Abstract
Background:
HIV is associated with alterations in androgen hormone levels and sex hormone binding globulin (SHBG) in women. Higher SHBG has been associated with a lower risk of diabetes in the general population, but the contribution of HIV, androgen hormones, SHBG, and menopausal phase to diabetes is unclear.
Methods:
From April 2003 through February 2020, 896 women with HIV (WWH) and 343 women without HIV (WWOH) from the Women’s Interagency HIV Study with morning total testosterone (TT), dehydroepiandrosterone sulfate (DHEAS), and SHBG levels were followed to assess for incident diabetes. Parametric regression models were used with age as the time scale and relative times (RT) as the measure of association of hormone level and menopausal phase with incident diabetes. Analyses incorporated time-dependent androgen hormone, SHBG levels, menopausal phase and were adjusted for race/ethnicity, enrollment year, smoking status, BMI, HCV status, and HIV-related factors.
Results:
128 (14%) WWH and 47 (14%) WWOH developed diabetes. In WWH, a doubling of SHBG and DHEAS were associated with 7%(RT=1.07[95%CI:0.82,1.40] and 15%(RT=1.15[95%CI:0.95,1.39]) longer times to diabetes, respectively; in WWOH, a doubling of SHBG and DHEAS were associated with 84%(RT=1.84[95%CI:0.89,3.82]) and 41%(RT= 1.41[95%CI:0.82,2.44]) longer times to diabetes. TT was not associated. In WWH, later menopausal phase was associated with shorter times to diabetes.
Conclusions:
Despite alterations in androgen hormone and SHBG levels in HIV, regardless of HIV status, higher SHBG and DHEAS were associated with non-statistically significant slower progression to diabetes. The menopausal transition may be a better hormonal indicator of diabetes risk in WWH.
Keywords: Women, HIV, Diabetes, Sex Steroids, Menopause
Introduction
Women with HIV (WWH) are at increased risk for developing diabetes mellitus (DM) compared to women without HIV (WWOH), and the extent of this excess risk is influenced by how DM is defined [1]. Given the rise in prevalence of weight gain and metabolic syndrome in people with HIV (PWH), DM has become an increasing health concern, especially in WWH. Prior studies from the Women’s Interagency HIV Study (WIHS) have found that certain antiretroviral therapies [2], hepatitis C and opiate use [3], lymphocyte count and CD4 T cell count [4], visceral adiposity [5], and certain genetic factors [6] are associated with insulin resistance or incident DM. However, whether sex hormone binding globulin (SHBG) and androgen hormone levels independently contribute to DM risk in PWH is unclear, despite evidence in the general population that androgen hormones, such as testosterone, and SHBG have been found to differentially modulate the status and risk of DM. Higher SHBG levels are protective against the development of DM [7–10], with some studies suggesting that the association between SHBG and DM may be stronger in women as compared to men [7,8]. As WWH are now transitioning to menopause, a time of alterations in androgen hormones, it is imperative to understand their contribution to disorders in glucose metabolism.
Women with HIV may be particularly vulnerable to alterations in androgen hormone and SHBG levels, especially during the transition to menopause. WWH have been found to have lower dehydroepiandrosterone sulfate (DHEAS), a marker of adrenal androgen secretion, and higher SHBG compared to WWOH [11]. Androgen hormone and SHBG levels also change with age and menopausal status [12,13]; menopause has been associated with lower SHBG levels, whose concentration is mainly related to the level of estrogen and androgens [14]. DHEAS levels have also been found to increase during the menopausal transition [13]. Body composition changes, which occur during the transition to menopause, result in alterations in adiposity [15–17] that may also influence androgen hormone and SHBG levels. Due to declining ovarian production of androgen hormones during the menopausal transition, such as estrogen, androgen hormone production may predominantly occur in adipose tissue via the aromatization of androstenedione [18,19]. Understanding the contribution of the menopausal transition to the relationship of androgen hormones, SHBG levels and incident DM within the setting of HIV is of clinical importance, especially as over half of WWH in the US are over the age of fifty.
Using longitudinal data from the Women’s Interagency HIV Study (WIHS), a cohort of middle-aged women with and without HIV, we quantified the relationship of androgen hormones, TT and DHEAS, as well as SHBG levels, and the menopausal transition with incident DM. We hypothesized that the association of androgen hormones, SHBG, and the menopausal transition on incident DM would differ by HIV status.
Methods
Study design
The WIHS (now part of the MACS-WIHS Combined Cohort Study, or MWCCS) was a prospective cohort study of WWH and demographically similar WWOH who were enrolled over four recruitment waves (1994–1995, 2000–2001, 2011–2012, 2013–2015) [20]. Full details of recruitment, retention, and demographics have been published elsewhere [21]. Women were seen every six months for anthropometry, specimen collection, and administration of survey instruments.
