Factors associated with D-dimer levels in HIV-infected individuals

Abstract

Background: Higher plasma D-dimer levels are strong predictors of mortality in HIV+ individuals. The factors associated with D-dimer levels during HIV infection, however, remain poorly understood.

Methods: In this cross-sectional study, participants in three randomized controlled trials with measured D-dimer levels were included (N = 9,848). Factors associated with D-dimer were identified by linear regression. Covariates investigated were: age, gender, race, body mass index, nadir and baseline CD4+ count, plasma HIV RNA levels, markers of inflammation (C-reactive protein [CRP], interleukin-6 [IL-6]), antiretroviral therapy (ART) use, ART regimens, co-morbidities (hepatitis B/C, diabetes mellitus, prior cardiovascular disease), smoking, renal function (estimated glomerular filtration rate [eGFR] and cystatin C) and cholesterol.

Results: Women from all age groups had higher D-dimer levels than men, though a steeper increase of D-dimer with age occurred in men. Hepatitis B/C co-infection was the only co-morbidity associated with higher D-dimer levels. In this subgroup, the degree of hepatic fibrosis, as demonstrated by higher hyaluronic acid levels, but not viral load of hepatitis viruses, was positively correlated with D-dimer. Other factors independently associated with higher D-dimer levels were black race, higher plasma HIV RNA levels, being off ART at baseline, and increased levels of CRP, IL-6 and cystatin C. In contrast, higher baseline CD4+ counts and higher high-density lipoprotein cholesterol were negatively correlated with D-dimer levels.

Conclusions: D-dimer levels increase with age in HIV+ men, but are already elevated in women at an early age due to reasons other than a higher burden of concomitant diseases. In hepatitis B/C co-infected individuals, hepatic fibrosis, but not hepatitis viral load, was associated with higher D-dimer levels.

Figure 1. Demographics and D-dimer levels.

(a) SMART/ESPRIT/SILCAAT; adjusted for demographics, HIV-specific variables and biomarkers of inflammation. (b) SMART/ESPRIT; as in (a) and also adjusted for co-morbidities (CVD, DM and hepatitis B/C) and eGFR. (c) SMART only; as in (b) and also adjusted for smoking, cystatin C and cholesterol levels.

Figure 2. Figure 2. D-dimer levels across age groups stratified by gender (a).

(a) SMART/ESPRIT/SILCAAT; adjusted for demographics, HIV-specific variables and biomarkers of inflammation.

Figure 3. HIV-specific variables and D-dimer levels.

(a) SMART/ESPRIT/SILCAAT (N = 9848; 821 of whom were off ART at baseline); adjusted for demographics, HIV-specific variables and biomarkers of inflammation. (b) SMART/ESPRIT (N = 6928); as in (a) and also adjusted for co-morbidities (CVD, DM and hepatitis B/C) and eGFR. (c) SMART (N = 4488); as in (b) and also adjusted for smoking, cystatin C and cholesterol levels.

Figure 4. Biomarkers of Inflammation and D-dimer levels.

(a) SMART/ESPRIT/SILCAAT (N = 9848); adjusted for demographics, HIV-specific variables and biomarkers of inflammation. (b) SMART/ESPRIT (N = 6928); as in (a) and also adjusted for co-morbidities (CVD, DM and hepatitis B/C) and eGFR. (c) SMART (N = 4488); as in (b) and also adjusted for smoking, cystatin C and cholesterol levels.

Figure 5. Correlation between D-dimer and IL-6 levels*.

* Plotted values refer to log₁₀ transformed levels of units of measurement based on the molecular masses of D-dimer and IL-6 (nmol/L for D-dimer and fmol/L for IL-6).