Health-related quality of life and associated factors among people living with HIV/AIDS in Sichuan, China: A cross-sectional study

Purpose This study aimed to explore health-related quality of life (HRQoL) and its associated factors among people living with HIV/AIDS (PLWH) in Sichuan, China. Methods A total of 401 PLWH were recruited from the city of Panzhihua between August 2018 and January 2019. Demographic characteristics and disease-related data were collected by self-administered questionnaires and medical system records. Health-related quality of life (HRQoL) was measured by the medical outcome study HIV health survey (MOS-HIV), which measured ten subdimensions and two summarized dimensions, the physical health summary score (PHS) and the mental health summary score (MHS). Logistic regression models were used to explore the variables independently associated with quality of life. Results The PHS and MHS measured by MOS-HIV were 53.66 ± 6.80 and 51.31 ± 7.66, respectively. Younger age, higher educational level, no methadone use, higher CD4 lymphocyte counts, less symptom counts and heathy BMI significantly were associated with higher HRQOL in the univariate χ2-test analysis. Education level was found to have a significant influence on patients' quality of life, both in physical health (P = 0.022) and mental health (P = 0.002) dimensions. Younger age (P = 0.032), higher CD4 lymphocyte counts (P = 0.007), less symptom counts (P < 0.001) and health BMI level (P < 0.001) were positively related to the PHS of quality of life in the multivariable logistic regression model. Conclusion The HRQoL of PLWH in Sinchuan Province was relatively low. Age, educational level, methadone use, CD4 lymphocyte counts, symptom counts and BMI were positively related to quality of life. This study indicates that health caregivers should pay more attention to comorbidity issues and mental health in PLWH, especially for those with lower education levels, unhealthy body mass index, more symptomatic presentation and older age.


Introduction
The human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS) pandemic is a serious global challenge (1). According  relevant and holistic interventions to improve the general wellbeing and overall HRQoL of PLWH (32). In particular, HRQoL in PLWH is critical for monitoring the impact of drug therapy on disease progression (33). Thus, the aim of this study was to investigate HRQoL status and associated factors in PLWH by using the MOS-HIV.

Study design and participants
This was a cross-sectional study conducted from August 2018 to January 2019 in Panzhihua city, Sichuan Province, China. HRQoL was used as the main indicator, and the sample content was designed to be 328 with reference to the relevant literature and the mean survey formula. The required sample size was finally determined to be 410 cases because 20% of invalid questionnaires were considered. Regarding the sites for conducting the sampling, 2 districts or counties in Panzhihua were randomly selected to conduct the survey. Finally, Renhe District and Miyi County were chosen. The researcher inquired about the number of PLWH in different districts or counties of Panzhihua City in that year to calculate the relative proportion to determine the sample size assigned to the two sites. And the convenient sampling sites were selected as the Fourth People's Hospital of Renhe District and the People's Hospital of Miyi County in Panzhihua City, respectively, both of which were designated as PLWH getting free medications, testing and consulting. Participants were recruited through convenience sampling. The inclusion criteria of participants were (1) HIV seropositivity, (2) age older than 17 years, (3) willing to fill out the study questionnaire. People with mental illness or cognitive impairment were excluded.

Ethical considerations
The study was approved by the West China Hospital Medical Ethics Committee (study ID#: 20170430). The objectives and procedures of the study were verbally explained to participants. Before the study, participants signed written informed consent.

Data collection
All data were collected through paper questionnaires. First, the investigators received training about the administration of the questionnaires. Second, all participants were informed of the purpose, content and potential risks of the study before the investigation. Participants independently completed the anonymous questionnaires. If the participants had difficulty understanding or reading the questionnaire, the investigators explained it in detail and recorded the answer if inquired. After the questionnaire was completed by participants recruited by the local CDC, the investigators immediately completed the questionnaire in person to verify whether . /fpubh. .
the respondent had answered all the questions. If the patient refused to answer and the incomplete part of the questionnaire was >20%, the questionnaire was discarded. Each participant was paid 50 (equivalent to $7) as compensation for transportation.