Women were eligible for inclusion into the current analysis if they had available TT, DHEAS, and SHBG measurements as well as menopausal phase categorized at or after April 2003 when a sex steroid ancillary study in the WIHS was established. The first visit with TT, DHEAS, and SHBG data measured from a morning blood draw and with menopausal phase categorized was considered the index visit. Of the 1,674 who met criteria, 435 were excluded for the following reasons: HIV seroconversion (n=15), had prevalent DM at or before the index visit (n=319), were pregnant or had missing pregnancy data at the index visit (n=23), did not have follow-up after the index visit (n=23), were not in the study age range (25–80 years for WWH and 25–67 years for WWOH, n=6), or were missing data for at least one covariate (n=49). Our final study population of 1,239 women were followed through February 2020.
Study measurements
TT, DHEAS, and SHBG levels were measured from frozen sera stored at −80°C using automated chemiluminescent immunoassay on Siemens IMMULITE (Siemens Medical Solutions, Inc, PA USA) or, after 2015, the Beckman Automated DxI (Beckman Coulter Inc, MN USA). Lower limits of detection for the Siemens immunoassay were 0.20 ng/dL, 15 μg/dL, and 2 nmol/L for TT, DHEAS, and SHBG, respectively and were 0.10 ng/dL, 2 μg/dL, and 0.33 nmol/L for the Beckman for TT, DHEAS, and SHBG, respectively. Method comparison data were used to adjust all results to the Siemens assay as the majority of samples (83%) were run using that platform.
Menopausal status was defined using Anti-Müllerian hormone (AMH), an objective biomarker of ovarian reserve which has been shown to be a strong predictor for the age of final menstrual period among WWH from the WIHS and has been used to determine menopausal phase in prior analyses [22–25]. As previously described [26], we categorized menopausal phase as follows: pre-menopause, defined as ≥5 years before AMH became undetectable; early perimenopause, defined as 0 to <5 years before an undetectable AMH; late perimenopause, defined as 0 to up to 5 years after AMH became undetectable; and menopause, defined as >5 years after AMH became undetectable. Menopausal categories (pre-menopause, early perimenopause, late perimenopause, and menopause) were based upon the STRAW +10 criteria [27].
Incident DM was determined using a standardized definition established in WIHS and adapted from the American Diabetes Association criteria, i.e., if: anti-diabetic medication use was reported, a HgbA1c value ≥6.5% and subsequently confirmed by either a fasting glucose (FG) ≥126 mg/dl or use of anti-diabetic medication, a FG ≥126 mg/dl and subsequently confirmed by either a second FG ≥126 mg/dl, report of antidiabetic medication, or HgbA1c ≥6.5%, or if DM was reported and was subsequently confirmed by use of an anti-diabetic medication or by a HgbA1c ≥6.5% or a FG >126 mg/dl [28].
Statistical Analysis
Univariable and multivariable parametric failure time regression models using the Generalized Gamma distribution were used to assess the relationship between androgen hormone levels and SHBG and the menopause category with the timing of incident DM separately in WWH and in WWOH [29]. Relative times (RT) were used to quantify the magnitude of the relationship between androgen hormone levels and SHBG with incident DM. Specifically, a relative time represents the percentage increase (RT>1) or percent decrease (RT<1) in the time to develop DM per doubling of androgen hormone or SHBG level (androgen hormone levels and SHBG level were log2 transformed in analyses). More precisely, the relative times are the ratios of the percentiles of the distributions of these times. Our models assumed proportional times, the special case in which these ratios are constant for all percentiles. Parametric models were used as they do not assume proportional hazards. The goodness of fit of the parametric failure time model was assessed by HIV status by comparing the non-parametric Kaplan-Meier curves to the survival curves estimated by the parametric models. The time scale for our time-to-event analyses was age, with age 25 as the origin (time 0) of the time at risk. Twenty-five was selected as the origin as there were no incident DM events before age 25 and a very small number of at-risk participants had exposure data before 25 years of age. Those who were younger than 25 years at the index visit entered the analyses at time 0 (i.e., no late entry). Those who were older than 25 years at the index visit contributed to years after the index visit and were treated as late entries with age minus 25 at the index visit as the entry time. Time was measured to either age of incident DM (an event) or age at the last study visit if DM did not occur (a right censored observation). WWH without DM at age 80 and WWOH without DM at age 67 were censored at 55 (80 minus 25) and 42 (67 minus 25), respectively. Time intervals (as opposed to one record per participant) were used to allow androgen hormone, SHBG levels, and menopause category as well as covariates measured between the index visit and incident DM to be incorporated as time-varying in analyses. Participants with only the index visit with androgen hormone and SHBG data contribute only one time interval (i.e., one record) to analyses.
Other covariates were race/ethnicity, enrollment cohort, body mass index (time-varying), smoking status (time-varying, current versus non-current), and hepatitis C virus (HCV) infection status (yes or no). All analyses were stratified by HIV status to allow the associations of androgen hormones, SHBG, and menopause category with incident DM to differ by HIV status. CD4 cell count, HIV RNA level, and ART use (yes or no) were included as time-varying in the models including WWH.
