Aging,Frailty,Disability & Immune Senescence

Abstract P02

Antiviral Therapy 2019; 24 Suppl 1:A39

Validation of a measurement tool for estimating step counts under free-living conditions

C Wu ¹ , K Thompson ² , R Latif ¹ , A Guerson Gil ¹ , A Oke ¹ , F Okoro ¹ , DP Kotler ¹

¹Jacobi Medical Center, Albert Einstein College of Medicine, Bronx, NY, USA; ²Coalition on Positive Health Empowerment, New York, NY, USA

Background: The recognition of frailty as a clinical entity has increased interest in the assessment of skeletal muscle function, which is predictive of clinical outcomes. Advances in low cost physiological monitoring devices have led to the possibility of making objective measures of physical activity available for clinical or research use. Developing a tool for assessing physical activity during free-living conditions may be important as an indicator of a change in behaviour. However, there are uncertainties about the precision and accuracy of various monitoring devices.

Aim: We estimated the accuracy and precision of a pedometer P and an activity monitor (AM) worn on the torso or on the wrist under laboratory conditions. We compared total daily and exercise step counts using the P and an AM and examined the effects of device type and location on the body on total daily step counts under free-living conditions. We also performed power calculations for a hypothetical clinical trial, using Cohen's d and measurements of total daily and exercise step counts

Methods: Studies were performed in nine healthy adults who wore a pedometer (P; Omron Alvita HJ-327T) and an activity monitor (AM; Letscom Fitness Tracker ID130Plus ColorHR), both with Bluetooth capability. Data and statistical analyses were by paired t-test, oneway ANOVA and Bland–Altman analyses.

Results: Immediate test-retest differences averaged 1% for steps and time during a 400-metre hall walk while delayed (>4 weeks) test-retest results averaged −2% and average times −5%. Total daily step counts averaged 1,215 steps higher on a wrist-worn AM than on a torso-worn P, while step counts during exercise, defined as 60 or more steps/min for 10 or more min, were similar using the two devices (6 measures per subject). Separate studies were performed varying the placement of both the P and AM. Daily step counts on wrist-worn devices were significantly higher than when worn on the torso. The AM had higher counts than the P when both were worn on the wrist but not on the torso. In a hypothetical intervention that would result in a 10% increase in total or a 20% increase in exercise steps, the number needed to demonstrate statistical significance was estimated using Cohen's d and P data. Group sizes of 14 and 24 were needed for total daily and exercise steps, respectively, to provide statistical significance at the 0.05 level for a two-sided test at 80% power.

Conclusions: Both device sensitivity and location on the body affect step counts, with location having a stronger effect. Wearing the sensor on the torso may be the most accurate measurement. The power analysis demonstrates that the P measurement of changes in either total or exercise steps can demonstrate statistical significance with relatively small sample sizes.

Abstract P03

Antiviral Therapy 2019; 24 Suppl 1:A40

Pilot study assessing the Rotterdam Healthy Aging Score in a cohort of HIV-positive adults

M Ren ¹ , ² , L Szadkowski ¹ , ² , ³ , C Simon ¹ , ² , SL Walmsley ¹ , ² , ⁴

¹University Health Network; University of Toronto, Canada; ³Biostatistical Research Unit; ⁴Toronto General Hospital Research Institute

Purpose: A standard measure of healthy ageing would enable identification of factors predictive of health and facilitate evaluation of interventions. The Rotterdam Healthy Aging Score (HAS) is a validated multidimensional index constructed from five health domains (mental health, cognitive function, physical function, social support and quality of life). We describe the HAS distribution among a cohort of HIV-positive adults.

Methods: A prospective pilot study of 100 adults attending a tertiary HIV clinic, aged ≥40 on cART with suppressed HIV RNA. Participants completed questionnaires to calculate the HAS (range 0-14). Demographics, HAS and domain scores were compared by age and sex using the Kruskal–Wallis rank-sum or Fisher's exact test.