Measurements Sociodemographic characteristics
Sociodemographic characteristics included sex, age, ethnicity, educational level, marital status, religion, sexual orientation, religious beliefs, occupation, alcohol consumption, residence, smoking habits, drinking habits, BMI and per capita monthly household income. In this study, age was divided into three groups: young adults aged 18-45 years, middle-aged adults aged 45-60 years and elderly adults aged 65 years and above. Participants were classified into 4 categories according to the weight determination criteria for Chinese adults published by the National Health and Wellness Commission Participations, which were underweight (<18.5), healthy weight (18.5-24.9), overweight (25-27.9), and obese (≥28).

Disease-related characteristics
We chose methadone use, infection route, disclosure status, CD4 cell count, symptom counts, duration since HIV diagnosis, duration of treatment, viral load and the 30 most common symptoms for now according to the literature review and clinical experience. Years since HIV diagnosis, duration of treatment, CD4 cell count and viral load from database and clinic records. Disclosure status, infection route, and symptoms were obtained through patient self-reporting. If there were missing data in the database or clinic records, the investigator asked the patient for relevant information.
Health-related quality of life HRQoL was assessed by the Medical Outcomes Study HIV Survey (MOS-HIV), which is a brief, comprehensive measure of health-related quality of life used extensively in HIV/AIDS (33). It consists of 35 items and measures 10 dimensions, including general health, physical function, role function, social function, cognitive function, pain, mental health, energy/fatigue, health distress, and quality of life. Of these 10 domains, 8 were multi-item and 2 (Social Function and Quality of Life) consisted of a single item. Ten subdimension scores and two summary dimensions (physical health summary score, PHS and mental health summary score, MHS) were obtained by adding the raw item scores of the respective scales and then transforming them into a 0-100 scale, with higher scores indicating better functioning and wellbeing. The simplified Chinese version of the MOS-HIV questionnaire has been reported to have good validity and reliability (34,35). The Cronbach's α of the PHS and MHS scales was 0.87 and 0.89, respectively. The Cronbach's α for individual dimensions was >0.70 (34).

Statistical analyses
Data were inputted by two researchers through EpiData 3.1, and all statistical analyses were performed using SPSS 24.0. A Pvalue of <0.05 was considered statistically significant. Descriptive statistics analysis was performed by mean ± standard deviation (SD) and median, interquartile range (IQR), frequency, and constituent ratio, as appropriate. We divided PHS and MHS scores into dichotomous variables by 50, in which scores above or below 50 can be considered better or worse HRQoL (36). Univariate analysis χ 2 -test for categorical data and Spearman's correlation for continuous data were performed to examine the quality of life of PLWH with different demographic characteristics and disease-related characteristics. Intragroup comparisons for multiple categorical variables were also performed by Fisher's exact test. Binary logistic regression analysis was conducted to explore factors associated with HRQoL in PLWH. Because there is no universal consensus in choosing predictor variables in a multivariable logistic regression analysis, we included all the variables with P ≤ 0.2 in the univariate analysis and all clinically significant variables regardless of their P-values, based on the literature evidence. The Hosmer-Lemeshow goodness-offit statistic with a P-value > 0.05 was considered a well-fitting regression model, and the percentage of the variability predicted by the model was explained by the Nagelkerke R 2 .

Results
Sociodemographic and disease-related characteristics of participants Finally, 410 participants were enrolled in the study. However, A total of 9 questionnaires with missing values >20% were excluded, and the final number of valid questionnaires collected was 401. Among the participants, 64.7% (n = 258) were male and 35.2% (n = 141) were female. Their mean (SD) age was 50.02 (15.06) years. Most participants were of Han ethnicity (n = 340, 85.4%), married (n = 227, 58.2%), and employed (n = 304, 76.9%). Nearly half of the participants were living in rural areas (n = 209, 53.3%). The sociodemographic characteristics of the participants and the HRQoL in different groups are presented in Table 1.

HIV-related variables
The mean duration after being diagnosed with HIV was 3.7 (SD = 3.3) years, ranging from 1 month to 30 years. All participants received ART with an average duration of 2.8 (SD = 2.4) years. Most of the participants' (n = 284, 72.6%) immune systems were weakened by HIV, while more than half of the patients had viral suppression (n = 249, 75.2%) and disclosed the HIV-infected condition (n = 287, 74.1%). The most common form of infection was sexual contact (n = 184, 47.1%). Most participants reported that they had more than one symptom, and only 14.5% (n = 58) of people indicated that they were not bothered by HIV-related symptoms ( Table 2).  The bold values indicate that the P value is less than 0.05 and statistically significant.