Results
Table 1 shows the demographic and clinical characteristics by HIV status of the 1,239 women included in the analysis. At the index visit, over half of women identified as Black. Women with HIV were older (median: 41 versus 36 years), were less likely to be premenopausal (47% versus 65%), reported less current smoking (44% versus 53%), had a slightly lower BMI (27 kg/m2 versus 28 kg/m2), and were more likely to have HCV infection (20% versus 11%) compared to WWOH. Among WWH, the median CD4 count was 434 cells/mm3 [IQR 267, 632] and median HIV RNA level was 190 copies/mL [IQR 80, 5,650]. Seventy one percent of women reported being on antiretroviral therapy.
Table 1.
Demographic and clinical characteristicsa of 1239 WIHS women by HIV status at the index visit
| WWH n=896 |
WWOH n=343 |
|
|---|---|---|
| Age | 41 [35, 47] | 36 [30, 44] |
| Race/Ethnicity | ||
| Black | 491 (55%) | 197 (57%) |
| White | 113 (13%) | 44 (13%) |
| Hispanic | 292 (33%) | 102 (30%) |
| Menopause status | ||
| Premenopausal | 421 (47%) | 222 (65%) |
| Early Perimenopause | 190 (21%) | 64 (19%) |
| Late Peri/Early Menopause | 166 (19%) | 41 (12%) |
| Menopause | 119 (13%) | 16 (5%) |
| 1994 – 1995 Enrollment Period | 499 (56%) | 142 (41%) |
| Currently Smoking | 393 (44%) | 182 (53%) |
| BMI (kg/m2) | 27 [23, 31] | 28 [24, 34] |
| HCV seropositive | 177 (20%) | 38 (11%) |
| CD4 cell count (cells/mm3) | 434 [267, 632] | 948 [724, 1214] |
| HIV RNA (copies/ml) | 190 [80, 5650] | NA |
| Current ART use | 632 (71%) | NA |
| Incident DM event during follow-up | 128 (14%) | 47 (14%) |
| Number of person-years of follow-up | 8,811 | 3,563 |
n (%) or Median [IQR]
Figure 1 shows the distribution of TT, DHEAS, and SHBG values at the index visit by HIV status. Compared to WWOH, WWH were nearly twice as likely to have initial undetectable TT levels (43% versus 22%) and had much lower detectable TT levels (median of 33 ng/dL versus median of 40 ng/dL). Women with HIV were also more likely to have an undetectable DHEAS level (10% versus 1%) and had lower detectable DHEAS levels overall (median of 77 μg/dL versus median of 113 μg/dL). The median SHBG was 61 nmol/L [IQR 42, 91] among WWH and 47 nmol/L [IQR 34, 70] among WWOH.
Figure 1.
Androgen hormone and SHBG levels among 896 WWH (left panel) and 343 WWOH (right panel)
One hundred twenty-eight (14%) of the 896 WWH developed incident DM over 8,811 person-years of follow-up (incidence rate of 1.5 per 100 person-years). Among the 343 WWOH, 47 (14%) developed incident DM over 3,563 person-years of follow-up (incidence rate of 1.3 per 100 person-years). Supplemental Figure 1 shows Kaplan-Meier plots of the cumulative incidence of diabetes by HIV status as well as the appropriateness of fit of the Generalized Gamma regression models to the data. WWH and WWOH have distributions of times to incident diabetes (p-value =0.95). Six hundred and twenty WWH (69%) contributed one time interval to the analyses, 195 (22%) contributed two time intervals, 64 (7%) contributed three, and 17 (2%) contributed four time intervals for a total of 1,270. Similarly, 231 (67%) of WWOH contributed one time interval to the analysis, 77 (22%) contributed two, 26 (8%) contributed three, and 9 (3%) contributed four for a total of 499.
Unadjusted associations of TT, DHEAS, and SHBG with incident diabetes by HIV status
In the unadjusted models (Figure 2), there was no association of TT levels with time to DM [RT=1.00 per doubling of TT (95% CI 0.90, 1.10) in WWH and 0.99 per doubling (95% CI 0.79, 1.24) in WWOH]. Greater DHEAS levels were associated with longer time to DM [14% longer time per doubling (RT=1.14, 95% CI 1.00, 1.29) in WWH and 70% longer times per doubling of DHEAS (RT=1.70, 95% CI 0.98, 2.97) in WWOH]. Greater SHBG levels were associated with longer time to DM [8% longer times per doubling (RT=1.08, 95% CI 0.93, 1.25) in WWH and 48% longer times per doubling (RT=1.48, 95% CI 1.01, 2.17) in WWOH]. Heterogeneity of effect was assessed using a multiplicative interaction term between HIV and hormone in the unadjusted models. None of the interactions reached statistical significance. Despite these results, the multivariable models were stratified by HIV status due to biologic plausibility and to allow covariates to have different effects by HIV status on incident DM.
