Cardio-metabolic risk profile of people living with HIV: Is retinal microvascular geometric morphology a marker of effect?

Kgokane, Boipelo Mirriam Lepetjia (2022-04)

Thesis (MSc)--Stellenbosch University, 2022.

Thesis

Background and Aim Cardiovascular disease in people living with HIV has become of great concern. HI-viral factors, ART-toxicity and HIV/ART-associated cardiometabolic adverse effects have been implicated in the development of cardiovascular disease. Retinal microvascular geometric features may be potential useful markers of these effects. We aimed to investigate whether altered retinal microvascular geometric features are markers of HIV, ART and/or HIV/ART-associated cardiometabolic effects in a study population from the Western Cape Province. Methods The study followed a cross-sectional (HIV-free: n = 88 and HIV+ART: n = 122) and longitudinal (baseline vs. 18-month follow-up for HIV+ART only: n = 82) study design. Volunteering participants were recruited from health care clinics. Demographic, lifestyle, socioeconomic and anthropometric data were collected. Fasting blood and urine samples were collected and transported to the National Health Laboratory Services for biochemical analyses. Retinal images were obtained (Canon CR-2 camera) and vessel features quantified (MONA REVA 2.1.1 software). Linear stepwise regression (cross-sectional) and linear mixed model (longitudinal) analyses were applied to elucidate independent associations and statistical significance of p < 0.05. Results Population characteristics: The study population was relatively young (HIV-free:44.06±11.09 and HIV+ART:40.35±8.94years) and mostly women (HIV-free:80.7% and HIV+ART:63.1%). The baseline median/mean viral load (VL), CD4 cell count and ART-duration were 50 (10 to 675032) copies mRNA/mL, 539.92±237.16 cells/mm3 and 166 (1 to 707) weeks respectively. Cardiometabolic results: Body mass index (BMI) (24.50±6.65 vs. 28.25±7.68kg/m2, p < 0.001) was significantly lower in HIV+ART vs. HIV-free. ∆BMI in HIV+ART was significantly correlated with average arterial tree diameter (r = 0.323, p < 0.05), total length of skeletonised tree (r = 0.355, p < 0.01) and arteriolar branching angle (r = 0.234, p < 0.05). High density lipoprotein-cholesterol (1.59±0.74 vs. 1.39±0.45mmol/L, p = 0.019) and gamma-glutamyl transferase (GGT) (43.5 (14 to 494) vs. 27.0 (11 to 814)U/L, p < 0.001) were significantly higher in HIV+ART vs. HIV-free, but decreased in HIV+ART (Baseline vs. Follow-up HDL:1.62±0.77 vs. 1.44±0.64mmol/L, p = 0.017 and GGT:45 (14 to 494) vs. 41.50 (14 to 219)U/L, p = 0.004). HDL was significantly correlated with central retinal venular equivalent (CRVE) (r =-0.195, p < 0.01) and GGT with venular branching optimality (r = 0.180, p < 0.05). HIV and ART results: Cross-sectionally, HIV+ART status independently associated with CRVE (-0.146 (- 0.280 to -0.012), p = 0.033) and arteriolar and venular mother branch (D0), first daughter branch (D1) and second daughter branch (D2) (p < 0.05, respectively). VL (-0.198 (-0.025 to -0.001), p = 0.037) and ART- duration (0.188 (0.001 to 0.024), p = 0.047) were independently associated with arteriolar-venular ratio (AVR). Longitudinally, VL independently associated with CRVE (0.096 (0.017 to 0.175), p = 0.018) and AVR (-0.003 (-0.0006 to 0.000003), p = 0.046). CD4 cell count was independently associated with number of branchpoints 0.042 (-0.002 to 0.086), p=0.006) and endpoints (3.0 (0.750 to 5.250), p=0.010). HIV duration independently associated with lacunarity (-0.0080 (-0.0150 to -0.0010), p=0.036) and fractal analyses (0.011 (0.0001 to 0.021), p=0.045). 2nd-line ART was independently associated with CRVE (8.58 (0.35 to 16.81), p=0.041) and ART-duration with fractal analysis (-0.022 (-0.037 to -0.008), p=0.003). Discussion and conclusion HIV+ART appeared to have a more favourable cardiovascular risk profile vs. HIV-free. Various markers of HIV/ART and HIV-ART-associated cardiometabolic risk factors were associated with retinal vessel features and associations appeared mostly favourable/cardioprotective. These results indicate that retinal vessel geometric features may be potential markers of the effects of HIV/ART and/or associated cardiometabolic risk factors in the current study population.

