| |
Long-term evolution of comorbidities and their disease burden in individuals with and without HIV as they age: analysis of the prospective AGEhIV cohort study
|
| |
| |
Download the PDF here
Download the PDF here
.... PLWH: more comorbidities, more disability, lower CD4 nadir/LDL/CHOL/CMV matter - all associated with increased death rates, particularly non-AIDS cancers
The larger comorbidity prevalence in HIV-positive participants aged 50-55 years on effective antiretroviral treatment than in HIV-negative participants increased similarly as participants aged and was associated with an increased risk of death, particularly of non-AIDS malignancies. Our findings reinforce the need for strategies to optimise prevention, screening, and early intervention.
In this prospective, longitudinal cohort study, we showed that HIV-positive participants aged 50-55 years at the time of enrolment had consistently higher mean numbers of comorbidities at all timepoints during 6 years of follow-up than did HIV-negative participants, but without a difference in the rate at which these increased. Having a higher number of comorbidities overall was strongly associated with increased risk of dying, but not with dropping out of the study.
The burden of living with comorbidity in terms of DALYs (Mean disability-adjusted life-years) was higher in people with HIV at enrolment, and DALYs increased more rapidly over time in HIV-positive participants than in HIV-negative participants. The increase in DALYs was largely explained by a higher rate of comorbidity-associated mortality in HIV-positive participants, with a large proportion of deaths being related to non-AIDS malignancies.
Summary
Background
People with HIV generally have more ageing-associated comorbidities than those without HIV. We aimed to establish whether the difference in comorbidities and their disease burden changes with ageing.
Methods
In this prospective, longitudinal cohort study, we assessed comorbidities commonly associated with ageing every 2 years in 596 HIV-positive and 550 HIV-negative participants. HIV-positive participants were recruited from the HIV outpatient clinic of the Amsterdam University Medical Centres (Amsterdam, Netherlands). HIV-negative participants were recruited from the sexual health clinic and the Amsterdam Cohort Studies at the Public Health Service of Amsterdam (Amsterdam, Netherlands). Inclusion criteria were participants aged 45 years or older and, for HIV-negative participants, a documented HIV-negative antibody test. The mean number of comorbidities present over time was compared between groups by use of Poisson regression, accounting for dropout and death through joint survival models. Mean disability-adjusted life-years (DALYs) accrued during 2-year intervals were compared between groups by use of an exponential hurdle model.
Findings
Between Oct 29, 2010, and Oct 9, 2012, participants were enrolled and then prospectively followed up until their last visit before Oct 1, 2018. 1146 participants were followed up for a median 5⋅9 years (IQR 5⋅7-6⋅0), during which 231 participants (20⋅2%) dropped out: 145 (24⋅3%) of 596 HIV-positive and 86 (15⋅6%) of 550 HIV-negative. 38 (3⋅3%) of 1146 participants died: 31 (5⋅2%) of 596 HIV-positive and seven (1⋅3%) of 550 HIV-negative.
24 HIV-positive and two HIV-negative participants died from ageing-associated comorbidities. 15 HIV-positive participants versus one HIV-negative participant died from non-AIDS malignancies.
At inclusion, mean number of comorbidities was higher in HIV-positive participants (0⋅65) than in HIV-negative participants (0⋅32; p<0⋅0001). Mean number of comorbidities increased at similar rates over time: rate ratio (RR) per year for HIV-positive participants 1⋅04 (95% CI 1⋅00-1⋅08), RR per year for HIV-negative participants 1⋅05 (1⋅01-1⋅08; pinteraction=0⋅78).
Number of comorbidities was associated with an increased risk of death (hazard ratio 3⋅33 per additional comorbidity, 95% CI 2⋅27-4⋅88; p<0⋅0001).
HIV-positive participants had higher increases in mean DALYs than HIV-negative participants (0⋅209 per year, 95% CI 0⋅162-0⋅256 vs 0⋅091 per year, 0⋅025-0⋅157; pinteraction=0⋅0045). This difference was reduced when deaths were excluded in establishing DALYs (0⋅127, 0⋅083-0⋅171 vs 0⋅066, 0⋅005-0⋅127; pinteraction =0⋅11).
Interpretation
The larger comorbidity prevalence in HIV-positive participants aged 50-55 years on effective antiretroviral treatment than in HIV-negative participants increased similarly as participants aged and was associated with an increased risk of death, particularly of non-AIDS malignancies. Our findings reinforce the need for strategies to optimise prevention, screening, and early intervention.
Our burden of disease analysis showed a more rapid increase in DALYs in people with HIV, primarily due to increased mortality associated with various comorbidities, mostly non-AIDS malignancies. This increase in DALYs was observed most in people with HIV with low nadir CD4 counts or exposure to toxic NRTIs. None of the participants died from AIDS-related causes. Large cohort studies have shown non-AIDS-defining malignancies to be an important cause for morbidity and mortality in people with HIV and to be associated with increased age, low CD4 cell count nadir, extended periods of untreated HIV infection, and previous use of toxic NRTIs. Our data might offer support to the concept that HIV-infection-induced severe immune deficiency and previous exposure to toxic NRTIs could result in an accentuation of ageing (ie, there was a point before follow-up at which these events led to an increased risk of comorbidities), but that this accentuation does not increase over time after virological suppression.
Our results have important consequences for clinical care and prevention strategies, and they bring awareness of comorbidities to both people with HIV and HIV doctors alike. Having comorbidities comes with disability, risk of polypharmacy, and risk of premature mortality. Despite awareness about the increased risk of comorbidities, uptake of preventive strategies is not optimal and should be improved, such as for cardiovascular risk management.
None of the cancer screening programmes, except those concerning anogenital cancers, include indications for increased screening in people with HIV, despite increasing evidence for increased cancer risk in these individuals.
