Adequate attainment of peak bone mass by young adulthood decreases the risk of osteoporosis and fragility fractures later in life,1 and it can be optimized with adequate intakes of calcium, phosphorus, and vitamin D, sunlight exposure, and regular physical activity.1 Bone mineralization may be adversely affected by delays in growth and puberty, nutrient malabsorption, and chronic infections and inflammation, including HIV. Children with perinatally acquired HIV (PHIV) have lower bone mineral density (BMD) compared with healthy children2 and perinatally HIV-exposed uninfected (PHEU) children.3 There remains a paucity of information regarding factors underlying the observed low BMD in PHIV children.
It is likely that low vitamin D status, as reflected by low serum levels of 25-hydroxy vitamin D [25(OH)D] and suggested by elevated parathyroid hormone (PTH) levels, contributes to suboptimal bone accumulation in PHIV children. Studies in HIV-infected adults report low 25(OH)D in 13%–77% of participants4,5 and associations of low 25(OH)D with bone loss.6 Among the few studies conducted in HIV-infected children and adolescents, the prevalence of low 25(OH)D ranged from 21% to 39% in PHIV adolescents7–11 and ≥50% in adolescents behaviorally infected by HIV.12,13 In PHIV adolescents, lumbar spine (LS) BMD was lower in those with both low 25(OH)D and high PTH.11 In HIV-infected adults, 25(OH)D levels were lower among those initiating antiretroviral therapy (ART) with efavirenz (EFV) compared with those without initiating ART without EFV,14 and baseline PTH levels were higher in HIV-infected adolescents receiving ART with tenofovir (TDF) compared with those not receiving ART with TDF.12
Few studies have evaluated the relationship of low 25(OH)D and high PTH on bone outcomes in large cohorts of PHIV and PHEU children.11 The primary objectives of this cross-sectional study were to (1) evaluate associations of low 25(OH)D and high PTH concentrations with total body (TB) and spine BMD and bone mineral content (BMC) and (2) assess distributions of 25(OH)D and PTH by HIV status and exposure to EFV and TDF.
The Adolescent Master Protocol of the NIH-supported Pediatric HIV/AIDS Cohort Study (PHACS) evaluates outcomes of HIV infection and ART among PHIV compared with PHEU preadolescents and adolescents across 15 clinical sites. The protocol was approved by the Institutional Review Boards at each site and the Harvard T.H. Chan School of Public Health. Informed consent from the parent(s) or guardian(s) and assent from participants were obtained per local institutional review board guidelines.
Sociodemographic and Clinical History
Clinical and laboratory data, including sociodemographics and Tanner staging, were obtained as previously described.15 Tanner stage was ascertained by inspection of breasts and pubic hair for females and of genitalia and pubic hair for males at semiannual visits until Tanner stage 5 was reached. For boys and girls, the more advanced stage of the 2 respective pubertal components was used for classification if there was discordance between the examined body sites (eg, breast vs. pubic hair in females). Weight, height, and body mass index were expressed as z-scores.16 Season at the time of blood draw was categorized as winter, spring, summer, or fall based on solstice dates. Latitude was considered both as a categorical [sites classified as northern (≥39 degree) or southern (<39 degree)] and a continuous variable.
Because circulating levels of both 25(OH)D and BMD vary by race, we included genetic ancestry informative markers to adjust for residual confounding in addition to self-reported race and ethnicity. A panel of 41 single-nucleotide polymorphisms ancestry informative markers was used to determine continental ancestry, which was estimated by comparing each child's genotype to allele frequencies in a reference set of 3517 individuals.17 Reference populations were originally grouped into the 7 world regions: Europe, Africa, America, Central/South Asia, South/West Asia, East Asia, and Oceania. We determined the percentage of each region present within an individual, which totals 100%.18,19 Because of the small number of Asians in the study, we combined Europe, Central/South Asia, and South/West Asia (Europe/CSW Asia) by adding the regional ancestry percentage.
TB-BMD including head, and LS-BMD, TB-BMC including head (TB-BMC) and excluding head (TBLH-BMC), LS-BMC, percent body fat (%), and extremity lean mass (g) were measured on a Lunar or Hologic dual energy x-ray absorptiometry (DXA) scanner (General Electric Healthcare, Buckinghamshire, United Kingdom or Hologic, Inc, Bedford, MA) as previously described.3 A phantom was circulated to each clinical site to standardize results. Scans were sent to the Body Composition Analysis Center at the Tufts University School of Medicine for central analysis and standardization. Hologic scans were analyzed using Hologic QDR version 12.3 and APEX version 3.3. Lunar scans were analyzed using Prodigy Advance enCORE 2005 version 9 and enCORE 2011 version 13.6. All scans were analyzed by a single technician blinded to the HIV status of the participants. Bone age (BA) was assessed using x-rays of the left hand and wrist by a radiologist at each clinical site blinded to HIV status, and TB and LS BMD z-scores for age and sex (BMDz) were calculated using normative BMD data.3,20 For children between Tanner stages 1 and 4 with a chronologic age (CA) that differed by more than 1 SD from the BA, the BA was used instead of the CA in the TB-BMD and LS-BMD calculations.3 For children at Tanner stage 5, CA was used. Children between Tanner stages 1 and 4 were excluded from analyses of BMDz if their BA was missing.
For this cross-sectional analysis, we selected a repository blood specimen (see Table, Supplemental Digital Content 5, https://links.lww.com/QAI/B45) drawn within 1 year of a DXA scan (generally the first DXA scan). If no specimen was available within 1 year of the DXA or no DXA was completed, we selected the earliest available repository specimen. Measurements of laboratory variables were performed in batch at the USDA-Agricultural Research Service WHNRC, (Davis, CA). 25(OH)D was measured by enzyme-linked immunosorbent assay (Immunodiagnostic System, Inc, Gaithersburg, MD). Serum intact PTH was measured using a solid phase, 2-site chemiluminescent enzyme-labeled immunometric assay (Intact-PTH; Siemens Medical Solutions Diagnostics, Tarrytown, NY). Calcium, phosphate, and creatinine were determined using a clinical chemistry analyzer [Cobas Integra 400 Plus (04469658), Roche Diagnostics, Corp, Indianapolis, IN]. Samples were assayed in duplicate, and the mean of each pair was calculated. Low 25(OH)D status was defined as ≤20.0 ng/mL and deficient 25(OH)D as <12 ng/mL.21 High PTH was defined as >65 pg/mL. (http://labmed.ucsf.edu/labmanual/db/resource/Immulite_2000_Intact_PTH.pdf).
Database records from children with high PTH and/or high phosphate levels (>5.4 mg/dL) were reviewed for evidence of concurrent clinical diseases (pancreatitis, chronic kidney disease, inflammatory bowel disease, neonatal renal failure, and nephrolithiasis).
Dietary Recall and Physical Activity
The Block Dietary Questionnaire and Physical Activity Screener for Children and Adolescents were administered to assess dietary intake including supplement use and physical activity, respectively, over the past week (Nutriquest; Block Dietary Data Systems, Berkeley, CA). The recommended daily dietary allowance for calcium was ≥1100 mg for those aged 4–8 years and ≥1300 mg for those aged 9–18 years, and vitamin D was ≥600 IU/d for those aged 1–18 years of age.22 Minutes of vigorous physical activity were categorized into >75th percentile (>25.2 min/d) or ≤75th percentile (≤25.2 min/d) based on our data distribution.
The distribution of socioeconomic and clinical covariates by HIV, 25(OH)D, and PTH status was compared using Wilcoxon test or Kruskal–Wallis for continuous and χ2 for categorical variables. A variable was constructed for combinations of low 25(OH)D and high PTH.
Linear regression models were fit with the robust variance to evaluate differences in bone outcomes by 25(OH)D and PTH status in PHEU and PHIV separately. Effect modification of 25(OH)D by HIV status was tested in models that included both PHIV and PHEU children. Potential confounders were chosen a priori based on the literature review and expert knowledge. Adjusted models of bone outcomes included age at vitamin D measure, black race, continental ancestry (African, Europe/CSW Asia, and other), height z-score, extremity lean mass, and vigorous activity >75th percentile. We additionally adjusted for percent body fat when 25(OH)D status was the exposure and for Tanner stage and sex when the outcome was TB-BMC, TBLH-BMC, or LS-BMC. We did not include season and latitude as potential confounders because we did not hypothesize an effect on bone other than through 25(OH)D. Missing indicators were included for those missing information about race and vigorous activity.
