Health-related quality of life of perinatally HIV-infected young people: a longitudinal study

ABSTRACT

Young people perinatally infected with HIV (pHIV) are at risk of a lowered health-related quality of life (HRQOL). Previous evaluation of the NeurOlogical, VIsual and Cognitive performance in HIV-infected Children (NOVICE)-cohort showed no difference in HRQOL between pHIV and matched HIV-uninfected controls (HIV-), yet a higher percentage of pHIV had impaired HRQOL. The aim of this study is to compare the change over time in HRQOL of pHIV to HIV- over a 5-year period. We used the Pediatric Quality of Life Inventory (PedsQL)™ 4.0 to repeat HRQOL assessment. High PedsQL scores indicate good HRQOL. Fifteen/33 (45.5%) pHIV and 17/37 (45.9%) HIV- completed both assessments. At the first assessment, the mean age was 13.1 years (range 8.0–18.4). PHIV scored higher than HIV- on Emotional functioning and on Total scale score. After five years, the mean age was 17.6 years (range 12.1–22.8). PHIV scored higher than HIV- on all scales, except Social functioning. PHIV did not differ significantly from the Dutch norm on either time-point. LMEM showed no difference in change over time for any of the PedsQL scales. In this study, young people with pHIV receiving high-quality health care, including monitoring of HRQOL, remain to experience a good HRQOL.

KEYWORDS:

• Perinatal HIV infection
• health-related Quality of life
• children
• adolescents

Introduction

Mother-to-child transmission of HIV has gradually decreased worldwide, yet still as much as 160,000 new pediatric HIV infections occurring in children between 0 and 14 years old in 2018 (UNAIDS, 2019). In 2018, an estimated 1.7 million children under 15 years were living with HIV. In the Netherlands, most perinatally HIV-infected children (pHIV) were born outside of the Netherlands (±67%) with an increasing proportion having been internationally adopted by adoptive parents (van Sighem et al., 2018). With improving healthcare, young people with pHIV are increasingly reaching adulthood and health-related quality of life (HRQOL) has become an important outcome measure (Haverman et al., 2017; World Health Organisation, 1996).

Young people with pHIV are, similar to all young people living with a chronic condition, at risk of a lowered HRQOL (Cuellar-Flores et al., 2019, Jul 8; Nkwata et al., 2017; Rajeshwari, 2019). HRQOL is a multidimensional concept that refers to the impact of health and illness on an individual’s quality of life (QOL), which encompasses not only physical aspects but also social and emotional aspects (Eiser & Morse, 2001). HRQOL measurement is important to monitor the wellbeing of individuals with pHIV and is generally used as a metric for the evaluation of effectiveness and quality of (HIV) care (Haverman et al., 2017).

Cross-sectional studies investigating HRQOL in young people with pHIV have reported contrasting results (Cohen, ter Stege, et al., 2015; Cuellar-Flores et al., 2019, Jul 8; Nkwata et al., 2017; Rajeshwari, 2019; Rydstrom et al., 2016; Tan et al., 2017). Our previous study of pHIV children (aged 8–18 years; the Neurological, Visual and Cognitive performance in HIV-infected children (NOVICE)-cohort study), showed a higher percentage of lowered HRQOL in pHIV when compared with healthy controls matched for age, sex and socio-economic status (SES; Cohen, ter Stege, et al., 2015). Lowered HRQOL was defined as a deviation of ≥1 SD below the mean of the norm population. These results suggested that pHIV children were still more vulnerable to a lowered HRQOL.