Results: Median (IQR) age was 56 (50-62), 81 (81%) were male, and 50 (51.5%) born in Canada. Participants aged ≥61 were more often Caucasian (75%) compared with those 40-50 (47%) and 51-60 (59%; P=0.056). Women were more often Black (53%) compared with men (9%; P<0.001). Participants ≥61 had longer HIV duration (25y; P=0.001) and lower CD4 nadir (152 cells/mm³; P=0.05) compared with 40-50 (16y, 267 cells/mm³) or 51-60 (22y, 190 cells/mm³). Median (IQR) CD4 was 574 (417-790) cells/mm³. Median (IQR) HAS was 12 (10-13) and 38 (38.4%) achieved a score >12 (considered healthy ageing). HAS did not vary by age group (P=0.72) or sex (P=0.49). Younger participants were more likely to have low mental health scores (27% for age 40-50 and 27% for age 51-60) compared with those ≥61 (6%; P=0.04). Women were more likely to have low pain scores (that is, experience greater pain, 16%) compared with men (1%; P=0.02).

Conclusions: HAS scores ranged from 4-14 in this cohort of older treated HIV adults with 38% attaining the ‘healthy’ range. The HAS requires further study for its ability to discriminate health outcomes and determine health domains that dominate poor scores.

Abstract P04

Antiviral Therapy 2019; 24 Suppl 1:A41

Multimorbidity in elderly subjects according to year of diagnosis of HIV-infection – a cross-sectional DATAIDS Cohort Study

M Demontès ¹ , S Eymard Duvernay ² , C Allavena ³ , T Jovelin ⁴ , J Reynes ⁵ , M Hentzien ⁶ , I Ravaux ⁷ , P Delobel ⁸ , D Rey ⁹ , T Ferry ¹⁰ , O Robineau ¹¹ , P Pugliese ¹² , C Duvivier ¹³ , P Krolak-Salmon ¹⁴ , A Gagneux-Brunon ¹⁵ , A Makinson ¹⁶ , the Dat'AIDS Study Group

¹Memory Clinical and Research Center of Lyon (CMRR), Hospital of Charpennes, University Hospital of Lyon, France. Institute of Aging I-Vie, University Hospital of Lyon, France; ²Infectious and Tropical Diseases Department, University Hospital Montpellier, Montpellier, France University Montpellier, UMI233-IRD/U1175-INSERM, Montpellier, France; ³Infectious Diseases Department, University of Nantes, CHU Hôtel Dieu, Nantes, France; ⁴Infectious Diseases Department, University of Nantes, CHU Hôtel Dieu, Nantes, France; ⁵Département des Maladies Infectieuses, CHU de Montpellier, unité Inserm U1175, Montpellier; ⁶Reims Teaching Hospitals, Robert Debré Hospital, Department of Internal Medicine, Infectious Diseases, and Clinical Immunology, Reims, France; ⁷Aix Marseille University, CNRS, IRD, INSERM, AP-HM, URMITE, IHU Méditerranée-Infection, Marseille, France; ⁸CHU de Toulouse, Service des Maladies Infectieuses et Tropicales, 31300, Toulouse, France; INSERM, UMR1043, 31300, Toulouse, France; ⁹Hôpitaux universitaires de Strasbourg, Nouvel Hôpital Civil, HIV Care Center, Strasbourg, France; ¹⁰Infectious and Tropical Diseases Department, Hospices Civils de Lyon, Claude Bernard Lyon 1 University, Lyon, France; ¹¹ Infectious Diseases Department, Gustave Dron Hospital, Tourcoing, France; ¹²Infectious Diseases Department, University of Nice, CHU L'Archet, Nice, France; ¹³APHP-Hôpital Necker-Enfants malades, Service de Maladies Infectieuses et Tropicales, Centre d'Infectiologie Necker-Pasteur, Paris, France, IHU Imagine, Paris, France. Institut Cochin - CNRS 8104 - INSERM U1016 - RIL Team: Retrovirus, Infection and Latency, Université de Paris, Paris, France. Institut Pasteur, Centre Médical de l'Institut Pasteur, Paris, France; ¹⁴Memory Clinical and Research Center of Lyon (CMRR), Hospital of Charpennes, University Hospital of Lyon, France. Institute of Aging I-Vie, University Hospital of Lyon, France; ¹⁵Service d'Infectiologie, CHU Saint-Etienne, Groupe Immunité des Muqueuses et Agents Pathogènes, Institut Presage, Université de Lyon, Lyon, France; ¹⁶Département des Maladies Infectieuses, CHU de Montpellier, unité Inserm U1175, Montpellier

