Lifetime commercial heterosexual behavior among HIV negative elderly men from rural Chengdu, China: a modified knowledge-attitude-practice perspective

Background China is facing big challenges to achieve the “90–90-90 targets”. The HIV prevalence of elderly (≥50 years) men have been steadily increasing in China, mainly through the sexual transmission route, but sexual behaviors of them are far from well-studied. In 2019, elderly men accounted for 59.2% of HIV/AIDS cases in Sichuan, China. Methods The research design is a cross-sectional study. Face-to-face interviews were conducted among 795 HIV negative elderly men from rural Chengdu, capital City of Sichuan. Bivariate and multivariate logistic regression models were applied to examine factors associated with commercial heterosexual behavior from a modified Knowledge-Attitude-Practice (KAP) perspective. Results 129 (16.23%) respondents admitted high-risk sexual behaviors, including 11.07% commercial heterosexual behavior, 6.16% extramarital, 2.89% casual and 0.25% homosexual behavior, and no one used condom consistently. 427(68.43%) had ever gotten HIV-related Health Education (HRHE), mainly through mass media (70.49%). The HIV-related knowledge awareness rate was only 31.41%. Migration history (AOR =2.46,95% CI = 1.02–5.91), age(≥60 vs. 50–59, OR = 0.41, 95% CI = 0.19–0.91), receiving HRHE from mass media (OR = 0.37, 95%CI = 0.16–0.85), marital status (married vs. never married, OR = 0.04, 95%CI = 0–0.52), and undecided (AOR =0.02, 95%CI = 0.01–0.09) and objection (AOR =0.04, 95%CI = 0.01–0.1) attitude toward commercial sex were related to lifetime commercial heterosexual behavior. Conclusions High-risk sexual behaviors are common among elderly men from rural areas in Chengdu. Receiving HRHE from mass media and undecided and objection attitude toward commercial sex prevent elderly from being involving in commercial heterosexual behavior. According to the results, health facilities should continue to conduct systematic interventions, paying more attention to 50–59 years old group. Sex and condom use need to be talked in public. Working with mass media, health facilities give elderly men education not only focusing on HIV/AIDS, but also on knowledge and skills of condom use.


Introduction
Benefiting from a scale-up of antiretroviral therapy (ART) and prevention of mother-to-child transmission in the past two decades, annual incidences and mortality rates of human immunodeficiency virus/Acquired Immune Deficiency Syndrome (HIV/AIDS) have been declining [1,2]. The declining leads to ambitious calls to end the HIV/AIDS epidemic as a public health threat by 2030 [3]. The Joint United Nations Programme on HIV/ AIDS (UNAIDS) proposed the "90-90-90 targets", which aims to have 90% of people living with HIV (PLHIV) knowing their status, 90% of those diagnosed receiving sustained ART, and 90% of those receiving ART achieving viral suppression by 2020 [4].
China is facing big challenges to achieve the "90-90-90 targets" [5,6]. One of the biggest challenges is that the HIV prevalence of elderly (≥50 years old) men has been steadily increasing, mainly through the sexual transmission route [7][8][9][10][11][12]. Due to immune functional decline and increased vulnerability to infections [13], HIV infected elderly men quickly progress into AIDS even deaths [9,13]. Being considered as sexually inactive [14], elderly men had been excluded from the priority of HIV/AIDS prevention programs [15][16][17][18], and are not recognized and listed as a key HIV prevention group until 2017 for the first time in China [19]. On one hand, current elderly men are from 20 to 50-year-old group, who are previously sexually active and might be HIV infected as clients of female commercial sex workers (FCSW) [20][21][22][23][24][25] and men who have sex with men (MSM) [26][27][28][29][30], especially when they migrate [31][32][33] but not detected. On the other hand, they could be a newly infected group. Compared with other age groups, elderly men have minimal knowledge about HIV [34], condom use among them is even lower [25,35,36].
Located in the southwest of China, Sichuan is one of the few provinces heavily affected by the HIV/AIDS epidemic where heterosexual transmission increased from 65.3% of total cases in 2008 to 98.2% in 2019 [37]. People over 50 years old in Sichuan accounted for less than 5% of HIV/AIDS cases in 2005, but over 40% in 2017, and 59.2% in 2019, meanwhile the proportion of elderly men in the population is less than 40% [5,37,38] .One of fifth PLHIV in Sichuan are from Chengdu [5,37], the capital city. Based on case report system at population level [9,10,37], heterosexual transmission has been widely accepted as the major route for HIV epidemics in China. Why are elderly men involved in high risky sexual behaviors? Since the 1980s, rural-to-urban migration have been part of their lives for men from rural areas in China, and increased the possibility of being involved in high risky sexual behaviors [31][32][33]. According to knowledge-attitude-practice (KAP) model [25, 34-36, 39, 40] which has been applied in HIV prevention since 1990s, the hypothesis flows are the following: elderly men are involved in high risky sexual behaviors because they don't believe these behaviors are highly risky. Why don't they believe? Because they don't have enough knowledge. In order to help them to avoid high risky sexual behaviors, it is necessary to give them health education to improve knowledge level. Though KAP model has been proved not as effectively as expected [25], due to their minimal knowledge about HIV [34], current comprehensive HIV prevention strategies in Sichuan is based on KAP [5] .
Sexual behaviors of elderly men in China are far from well-studied. In 2019, 1 year before 2020 when the "90-90-90 targets" are expected to be achieved, our team conducted a cross-sectional investigation to understand sexual behaviors among elderly men in Chengdu, China where HIV prevalence of them have been increasing in recent years and HIV prevalence of MSM has been consistently high [5].