Symptoms
The majority of participants suffered fewer than 10 symptoms ( Table 2). The top 5 symptoms reported by patients were insomnia (32.90%), fatigue (29.00%), forgetfulness (28.10%), joint pain (24.60%) and dry mouth (24.40%). Most participants' symptoms were light ( Table 3). The least reported symptom was mouth ulcers. We found that anxiety and depression were not the most frequent symptoms among PLWH. In this study, the number of patients who reported having symptoms of anxiety and depression was 68 (16.6%) and 46 (11.3%), respectively.

HRQoL of participants
The Cronbach's α coefficient for the MOS-HIV in this study was 0.659. According to the questionnaire, the HRQoL of the participants in terms of PHS, MHS and 10-dimensional scores are presented in Table 4. The mean physical health summary (PHS) and mental health summary (MHS) scores were 53.66 (6.81) and 51.31 (7.66), respectively, and there were 300 (74.8%) participants with good PHS and 255 (63.6%) participants with good MHS based on the standardized mean score of 50 (36). Of the 10 dimensions, the highest mean was found in the role functioning dimension (87.39 ± 29.37), while the lowest mean was found in the general health perceptions subscale (61.51 ± 21.37). There was no significant correlation between PHS and MHS (Spearman's correlation = 0.20; P = 0.017) and no significant correlation between PHS and MHS rank material, which was divided by 50 to be considered a good or poor health situation (Pearson chi-square = 2.99, P = 0.084).
/fpubh. .  were reported by people with higher education levels (p = 0.001) and fewer symptoms. However, there was no correlation between disclosure, sex, sexual orientation and HRQoL score. The association of the predictor variables with the PHS and MHS categories were explored using multivariate analysis and multivariable logistic regression analyses and are presented in Table 5.
Only educational level and symptom counts were both significant in PHS and MHS in the univariable analysis and multivariable logistic regression analyses. However, there was no significant correlation between PHS and MHS by Spearman's correlation or χ 2 -test. Youth population, higher education, health BMI level, CD4 level ≥500 cells/mm 3 , fewer symptoms, not using

Median (IQR) Mean (SD) Summary scores
Physical health summary score 55.1 (49.9-58.7) 53.6 (6. methadone and shorter time receiving ART were associated with good PHS in the univariable analysis, whereas only education level and symptom counts were associated with MHS. Some variables significantly associated with PHS and MHS in the univariate analyses were not significant in the multivariable models ( Table 4). The Hosmer-Lemeshow goodness-offit test p-values were 0.628 and 1 for the PHS and MHS models constructed, respectively, suggestive of well-fitting models. It is assumed that 33.2% of the variability in the PHS (Nagelkerke R 2 = 0.33) and 7% of the variability in the MHS (Nagelkerke R 2 = 0.07) is explained by these models.