Log binomial regression models were fit to obtain the unadjusted and adjusted prevalence ratios (aPR) of low 25(OH)D and high PTH in PHIV compared with PHEU children and for current TDF and EFV use compared with PHEU children. When the log binomial model did not converge, a Poisson model was fit.23 To estimate average differences in 25(OH)D and PTH levels by the above groups, we fit linear regression models with a robust variance estimator. All models were adjusted for black race and continental origin. 25(OH)D models were additionally adjusted for season and latitude. PTH levels may be higher in TDF users possibly because of phosphate wasting. Thus, as a secondary aim, we fit linear regression models to determine the relationship of PTH with calcium levels by HIV and TDF status. PHIV children without antiretroviral information or not using antiretrovirals at the 25(OH)D assessment were excluded from analyses of EFV or TDF. Analyses were performed in SAS version 9.4 (SAS Institute, Cary, NC).
Of the 448 PHIV and 226 PHEU children with an entry visit in Adolescent Master Protocol, 25(OH)D was measured in 426 and 219, respectively. Continental ancestry was available, as well as 25(OH)D, for 412 PHIV and 207 PHEU children. In the final data set, 394 of 412 PHIV and 199 of 207 PHEU had a DXA scan within 1 year of the 25(OH)D measure, with a median (interquartile range) of 0 days (1–39 days).
The distribution and unadjusted comparison of socioeconomics, lifestyle characteristics, anthropometrics, and laboratory values between PHIV and PHEU children are shown in Table 1. The prevalence of low 25(OH)D was 40% overall and higher in PHIV compared with PHEU (42% vs. 34%). Average PTH concentrations were higher in PHIV, and 9% of PHIV and 2% of PHEU had high PTH. Among PHIV with high PTH and without low 25(OH)D, the range of 25(OH)D levels was 21.1–25.5 ng/mL in 10 children, 29.4–32.6 ng/mL in 4 children, and 52.4 ng/mL in 1 child.
Diagnoses Among Children With High PTH and/or Phosphate Levels
Among the 50 children with high PTH and/or high phosphate, only 1 child had a relevant clinical diagnosis other than HIV (ie, chronic kidney disease). That child was included in all analyses. None of the PHEU children had a relevant diagnosis.
Association of 25(OH)D and PTH With Bone
Among PHIV children (Table 2), those with low 25(OH)D, compared with those without, had adjusted TB-BMD z-scores that were on average 0.38 SD lower [95% confidence interval (95% CI), −0.60 to −0.16] and TB-BMC levels that were 59.1 g lower (95% CI, −108.3 to −9.8). In an unadjusted analysis, this represents a 5.9% lower TB-BMC. Similar results were observed for TBLH-BMC. Among PHEU children, adjusted TB-BMD z-scores were on average 0.34 SD lower (95% CI, −0.64 to −0.03) in children with low 25(OH)D, but there were no differences in TB-BMC and TBLH-BMC. In models including PHIV and PHEU children, there was no strong indication of effect modification of HIV status by low 25(OH) for any of the bone outcomes (P > 0.29), but power may be limited.
The average adjusted difference in LS-BMD z-scores between those with versus without low 25(OH)D was −0.21 SD (95% CI, −0.42 to 0.0) for PHIV and 0.10 SD (95% CI, −0.30 to 0.51) for PHEU children. There were no differences in LS-BMC by 25(OH)D status in PHIV or PHEU children.
There was no apparent difference in any bone outcome by PTH status among PHIV children (Table 2). The average difference in TB-BMD z-scores was −0.02 (95% CI, −0.37 to 0.42) between those with high versus normal PTH. However, TB-BMD z-scores were on average 0.48 SD lower (95% CI, −0.92 to −0.03) in children with low 25(OH)D and high PTH and 0.28 SD units lower (95% CI, −0.51 to −0.05) in those with low 25(OH)D and normal PTH, compared with the reference group (25(OH)D > 20 ng/mL and PTH ≤ 65 pg/mL) (See Table, Supplemental Digital Content 1, https://links.lww.com/QAI/B45). LS-BMC was on average 3.1 g lower (95% CI, −6.2 to 0.05) in PHIV with low 25(OH)D and high PTH compared with the reference.
Differences in Characteristics by 25(OH)D and PTH Status in PHIV and PHEU Children
Characteristics associated with low 25(OH)D for PHIV and PHEU children combined included older age, female sex, black race, African ancestry, blood drawn in winter or spring, northern latitude, Tanner stage >1, higher percent body fat, lower calcium and phosphate, and higher PTH and creatinine (Table 3) (See Table, Supplemental Digital Content 2, https://links.lww.com/QAI/B45). Among the PHIV, children with low 25(OH)D had lower CD4 counts and were more likely to receive EFV.
Factors associated with high PTH were older age, African ancestry, Tanner stage >1, lower 25(OH)D and calcium, and higher creatinine. Among PHIV children, those with high PTH had a lower frequency of EFV use, a higher frequency of TDF use, and higher HIV viral load (Supplemental Digital Content 3, https://links.lww.com/QAI/B45) (48 of 401 received both TDF and EFV).