Figure 2.
Unadjusted relative times and 95% confidence intervals to diabetes per doubling of androgen hormone and SHBG by HIV status. P-values reported are from a comparison of relative times by HIV status.
Adjusted associations of TT, DHEAS, SHBG, and menopausal phase with incident diabetes by HIV status
After adjustment for demographic and behavioral factors and menopausal status, the association of TT with time to DM remained nearly the same as before adjustment in WWH [0.99 (95% CI 0.83, 1.18)] and in WWOH [0.99 (95% CI 0.58, 1.67)] (Table 2). The positive association per doubling of DHEAS and time to DM remained non-statistically significant [RT=1.15 (95% CI 0.95, 1.39) in WWH and RT=1.41 (95% CI 0.82, 2.44) in WWOH]. The adjusted positive associations per doubling of SHBG also remained similar to the unadjusted values [RT=1.07 (95% CI 0.82, 1.40) in WWH and RT=1.84 (95% CI 0.89, 3.82) in WWOH].
Table 2.
Adjusted relative times to diabetes mellitus by HIV status using separate models with TT, DHEAS, and SHBG
| HIV+ | HIV− | HIV+ | HIV− | HIV+ | HIV− | |
|---|---|---|---|---|---|---|
| Androgen hormone | ||||||
| TT (per doubling) | 0.99 (0.83, 1.18) |
0.99 (0.58, 1.67) |
||||
| DHEAS (per doubling) | 1.15 (0.95, 1.39) |
1.41 (0.82, 2.44) |
||||
| SHBG (per doubling) | 1.07 (0.82, 1.40) |
1.84 (0.89, 3.82) |
||||
| Menopause Transition (vs. premenopausal) |
||||||
| Early Perimenopause | 0.58 (0.27, 1.26) |
0.14 (0.01, 1.37) |
0.61 (0.29, 1.29) |
0.20 (0.03, 1.56) |
0.58 (0.27, 1.24) |
0.15 (0.02, 1.13) |
| Late Peri/Early Menopause | 0.42 (0.14, 1.24) |
0.07 (0.004, 1.23) |
0.46 (0.16, 1.31) |
0.11 (0.01, 1.41) |
0.42 (0.14, 1.24) |
0.09 (0.001, 1.08) |
| Menopausal | 0.17 (0.03, 0.93) |
0.22 (0.01, 5.05) |
0.19 (0.04, 1.00) |
0.31 (0.02, 5.20) |
0.17 (0.03, 0.94) |
0.25 (0.01, 4.18) |
| Race/Ethnicity (vs. Black) | ||||||
| Hispanic | 0.66 (0.42, 1.03) |
0.41 (0.12, 1.38) |
0.64 (0.41, 1.00) |
0.47 (0.15, 1.47) |
0.67 (0.43, 1.04) |
0.49 (0.17, 1.44) |
| White | 2.11 (0.97, 4.66) |
0.31 (0.06, 1.77) |
2.06 (0.93, 4.54) |
0.34 (0.07, 1.63) |
2.15 (0.97, 4.73) |
0.41 (0.09, 1.76) |
| 2001–2002 enrolment (vs 1994–1995) |
0.41 (0.22, 0.77) |
0.66 (0.24, 1.81) |
0.41 (0.22, 0.77) |
0.62 (0.24, 1.62) |
0.41 (0.22, 0.77) |
0.61 (0.24, 1.56) |
| BMI (per kg/m2) | 0.92 (0.88, 0.98) |
0.88 (0.80, 0.97) |
0.93 (0.88, 0.98) |
0.89 (0.81, 0.98) |
0.93 (0.88, 0.98) |
0.90 (0.83, 0.97) |
| Current Smoking | 0.62 (0.39, 0.99) |
0.46 (0.15, 1.36) |
0.61 (0.38, 0.99) |
0.40 (0.13, 1.20) |
0.62 (0.39, 0.99) |
0.44 (0.17, 1.18) |
| HCV infected | 0.68 (0.38, 1.25) |
1.01 (0.16, 6.30) |
0.81 (0.44, 1.49) |
1.59 (0.33, 7.72) |
0.65 (0.35, 1.23) |
0.58 (0.09, 3.60) |
| CD4 cell count (per 50 cells/mm3) |
1.00 (0.96, 1.04) |
1.00 (0.96, 1.04) |
1.00 (0.97, 1.04) |
|||
| HIV RNA (per log10 copies/ml) |
0.90 (0.72, 1.13) |
0.92 (0.74, 1.16) |
0.90 (0.72, 1.13) |
|||
| Current ART use | 0.84 (0.51, 1.39) |
0.85 (0.51, 1.40) |
0.84 (0.51, 1.38) |
In WWH, RT to incident DM were less than one and decreased as menopausal phase progressed (Table 2). In the multivariable models for TT, we found that being in early perimenopause, late perimenopause/early menopause, and menopause was associated with 42%, 58%, and 83% shorter times to DM, respectively, when compared to women who were premenopausal. Similar estimated effects of menopausal phase were also found in the models of DHEAS and SHBG exposure. Being premenopausal was also associated with longer times to DM in WWOH. Greater BMI, report of current smoking, and being of Hispanic ethnicity were associated with shorter times to DM; having enrolled in the cohort during 2001–2002 (versus 1994–1995) was associated with shorter time to DM in both women with and without HIV, but the association only reached statistical significance in WWH.