Background: The number of people ageing with HIV is growing. We assessed prevalence of multimorbidity (MM) according to year of HIV diagnosis in an HIV geriatric population.

Methods: We performed a cross-sectional study of MM involving patients from the Dat'AIDS French multicentric cohort, selecting patients over 70 years old. MM was defined as at least 3 concomitant comorbidities including either high blood pressure, diabetes mellitus, osteoporosis, non-AIDS cancer, chronic renal failure, cardio and cerebrovascular disease, obesity, cachexia or hypercholesterolaemia. We defined three groups of HIV-diagnosis period: between 1983 and 1996, 1997–2006 and 2007–2018. Logistic regression models evaluated the association between MM and predictors, including calendar period of HIV-diagnosis. A secondary analysis evaluated MM as a continuous outcome and a sensitivity analysis excluded subjects with less than 200 cells/mm³.

Results: Starting 1 January 2017 and ending on 29 September 2018, 2,476 patients were included. Median age was 73 years old, 75% were men. MM prevalence was 71% in our population (median number of comorbidities three). High blood pressure and hypercholesterolaemia were the most prevalent comorbities. After adjustment for age, gender, smoking status, HCV, HBV coinfection, group of exposure, nadir CD4, and CD4:CD8 ratio, association between MM and calendar period of diagnosis was not statistically significant (P=0.169). MM was significantly associated with older age, CD4/CD8 ratio <0.8 and nadir CD4 cells <200 cells/ml. Similar results were found with secondary and sensitivity analysis.

Conclusions: In this study of elderly persons living with HIV, MM prevalence was high and increased with age, low CD4/CD8 ratio and nadir CD4 cells <200 mm³ but was not associated with calendar period of HIV-diagnosis. Duration of HIV-diagnosis should not be a criteria for selecting a population at risk of MM.

Abstract P05

Antiviral Therapy 2019; 24 Suppl 1:A42

Immune activation and chronic inflammation: is there an additional effect of HIV in a geriatric population?

D Sauce ¹ , V Pourcher ² , T Ferry ³ , L Slama ⁴ , C Allavena ⁵

¹Sorbonne Université, INSERM, Paris, France; ²Assistance Publique-Hopitaux de Paris, Sorbonne Université, Paris, France; ³Hospices Civils de Lyon, Inserm U1111, CNRS UMR5308, ENS de Lyon, UCBL1, Lyon, France; ⁴Assistance Publique-Hopitaux de Paris, Hotel Dieu Hospital, Paris, France; ⁵Hotel Dieu CHU Nantes, Nantes, France

Introduction: HIV infection has become a chronic disease, with a lower mortality, but a consequent increase in age-related non-infectious comorbidities. Metabolic disorders have been linked to the effect of combined antiretroviral therapy as well to the effects of immune activation and chronic inflammation. Whereas it is known that ageing is intrinsically associated with hyper-inflammation and immune system deterioration, the relative impact of chronic HIV infection on such inflammatory and immune activation has not yet been studied on an ageing HIV-infected population.