Since this study, five studies have found a lower HRQOL in young people with pHIV compared to controls without HIV (Cuellar-Flores et al., 2019, Jul 8; Das et al., 2017; Enimil et al., 2016; García-Alonso et al., 2015; Nkwata et al., 2017). The majority of these studies have been performed in non-western regions and included pHIV participants receiving suboptimal treatment. Since poorer disease outcomes are associated with reduced HRQOL (Bunupuradah et al., 2012; Lee et al., 2006; Nachman et al., 2012; Oberdorfer et al., 2008; Rajeshwari, 2019), we cannot generalize these results to well-treated pHIV young people in the Netherlands who have full access to high-profile healthcare (van Sighem et al., 2019). Indeed, one of the five studies was a European study comparing HRQOL of (well treated and monitored) pHIV to healthy controls and reports similar HRQOL in both groups (Rydstrom et al., 2016).

In adults, HRQOL of people living with HIV in the Netherlands is found to be lower compared to the general Dutch population (Engelhard et al., 2018). Furthermore, a study of Spanish young adults (YAs) with pHIV aged 18–29 years also reported lower HRQOL compared to healthy peers (Cuellar-Flores et al., 2019, Jul 8). While growing up into young adulthood, new factors come into play that may influence the HRQOL of pHIV, such as sexuality, life-long adherence to medication and cognitive impairments (Mukherjee & Lodha, 2019; Rydstrom, 2019; ter Haar et al., 2019). Consequently, HRQOL can be subject to change during lifetime, and particularly during the development into adolescence and young adulthood (Meade, 2016; Palacio-Vieira et al., 2008). Until now, no studies have longitudinally investigated the trajectory of HRQOL in young people with pHIV growing up. Consequently, it is unknown whether this is different from that of healthy peers.

The aim of our study is to investigate the longitudinal trajectory of HRQOL in young people with pHIV and compare it to that of HIV-uninfected peers.

To this aim, all previously examined pHIV children and adolescents of the NOVICE cohort will be reassessed after a 5-year period. We will compare their HRQOL on both first and second assessment to a matched HIV-uninfected control group and the general Dutch population. Following, we will compare the change over time in the HRQOL of pHIV and HIV-uninfected controls.

Materials and methods

Participants and procedure

Between February 2017 and July 2018, we approached all participants of the NeurOlogical, VIsual and cOgnitive performance in HIV-infected Children (NOVICE)-cohort study to join a second assessment. NOVICE is a prospective, observational cohort study that investigates the potential effect of HIV infection and the use of combination antiretroviral therapy (cART) in children at the Amsterdam University Medical Centers (Amsterdam UMC), Academic Medical Center, the Netherlands (Blokhuis et al., 2016, 2017; Cohen et al., 2016; Cohen, Ter Stege, et al., 2015; Cohen, ter Stege, et al., 2015; ter Haar et al., 2019; Van Dalen et al., 2016; Van den Hof et al., 2019).

At the first assessment, pHIV children between 8 and 18 years old were recruited from the outpatient clinic of the Amsterdam UMC (Cohen, ter Stege, et al., 2015), together with HIV-uninfected controls from similar communities, and frequency matched regarding age, sex, ethnicity, and socioeconomic status (SES). SES was defined as the level of parental education and parental occupational status. Parental education was scored according to the International Standard Classification of Education (UNESCO, 2012). Occupational status was defined as 0, 1, or 2 caregivers with a paid job (Cohen, ter Stege, et al., 2015).

Between February 2017 and July 2018, we approached all NOVICE participants and included those who consented to participate in a second assessment. We obtained written informed consent from participants older than 12 years and from all parents or legal guardians of participants younger than 18 years old. The ethics committee of the Amsterdam University Medical Center approved the study protocol. This study was registered with the Netherlands Trial Register (ID NL6813).

Demographic and human immunodeficiency virus- and treatment-related characteristics

Data on sex, ethnicity, adoption status, and SES were used from the first assessment (Cohen, ter Stege, et al., 2015). We newly collected data on age and cART use during follow-up. We defined cART use as the daily use of a minimum of three antiretroviral drugs, from a minimum of two drug classes. For PHIV+ participants, we performed laboratory testing of HIV RNA viral loads (VLs) and CD4+ T-cell counts at the second assessment.