Prevalence of Low 25(OH)D in PHIV Compared With PHEU Children
The aPR (95% CI) of low 25(OH)D was 1.00 (0.81 to 1.24) in PHIV relative to PHEU children (Table 4, model 1B), 1.30 (0.98 to 1.74) in PHIV receiving EFV compared with PHEU children (model 2B), and 0.95 (0.76 to 1.18) in PHIV not receiving EFV compared with PHEU children (Table 4, model 2B). The opposite trend was observed for PHIV receiving TDF compared with PHEU (aPR, 0.77; 95% CI, 0.56 to 1.06, model 3B). Differences in mean 25(OH)D levels by HIV, and by EFV and TDF use, suggest similar results (Table 4, models 7B–9B).
Prevalence of High PTH in PHIV Compared With PHEU Children
PHIV had a 3.17 (1.25 to 8.06) times greater adjusted prevalence of high PTH than PHEU children (Table 4, model 4B). The average difference between groups was 5.26 pg/mL (2.35 to 8.17). PHIV who were not receiving EFV (Table 4, model 5B) had a 3.67 times higher prevalence of high PTH than PHEU children (1.44 to 9.36). The prevalence of high PTH (Table 4, model 6B) was 5.50 (1.95 to 15.49) times higher in PHIV receiving TDF compared with PHEU children and 2.64 times higher in PHIV not receiving TDF (1.01 to 6.94) than in PHEU children. Compared with PHEU, PTH concentrations were on average 5.54 pg/mL higher in PHIV not receiving EFV (Table 4, model 11B) and 11.65 pg/mL higher in those receiving TDF (Table 4, model 12B), compared with PHEU children.
Relationship Between Calcium and PTH by HIV Status and TDF Use
Serum calcium and PTH were negatively associated in PHIV not receiving TDF (slope, −10.5; P = 0.001) and PHEU (slope, −7.6; P = 0.002) but not associated among PHIV receiving TDF (slope, −3.8; P = 0.56) (See Figure, Supplemental Digital Content 4, https://links.lww.com/QAI/B45).
In this multicenter cohort of PHIV and PHEU children in the United States, 40% had vitamin D deficiency overall, and the prevalence did not differ by HIV status after careful adjustment for confounding. In both PHIV and PHEU children, low 25(OH)D was associated with lower TB-BMD but not with LS-BMD in either cohort. PTH levels were highest in PHIV children receiving TDF. Although PTH levels were, on average, higher in PHIV children with low 25(OH)D, high PTH was only associated with lower TB-BMD when accompanied by low 25(OH)D. The relationship between calcium and PTH was weak in PHIV children receiving TDF.
Several randomized clinical trials have been done examining the effects of vitamin D supplementation on healthy children and adolescents.24 A meta-analysis of 6 such randomized clinical trials found that children with normal baseline 25(OH)D concentrations (>18 ng/mL) did not have benefits in BMD accrual from supplementation but that there was a significant positive effect of supplementation on TB and a borderline significant effect on spine BMD among those with low levels of 25(OH)D at baseline.25 Four cross-sectional studies of the relationship of 25(OH)D as a continuous variable with BMD in normal children found positive associations with whole-body BMD,26–29 2 with LS,26,29 1 with total hip,29 and 1 with the forearm.28 No associations were found in 5 other cross-sectional studies.30–34 We previously found that TB-BMD was greater in PHIV children using vitamin supplements.35 In a study of vitamin D and calcium supplementation over 2 years in PHIV children, 25(OH)D levels increased, but this did not significantly increase TB or spine BMC or BMD measures over time.7 In the aforementioned study, those who advanced through puberty during the study period had a greater increase in TB-BMC and LS-BMC in the supplemented group, but this result did not achieve significance. Thus, our finding that low 25(OH)D was associated with decreased TB BMD and BMC is supported by some, but not all, studies. Because the effect sizes on BMD from vitamin D supplementation appear to be modest, and many prior studies have used low doses of vitamin D supplementation (as little as 133 IU/day) and included children who at baseline were vitamin D replete, further adequately powered studies examining the effects of vitamin D supplementation to target adequate 25(OH)D serum concentrations across puberty, particularly in children with PHIV with baseline vitamin D deficiency or insufficiency, are needed.