Objectives and methods: The objectives were to assess blood markers of immune activation and inflammation, using ultrasensitive techniques (Luminex & Simoa), in HIV-infected patients aged 75 years and older with no or one comorbidity (hypertension, renal disease, neoplasia, diabetes mellitus, cardiovascular disease, stroke, dyslipidaemia and osteoporosis), in comparison with age-adjusted HIV-uninfected individuals aged 75 years and older (control group), to identify whether biomarkers were associated with comorbidities. 28 inflammatory and immune biomarkers were performed in a centralized laboratory (INSERM U1135, Paris). Wilcoxon non-parametric tests were used to compare the levels of each biomarkers between control and HIV+ groups; logistic regression to identify biomarkers associated with comorbidity in the HIV+ group and principal component analysis (PCA) to determine clusters associated with a group or a specific comorbidity.

Results: 111 HIV+ subjects of whom 39 without comorbidity with a median age of 81 years (IQR 78–84) and a median duration of HIV infection and antiretroviral therapy of 18.2 (12–23) and 15.9 (9–19) years, respectively, were included from the Dat'AIDS cohort and compared with a control group of 63 HIV-negative subjects (median age of 83.4 years [IQR 78–89]). In the HIV+ group, four biomarkers were associated with the risk of comorbidity: MCP-1 OR (CI 95%) 0.78 (0.68, 0.91), NFL 1.42 (1.08, 1.87), neopterin 1.99 (1.33, 2.97) and sCD14 1.01 (1.00, 1.02). Six biomarkers (IL-1B, IL-7, IL-18, neopterin, sCD14 and FABP) were significantly higher in the HIV+ group compared with the control group, 11 biomarkers (MPO, IL-1RA, TNFR1, IFN gamma, MCP-1, TNF-R2, IL-22, usCRP, fibrinogen IL-6 and NFL) were lower. Despite those differences, PCA analysis did not reveal clustering between healthy control and HIV-infected patients. Similarly, PCA did not discriminate between the absence or the presence of comorbidity within HIV+ group.

Conclusions: No specific inflammatory or immune profile has been identified in HIV+ group according to comorbidity status. In this highly selected geriatric HIV population, HIV infection does not seem to have an additional impact on age-related inflammation and immune disorder. In the ageing HIV population, prevention and treatment of comorbidities could have limited both immune activation and inflammation. In order to validate this preliminary hypothesis, other studies need to be performed on a larger number of geriatric HIV-infected subjects.

Abstract P06

Antiviral Therapy 2019; 24 Suppl 1:A43

Vitamin D deficiency and frailty phenotype in HIV-infected men

FRP Bhondoekhan ¹ , L Zhang ² , B Genberg ¹ , FJ Palella Jr ³ , TT Brown ⁴ , MD Witt ⁵ , AN Hoofnagle ⁶ , AG Abraham ^1,3

¹Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA; ²University of Michigan, Ann Arbor, MI, USA; ³Northwestern University Feinberg School of Medicine, Chicago, IL, USA; ⁴Johns Hopkins University School of Medicine, Baltimore, MD, USA; ⁵Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, Torrance, CA, USA; ⁶University of Washington, Seattle, WA, USA

Objective: Frailty is a geriatric syndrome characterized as a state of diminished reserve and increased vulnerability to stressors due to dysregulation in multiple physiological systems. More HIV-infected persons are experiencing ageing-related diseases, including frailty, with the introduction of highly active antiretroviral therapy (HAART). Vitamin D deficiency, which has been linked to increased comorbidities such as osteoporosis and metabolic syndromes, may interfere with immune restoration following HAART and accelerate the onset of frailty in HIV-infected populations. We examined the association between vitamin D and frailty among HIV-infected men from the Multicenter AIDS Cohort Study.