Historical HIV VL and CD4+ T-cell counts, Center for Disease Control (CDC) clinical stage category and cART treatment history were derived from the Dutch HIV monitoring foundation (HMF) database (www.hiv-monitoring.nl). Clinical, immunological, and virological data from before moving into the Netherlands were registered, and recorded as “missing” when not traceable.

Health-related quality of life

Health-related quality of life (HRQOL) was measured using the Dutch child self-report version of the generic Pediatric Quality of Life Inventory™ 4.0 (PedsQL™) (Varni et al., 2003), available for children aged 8–12, 13–18 years and for young adults aged 19–30 years (Engelen et al., 2009; Limperg et al., 2014). The generic PedsQL™ inventory is developed to measure HRQOL in healthy and (chronically or acutely) ill children.

It includes 23 items, divided over four subscales: physical functioning (eight items), emotional functioning (five items), social functioning (five items), and school functioning (five items). The subscales are summarized into the psychosocial health summary score (including the emotional, social, and school functioning subscales) and a total score (including all subscales).

Items are rated on a five-point Likert scale (0 = “never a problem”, 1 = “almost never a problem”, 2 = “sometimes a problem”, 3 = “often a problem”, and 4 = “nearly always a problem”) based on the preceding week. Each answer is reversed scored and rescaled to 0–100 scale (0 = 100, 1 = 75, 2 = 50, 3 = 25 and 4 = 0). Higher scores on the PedsQL indicate better reported HRQOL.

Dutch normative data of the PedsQL™ are available for all three age groups (Engelen et al., 2009; Limperg et al., 2014). To be able to combine all three age groups for further analyses, we calculated z-scores using the mean and standard Dutch norm population.

Several studies have used this instrument to measure HRQOL in similar populations (Banerjee et al., 2010; Bomba et al., 2010; Boon-Yasidhi et al., 2016; Das et al., 2010; Gopakumar et al., 2018; Lang et al., 2014; Nkwata et al., 2017; Punpanich et al., 2010; van Elsland et al., 2019; Xu et al., 2010). The PedsQL™ has good validity and reliability (Engelen et al., 2009; Limperg et al., 2014; Varni et al., 2003; Varni & Limbers, 2008).

In this study, internal consistency reliability was α > 0.80 for the total, psychosocial, and emotional functioning scale, a > 0.70 for the social functioning scale, and α > 0.60 for physical and school functioning (Cronbach, 1947).

Statistical analysis

Selective dropout was explored in the NOVICE cohort using t-tests to compare sociodemographic characteristics at the first assessment of those who participated in the second assessment and of those who refrained from participation.

For participants who completed two assessments, we compared sociodemographic characteristics (as reported in Table 1) between pHIV and HIV-uninfected controls by using the unpaired t-test or Mann–Whitney U-test, and Fisher’s exact test for categorical data.

Table 1. Characteristics of participants who completed two assessments.

Characteristics		pHIV (n = 15)	HIV− (n = 17)	P-value
Follow-up rate		45.5%	45.9%	.999
Male sex		7 (46.7%)	7 (41.2%)	.755
Age at first assessment, years, mean (SD)		12.8 (3.09)	13.4 (2.4)	.529
Age at second assessment, years, mean (SD)		17.3 (3.20)	17.9 (2.4)	.538
Region of birth				<.001
The Netherlands		2 (13.3%)	16 (94.1%)
Sub-Sahara Africa		11 (73.3%)	1 (5.9%)
Other*		2 (13.3%)	0 (0%)
Ethnicity				.249
Black		11 (73.3%)	13 (76.5%)
Caucasian		0 (0%)	2 (11.8%)
Mixed		3 (20%)	2 (11.8%)
Other		1 (6.7%)	0 (0%)
Adoption status				.035
Not adopted		10 (66.7%)	17 (100%)
Adopted		2 (13.3%)	0 (0%)
Foster care		3 (20.0%)	0 (0%)
ISCED level of parent, median (IQR)		5 (3-6)	5 (2.5-6)	.251
Age at HIV diagnosis, years, median (IQR)	15	2.46 (1.13–6.02)
Route of transmission	15
Perinatal		15 (100%)
Unknown		0
CDC category	15
N or A		6 (40%)
B		6 (40%)
C		3 (20%)
Zenith HIV viral load, copies/mL, median (IQR)	12	3,79,788 (60,564–11,42,658)
HIV viral load at second assessment	15
Undetectable^a		13 (87%)
Detectable		2 (13.3%)
Nadir CD4+ T-cell z-score, mean (SD)	13	−0.79 (0.68)
Current cART use^b	15	15 (100%)
Age at cART initiation, years, median (IQR)	12	4.61 (1.47–7.60)
Duration of cART use, years, median (IQR)	12	14.49 (8.22–18.00)

