen breakdown and correlates with bone resorption. Osteocalcin is synthesized and secreted by osteoblasts and correlates with osteoblast function and bone formation. Procollagen type 1 N-terminal propeptide (P1NP) is a marker of bone formation. As collagen is synthesized, P1NP is cleaved from procollagen and released into circulation. If the balance between formation and resorption is disturbed, a net reduction in bone mass may result. In adults, high bone turnover, reflected in increased bone biomarkers, is usually associated with net bone loss, as bone formation occurs more slowly than bone resorption [17–19]. We aimed to determine the relative contribution of HIV infection to low BMD after adjustment for demographic, socioeconomic and laboratory variables. Furthermore, we aimed to describe markers of bone metabolism in the HIV-positive group compared with the HIV-negative group and to determine whether the effect of HIV on BMD is mediated through changes in bone turnover. Finally, we aimed to investigate the independent effects of HIV acquisition risk group and exposure to ART on BMD. Our hypotheses are that low BMD is more prevalent in HIV-positive patients after adjusting for
differences in age, sex, smoking, socioeconomic status and BMI and that treatment factors are associated with lower BMD.
Methods Understanding the pathology of bone disease in HIVinfected individual (HIV UPBEAT) is a prospective, observational cohort study that enrolled HIV-positive and HIV-negative participants from similar demographic backgrounds. HIV-positive participants were recruited from the Infectious Diseases clinics at the Mater Misericordiae University Hospital, Dublin, Ireland, whereas HIV-negative participants were recruited from the local area. Participants over 18 years of age were eligible for inclusion. There were no exclusion criteria. All participants gave written, informed consent and approval was granted by the local research ethics committee. Baseline study visits occurred between February 2011 and July 2012 and included collection of demographic, clinical and treatment information. Change in any component of ART regimen was considered a new regimen. Socioeconomic status (current employment, income compared to the average industrial wage), any history of fracture and falls history were collected. Fasting blood tests including full blood count, urea and electrolytes, liver function tests [bilirubin, alkaline phosphatase (ALP), alanine aminotransferase (ALT), gamma-glutamyltransferase (GGT), albumin], bone profile (corrected calcium, phosphate), lipid profile, glucose, thyroid function tests, were analysed by routine methodologies. Estimated glomerular filtration rate (eGFR) was calculated using the abbreviated Modification of Diet in Renal Disease equation [20]. HIV and hepatitis B and C serology were assessed in the HIV-negative group. Intact parathyroid hormone (PTH) was measured using sandwich immunoassay with chemiluminescent detection (Abbott, Abbott Park, Illinois, USA); serum and plasma samples were stored at 808C. Serum 25 hydroxy vitamin D (OH)D was measured by liquid chromatography mass spectrometry. After baseline visit completion, stored plasma samples were utilized for bone biomarker analyses. CTX-1, osteocalcin [N-mid osteocalcin (1–43)] and P1NP were measured using electrochemiluminescence immunoassays (Roche Diagnostics, Basel, Switzerland). Coefficients of variation for all assays were within acceptable limits. All participants were scanned on a Lunar Prodigy Dual-energy X-ray absorptiometry scanner (GE Medical Systems, Madison, Wisconsin, USA) by a single, trained operator (A.G.C.) as per manufacturer’s recommendations. Total body scan (total body densitometry, body composition), dual femur (femoral neck, total hip) and lumbar spine (lumbar spine) scans were obtained using a standardized protocol. Encore software version 12.10 (GE Healthcare, Madison, Wisconsin, USA) was used for acquisition and analysis.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
HIV infection, demographics and body mass index Cotter et al.
Low BMD was defined as T-score less than 1.0 in those older than 40 years or Z less than 2.0 in those younger than 40 years, respectively. Differences between the HIV-positive and HIV-negative groups were assessed for significance using Wilcoxon Mann-Whitney (continuous variables) and chi-squared or Fisher’s exact tests (categorical variables), whereas associations between BMD and the clinical, demographic and laboratory parameters were assessed using Wilcoxon Mann-Whitney tests, analysis of variance, chi-squared tests and Spearman correlation where appropriate. Bone biomarkers were correlated using Spearman correlation. Univariate followed by multivariable linear regression was performed to identify variables independently associated with BMD at each site. Models were constructed to include relevant clinical or demographic variables in this order: HIV status (positive; negative), age (per 5-year increment), sex (male; female), ethnicity (African; non-African), smoking status (current smoker; noncurrent smoker) and education status (third level; primary/secondary level completed) (model 1). The additional effect of BMI was determined by adding BMI at this point (model 2) and finally laboratory variables that were significantly associated with BMD on univariate testing (P < 0.05) were added (model 3). A HIV-specific subanalysis was performed to identify independent associations between HIV-specific variables [current CD4þ T-cell count, nadir CD4þ T-cell count, HIV RNA