It is unclear why we found no associations with LS-BMD/BMC in either group or with TB-BMC in PHEU children. In contrast to these studies, we evaluated the effect of categorically low 25(OH)D compared with 25(OH)D as a continuous variable. This is based on clinical evidence that the relationship between 25(OH)D and BMD may not be linear and that those at the lowest levels of 25(OH)D may benefit most from supplementation.21 The lack of association between high PTH and bone outcomes in our study is possibly a result of competing reasons for higher PTH values, including low 25(OH)D, additional requirements for calcium and phosphate during rapid bone accrual, and effects of TDF on PTH.12 This is supported by our finding of lower TB-BMD when high PTH was accompanied by low 25(OH)D.
Vitamin D deficiency is common in the United States.36–38 The 40% prevalence of low 25(OH)D in our cohort was similar to those of cohorts of healthy children in the northeastern United States37 and PHIV children in New York City.7 Our prevalence of vitamin D deficiency was higher than the 24% observed in 6- to 18-year-old children in the National Health and Nutrition Examination Survey representative US sample of this age group,38 but lower than the ≥50% reported in behaviorally HIV-infected adolescents.12,13 Differences across studies are likely attributable to age, race, prevalence of obesity, latitude, sun exposure, season of sampling, and dietary intake.
While low 25(OH)D was equally prevalent in our PHIV and PHEU children overall after adjustment for known confounders,37,39 low 25(OH)D concentrations may be more common in PHIV receiving EFV than in PHEU children, although our power to detect a significant difference between groups may be too low. This is consistent with adult studies14 and may be explained by in vitro evidence of EFV induction of the P450 enzyme CYP24,40 which converts 25(OH)D to its inactive form, calcitroic acid. PHIV children had higher vitamin D intake than did PHEU, but HIV providers may check 25(OH)D levels in their PHIV patients and recommend supplementation.
Calcium is important to optimize bones and tooth mineralization and for catalytic and mechanical functions throughout the body. The body tightly regulates ionized calcium levels. When the calcium supply is insufficient, PTH concentrations increase. PTH increases serum calcium through a variety of mechanisms. PTH mobilizes calcium adsorbed to the bone surface by breaking down bone mineral and matrix, and it stimulates conversion of 25(OH)D to 1,25-dihydroxyvitamin D (1,25(OH)(2)D), which increases gut calcium absorption and renal calcium reabsorption. During the rapid linear growth and bone accrual of puberty, there is an increased demand for calcium and phosphate, and PTH levels increase.41,42 PTH measurements also provide an assessment of the level of “stress” on the system as a result of low 25(OH)D concentrations, although the degree of serum PTH suppression may not determine optimal vitamin D status in children.43
TDF use has been previously associated with increases in serum concentrations of PTH,12,44 with multiple mechanisms implicated. TDF appears to have effects on both bone and hormones, such as fibroblast growth factor-23 and 1,25(OH)2D, that may indirectly increase PTH.45–47 TDF use can also result in renal tubular dysfunction stimulating the production of PTH and in renal phosphate wasting,44 which may, in part, be PTH mediated.12 In HIV-infected adolescents, baseline PTH levels were higher in TDF users regardless of 25(OH)D status. With vitamin D supplementation, PTH levels decreased in the TDF group but did not change in those not receiving TDF.12 In our cohort, PTH levels were higher in those with low 25(OH)D, highest at Tanner stages 3–4, the time of most rapid bone accrual, and higher among TDF users. PTH concentrations were not strongly associated with serum calcium among TDF users, suggesting other mechanisms for elevated PTH.
Our study has several limitations. Although 25(OH)D is a stable vitamin D metabolite with a biological half-life of 2–3 weeks, levels vary by season; thus, one measurement may not represent the average yearly level.48 However, children with 25(OH)D deficiency at one time of the year tend to be deficient at other times during the year,49 which favors using the ≤20 ng/mL cutoff recommended by the Global Consensus Recommendations on Prevention and Management of Nutritional Rickets.21 Although imperfect, we adjusted for season when evaluating the prevalence or differences in 25(OH)D by subgroups. We recognize that the 20 ng/mL cutoff is based upon what is needed to prevent rickets and osteomalacia,21 but concentrations required for optimal bone and immunological health are likely higher.50 Using this lower threshold, we could examine characteristics of those most likely to have clinically significant deficiency. Because this was a cross-sectional study where 25(OH)D, PTH, and DXA parameters were measured within 1 year of each other, we could not establish a temporal relationship between low 25(OH)D or high PTH and subsequent bone accrual. Although this is an important limitation, BMD ranking tracks well over time in healthy children such that those who ranked among the lowest earlier in childhood also ranked among the lowest later in adolescence and those who ranked highest generally remained high.51 When evaluating prevalence of high PTH, confidence intervals were wide because of a few children having high PTH. Finally, our findings might not be generalizable to populations other than those in the United States where there is a difference in the distribution of continental ancestry and nutritional status.