Methods: Levels of 25-hydroxyvitamin D (25[OH]D) and 1,25-dihydroxyvitamin D (1,25[OH]₂D) were measured in serum from 625 HIV-infected men, collected 2 years post-HAART between 1999–2008. Vitamin D deficiency was defined as 25[OH]D <20 ng/ml. 1,25[OH]₂D was analysed as tertiles. Frailty was assessed at 6-month intervals between 1994–2018 using the Fried frailty phenotypic criteria with those meeting ≥3 of the 5 criteria considered frail and 1 or 2 criteria prefrail. Discrete-time multistate models examined factors associated with transitioning from non-frail to prefail to frail. Models adjusted for baseline (race, season, HCV coinfection, enrolment after 2001, enrolment centre, education, AIDS diagnosis) and time updated covariates (age, body mass index, depression, kidney function, cumulative pack-years smoking, diabetes, CD4⁺ T-lymphocyte/ mm³, HIV RNA and cumulative HAART use).

Results: HIV-infected men had a median of 24 frailty measures (IQR: 18–32) and median follow-up of 14.9 years (IQR: 11.8–18.6). At baseline, vitamin D deficiency prevalence was 41%, 60% of men were non-frail, 27% prefrail and 5% frail. Across follow-up, 6% (824) of visits were characterized as frail, 30% (4,219) prefrail and 51% (7,050) non-frail. Vitamin D deficiency had no effect on the probability of transitioning to prefrail or frail. However, non-frail men with 1,25[OH]₂D values in the smallest tertile (median: 32.9 ng/ml, IQR: 28.0–36.2 ng/ml) had 2.74 (95% CI: 1.10, 6.84) times the risk of becoming frail compared with men with 1,25[OH]₂D values in the highest tertile (median: 60.3 ng/ml, IQR: 55.5–67.6 ng/ml). There was an elevated risk of transitioning into a higher frail state (prefrail or frail) with older age, depression and diabetes among non-frail men. AIDS diagnosis and low kidney function were associated with increased risk of becoming frail, regardless of non-frail or prefrail state. Among non-frail men, CD4≥200 cells/mm³ conferred protection against becoming frail (RRR for 25[OH]D: 0.13, 95% CI: 0.04, 0.46, and RRR for 1,25[OH]₂D: 0.14, 95% CI: 0.04, 0.48), while BMI (25–29.9 kg/m²), college education and HIV RNA <50 copies/ml decreased the risk of becoming prefrail.

Conclusions: Vitamin D deficiency was not associated with transitions to frailty, while the active vitamin D metabolite, 1,25[OH]₂D, captured an increased risk of becoming frail among non-frail men. Risk factors contributed in varying degrees to increase or decrease the risk of transitions into frailty, consistent with prior work.

Abstract P07

Antiviral Therapy 2019; 24 Suppl 1:A44

CD4/CD8 ratio is a better indicator of acute phase inflammation than absolute CD4 count during virally suppressed HIV infection

R Rousseau ¹ , L Szadkowski ² , MF Saikali ³ , R Nadeem ⁴ , F Malazogu ⁴ , S Huibner ⁴ , CL Cummins ³ , R Kaul ^1,4,5 , SL Walmsley ^4,6

¹University of Toronto, Department of Immunology, Toronto, Canada; ²University Health Network, Toronto General Hospital, Biostatistical Research Unit, Toronto, Canada; ³University of Toronto, Leslie Dan Faculty of Pharmacy, Toronto, Canada; ⁴University of Toronto, Department of Medicine, Toronto, Canada; ⁵University Health Network, Toronto General Hospital, Immunodeficiency Clinic, Toronto, Canada; ⁶Toronto General Hospital Research Institute, Toronto, Canada

Purpose: Life expectancy of treated HIV infection approaches that of the general population. Persistent increased mortality and non-AIDS comorbidity is driven by residual immune activation/inflammation. Absolute CD4 count (CD4 count) and CD4/CD8 ratio (CD4/CD8) both correlate with poorer clinical outcomes in prospective cohorts. We sought to identify which parameter best correlates with biomarkers of acute phase immune activation.