Discussion
In our large cohort of women with and without HIV, consistent with prior studies [11,30], we found that WWH have lower levels of androgen hormones TT and DHEAS and higher levels of SHBG than WWOH. Higher SHBG levels and higher DHEAS levels were non-significantly associated with longer time to DM progression in both WWH and WWOH, but the magnitude of association was slightly greater in WWOH, which could suggest that HIV alters the effects of SHBG and DHEAS on DM. Regardless of HIV status, TT levels were not associated with time to DM. Compared to premenopause, being in early perimenopause, late perimenopause, or menopause was associated with shorter times to DM in WWH.
Our findings that HIV may alter androgen hormone and SHBG levels in women are consistent with prior studies in the WIHS limited to premenopausal women [11,30] and corroborate the findings of a small study from before the era of highly active antiretroviral therapy (HAART) [31]. In the latter study, WWH were categorized as nonwasting, early wasting, and late wasting (as determined by ideal body weight) and were compared to healthy controls. SHBG levels were higher among WWH than healthy controls regardless of wasting status [31]. By contrast, DHEAS and free testosterone were higher only in WWH who had late wasting when compared to healthy controls [31]. These studies suggest that HIV infection is associated with altered SHBG, and lean mass may be associated with androgen hormone levels in WWH.
While SHBG has been associated with DM risk in several epidemiologic studies, the complex biologic mechanisms that underlie this risk are not fully understood [8,32]. SHBG binds circulating androgen hormones such as testosterone and 17-beta-estradiol, mediating metabolic disease through modulation of the bioavailability of androgen hormones, but SHBG may also function as an independent biomarker of metabolic derangements [33]. Similar to our findings, results from a study within the large multisite Diabetes Prevention Program (DPP) that included a select US population of people with obesity at risk for DM found that higher SHBG was non-statistically significantly associated with lower incident DM in women regardless of menopausal status [34]. That study concluded that traditional risk factors such as glucose intolerance and being overweight may be more important than SHBG as a marker of diabetes risk [34]. Interestingly, our study cohort is also mostly overweight or obese and at high risk for DM. We also found that regardless of HIV status, greater BMI was significantly associated with shorter time to DM in models that adjusted for menopausal status and separately for each androgen hormone and SHBG.
Our findings that TT was not associated with incident DM is similar to the findings of other general population studies. Interestingly, the DPP trial showed that the direction of the association of testosterone with incident DM was in a negative direction in post-menopausal women but in a positive direction in premenopausal women [34]. Our study included women in pre- and post-menopause at the index visit and could partly explain the lack of an association of TT with DM risk. Furthermore, we noted a higher proportion of undetectable TT and DHEAS levels as women progressed to menopause. The DPP study also showed that greater DHEAS was non-statistically significantly associated with lower DM incidence in postmenopausal women, but in premenopausal women the association shifted from the null to a positive direction when the model was additionally adjusted for waist circumference, a marker of visceral obesity. In our study, we observed that in WWOH, the association of higher DHEAS with incident DM was not statistically significant after adjustment for demographic and behavioral factors and BMI. Taken together, this suggests that studies of androgen hormones and DM in women should consider both menopausal status and body composition. Given that the magnitude of association was higher in WWOH than in WWH, it could suggest that HIV may differentially modulate the associations between body composition and androgen hormone levels.
We found that in WWH, even after accounting for chronologic age, being in the early perimenopausal, late perimenopausal, and menopausal phases was associated with shorter times to DM when compared to premenopause. While unmeasured age-related confounders might have played a role, a more likely potential mechanism for this may be through the aromatization of estrogens in the adipose tissue post-menopause. In postmenopausal women, DM has been associated with endogenous estrogens [7,35]. Among obese postmenopausal women in the general population, higher estrogen levels, especially estrone, have been found, and are felt to be due to conversion of androstenedione and testosterone in the adipose tissue [18,36]. Our cohort includes a high prevalence of overweight and obese women, which may have led to overall higher estrogen levels, affecting the associations of androgens with incident DM in our analysis. These findings suggest that further investigation of the contribution of chronologic and ovarian aging, adiposity, and androgen hormones with DM risk is needed.