Children gain more than half of their peak bone mass during adolescence with the greatest increase following the pubertal growth spurt, highlighting the need to identify modifiable factors during this critical period that could improve bone accrual. This study afforded a unique opportunity to further quantify the risk factors for poor bone health, specifically low 25(OH)D. This may lead to novel vitamin D supplementation trials that could ameliorate deficits and improve bone health at critical developmental stages.
The authors thank the children and families for their participation in PHACS and the individuals and institutions involved in the conduct of PHACS. (Principal Investigator: George Seage; Project Director: Julie Alperen) and the Tulane University School of Medicine (HD052104) (Principal Investigator: R.B.V.; Co-Principal Investigators: Kenneth Rich, Ellen Chadwick; Project Director: Patrick Davis). Data management services were provided by Frontier Science and Technology Research Foundation (PI: Suzanne Siminski), and regulatory services and logistical support were provided by Westat, Inc (PI: Julie Davidson). ssThe following institutions, clinical site investigators, and staff participated in conducting PHACS AMP and AMP Up in 2015, in alphabetical order: Ann & Robert H. Lurie Children's Hospital of Chicago: Ram Yogev, Margaret Ann Sanders, Kathleen Malee, Scott Hunter; Baylor College of Medicine: William Shearer, Mary Paul, Norma Cooper, Lynnette Harris; Bronx Lebanon Hospital Center: Murli Purswani, Mahboobullah Baig, Anna Cintron; Children's Diagnostic & Treatment Center: Ana Puga, Sandra Navarro, Patricia A. Garvie, James Blood; Children's Hospital, Boston: Sandra K. Burchett, Nancy Karthas, Betsy Kammerer; Jacobi Medical Center: Andrew Wiznia, Marlene Burey, Molly Nozyce; Rutgers—New Jersey Medical School: Arry Dieudonne, Linda Bettica; St. Christopher's Hospital for Children: J.S.C., Maria Garcia Bulkley, Latreaca Ivey, Mitzie Grant; St. Jude Children's Research Hospital: Katherine Knapp, Kim Allison, Megan Wilkins; San Juan Hospital/Department of Pediatrics: Midnela Acevedo-Flores, Heida Rios, Vivian Olivera; Tulane University School of Medicine: M.S., Medea Gabriel, Patricia Sirois; University of California, San Diego: S.A.S., Kim Norris, Sharon Nichols; University of Colorado Denver Health Sciences Center: Elizabeth McFarland, Juliana Darrow, Emily Barr, Paul Harding; University of Miami: Gwendolyn Scott, Grace Alvarez, Anai Cuadra.
1. Heaney RP, Abrams S, Dawson-Hughes B, et al. Peak bone mass. Osteoporos Int. 2000;11:985–1009.
2. Jacobson DL, Lindsey JC, Gordon CM, et al. Total body and spinal bone mineral density
across tanner stage in perinatally HIV-infected and uninfected children and youth in PACTG 1045. AIDS. 2010;24:687–696.
3. DiMeglio LA, Wang J, Siberry GK, et al. Bone mineral density
in children and adolescents with perinatal HIV infection. AIDS. 2013;27:211–220.
4. Rodriguez M, Daniels B, Gunawardene S, et al. High frequency of vitamin D deficiency in ambulatory HIV-Positive patients. AIDS Res Hum Retroviruses. 2009;25:9–14.
5. Stone B, Dockrell D, Bowman C, et al. HIV and bone disease. Arch Biochem Biophys. 2010;503:66–77.
6. Yin MT, Lu D, Cremers S, et al. Short-term bone loss in HIV-infected premenopausal women. J Acquir Immune Defic Syndr. 2010;53:202–208.
7. Arpadi SM, McMahon DJ, Abrams EJ, et al. Effect of supplementation with cholecalciferol and calcium on 2-y bone mass accrual in HIV-infected children and adolescents: a randomized clinical trial. Am J Clin Nutr. 2012;95:678–685.
8. Ross AC, Judd S, Kumari M, et al. Vitamin D is linked to carotid intima-media thickness and immune reconstitution in HIV-positive individuals. Antivir Ther. 2011;16:555–563.