Methods: We enrolled 83 consenting ART-treated HIV-positive participants with viral load suppression and diverse CD4 count and CD4/CD8 responses from a tertiary care HIV site in Toronto, Canada. In a cross-sectional study, blood plasma and peripheral blood mononuclear cells were isolated and stored for batched measurement of soluble immune activation markers by ELISA (CRP, TNF, IFNγ, sCD14, D-Dimer, I-FABP, MCP-1, VCAM, ICAM), and co-expression of HLA-DR and CD38 on CD8 T-cells (CD8 activation). Multiple log-linear regression models adjusted for age, sex, duration of ART and recent infection/vaccination were used to estimate the association of each outcome with: (1) CD4 count and (2) CD4/CD8.

Results: CD4 count and CD4/CD8 were strongly correlated (0.76; P<0.0001). A 0.1 increase in CD4/CD8 was associated with a 3% decrease in plasma TNF (P<0.01) and a 5% decrease in plasma CRP (P=0.05). Higher CD4 count (per 100 cells/mm³) was associated with a 2% decrease in sCD14 (P=0.02) while no association was observed with plasma TNF or CRP levels. Neither CD4 count nor CD4/CD8 were associated with CD8 activation.

Conclusions: CD4/CD8 and CD4 count are strongly linked, but CD4/CD8 correlates more strongly with two soluble markers of acute phase immune activation (TNF, CRP). CD4 count was more strongly associated with a marker of a distinct immune activation pathway (sCD14; TLR4 activation by LPS). Novel therapies to reduce non-AIDS comorbidity by targeting residual immune activation pathways warrant exploration in the context of persistently low CD4/CD8.

Abstract P08

Antiviral Therapy 2019; 24 Suppl 1:A45

The association of dietary protein with frailty risk in adults living with HIV in the PROSPER-HIV Study

AL Willig ¹ , D Long ¹ , L Yang ¹ , CH Davey ² , AR Webel ² , the PROSPER-HIV collaborators

¹University of Alabama at Birmingham, Birmingham, AL, USA; ²Case Western Reserve University, Cleveland, OH, USA

Introduction: Despite significant gains in lifespan due to antiretroviral therapy (ART) use, people living with HIV (PLWH) experience higher risk for age-related comorbidities compared with HIV-uninfected adults. In particular, frailty presents at a higher prevalence and earlier age in PLWH compared with the general population. PLWH age 50–65 in the CNICS cohort had a physical function status equivalent to 80+ year old adults without HIV. Nutrition interventions, specifically additional dietary protein for muscle maintenance, may reduce frailty risk in this vulnerable population, but PLWH have unique nutrition requirements that could impact the role of nutrition in disease prevention. Our objective was to determine how dietary factors may influence the frailty status of PLWH.

Methods: As part of the multisite, observational PROSPER-HIV study, adults (≥18 years) living with HIV, who are on HIV antiretroviral therapy, are virally suppressed, and part of the CNICS cohort completed three (two weekdays and one weekend) 24-h diet recalls. Dietary data were analysed using NDSR Nutrition Analysis Software to determine average consumption of dietary protein (g/day) and omega-3 fatty acids (g/day). Frailty risk was assessed using the following measures of physical function: hand-grip strength of the dominant hand (measured using the Jamar Hand grip Dynamometer); 5-time repeated sit/stand chair test; and the Short Physical Performance Battery (SPPB) consisting of timed gait speed, repeated chair stands, and balance tests, each assigned with a score 0–4. SPPB summary scores ranged 0 (frail) to 12 (not frail). Models were adjusted for age and present gender of participants.