Our study has several strengths. This is the first study evaluating the association of androgen hormone levels with incident DM in a large, decades-long, and ethnically diverse cohort of nationally representative WWH and a comparison group of WWOH who have similar demographics and risk behaviors. Whereas many similar analyses rely on cross-sectional data, we used repeated measures of androgen hormones collected from a morning blood draw, an objective biomarker AMH to define menopausal phase, and incident DM data that were derived from fasting blood draws over an average of 10 years of follow-up.
There are limitations to our study. Seventy percent of our measures of androgen hormone and SHBG levels were drawn in the premenopausal or early perimenopausal phase and may not have fully captured expected changes across the menopausal transition, and many participants only had one measurement of androgen hormone or SHBG level. In addition, while antiretroviral therapy (ART) use was included as a covariate in our analysis, we did not distinguish by ART class despite differential effects of ART classes on body composition [25,37,38], which could have affected the development of incident DM. Assessment of testosterone levels in women may be best achieved by mass spectrometry measurements, which were not performed in this study. Last, we were not able to examine the association of androgen hormones and SHBG with DM in postmenopausal women given the small number of incident diabetes events in a limited number of postmenopausal women over follow-up.
Conclusion
We found that WWH have lower levels of androgen hormones TT and DHEAS and higher levels of SHBG than WWOH. However, the overall contribution of androgen hormones and SHBG to DM incidence in WWH is small or null, especially for TT, relative to traditional risk factors for DM. The menopausal transition may be a better hormonal indicator of diabetes risk in WWH. Whether the magnitude of the associations of androgen hormones and SHBG is altered in post-menopausal WWH warrants further study.
Supplementary Material
Key points:
Despite alterations in androgen hormone and SHBG levels in HIV, regardless of HIV status, higher SHBG and DHEAS were associated with non-statistically significant slower progression to diabetes. The menopausal transition may be a better hormonal indicator of diabetes risk in WWH.
Funding
The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health (NIH). MWCCS (Principal Investigators): Atlanta CRS (Ighovwerha Ofotokun, Anandi Sheth, and Gina Wingood), U01-HL146241; Baltimore CRS (Todd Brown and Joseph Margolick), U01-HL146201; Bronx CRS (Kathryn Anastos, David Hanna, and Anjali Sharma), U01-HL146204; Brooklyn CRS (Deborah Gustafson and Tracey Wilson), U01-HL146202; Data Analysis and Coordination Center (Gypsyamber D’Souza, Stephen Gange and Elizabeth Topper), U01-HL146193; Chicago-Cook County CRS (Mardge Cohen and Audrey French), U01-HL146245; Chicago-Northwestern CRS (Steven Wolinsky, Frank Palella, and Valentina Stosor), U01-HL146240; Northern California CRS (Bradley Aouizerat, Jennifer Price, and Phyllis Tien), U01-HL146242; Los Angeles CRS (Roger Detels and Matthew Mimiaga), U01-HL146333; Metropolitan Washington CRS (Seble Kassaye and Daniel Merenstein), U01-HL146205; Miami CRS (Maria Alcaide, Margaret Fischl, and Deborah Jones), U01-HL146203; Pittsburgh CRS (Jeremy Martinson and Charles Rinaldo), U01-HL146208; UAB-MS CRS (Mirjam-Colette Kempf, Jodie Dionne-Odom, Deborah Konkle-Parker, and James B. Brock), U01-HL146192; UNC CRS (Adaora Adimora and Michelle Floris-Moore), U01-HL146194. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute (NHLBI), with additional co-funding from the Eunice Kennedy Shriver National Institute Of Child Health & Human Development (NICHD), National Institute On Aging (NIA), National Institute Of Dental & Craniofacial Research (NIDCR), National Institute Of Allergy And Infectious Diseases (NIAID), National Institute Of Neurological Disorders And Stroke (NINDS), National Institute Of Mental Health (NIMH), National Institute On Drug Abuse (NIDA), National Institute Of Nursing Research (NINR), National Cancer Institute (NCI), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute on Minority Health and Health Disparities (NIMHD), and in coordination and alignment with the research priorities of the National Institutes of Health, Office of AIDS Research (OAR). MWCCS data collection is also supported by UL1-TR000004 (UCSF CTSA), UL1-TR003098 (JHU ICTR), UL1-TR001881 (UCLA CTSI), P30-AI-050409 (Atlanta CFAR), P30-AI-073961 (Miami CFAR), P30-AI-050410 (UNC CFAR), P30-AI-027767 (UAB CFAR), P30-MH-116867 (Miami CHARM), UL1-TR001409 (DC CTSA), KL2-TR001432 (DC CTSA), and TL1-TR001431 (DC CTSA).
The study was also supported by the National Institute of Allergy and Infectious Diseases (K24 AI 108516 [PCT]) and the National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK 109823 [PCT]).