9. Eckard AR, Judd SE, Ziegler TR, et al. Risk factors for vitamin D deficiency and relationship with cardiac biomarkers, inflammation and immune restoration in HIV-infected youth. Antivir Ther. 2012;17:1069–1078.
10. Rutstein R, Downes A, Zemel B, et al. Vitamin D status in children and young adults with perinatally acquired HIV infection. Clin Nutr. 2011;30:624–628.
11. Sudjaritruk T, Bunupuradah T, Aurpibul L, et al. Hypovitaminosis D and hyperparathyroidism: effects on bone turnover and bone mineral density
among perinatally HIV-infected adolescents. AIDS. 2016;30:1059–1067.
12. Havens PL, Stephensen CB, Hazra R, et al. Vitamin D3 decreases parathyroid hormone in HIV-infected youth being treated with tenofovir: a randomized, placebo-controlled trial. Clin Infect Dis. 2012;54:1013–1025.
13. Stephensen CB, Marquis GS, Kruzich LA, et al. Vitamin D status in adolescents and young adults with HIV infection. Am J Clin Nutr. 2006;83:1135–1141.
14. Brown TT, McComsey GA. Association between initiation of antiretroviral therapy with efavirenz and decreases in 25-hydroxyvitamin D. Antivir Ther. 2010;15:425–429.
15. Jacobson DL, Patel K, Siberry GK, et al. Body fat distribution in perinatally HIV-infected and HIV-exposed but uninfected children in the era of highly active antiretroviral therapy: outcomes from the pediatric HIV/AIDS cohort study. Am J Clin Nutr. 2011;94:1485–1495.
16. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States. Adv Data. 2000;314:1–27.
17. Nievergelt CM, Maihofer AX, Shekhtman T, et al. Inference of human continental origin and admixture proportions using a highly discriminative ancestry informative 41-SNP panel. Investig Genet. 2013;4:13.
18. Brummel SS, Singh KK, Maihofer AX, et al. Associations of genetically determined continental ancestry with CD4+ count and plasma HIV-1 RNA beyond self-reported race and ethnicity. J Acquir Immune Defic Syndr. 2016;71:544–550.
19. Spector SA, Brummel SS, Nievergelt CM, et al. Genetically determined ancestry is more informative than self-reported race in HIV-infected and -exposed children. Medicine (Baltimore). 2016;95:e4733.
20. Ellis KJ, Shypailo RJ. Body Composition Comparison Data for Children. Houston, TX: Children's Nutrition Research Center. Baylor College of Medicine, Body Composition Web Site; 2001.
21. Munns CF, Shaw N, Kiely M, et al. Global consensus recommendations on prevention and management of nutritional rickets. J Clin Endocrinol Metab. 2016;101:jc20152175.
22. Ross AC, Manson JE, Abrams SA, et al. The 2011 report on dietary reference intakes for calcium and vitamin D from the institute of medicine: what clinicians need to know. J Clin Endocrinol Metab. 2011;96:53–58.
23. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol. 2005;162:199–200.
24. Winzenberg TM, Powell S, Shaw KA, et al. Vitamin D supplementation for improving bone mineral density
in children. Cochrane database Syst Rev. 2010;10:Cd006944.
25. Winzenberg T, Jones G. Vitamin D and bone health in childhood and adolescence. Calcif Tissue Int. 2013;92:140–150.
26. Lee YA, Kim JY, Kang MJ, et al. Adequate vitamin D status and adiposity contribute to bone health in peripubertal nonobese children. J Bone Miner Metab. 2013;31:337–345.
27. Mayranpaa MK, Viljakainen HT, Toiviainen-Salo S, et al. Impaired bone health and asymptomatic vertebral compressions in fracture-prone children: a case-control study. J Bone Miner Res. 2012;27:1413–1424.
28. Mouratidou T, Vicente-Rodriguez G, Gracia-Marco L, et al. Associations of dietary calcium, vitamin D, milk intakes, and 25-hydroxyvitamin D with bone mass in Spanish adolescents: the HELENA study. J Clin Densitom. 2013;16:110–117.
29. Pekkinen M, Viljakainen H, Saarnio E, et al. Vitamin D is a major determinant of bone mineral density
at school age. PLoS One. 2012;7:e40090.
30. Ceroni D, Anderson de la Llana R, Martin X, et al. Prevalence of vitamin D insufficiency in swiss teenagers with appendicular fractures: a prospective study of 100 cases. J Child Orthop. 2012;6:497–503.