Results: To date, 170 PLWH have been enrolled in the PROSPER-HIV study. Participants’ mean age was 53.5 years; 65% self-reported Black race; 26% were female; 1% transgender. Mean waist and hip circumferences were 101.9 cm and 106.3 cm, respectively. Participants consumed an average 78 g/day dietary protein, well below the recommended daily consumption of 1.2 g/kg/body weight. Mean hand grip strength was 33.1 ±11.4 kPa; and participants completed the chair sit/stand test in 11.9 ±3.8 s. The majority scored as non-frail (score 10–12) on the SPPB test. After controlling for age and present gender, higher dietary protein intake was correlated with hand grip strength (r=0.28; P<0.01). No association was observed with chair sit/ stand testing or overall SPPB performance.

Conclusions/discussion: In this cross-sectional analysis, dietary protein intake was associated with hand grip strength, a marker of frailty risk, among PLWH. Future investigations, including the 5-year PROSPER-HIV study, should include longitudinal assessments of dietary protein intake and physical function status to determine whether protein intake influences frailty risk over time, and what dose of dietary protein is needed to maintain adequate physical function status among PLWH.

Acknowledgements: This study was supported by research grants from the National Institute of Nursing Research (5R01NR018391) and from the National Institute of Allergy and Infectious Diseases (R24AI067039).

Abstract P09

Antiviral Therapy 2019; 24 Suppl 1:A46

Failure to restore CD4 cell count with combination antiretroviral therapy is associated with increased systemic inflammation

R Rousseau ¹ , L Szadkowski ² , C Kovacs ^3,4 , MF Saikali ⁵ , R Nadeem ⁶ , F Malazogu ⁶ , S Huibner ⁶ , CL Cummins ⁵ , R Kaul ^1,6,7 , SL Walmsley ^6,8

¹University of Toronto, Department of Immunology, Toronto, Canada; ²University Health Network, Biostatistical Research Unit, Toronto, Canada; ³Maple Leaf Medical Clinic, Toronto, Canada; ⁴University of Toronto, Department of Internal Medicine, Toronto, Canada; ⁵University of Toronto, Leslie Dan Faculty of Pharmacy, Toronto, Canada; ⁶University of Toronto, Department of Medicine, Toronto, Canada; ⁷University Health Network, Toronto General Hospital, Immunodeficiency Clinic, Toronto, Canada; ⁸Toronto General Hospital Research Institute, Toronto, Canada

Purpose: Systemic immune activation is thought to drive non-AIDS comorbidity despite suppressive antiretroviral therapy (ART). Immunological non-responders (INR) do not restore blood CD4 T-cells to >350/mm³. We sought to characterize the inflammatory profile of INRs, including the level of T-cell activation and inflammation-associated soluble markers linked to non-AIDS comorbidity. We hypothesized that co-expression of HLA-DR and CD38 on CD8 T-cells (CD8 activation) would be elevated in INRs.

Methods: 101 consenting adults were enrolled in a cross-sectional study at HIV care sites in Toronto, Canada. Virally suppressed HIV-positive participants were categorized as INR (n=36), or CD4 restorers (‘CR'; >350/mm³; n=47). An HIV-negative group was included (n=18). Peripheral blood mononuclear cells and blood plasma were cryopreserved before measurement of T-cell activation by flow cytometry (co-expression of HLA-DR and CD38) and soluble immune activation-related markers by ELISA (CRP, TNF, IFNg, sCD14, D-Dimer, I-FABP, MCP-1, VCAM, ICAM). Multiple log-linear regression models were used to estimate the association between INR phenotype and markers after adjusting for age, sex, duration of ART and recent infection/vaccination.

Results: HIV-positive individuals had median (IQR) infection 16 (9–26) years, and ART duration 12 (6–21) years, with no difference between groups. In adjusted log-linear regression models, INRs had 35% higher CD8 activation (P=0.02), 54% higher CD4 presentation of HLA-DR (P<0.001), and 20% higher plasma VCAM (P<0.01) compared with CRs. HIV-positive individuals had higher median (IQR) CD8 (2.96 [1.88–4.46]) compared with HIV-negatives (1.82 [1.17–2.81]; P<0.01). INR phenotype was not associated with CRP, TNF, IFNg, sCD14, D-Dimer, I-FABP, MCP-1, ICAM or CD8%DR+.