The authors gratefully acknowledge the contributions of the study participants and dedication of the staff at the MWCCS sites.
Disclosures
Rebecca A. Abelman: No disclosures.
Michael Schneider: No disclosures.
Christopher Cox: No disclosures.
Geralyn Messerlian: No disclosures.
Mardge Cohen: No disclosures.
Deborah Gustafson: No disclosures.
Michael Plankey: No disclosures.
Anjali Sharma: No disclosures.
Jennifer Price: No disclosures.
Carl Grunfeld: No disclosures.
Phyllis Tien: Grant support from Merck.
References
- 1.Tien PC, Schneider MF, Cox C, et al. Association of HIV infection with incident diabetes mellitus: impact of using hemoglobin A1C as a criterion for diabetes. Journal of acquired immune deficiency syndromes (1999) 2012; 61:334–340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Tien PC, Schneider MF, Cole SR, et al. Antiretroviral therapy exposure and incidence of diabetes mellitus in the Women’s Interagency HIV Study. AIDS 2007; 21:1739–1745. [DOI] [PubMed] [Google Scholar]
- 3.Howard AA, Hoover DR, Anastos K, et al. The Effects of Opiate Use and Hepatitis C Virus Infection on Risk of Diabetes Mellitus in the Women’s Interagency HIV Study. JAIDS Journal of Acquired Immune Deficiency Syndromes 2010; 54:152–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Johnston CD, Hoover DR, Shi Q, et al. White Blood Cell Counts, Lymphocyte Subsets, and Incident Diabetes Mellitus in Women Living With and Without HIV. AIDS Research and Human Retroviruses 2020; 36:131–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Glesby MJ, Hanna DB, Hoover DR, et al. Abdominal fat depots, insulin resistance, and incident diabetes mellitus in women with and without HIV infection. AIDS 2018; 32:1643–1650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Frasco MA, Karim R, Van Den Berg D, et al. Antiretroviral therapy modifies the genetic effect of known type 2 diabetes-associated risk variants in HIV-infected women. AIDS 2014; 28:1815–1823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ding EL, Song Y, Malik VS, Liu S. Sex differences of endogenous sex hormones and risk of type 2 diabetes: a systematic review and meta-analysis. JAMA 2006; 295:1288–1299. [DOI] [PubMed] [Google Scholar]
- 8.Ding EL, Song Y, Manson JE, et al. Sex hormone-binding globulin and risk of type 2 diabetes in women and men. N Engl J Med 2009; 361:1152–1163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Brand JS, van der Tweel I, Grobbee DE, Emmelot-Vonk MH, van der Schouw YT. Testosterone, sex hormone-binding globulin and the metabolic syndrome: a systematic review and meta-analysis of observational studies. Int J Epidemiol 2011; 40:189–207. [DOI] [PubMed] [Google Scholar]
- 10.Liang J, Peng Q, Yang X, Yang C. The association between serum testosterone levels and metabolic syndrome among women. Diabetol Metab Syndr 2021; 13:26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Karim R, Mack WJ, Kono N, et al. Gonadotropin and sex steroid levels in HIV-infected premenopausal women and their association with subclinical atherosclerosis in HIV-infected and -uninfected women in the women’s interagency HIV study (WIHS). J Clin Endocrinol Metab 2013; 98:E610–618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Burger HG, Dudley EC, Cui J, Dennerstein L, Hopper JL. A Prospective Longitudinal Study of Serum Testosterone, Dehydroepiandrosterone Sulfate, and Sex Hormone-Binding Globulin Levels through the Menopause Transition. 2000; 85:7. [DOI] [PubMed] [Google Scholar]
- 13.Crawford S, Santoro N, Laughlin GA, et al. Circulating dehydroepiandrosterone sulfate concentrations during the menopausal transition. J Clin Endocrinol Metab 2009; 94:2945–2951. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Burger HG. Androgen production in women. Fertil Steril 2002; 77 Suppl 4:S3–5. [DOI] [PubMed] [Google Scholar]
- 15.Carr MC. The emergence of the metabolic syndrome with menopause. J Clin Endocrinol Metab 2003; 88:2404–2411. [DOI] [PubMed] [Google Scholar]
- 16.Donato GB, Fuchs SC, Oppermann K, Bastos C, Spritzer PM. Association between menopause status and central adiposity measured at different cutoffs of waist circumference and waist-to-hip ratio. Menopause 2006; 13:280–285. [DOI] [PubMed] [Google Scholar]
- 17.Greendale GA, Sternfeld B, Huang M, et al. Changes in body composition and weight during the menopause transition. JCI Insight 2019; 4:e124865, 124865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Baglietto L, English DR, Hopper JL, et al. Circulating steroid hormone concentrations in postmenopausal women in relation to body size and composition. Breast Cancer Res Treat 2009; 115:171–179. [DOI] [PubMed] [Google Scholar]