31. Cheng S, Tylavsky F, Kroger H, et al. Association of low 25-hydroxyvitamin D concentrations with elevated parathyroid hormone concentrations and low cortical bone density in early pubertal and prepubertal finnish girls. Am J Clin Nutr. 2003;78:485–492.
32. Outila TA, Karkkainen MU, Lamberg-Allardt CJ. Vitamin D status affects serum parathyroid hormone concentrations during winter in female adolescents: associations with forearm bone mineral density
. Am J Clin Nutr. 2001;74:206–210.
33. Stein EM, Laing EM, Hall DB, et al. Serum 25-hydroxyvitamin D concentrations in girls aged 4-8 y living in the southeastern United States. Am J Clin Nutr. 2006;83:75–81.
34. Talwar SA, Swedler J, Yeh J, et al. Vitamin-D nutrition and bone mass in adolescent black girls. J Natl Med Assoc. 2007;99:650–657.
35. Jacobson DL, Spiegelman D, Duggan C, et al. Predictors of bone mineral density
in human immunodeficiency virus-1 infected children. J Pediatr Gastroenterol Nutr. 2005;41:339–346.
36. Rajakumar K, Moore CG, Yabes J, et al. Effect of vitamin D3 supplementation in black and in white children: a randomized, placebo-controlled trial. J Clin Endocrinol Metab. 2015;100:3183–3192.
37. Gordon CM, DePeter KC, Feldman HA, et al. Prevalence of vitamin D deficiency among healthy adolescents. Arch Pediatr Adolesc Med. 2004;158:531–537.
38. Karalius VP, Zinn D, Wu J, et al. Prevalence of risk of deficiency and inadequacy of 25-hydroxyvitamin D in US children: NHANES 2003-2006. J Pediatr Endocrinol Metab. 2014;27:461–466.
39. Weng FL, Shults J, Leonard MB, et al. Risk factors for low serum 25-hydroxyvitamin D concentrations in otherwise healthy children and adolescents. Am J Clin Nutr. 2007;86:150–158.
40. Landriscina M, Altamura SA, Roca L, et al. Reverse transcriptase inhibitors induce cell differentiation and enhance the immunogenic phenotype in human renal clear-cell carcinoma. Int J Cancer. 2008;122:2842–2850.
41. DeBoer MD, Weber DR, Zemel BS, et al. Bone mineral accrual Is associated with parathyroid hormone and 1,25-dihydroxyvitamin D levels in children and adolescents. J Clin Endocrinol Metab. 2015;100:3814–3821.
42. Abrams SA, Griffin IJ, Hawthorne KM, et al. Relationships among vitamin D levels, parathyroid hormone, and calcium absorption in young adolescents. J Clin Endocrinol Metab. 2005;90:5576–5581.
43. Hill KM, McCabe GP, McCabe LD, et al. An inflection point of serum 25-hydroxyvitamin D for maximal suppression of parathyroid hormone is not evident from multi-site pooled data in children and adolescents. J Nutr. 2010;140:1983–1988.
44. Woodward CL, Hall AM, Williams IG, et al. Tenofovir-associated renal and bone toxicity. HIV Med. 2009;10:482–487.
45. Razzaque MS. The FGF23-Klotho axis: endocrine regulation of phosphate homeostasis. Nat Rev Endocrinol. 2009;5:611–619.
46. Grigsby IF, Pham L, Mansky LM, et al. Tenofovir treatment of primary osteoblasts alters gene expression profiles: implications for bone mineral density
loss. Biochem Biophys Res Commun. 2010;394:48–53.
47. Grigsby IF, Pham L, Gopalakrishnan R, et al. Downregulation of Gnas, Got2 and Snord32a following tenofovir exposure of primary osteoclasts. Biochem Biophys Res Commun. 2010;391:1324–1329.
48. Harris SS, Dawson-Hughes B. Seasonal changes in plasma 25-hydroxyvitamin D concentrations of young American black and white women. Am J Clin Nutr. 1998;67:1232–1236.
49. Benitez-Aguirre PZ, Wood NJ, Biesheuvel C, et al. The natural history of vitamin D deficiency in african refugees living in Sydney. Med J Aust. 2009;190:426–428.
50. Holick MF, Binkley NC, Bischoff-Ferrari HA, et al. Evaluation, treatment, and prevention of vitamin D deficiency: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2011;96:1911–1930.
51. Kalkwarf HJ, Gilsanz V, Lappe JM, et al. Tracking of bone mass and density during childhood and adolescence. J Clin Endocrinol Metab. 2010;95:1690–1698.