Agtergrond en Doelstelling Kardiovaskulêre siekte in MIV-positiewe individue wek groot kommer. MI-virale faktore, antiretrovirale terapie (ART)-toksisiteit en MIV/ART-geassosieerde kardiometaboliese newe-effekte word geïmpliseer. Retinale mikrovaskulêre geometriese patrone mag potensieël nuttige merkers van hierdie effekte wees. Die doel van hierdie studie was om te ondersoek of veranderinge in retinale mikrovaskulêre geometriese patrone as merkers van MIV, ART en/of MIV/ART-geassosieerde kardiometaboliese effekte in ‘n studie populasie van die Wes-Kaap beskou kan word. Metodes Die studie het ‘n deursnee (MIV-vry: n = 88 en MIV+ART: n = 122) en longitudinale (basislyn vs. 18-maande opvolg vir slegs MIV+ART: n = 82) studie-ontwerp gevolg. Vrywillige deelnemers was vanaf gesondheidsorg klinieke gewerf. Demografiese, lewenstyl, sosio-ekonomiese en antropometriese inligting is ingewin. Vastende bloed- en urinemonsters was versamel en na die NHLS vervoer vir biochemiese ontledings. Retinale beelde was vasgelê (Canon CR2-kamera) en bloedvat patrone was gekwantifiseer (MONA REVA 2.1.1 sagteware). Liniêre, stapsgewyse regressie (deursnee studie) en liniêre gemengde model (longitudinale studie) analises was aangewend om onafhanklike assosiasies te ondersoek (statistiese beduidenheid: p<0.05). Resultate Populasie eienskappe: Die studie populasie was relatief jonk (MIV-vry:44.06±11.09 en MIV+ART:40.35±8.94 jaar) en het meestal uit vroue bestaan (MIV-vry:80.7% and MIV+ART:63.1%). Die basislyn mediaan/gemiddelde virale lading (VL), CD4 seltelling en ART-duur was onderskeidelik 50 (10 tot 675032) kopieë bRNA/mL, 539.92±237.16 selle/mm3 and 166 (1 to 707) weke. Kardiometaboliese resultate: Liggaam-massa-indeks (BMI) (24.50±6.65 vs. 28.25±7.68kg/m2, p<0.001) was betekenisvol laer in MIV+ART vs. MIV-vry. ∆BMI in MIV+ART was betekenisvol gekorreleer met die gemiddelde arteriële boom deursnit (r=0.323, p<0.05), totale lengte van die skelet boom (r=-0.355, p<0.01) en arteriolêre vertakkingshoek (r=0.234, p<0.05). HDL-cholesterol (1.59±0.74 vs. 1.39±0.45mmol/L, p=0.019) en GGT (43.5 (14 to 494) vs. 27.0 (11 to 814) U/L, p<0.001) was betekenisvol hoër in MIV+ART vs. MIV-vry, maar laer in MIV+ART (Basislyn vs. Opvolg HDL:1.62±0.77 vs. 1.44±0.64mmol/L, p=0.017 en GGT:45 (14 to 494) vs. 41.50 (14 tot 219)U/L, p=0.004). HDL het betekenisvol gekorreleer met CRVE (r=-0.195, p<0.01), en GGT met venulêre vertakking (r=0.180, p<0.05). MIV en ART resultate: Deursnee studie: MIV+ART status was onafhanklik geassosieer met CRVE (-0.146 (-0.280 tot -0.012), p=0.033), en arteriolêre en venulêre D0, D1 en D2 (p<0.05 onderskeidelik). VL (-0.198 (-0.025 tot -0.001), p = 0.037) and ART-duur (0.188 (0.001 tot 0.024), p=0.047) het onafhanklik met AVR geassosieer. Longitudinale studie: VL was onafhanklik geassosieer met CRVE (0.096 (0.017 tot 0.175), p=0.018) en AVR (-0.003 (-0.0006 tot 0.000003), p=0.046). CD4 telling was onafhanklik geassosieer met die aantal takpunte (0.042 (-0.002 tot 0.086), p=0.006) en eindpunte (3.0 (0.750 to 5.250), p=0.010). MIV-duur was onafhanklik geassosieer met lakunariteit (-0.0080 (-0.0150 tot -0.0010), p=0.036) en fraktale analises (0.011 (0.0001 tot 0.021), p=0.045). Tweede-lyn ART was onafhanklik geassosieer met CRVE (8.58 (0.35 tot 16.81), p=0.041) en ART-duur met fraktale analise (-0.022 (-0.037 tot -0.008), p=0.003). Bespreking en gevolgtrekking MIV+ART individue blyk oor die algemeen ‘n meer gunstige kardiovaskulêre profiel te toon as mense sonder MIV. Verskeie merkers van MIV/ART en MIV-ART-geassosieerde kardiometaboliese risiko faktore was met retinale bloedvat eienskappe geassosieer, en die assosiasies was meestal gunstig / kardiobeskermend. Hierdie resultate dui daarop dat retinale bloedvat geometriese eienskappe as potensiële merkers van die effekte van MIV/ART en/of geassosieerde kardiometaboliese risiko faktore in die huidige studie populasie beskou kan word.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/124674
This item appears in the following collections: