Vitamin D deficiency extremely common among HIV-positive patients in diverse regions

New analyses have found widespread vitamin D insufficiency among American, Italian and Swiss cohorts of HIV-positive patients. These data were presented in a poster session and related discussion at the 17^th Conference on Retroviruses and Opportunistic Infections (CROI).

Vitamin D deficiency was consistently less frequent in Caucasians than in other races. Otherwise, the studies were not entirely consistent as to other risk factors, such as duration or type of antiretroviral treatment.

At the CROI discussion session on Wednesday, moderator Peter Reiss from the University of Amsterdam began by noting that vitamin D deficiency can result in bone density loss, cardiovascular disease, diabetes and insulin resistance, kidney disease, and other metabolic conditions commonly seen in patients with HIV.

Vitamin D levels are generally assessed as blood levels of 25-hydroxyvitamin D, also expressed as 25(OH)D. According to a widely accepted reference scale, 25-hydroxyvitamin D levels are:

• "sufficient" if ≥75 nmol/l (≥30 ng/ml),
• "insufficient" if between 50-75 nmol/l (20-30 ng/ml), and
• "deficient" if <50 nmol/l (<20 ng/ml).

However, Reiss pointed out that these values are not universally used (even within the studies in this session), that they do not account for seasonal variation or ethnic differences (as darker skin is less able to produce vitamin D, and levels are higher in the summer months), and that optimal levels for health have not been well established.

Despite these uncertainties, these newly reported studies were consistent with previous reports in identifying very high rates of vitamin D insufficiency or deficiency among HIV-positive people. (See, for instance, this Dutch study and these US and UK studies reported at IAS 2009.)

United States: the SUN study

Christine Dao from the Centers for Disease Control presented data from the SUN study, a prospective observational cohort of 700 HIV-positive adults enrolled at clinics in four US cities from March 2004 to June 2006.

The findings were based on 672 participants who had had baseline serum 25-hydroxyvitamin D determinations, and who were not taking vitamin D supplements. The cohort was 77% male, 30% black and 10% Hispanic, median age was 41 years, median CD4 count 471 cells/mm³, and most (74%) had viral loads < 400 copies/ml.

In this cohort, 71.6% (95% confidence interval [CI] 68.1 to 74.9) were deemed vitamin D insufficient, defined as serum 25-hydroxyvitamin D levels <30 ng/ml. In multivariable analysis, sex, age and bone mineral density (BMD) had no association with vitamin D levels. The following factors were independently associated with higher risk of insufficiency:

• black race (adjusted odds ratio [aOR] = 4.50, 95% confidence interval [CI] 2.59 to 7.85),
• Hispanic ethnicity (aOR = 2.78, 95% CI 1.31 to 5.90),
• lower exposure to ultraviolet light, as estimated from National Weather Service data for the month of sampling (aOR = 1.28, 95% CI 1.17 to 1.40),
• hypertension (aOR = 1.88, 95% CI 1.10 to 3.22),
• lack of exercise (aOR = 3.14, 95% CI 1.80 to 5.47), and
• exposure to efavirenz (aOR = 1.98, 95% CI 1.18 to 3.34).

Lower odds of vitamin D insufficiency were seen in patients with renal (kidney) insufficiency (GFR <90 mL/min/1.73m²) (aOR = 0.55, 95% CI 0.36 to 0.83) and exposure to ritonavir (aOR = 0.56, 95% CI 0.35 to 0.89).

United States: the WIHS

The Women's Interagency Health Study (WIHS) is a longitudinal study of women with and at risk for HIV. The substudy reported here looked at a cross-section of 609 WIHS participants (480 HIV-positive, 122 HIV-negative) from Chicago and New York. Vitamin D deficiency (defined as 25-hydroxyvitamin D levels ≤ 20 ng/ml) was found in 60% of these women, and insufficiency (20 to ≤30 ng/ml) in 23.5%.

Age, HIV status, and CD4 count were not predictive of vitamin D deficiency in this group. In multivariate analysis, black race was the only significant predictor of deficiency, with an odds ratio [OR] of 3.16 compared to white race (95% CI, 2.06 to 4.89).

Bacterial vaginosis was found in 19% of the study group and was strongly correlated with vitamin D level (r = –0.14, p<0.001). Risk of bacterial vaginosis increased with worsening deficiency: compared to sufficient vitamin D levels, the odds ratio was 2.12 for insufficiency and 3.55 for deficiency.

Italy: ICONA

Findings from an analysis of the Italian ICONA cohort were presented next. A total of 1505 plasma samples from 856 patients were analysed; 262 before ART initiation and 1243 after a median of 14 months of ART. The median age was 36 years, and most patients (93%) were from Italy.

In this cross-sectional sample, vitamin D insufficiency (defined as 25-hydroxyvitamin D levels <75 nmol/l) was found in 54% of the samples, and deficiency (levels <30 nmol/l) in 7%.

Levels varied seasonally (much lower in winter and spring). In multivariable analysis, older age increased risk of deficiency (odds ratio [OR] = 1.53 per 10 years older, 95% CI 1.11 to 2.09, p = 0.009). The following factors decreased risk:

• Caucasian origin (OR = 0.17, 95% CI 0.07 to 0.42, p = 0.0001),
• higher CD4 count (OR = 0.90 per 100 cells/mm³ higher, 95% CI 0.82 to 0.99, p = 0.04),
• higher body mass index (BMI) (OR = 0.90 per unit higher, 95% CI 0.83 to 0.98, p = 0.01), and
• type of ART (PI use decreased risk compared to NNRTI use: OR = 0.47, 95% CI 0.27 to 0.84, p = 0.01).

Swiss HIV Cohort Study

The third study presented was a retrospective analysis of 25-hydroxyvitamin D levels in stored serum from 211 Swiss HIV Cohort participants (75% male, 88% Caucasian, median age 37 years). Samples were taken at three time points: before initiating ART, and at twelve and eighteen months after starting ART. Seasonality was thus controlled, as the second sample was taken during the same season as the first, and the final sample in the opposite season (spring/fall, summer/winter).

At the time of the first (baseline) sample, vitamin D deficiency (defined as 25-hydroxyvitamin D levels <30 nmol/l) was found in 14% of participants in fall and 42% in spring. These levels were essentially unchanged twelve months after ART initiation (14% in fall, 47% in spring), but showed the expected seasonal change at 18 months (deficiency levels were higher in the spring and lower in the fall).

By multivariable analysis, apart from seasonal variation, the following factors increased vitamin D levels:

• Caucasian ethnicity (multivariable coefficient 14.1, 95% CI 6.0 to 22.1, p = 0.001), and
• duration since HIV diagnosis (6.4, 95% CI 1.2 to 11.7, p = 0.02).

Factors that decreased vitamin D levels were:

• injection drug use (-11.2, 95% CI -21.0 to -1.5, p = 0.02), and
• NNRTI use (-8.2, 95% CI -13.3 to -3.0, p = 0.002).

In a subset of 74 patients, the study researchers also looked at levels of 1,25(OH)₂D, the actual active molecule of vitamin D which the body produces from the parent molecule 25-hydroxyvitamin D. By multivariate analysis, higher 1,25(OH)₂D levels were found with higher BMI and with use of tenofovir; lower levels were found with higher CD4 cell count, hepatitis C infection, and a previous AIDS diagnosis.

Tanzania

Rather than assessing prevalence and risk factors for vitamin D deficiency, this study looked at health outcomes in Tanzanian women with low vitamin D levels. In the first two years of follow-up, compared to women with adequate vitamin D levels, women with low vitamin D status (defined as 25-hydroxyvitamin D <32 ng/ml had:

• a 45% higher chance of wasting (reaching a body mass index <18 kg/m² (incidence rate ratio [RR] = 1.45; 95% CI, 1.04 to 2.01),
• a 28% higher chance of acute upper respiratory infections (RR =1.28, 95% CI,1.05 to 1.55), and
• a 192% higher chance of thrush (RR = 2.92, 95% CI, 1.43 to 5.96).

Conclusions

These studies add to a growing body of evidence that insufficient or deficient vitamin D levels are extremely prevalent among people with HIV. While prevalence figures (and the cut-off values used to define them) vary, these studies reported insufficiency rates of 54% to 72%; figures which are generally consistent with other reports. Studies in women linked vitamin D deficiency with risk of bacterial vaginosis, thrush, and other health problems.

The single factor invariably associated with insufficiency or deficiency was non-Caucasian race. Otherwise, reported risk factors were not entirely consistent, although several studies identified NNRTI and/or efavirenz use.

Investigators agreed that vitamin D deficiency is prevalent among HIV-positive individuals, has harmful effects on health, and is easily addressable through supplementation. Remaining research questions include the link between deficiency and clinical health outcomes, the impact of supplementation, the best doses for supplementation, and closer comparisons of deficiency rates in people with HIV and in the general population, where deficiency is also common.

Further information

You can view poster abstract 750, poster abstract 751, poster abstract 752, poster abstract 753, and poster abstract 754 on the official conference website.

You can also view a webcast and slides of this session on the official conference website.