- 19.Gruber CJ, Tschugguel W, Schneeberger C, Huber JC. Production and Actions of Estrogens. N Engl J Med 2002; 346:340–352. [DOI] [PubMed] [Google Scholar]
- 20.D’Souza G, Bhondoekhan F, Benning L, et al. Characteristics of the MACS/WIHS Combined Cohort Study: Opportunities for Research on Aging With HIV in the Longest US Observational Study of HIV. Am J Epidemiol 2021; 190:1457–1475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Adimora AA, Ramirez C, Benning L, et al. Cohort Profile: The Women’s Interagency HIV Study (WIHS). Int J Epidemiol 2018; 47:393–394i. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Scherzer R, Greenblatt RM, Merhi ZO, et al. Use of antimüllerian hormone to predict the menopausal transition in HIV-infected women. American Journal of Obstetrics and Gynecology 2017; 216:46.e1–46.e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Freeman EW, Sammel MD, Lin H, Boorman DW, Gracia CR. Contribution of the rate of change of antimüllerian hormone in estimating time to menopause for late reproductive-age women. Fertil Steril 2012; 98:1254–1259.e1-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Broer SL, Eijkemans MJC, Scheffer GJ, et al. Anti-mullerian hormone predicts menopause: a long-term follow-up study in normoovulatory women. J Clin Endocrinol Metab 2011; 96:2532–2539. [DOI] [PubMed] [Google Scholar]
- 25.Abelman RA, Nguyen TTJ, Ma Y, et al. Body Composition Changes over the Menopausal Transition in Women With and Without HIV. Clinical Infectious Diseases 2023; :ciad165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Abelman RA, Nguyen TTJ, Ma Y, et al. Body Composition Changes Over the Menopausal Transition in Women With and Without Human Immunodeficiency Virus. Clin Infect Dis 2023; 77:265–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Harlow SD, Gass M, Hall JE, et al. Executive Summary of the Stages of Reproductive Aging Workshop + 10: Addressing the Unfinished Agenda of Staging Reproductive Aging. The Journal of Clinical Endocrinology & Metabolism 2012; 97:1159–1168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Diagnosis | ADA. Available at: https://diabetes.org/diabetes/a1c/diagnosis. Accessed 22 November 2022.
- 29.Cox C, Chu H, Schneider MF, Muñoz A. Parametric survival analysis and taxonomy of hazard functions for the generalized gamma distribution. Statist Med 2007; 26:4352–4374. [DOI] [PubMed] [Google Scholar]
- 30.Coburn SB, Dionne-Odom J, Alcaide ML, et al. The Association Between HIV Status, Estradiol, and Sex Hormone Binding Globulin Among Premenopausal Women in the Women’s Interagency HIV Study. Journal of Women’s Health 2022; :jwh.2021.0276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Grinspoon S, Corcoran C, Miller K, et al. Body composition and endocrine function in women with acquired immunodeficiency syndrome wasting. J Clin Endocrinol Metab 1997; 82:1332–1337. [DOI] [PubMed] [Google Scholar]
- 32.Muka T, Nano J, Jaspers L, et al. Associations of Steroid Sex Hormones and Sex Hormone-Binding Globulin With the Risk of Type 2 Diabetes in Women: A Population-Based Cohort Study and Meta-analysis. Diabetes 2017; 66:577–586. [DOI] [PubMed] [Google Scholar]
- 33.Simons PIHG, Valkenburg O, Stehouwer CDA, Brouwers MCGJ. Sex hormone–binding globulin: biomarker and hepatokine? Trends in Endocrinology & Metabolism 2021; 32:544–553. [DOI] [PubMed] [Google Scholar]
- 34.Mather KJ, Kim C, Christophi CA, et al. Steroid Sex Hormones, Sex Hormone–Binding Globulin, and Diabetes Incidence in the Diabetes Prevention Program. The Journal of Clinical Endocrinology & Metabolism 2015; 100:3778–3786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ding EL, Song Y, Manson JE, Rifai N, Buring JE, Liu S. Plasma sex steroid hormones and risk of developing type 2 diabetes in women: a prospective study. Diabetologia 2007; 50:2076–2084. [DOI] [PubMed] [Google Scholar]
- 36.Karim R, Mack WJ, Hodis HN, Roy S, Stanczyk FZ. Influence of age and obesity on serum estradiol, estrone, and sex hormone binding globulin concentrations following oral estrogen administration in postmenopausal women. J Clin Endocrinol Metab 2009; 94:4136–4143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Sax PE, Erlandson KM, Lake JE, et al. Weight Gain Following Initiation of Antiretroviral Therapy: Risk Factors in Randomized Comparative Clinical Trials. Clinical Infectious Diseases 2020; 71:1379–1389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Lake JE, Trevillyan J. Impact of Integrase inhibitors and tenofovir alafenamide on weight gain in people with HIV. Curr Opin HIV AIDS 2021; 16:148–151. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.