Risk Factors and Impact of Patterns of Co-Occurring Comorbidities in People Living With HIV

Davide De Francesco; Jonathan Underwood; Emmanouil Bagkeris; Jane Anderson; Ian Williams; Jaime H. Vera; Frank A. Post; Marta Boffito; Margaret Johnson; Patrick W.G. Mallon; Alan Winston; Caroline A. Sabin; on behalf of the Pharmacokinetic and Clinical Observations in PeoPle Over fiftY (POPPY) study

Disclosures

AIDS. 2019;33(12):1871-1880. 

In This Article

Results

Characteristics of Study Participants

A total of 1073 PLWH were recruited into the POPPY study: 699 older and 374 younger PLWH. Sociodemographic and HIV-related characteristics are summarized in Table 1. Participants were predominantly male (85.2%), of white ethnicity (84.1%), MSM (76.0%) with a median (interquartile range: IQR) age of 52 (47, 59). The median (IQR) CD4+ T-cell count was 624 (475, 811) cells/μl and 89.9% had a suppressed viral load (<50 copies/ml). Current smoking was reported by 24.9% of study participants, current recreational drug use and history of injection drug use were reported by 28.6 and 10.5%, respectively.

The numbers of participants with completed information on the five health outcomes were 886 for SF-36 physical and mental health scores, 1020 for functional impairment and 1073 for number of GP visits and hospitalization in the last year. The median (IQR) physical and mental health score were 52.4 (42.3, 56.3) and 51.0 (41.6, 57.2), respectively. Functional impairment was observed in 14.5% of PLWH, 65.0% had been hospitalized in the year preceding the study visit and the median (IQR) number of GP visits was 1 (1, 2).

Patterns of Comorbidities

The prevalence of individual comorbidities is reported in Supplementary Table 1, http://links.lww.com/QAD/B496. The six patterns obtained using the PCA are described in Figure 1, with full details reported elsewhere.[5] The patterns were termed as follow: cardiovascular disease (CVD), sexually transmitted diseases (STDs), mental health problems, cancers, metabolic disorders and chest problems and other infection.

Figure 1.

Patterns of comorbidities and distribution of their severity scores in all study participants (n = 1073).

The distribution, with median and IQR, of the severity scores of the six patterns in study participants are reported in Figure 1. Highest severity scores were observed for the STDs pattern with a median (IQR) of 1.34 (0.01, 2.83). There was high variability on the metabolic and mental health pattern scores, with PLWH reporting either very low or relatively high scores. The distribution of CVD and cancer severity scores were similar, with only a small proportion of individuals reporting extremely high scores compared with the majority of PLWH. There was less variability in the severity scores for the chest/other infections pattern for which the median (IQR) score was 0.17 (0.03, 0.34).

Risk Factors of Patterns of Comorbidities

Older age [0.06 (0.04, 0.07) per each 10-year older, P < 0.001], greater BMI [0.03 (0.01, 0.05) per 5-kg/m2 increment, P = 0.009], longer time since HIV diagnosis [0.04 (0.02, 0.06) per 10-year increment, P < 0.001] and prior AIDS [0.08 (0.04, 0.12), P < 0.001] were independently associated with higher CVD severity scores (Table 2). MSM [1.45 (1.33, 1.57), P < 0.001] and those with a history of injection drug use [1.24 (0.64, 1.84), P < 0.001] had significantly higher severity scores for the STDs pattern compared with heterosexuals and those who never reported using injection drugs, respectively. Mental health severity scores were significantly associated with injection drug use [1.27 (0.87, 1.66), P < 0.001], longer time since HIV diagnosis [0.14 (0.07, 0.21) per 10-year increment, P < 0.001] and prior AIDS [0.15 (0.04, 0.26), P = 0.007]. Women also appeared to have higher mental health scores [0.12 (0.01, 0.23), P = 0.03] than men. Older age [0.06 (0.04, 0.07) per each 10-year older, P < 0.001], greater BMI [0.03 (0.01, 0.05) per 5-kg/m2 increment, P = 0.009], longer time since HIV diagnosis [0.04 (0.02, 0.06) per 10-year increment, P < 0.001] and prior AIDS [0.08 (0.04, 0.12), P < 0.001] were independently associated with higher CVD severity scores (Table 2). MSM [1.45 (1.33, 1.57), P < 0.001] and those with a history of injection drug use [1.24 (0.64, 1.84), P < 0.001] had significantly higher severity scores for the STDs pattern compared with heterosexuals and those who never reported using injection drugs, respectively. Mental health severity scores were significantly associated with injection drug use [1.27 (0.87, 1.66), P < 0.001], longer time since HIV diagnosis [0.14 (0.07, 0.21) per 10-year increment, P < 0.001] and prior AIDS [0.15 (0.04, 0.26), P = 0.007]. Women also appeared to have higher mental health scores [0.12 (0.01, 0.23), P = 0.03] than men.

The cancer and metabolic pattern scores were each significantly associated with older age (both P < 0.001) and greater BMI (P = 0.03 and 0.006, respectively). In addition, cancer scores were significantly higher in MSM [0.06 (0.02, 0.09), P = 0.005] than in heterosexuals and those with prior AIDS [0.15 (0.12, 0.19), P < 0.001], whilst the associations with nadir CD4+ T-cell count and time since HIV diagnosis did not reach statistical significance (P = 0.08 and 0.09, respectively). On the other hand, metabolic scores appeared to be significantly associated with longer time since HIV diagnosis [0.11 (0.05, 0.16) per 10-year increment, P < 0.001], prior AIDS (P = 0.05) and nadir CD4+ T-cell count (P = 0.07). Finally, older age (P < 0.001), prior AIDS (P < 0.001) and a longer time since HIV diagnosis (P = 0.001) were all independent predictors of chest/other infections severity scores. Neither smoking status nor alcohol consumption appeared to be significant risk factors for any of the six patterns.

Associations With Patient-reported Health Outcomes

The CVD pattern was associated with poorer physical health scores (P = 0.02), higher odds of functional impairment (P = 0.02) and hospitalization (P < 0.001) and higher number of GP visits (P < 0.001; Table 3). Higher STDs severity scores were associated with better physical health (P < 0.001) and a lower number of GP visits (P < 0.001). In addition, weaker evidence was found regarding their association with a lower risk of functional impairment (P = 0.08) and a higher risk of hospitalization (P = 0.12).

The mental health pattern was negatively associated with all the outcomes considered (all P < 0.001), in particular with poorer physical [β (95% CI): −2.62 (−3.40, −1.84)] and mental [β (95% CI): −5.46 (−6.19, −4.72)] health scores and increased odds of functional impairment [OR (95% CI): 1.88 (1.53, 2.32)]. Significantly poorer physical health scores (P = 0.03) and a higher risk of functional impairment (P = 0.01) were observed for the cancers pattern. Higher metabolic severity scores were associated with better physical (P = 0.04) and mental (P = 0.04) health and lower risk of functional impairment (P = 0.007).

Finally, the chest/other infections pattern was negatively associated with all the outcomes. Higher severity scores were particularly associated with poorer physical health [β (95% CI): 9.15 (−11.74, −6.56), P < 0.001], an almost 5-time higher risk of functional impairment [OR (95% CI): 4.83 (2.54, 9.17), P < 0.001] and a higher number of GP visits [IRR (95% CI): 1.45 (1.24, 1.69), P < 0.001]. Also significant were the associations with mental health (P = 0.03) and hospitalization (P = 0.02).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....