At cohort enrolment, 265 HIV-positive participants (44⋅5%) had one or more comorbidities compared with 142 HIV negative participants (25⋅8%; p<0⋅0001); 90 HIV-positive participants (15⋅1%) had two or more compared with 24 HIV-negative participants (4⋅4%; p<0⋅0001). Accordingly, the mean number of comorbidities per individual was 0⋅65 (95% CI 0⋅58-0⋅72) in HIV-positive participants and 0⋅32 (95% CI 0⋅27-0⋅37) in HIV-negative participants (p<0⋅0001).
The most common comorbidities present at enrolment among the total study population were kidney disease (192 of 1146, 16⋅8%), osteoporosis (110 of 1146, 9⋅6%), and COPD (107 of 1146, 9⋅3%); kidney disease and osteoporosis were significantly more prevalent in HIV-positive participants than in HIV-negative participants (appendix 2 p 11).
From inclusion to end of follow-up, there were 1437 study-assessed and 421 self-reported comorbidities. Of the self-reported comorbidities, 218 (51⋅8%) could be validated: 155 (59⋅2%) of 262 in HIV-positive participants and 63 (39⋅6%) of 159 in HIV-negative participants (p=0⋅0001). 14 (2⋅5%) of 550 HIV-negative participants did not consent to contacting their general practitioner for validation. Similar increases in mean number of comorbidities were observed over time in both HIV-positive (RR 1⋅04, 95% CI 1⋅00-1⋅08) and HIV-negative participants (1⋅05, 1⋅01-1⋅08; figure 1) without significant differences between groups (pinteraction=0⋅78). The period prevalence over time for individual comorbidities has been described (appendix 2 p 18). There were similar increases in mean number of comorbidities when stratifying HIV-positive participants by previous severe immunodeficiency, history of AIDS, and previous use of toxic NRTIs (appendix 2 p 19). HIV-positive participants with previous severe immunodeficiency, AIDS, and previous use of toxic NRTIs had more ageing-associated comorbidities than did HIV-positive participants without these characteristics at any point during follow-up. Each additional comorbidity was associated with an increased hazard of death, but not with dropout (HR 3⋅33 per additional comorbidity, 95% CI 2⋅27-4⋅88, p<0.001; figure 1).
HIV-positive participants had higher increases in mean DALYs [Mean disability-adjusted life-years (DALYs)] than HIV-negative participants (0⋅209 per year, 95% CI 0⋅162-0⋅256 vs 0⋅091 per year, 0⋅025-0⋅157; pinteraction=0⋅0045). This difference was reduced when deaths were excluded in establishing DALYs (0⋅127, 0⋅083-0⋅171 vs 0⋅066, 0⋅005-0⋅127; pinteraction =0⋅11).
DALYs reflect the disease burden in a population by combining morbidity severity and mortality.
The increase in DALYs in HIV-positive participants with a nadir CD4 count of larger than 200 cells per μL (0⋅185, 95% CI 0⋅105-0⋅264) was not significantly different compared with HIV-negative participants (0⋅091, 0⋅025-0⋅157; pinteraction=0⋅075). By contrast, the increase in DALYs in HIV-positive participants with nadir CD4 counts less than 200 cells per μL (0⋅221, 0⋅163-0⋅279) was significantly faster than in HIV-negative participants (0⋅091, 0⋅025-0⋅157; pinteraction=0⋅0039; appendix 2 p 21).
In HIV-positive participants both with (0⋅244, 0⋅168-0⋅320) and without (0⋅189, 0⋅129-0⋅248) a history of AIDS, the increase in DALYs was significantly faster than in HIV-negative participants (0⋅091, 0⋅025-0⋅157; pinteraction=0⋅0030 for HIV-positive participants with a history of AIDS and pinteraction=0⋅032 for HIV-positive participants without a history of AIDS; appendix 2 p 22). In HIV-positive participants who had never used toxic NRTIs, the DALY increase (0⋅173, 0⋅104-0⋅243) was not significantly different from HIV-negative participants (0⋅091, 0⋅025-0⋅157; pinteraction=0⋅093). The increase in DALYs in HIV-positive participants who had used toxic NRTIs (0⋅240, 0⋅178-0⋅303) was significantly faster than in HIV-negative participants (0⋅091, 0⋅025-0⋅157, pinteraction=0⋅0014; appendix 2 p 21).
38 (3⋅3%) of 1146 included participants died; 31 (5⋅2%) HIV-positive participants and seven (1⋅3%) HIV-negative participants. Of the 38 deceased participants, nine had no comorbidities at their previous study visits, 14 had one, and 15 had two or more. Among the 31 deaths in HIV-positive participants, none were AIDS-related but 24 were comorbidity-related. Of the seven deaths among HIV-negative participants, two were comorbidity-related (table 3). The median age of HIV-positive participants who died was higher than in HIV-positive participants who did not die; HIV-positive participants who died also had more pack years of smoking and were more frail (table 4). Furthermore, HIV-positive participants who died had lower mean nadir CD4 count at enrolment and more often had a history of AIDS and use of toxic NRTIs (table 4). Deceased HIV-positive participants and deceased HIV-negative participants had similar age, smoking status, and frailty; however, prevalence of cytomegalovirus co-infection and levels of LDL cholesterol and total cholesterol were higher in deceased HIV-positive participants (table 4). During the study period, 231 (20⋅2%) participants dropped out; 145 (24⋅3%) of 596 HIV-positive participants and 86 (15⋅6%) of 550 HIV-negative participants. 103 (71⋅0%) of 145 and 51 (59⋅3%) of 86 withdrew consent and 42 (29⋅0%) of 145 and 35 (40⋅7%) of 86 were lost to follow-up.
|
|
| |
| |
|
|
|