Study site
With the coverage of sentinel surveillance improving [5] and huge number of residents, rural County A in Chengdu ranked one of the top five counties in the number of PLHIV, mainly from male group.

Study process
The research design is a cross-sectional study. According to the surveillance information from the web-based National HIV/AIDS Comprehensive Response Information Management System (CRIMS), public health workers from County A center for disease prevention and control (CDC) selected the top six towns in HIV infection number to be the study sites. We conducted two pilot investigations to test the questionnaire and to adjust the investigation procedure and manual before the formal investigation.
Male interviewers were trained through three steps: before pilot investigation, pilot investigation and at site. Six male interviewers, who were well trained and pasted the interviewers' test, conducted the investigation.

Participants and investigation
Multistage cluster random sampling and multistage probability proportional stratified systematic sampling were applied to recruit eligible respondents. The hierarchy in County A is county-town-village-village group. In four of six towns, multistage cluster random sampling was applied. Firstly, one village from each town was randomly selected, then one village group was randomly selected. All eligible elderly men in the village group were invited to participate the study. In the other two of six towns, multistage probability proportional stratified systematic sampling were applied. According to the total number of elderly men in each village group and their proportion in each town, the sampling size for each village group were confirmed.
Resident roster was used to sample the potential respondents.
The selection criteria included: (1) ≥50 years old; (2) male; (3) live in the current address for at least 6 months. After eligibility was determined and a written informed consent was completed, a 30-min face-to-face structured interview was conducted in a private room by one well-trained male interviewer to make sure the respondents felt comfortable to talk about their sexual behaviors. Local slangs were used to refer sexual behaviors. Completed questionnaires were inspected at the sites to ensure no missing items, wrong information or logical errors. All participants got a big bottle of dishwashing liquid (approximately valued 2 USD) for their participation (no matter completing the questionnaire or not). From June to July 2019, 802 men were recruited, and 797(99.38%) questionnaires were reliable.

HIV rapid testing
According to the sentinel surveillance routine procedure in the township health centers, HIV rapid testing was conducted through two ways. The first method was sampling a 3-5 ml cubital venous blood of each respondent at the investigation sites and sending it to a local laboratory. An enzyme-linked immunosorbent assay (ELISA) was used for HIV antibody test. The other method was using Dot immunocolloid gold rapid test (Yingke New Technology Co., Ltd.) at the investigation sites where immediate results will be showed. In total, 2 of 797 were fast tested HIV positive.

Measures
Based on the current HIV/AIDS programs strategies in Sichuan [5] and literature [25,35,36,39,40], a modified KAP model was applied. Attitudes were reflected by sexual attitude, self-identified vulnerability, and HIV-related sigma. High-risk sexual behaviors included commercial heterosexual behavior, extramarital sexual behavior, casual sexual behavior, and homosexual behavior. Demographic characteristics, migration experiences [31][32][33], HIV-related health education (HRHE), sexual identity, and sexual desire change were taken as controlling factors for KAP model.

HIV-related knowledge
According to United Nations General Assembly Special Session on HIV/AIDS (UNGASS) indicator [40], eight survey questions (China National Eight) (Cronbach's Alpha = 0.77) were used to assess HIV-related knowledge (detail in Table 3). Respondents received one point for each correctly answered question. The sum of the eight questions represented respondents' HIV related knowledge level. Respondents were categorized into two groups: 'know' (score greater than 5) and 'do not know' (score equal and less than 5).

Sexual attitude
"How do you think about visiting FCSW?" was asked to measure their attitude toward commercial heterosexual behavior. The response options were: (1) it is understandable as a common human behavior; (2) undecided; (3) could not accept; (4) strongly objection. Besides the above four response options, "admiring" was considered as one option toward "How do you think about extramarital sexual behavior". The response options toward "How do you think about homosexual behavior" were: (1) it is understandable and he has homosexual behavior; (2) it is understandable but not acceptable; (3) could not accept; (4) could not accept but be involved in; (5) strongly objection; (6) do not know what homosexual behavior is.

Data analysis
Seven hundred ninety-five of 797 (99.75%) respondents who were HIV negative were included in the current analyses. Frequencies of nominal variables, mean and standard deviation for interval variables were assessed. Commercial heterosexual behavior was the most common high-risk sexual behavior. T test, ANOVA and chisquare tests were used to examine the relationships between commercial heterosexual behavior and independent variables including HIV-related knowledge, attitude and controlling factors. Binary logistic regression with backward stepwise selection was applied to examine factors associated with commercial heterosexual behavior, including all factors with p < 0.10 in bivariate analyses. 624(78.49%) who heard about HIV/AIDS were included in the logistic analysis. Adjusted odds ratio (AOR) and 95% confidence intervals were calculated. Factors with AOR greater than one was categorized as risk factors, and less than one was categorized as protective factors.