Discussion
The results of the study showed that the total physical health score (PHS) and mental health score (MHS) of HIV/AIDS in Sichuan Province were 53.6 ± 6.8 and 51.3 ± 7.7, which were higher than those in previous studies surveyed in other areas of China (37)(38)(39)(40). The reason may be that some time has passed since the previous studies were conducted. The impact of this illness on their quality of life is substantially lower than a few years ago because of the significant improvement in mortality and morbidity in PLWH with the implementation of ART (21). With the implementation of the "Four Free and One Care" policy, PLWH do not need to pay for medications, testing and counseling, can even obtain relief money from the government. Antiviral treatment and the four-free care policy help PLWH in poor areas of China receive better treatment, which helps to control disease and improve their quality of life.
The dimensions that scored higher were role function, physical function and health distress, which is similar to other studies (21,41,42). This is probably because HIV/AIDS can be controlled like any other chronic disease with the implementation of ART (41). The rate of severe illness decreased, and patients  did not suffer significant impairment in physical or social functioning. The participants reported lower scores in the quality of life, energy/fatigue and general health perception dimensions, consistent with the results of previous studies (21,41,43). The lowest mean score was 61.52 for general health perception. The highest score in the energy/fatigue dimension is 95, and only in this dimension did no patient report a full score, 100. This result indicated that HIV-infected peoples' quality of life was still impaired and that each participant's energy was affected to varying degrees. Fatigue is the second most common symptom, with a 29% reporting rate. The causes of fatigue are complex. The disease itself, comorbidities, drug effects and even psychological factors may cause fatigue (41,44,45). We suggest that fatigue management and self-perception of health promotion are areas that need further attention for researchers. The study revealed that education level was significantly associated with both the PHS and MHS of PLWH. The vast majority (80.1%) of participants in our study had only received a junior high school education or less. A further 10.5% were illiterate. Consistent with the findings of previous studies (42, 46, 47), educational level was a protective factor of HRQoL. Moreover, having no formal education is a barrier to accessing health services and increasing PLWH vulnerability (48). This result was also confirmed in this study, even if PLWH with primary education had a much lower risk of low quality of life than illiterate PLWH. The impact of education level on PLWH is multifaceted. It plays an important role in the spread and prevalence of HIV (49) and is more likely to lead to risky sexual behavior (50). Well-educated PLWH were more likely to show good knowledge, positive attitudes toward HIV/AIDS (48) and good social support (47) and cognitive level (42, 46), which leads to better self-management and health outcomes. Although education level is difficult to change, we recommend that more health information and education be provided through brochures, bulletin boards, and public online platforms. More tailored patient education should be conducted during the patient's hospitalization and at follow-up visits.
Symptoms of HIV have also been reported to be associated with HRQoL in many previous studies (22,32,51,52), and more serious complications can lead to poor quality of life. In our study, we found that the number of symptoms was a risk factor for HRQoL, both in PHS and MHS. The most common symptoms reported by the participants in our study were sleep disturbances, fatigue, forgetfulness, joint pain and dry mouth. Sleep disturbances, fatigue, and forgetfulness have received extensive attention in previous studies (35,53,54). There are many factors that influence the symptoms, including the impact of AIDS on the immune system (53,54), the adverse effects of medication (35,55,56) and psychosocial factors (45,55,57). Many studies have revealed that .
/fpubh. . joint pain (58)(59)(60)(61)(62) and xerostomia are also common symptoms in PLWH (63, 64). In a cross-sectional survey of 195 PLWH conducted in Brazil, 40% of patients reported dry mouth symptoms (63). In a cross-sectional survey of 312 PLWH conducted in Italy, 34% reported joint pain (65). They have all been shown to have a negative impact on the QOL of PLWH (63, 65, 66). Possible causes are joint pain and salivary gland hypofunction due to ART (59,64,67). It has also been suggested that joint pain may be related to inflammatory responses mediated by inflammatory factors, but the exact mechanisms need to be further elucidated (68). This suggests that we need to include a full range of physical, psychological, and social care in future studies to reduce the impact of symptoms on the quality of life of PLWH and implement proven effective interventions in health services. Meanwhile, PLWH's joint pain and dry mouth need more attention by researchers and medical services team members. Age was an independent risk factor that was significantly associated with the physical health of PLWH, which was similar to other studies (69,70) but different from George's research (21). It may be that older PLWH experience a greater burden of age-related comorbidities, poorer social determinants of health and even ageism (71). Murzin's team (71) showed that ageism transcended multiple interactions and environments from dating to healthcare and community services to society because of social determinants, health provider issues and structural challenges. Therefore, we suggest that the government should pay more attention to older people, especially in the advent of an aging society.