Conclusions: INRs, compared with CRs (and controls), had higher disease-associated CD8 activation that could be assessed as a target of intervention to improve outcomes. VCAM (marking endothelial activation) and CD4 activation were elevated amongst INRs and require exploration as prognostic indicators.

Abstract P10

Antiviral Therapy 2019; 24 Suppl 1:A47

Fitness tracking wearable devices and a dedicated smart phone app (MySAwH App) to predict quality of life in PLWH: a multi-centre prospective study

G Guaraldi ¹ , M Orsini ² , A Caselgrandi ¹ , A Malagoli ¹ , F D'Imprima ³ , J Milic ^1,4 , F Ghinelli ³ , R Martoglia ⁵ , F Mandreoli ⁵ , D Ferrari ³ , G Liu ⁶ , M Bloch ⁷

¹Modena HIV Metabolic Clinic, University of Modena and Reggio Emilia, Italy; ²DataRiver Srl; ³University of Modena and Reggio Emilia, Modena, Italy; ⁴Clinical and Experimental Medicine PhD Program, University of Modena and Reggio Emilia, Modena, Italy; ⁵Department of Mathematics, University of Modena and Reggio Emilia, Italy; ⁶Faculty of Medicine, The Chinese University of Hong Kong; ⁷Holdsworth House Medical Practice, Sydney, NSW, Australia

Objective: My Smart Age with HIV (MySAwH) is a multi-centre prospective ongoing study with the intention of empowering people living with HIV (PLWH) 50+ years to develop healthy lifestyles and healthy ageing. MySAwH is based on collection of physical function data and patient-related outcomes through a dedicated smart-phone app (MySAwH App). Our objective was to describe health changes assessed with frailty index (FI), collected by health professionals, and with a self-generated health measure called intrinsic capacity (IC) index which explores 5 different health domains: locomotion, vitality, sensory, cognition and psychosocial factors. FI and IC were used to predict quality of life (QOL) at follow-up.

Methods: We included 261 PLWH who were recruited from Italy (128), Australia (100) and Hong Kong (33). Two scheduled visits were performed, at baseline and follow-up (9 months). Frailty index was measured with a 36-item FI which objectively detect the presence of health deficits, while 27-item IC index was self-assessed with fitness tracking wearable devices and a MySAwH app. Outcome variables were QOL and health score (HS), using EQ5D5L questionnaire.

Results: Mean age was 56.9 years; 88% patients were men. Median CD4 was 657 c/μl (480–817 IQR) and 98% of patients had undetectable HIV viral load. Mean FI at baseline and follow-up were 0.22 (±0.1 sd) and 0.20 (±0.09 sd), respectively, P<0.001. Mean IC at baseline and follow-up were 0.69 (±0.12 sd) and 0.71 (±0.12 sd), P=0.27. Median QOL at baseline and follow-up were 0.88 (0.8–1 IQR) and 0.90 (0.83–1 IQR), P<0.03. Mean HS at baseline and follow-up were 7.6 (±1.68 sd) and 7.63 (±1.56 sd), P<0.001. In a multivariate logistic model, positive predictors for a good health status at follow-up were IC at baseline (OR=6.74, 3.86–11.77) and recruitment site (Hong Kong [OR=1.25,1.01–1.54]). Positive predictors for QOL at follow-up were IC at baseline (OR=7.62, 4–14.51) and recruitment site (Hong Kong [OR=1.33,1.05–1.69]).

Conclusions: Utilizing MySAwH with smart phone technology, FI and IC are performative tools that can be used in research and clinical setting to describe, respectively, disease and health status in PLWH. IC score in comparison to FI displayed higher sensitivity to predict both QOL and self-perceived health in PLWH.