Values are reported as n (%), unless otherwise stated.

pHIV, perinatally human immunodeficiency virus infected; HIV−, HIV negative; IQR, interquartile range; ISCED, International Standard Classification of Education; SD, standard deviation; CDC, Centre for Disease Control and Prevention category (N or A = no clinical or minimal signs, B = moderate signs, C = AIDS).

^a Undetectable values are defined as <40 c/mL, including viral blips.

^b Number of participants on cART at the second assessment.

We calculated z-scores based on the Dutch national norm for all individual scale scores on the first and second assessment. While this study focusses only on those who participated in both assessments, we reassessed cross-sectional differences at the first assessment using one-sample t-tests. We used one-sample t-tests to assess cross-sectional differences at the second assessment between pHIV, HIV-uninfected controls and the Dutch norm population, respectively.

To assess the difference in HRQOL trajectories between HIV status and outcome variables over time, we used linear mixed-effects models. This model is able to account for intrasubject correlation in repeated measurements. The small number of participants did not allow for adjustment of covariables.

We performed all statistical analyses in IBM SPSS Statistics (version 25). We considered a two-sided P < .05 as statistically significant. We did not adjust for multiple comparisons, since the analyses were considered to be exploratory.

Results

Sociodemographic background and HIV-related parameters

The original NOVICE cohort consisted of 33 pHIV and 37 HIV-uninfected participants. Fifteen (45.5%) pHIV and 17 (45.9%) HIV-uninfected participants provided consent to complete the second assessment. Reasons not to participate were: unwillingness to participate (n = 19), inability to contact (n = 6), or relocation (n = 3). Selective drop-out analyses did not show statistically significant differences between sociodemographic variables of participants who participated and those who refrained from the second assessment.

Table 1 shows the characteristics of the pHIV and HIV-uninfected participants who completed both assessments. Participants had a mean age of 13.1 years (range 8.04–18.40) at baseline and 17.62 years (range 12.05–22.80) at follow-up. PHIV and HIV-uninfected participants did not significantly differ in age, sex, ethnicity, or SES. PHIV participants were more often adopted or in foster care (P = .035) and were more often born in sub-Saharan Africa, while the majority of HIV-uninfected participants was born to immigrant parents in the Netherlands (P < .001).

Cross-sectional comparison of HRQOL in pHIV versus HIV-uninfected participants and the Dutch healthy norm

First assessment

Cross-sectional reassessment of those who participated in both assessments showed significantly higher average Emotional functioning score in pHIV participants on the first assessment (mean z-score 0.643, SD 14.7) compared to HIV-uninfected controls (mean z-score −0.252, SD 1.26, P = .036), but not compared to the Dutch Healthy norm (P = .443). Other PedsQL scores were not statistically different between groups (see Table 2).

Table 2. Cross-sectional analyses of health-related quality of life at baseline in participants who completed two assessments.