This study showed a significant relationship between HRQoL PHS and CD4 lymphocyte counts. As CD4 lymphocyte counts increased, physical HRQoL improved. The findings were consistent with the results of other studies (31,72,73). A few studies have reported that lower CD4 counts may affect physical health (74-76). CD4+ cell count is an indicator of clinical progression and can reflect the impact of therapeutic efforts. Because of the "Four Frees and One Care" Policy, all HIV/AIDS patients in China can receive free ART in China. However, a relatively low CD4+ cell count of participants was also found in our study. This suggests that strategies to enhance medication adherence may be needed.
BMI is considered to be an important factor associated with patients' quality of life. In our study, both malnutrition and obesity may have contributed to the impairment of physical health. This is consistent with previous studies (77,78), and some studies showed that obesity or malnutrition affects adult PLWH muscle strength (79) and the risk of frailty (80,81) and comorbidities in PLWH followed up for 12 years (78). Even, BMI can be a predictor tool for death risk in Ethiopian adults living with HIV on ART (82). However, it is noteworthy that the PHS scores of the overweight group were the highest, even exceeding the healthy BMI level. Obesity may be due to a lack of physical activity and poor lifestyle (21), while underweight may be associated with disease consumption (77). We need more research to reveal whether the burden of being overweight will or will not affect patients' psychosocial functioning and how it could happen.
In addition, only 75.2% of participants in this study achieved viral suppression, which is much lower than the 92.4% in Beijing (78) but close to the figure for Liangshan, Sichuan (79). This may be attributed to the relatively short duration of HIV diagnosis and antiretroviral treatment in most of our participates. In addition, the comparison of data from two districts in China also illustrates the large differences in treatment effectiveness between regions. In particular, there is the largest base of HIV-infected patients in Sichuan Province. To achieve the "95 95 95" targets, there is a greater need for more targeted programs in less economically developed regions to explore critical influencing factors and efficient management interventions. Rationalize the allocation of limited resources to the whole process of HIV/AIDS management.
Although some studies have shown that women with HIV/AIDS have a higher quality of life than men (80, 81), we did not find any differences in quality of life between male and female HIV/AIDS patients. The difference might be due to the study population, sampling method, and measurement tools used to assess HRQoL among studies. A cross-sectional study of sex differences in quality of life among PLWH found that total physical health scores and mental health scores among PLWH were not statistically significant by sex (82).
Disclosure of HIV infection in this study was 71.6%, higher than other studies (21,83). The majority of participants chose to disclose to their spouses (58.9%) and children (46.7%), and a small proportion disclosed the disease to parents (19.6%) and friends (7.4%). This may be one of the reasons why the MHS score in this study was higher than that in other studies.
One thing that should not be overlooked in our study is that PLWH explain only 7% of the PHS, and only two factors are included in the final multivariate regression. The reason may be that the sample was a general adult PLWH population with higher psychological scores than those reported in other literature (84,85), and fewer people reported anxiety (17%) and depression (11%) among the symptom questionnaires, which indicates a relatively high level of mental health in this study sample. Second, nearly half of the PLWH in this sample had received only primary education and below, and their cognitive level of mental health might not be sufficient (86). Last, previous studies have shown that disclosure, sexual orientation and other factors have a significant impact on the mental health of PLWH (87), the level of disclosure in this sample population is higher than that in previous studies (21,83), and the sexual orientation is mostly heterosexual. It would be easier to tap into special populations such as men who have sex with men (MSM), sexual service providers or adolescent groups to identify the relevant influencing factors.
There were limitations in our study that should be acknowledged. First, it is a population with a relatively recent history of antiretroviral treatment but with many symptoms that seem more related to treatments used in the past. However, the type of treatment and regimens used are not available. We were unable to analyze whether the patient-reported symptom presentation was related. Second, the study was conducted in a single city in Sichuan. Future studies need to be conducted in other Sichuan provinces and even nationwide to verify whether there are differences in PLWH quality of life across geographic areas. Third, we could not determine the exact causality between HRQoL and the related factors due to the cross-sectional design. Therefore, there is a need for longitudinal research and multicenter studies to confirm our findings and explore their causality.

Conclusion
In conclusion, the HRQoL of PLWH in Panzhihua improved compared to previous studies. Alleviating symptoms of HIV and preventing comorbidities, especially insomnia and fatigue, are important for patient life treatment. We recommend that further research should be conducted to explore factors affecting quality of life so that we can develop more comprehensive interventions to prevent HIV/AIDS. Moreover, more care and support should be given to patients, especially for those with lower educational attainment, unhealthy body mass index, more symptomatic presentation and older age.

Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

Ethics statement
The studies involving human participants were reviewed and approved by Sichuan University Medical Ethics Committee. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Author contributions
HZ: writing manuscripts and data processing. FW: data entry and writing manuscript. YS, ZN, and HC: draft touch-ups and guidance on writing. All authors contributed to the article and approved the submitted version.