	Baseline				Follow-up
	pHIV	HIV-	P-value*	P-value**	pHIV	HIV-	P-value*	P-value**
Physical functioning	0.458 (1.12)	−0.328 (1.36)	.087	.862	0.691 (0.87)	−0.302 (0.90)	.004	.365
Emotional functioning	0.643 (1.02)	−0.252 (1.26)	.036	.443	0.645 (0.98)	−0.363 (0.89)	.005	.559
Social functioning	0.182 (0.91)	−0.157 (1.09)	.352	.991	0.410 (0.69)	0.011 (0.89)	.169	.179
School functioning	0.210 (0.97)	−0.748 (1.49)	.137	.206	0.033 (1.05)	−0.855 (1.23)	.037	.050
Total scale score	0.404 (1.03)	−0.379 (1.37)	.049	.978	0.604 (0.98)	−0.471 (0.94)	.004	.866
Psychosocial Health Summary score	0.373 (1.04)	−0.392 (1.31)	.065	.880	0.482 (1.00)	−0.498 (1.01)	.010	.845

Presented are mean (SD) z-scores of Pediatric Quality of Life Inventory (PedsQL)™ 4.0 scale scores.

*Comparison between z-scores of pHIV and HIV-uninfected participants.

**Comparison between z-scores of pHIV participants and Dutch healthy norm.

pHIV, perinatally human immunodeficiency virus infected; HIV−, HIV-uninfected participants; SD, standard deviation.

Second assessment

At the second assessment, pHIV-adolescents scored significantly higher than HIV-uninfected controls on all PedsQL scales except social functioning (see Table 2).

Longitudinal comparison of changes in HRQOL in pHIV versus HIV-uninfected participants over the study duration

Table 3 shows the HRQOL trajectories for each domain in pHIV and HIV-uninfected participants. Changes over time were not statistically different for either of the PedsQL scale scores.

Table 3. Longitudinal analysis of health-related quality of life in participants who completed both assessments.

	Group * Time
	B* (95% CI)	P-value
Physical functioning	0.442 (−0.51–1.30)	.640
Emotional functioning	0.195 (−0.85–1.24)	.826
Social functioning	0.405 (−0.40–1.21)	.876
School functioning	−0.070 (−1.14–1.00)	.719
Total Scale Score	0.291 (−0.73–1.32)	.703
Psychosocial Health Summary score	0.215 (−0.81–1.24)	.730

*B represents the mean difference between pHIV and HIV-uninfected participants. Higher scores indicate better HRQOL; a positive score therefore means better HRQOL in PHIV. CI, confidence interval.

Discussion

These results suggest that pHIV-adolescents receiving high-profile HIV care in the Netherlands remain to experience a good HRQOL when they grow-up into young adulthood. The longitudinal trajectory of HRQOL in pHIV-infected children and adolescents in the Netherlands is similar to that of healthy peers with similar socioeconomic backgrounds. These results are in agreement with previous studies showing good HRQOL of adequately treated young people (Cohen, ter Stege, et al., 2015; Rydstrom et al., 2016).

Rather unexpected, however, is that young people with pHIV experience higher HRQOL than SES-matched uninfected peers on nearly all domains. According to the authors, the most likely explanation for this finding is that these young people with pHIV receive extra guidance. In the Netherlands, the International guidelines (American and European (PENTA: Pediatric European Network on HIV Antiretroviral Therapy) pediatric guidelines are followed, which include the immediate start of cART after HIV diagnosis and monitoring of therapy adherence, effectiveness, and toxicities every 3–4 months (Bamford et al., 2018; Nederlandse Vereniging van HIV Behandelaren, 2020).

Studies in other pediatric populations have shown that periodic monitoring of HRQOL through patient-reported outcomes measures (PROMS) helps the clinician to discuss emotional and psychosocial functioning (Engelen et al., 2012; Haverman et al., 2014). This improves adolescents’ satisfaction with care and has a positive impact on psychosocial wellbeing (de Wit et al., 2008). In our center, pHIV children are asked to fill out several questionnaires via the online KLIK PROM portal, including the PedsQL, before every hospital visit (Haverman et al., 2014, 2013). Like pHIV patients, their SES-matched peers have a higher risk of a HRQOL below the Dutch national average due to their low SES, yet they do not receive extra guidance from professionals (Emerson et al., 2006; von Rueden et al., 2006). This could explain the higher HRQOL outcomes of pHIV compared to healthy peers. We therefore emphasize that monitoring of HRQoL continues to be an important aspect of high-quality healthcare for pHIV+ youth. To optimize the use of HRQOL monitoring, healthcare professionals need to get trained in how to interpret HRQOL scores.

One Spanish study in pHIV-infected young adults showed worse outcomes of HRQOL (Cuellar-Flores et al., 2019, Jul 8). This disparity in outcome might be due to the age difference; pHIV participants in the Spanish study were on average 23.36 ± 3.84 years old, whereas participants in our study were on average 17.62 (range 12.05–22.80) years old on the second assessment. It is known that Dutch young adults aged 18–19 years are at risk for HIV viral failure, especially during the transition from pediatric to adult care (Weijsenfeld et al., 2016). Accordingly, although both studies are performed in western countries with adequate HIV care services, more pHIV participants in the Spanish group had a detectable viral load when compared with the proportion of participants with a detectable viral load in our study. In addition, more pHIV participants fell in category C of the Centers for Disease Control and Prevention (CDC) classification. Therefore, it is important to continue to monitor the HRQOL of these young people with pHIV growing up into young adulthood.

There are some limitations to this study. Firstly, the small sample size of this study prohibits us from drawing firm conclusions about the differences in HRQOL between pHIV and healthy peers. There is a reduced change of detecting a significant difference between pHIV and the other groups and we can therefore not dismiss the possibility of an existing effect. Simultaneously, it reduces the likelihood that a statistically significant result reflects a true effect. Future research should confirm our findings in larger samples.

Secondly, while not all original participants from the NOVICE cohort were measured again in the second assessment, a selection bias might exist. It might be that only a subgroup of healthy controls with lower than average HRQOL was reassessed: within the healthy control group, those who refrained from the second participation scored significantly higher at the first assessment on three scales (school function, total scale score and psychosocial scale score) compared to those that participated in both assessments.

Finally, the distribution of scale scores is quite wide for pHIV as well as healthy controls, indicating a wide range in experience of HRQOL for both groups. Our previous study showed a higher chance of low HRQOL in pHIV children, despite a good HRQOL in the whole group on average (Cohen, ter Stege, et al., 2015). HIV and treatment variables as well as demographic and psychosocial factors have previously been found to predict HRQOL (Bomba et al., 2010; Cuellar-Flores et al., 2019, Jul 8; Gopakumar et al., 2018; Rajeshwari, 2019; Sanjeeva et al., 2019). These factors could explain the wide distribution of HRQOL scores and give options for risk assessment and intervention. As a consequence of the small number of participants in this study, no association analyses could be performed. Future studies with larger samples should include these predicting factors in the (longitudinal) analyses to better understand HRQOL in pHIV.

In this study, we used a generic HRQOL questionnaire to enable the comparison of outcomes of pHIV to healthy controls. However, generic HRQOL measures might not be sensitive enough to study the more pHIV-specific problems of young adults, such as stigma, self-esteem, treatment adherence, or sexual health (Bryant & Fernandes, 2011; den Daas et al., 2019; Costanza et al., 2007). As a result of using a generic questionnaire, we could not measure dimensions that are specific to our population. Perhaps the focus of future research should rather be on QOL, while these measurements often include a broader scope of factors such as spirituality and financial resources (Haverman et al., 2017).

In conclusion, this study suggests that the HRQOL of young people with pHIV, who receive high-quality healthcare including regular monitoring of HRQOL, remains good while growing up. We emphasize the importance of monitoring HRQOL of these young people with pHIV developing into young adulthood.

Acknowledgements

The authors thank all study participants, their parents, and HIV pediatric nurses at the outpatient department. This work was supported by AIDSfonds (grant number 2015009).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Aids Fonds: [Grant Number 2015009].