Basic information
Demographic characteristics and migration experiences are showed in Table 1. Four of 795 (0.50%) respondents admitted that they had used illicit drugs but not through injection. They only used Opium, reported none highrisk sexual behavior, and were tested as HIV negative.

Sexual identity and sexual desire change
Seven hundred ninety-two of 795(99.62%) reported that they are heterosexual, 2(0.25%) as bisexual, and 1 respondent (0.13%) reported that he had never heard of homosexuality. Five hundred forty-two of 795(68.2%) respondents reported that their sexual desire has declined, 19.2% reported no change, 11.9% had no sexual need, and 0.6% reported increasing desire after the age of 50.

HIV-related health education
Four hundred twenty-seven of 624(68.43%) respondents reported having ever gotten HRHE, mainly through mass media (TV, internet, and radio) (70.49%) and health facilities (CDC, hospitals, and township health centers) (58.31%). Before this investigation, 234 of 427(54.80%) received HRHE from village doctors and health workers from township health center. The details are showed in Table 2.

HIV-related knowledge level
The HIV-related knowledge awareness rate was 31.41%, and only 20 people (3.21%) scored eight of "China National Eight." The details are showed in Table 3.

Sexual attitudes
The majority (41.38%) of 795 respondents reported that they could not accept commercial heterosexual behavior, 29.69% held undecided standpoint, 18.36% said it is understandable as natural human behavior, and only 10.57% reported strong objection. The percentages of corresponding attitudes toward extramarital sexual behaviors were 42.64, 29.56%,12.45 and 14.21%. Moreover, 9 of 795(1.13%) reported that they admired the ones

HIV vulnerability and HIV-related stigma
Forty-eight of 624 (7.69%) reported that they were at risk of HIV infection. If their relatives/friends got HIV infected, 29.65% reported that they would be fearful and choose to stay far away, 23.88% would express sympathy but still choose to stay away, 22.60% would show sympathy and want to keep in touch, 12.66% stated they would treat him/her as usual, and only 11.22% would like to care for him/her more.

Sexual behaviors
Seven hundred eighty-four of 795(98.62%) reported that they had sexual partners in their lifetime. 129 (16.23%) respondents admitted high-risk sexual behaviors, among which the prevalence of commercial heterosexual behavior was 11.07%, extramarital sexual behavior 6.16%, casual sexual behavior 2.89%, and homosexual behavior 0.25%. 20 (2.52%) and 2(0.25%) were involved in two and three types of above high-risk sexual behaviors. One of the self-identified bisexual respondents reported that he was involved in of the above sex activities throughout his lifetime. Another self-identified bisexual respondent was involved in commercial heterosexual behavior and homosexual behavior.

Commercial heterosexual behavior
Eighty-eight of 795(11.07%) respondents admitted they have patronized FCSWs in their lifetime. Their first patronization happened at the age of 33.90 ± 13.43, mainly in other towns in County A (42.53%) where they worked outside of their hometown (83.82%).

Commercial heterosexual behavior related factors
In bivariate analysis, commercial heterosexual behavior was not related to HIV-related knowledge level, attitude  All correct 20 (3.21) toward extramarital sexual behavior, casual sexual behavior, and homosexual behavior, sexual identity, and sexual desire change(P > 0.05). The prevalence of commercial heterosexual behavior among the ones who had ever migrated was 13.55%, higher than among the ones who had never migrated (7.12%)(χ 2 = 7.734, P < 0.01). Factors associated with commercial heterosexual behavior in bivariate and multivariate analyses are showed in Table 4. The greatest magnitude of a protective factor was undecided (AOR =0.02, 95% CI = 0.01-0.09) and objection (AOR =0.04, 95% CI = 0.01-0.1) attitude toward commercial sex. The greatest risk factors were experiences of work out of town (AOR =2.46,95% CI = 1.02-5.91).
It is understandable that single respondents were more likely to be involved in commercial heterosexual behavior. Similar with other elderly men in the world [14], most of the respondents' sexual desire declined after the age of 50. The older, the less the respondents reported visiting FCSW, but visiting FCSW is not related to sexual desire change. This finding is in concord with the fact that visiting FCSW is not only to fulfill physical needs but emotional needs also [25,41]. People often think that the conservative attitude could be a restriction against commercial heterosexual behavior [42]. In our survey, it is true among the mild objectors who are against this behavior. Surprisingly statistics found no difference in participating numbers of the commercial heterosexual behavior between the following two groups: accepting it as a natural human behavior verses strong objection. Since the 1980s, sexual attitudes in China have changed from procreation within a family context to being an individual's responsibility as long as there is no foreseen negative impact on the wellbeing of others or the larger society [42]. However, due to convention [42] and stigma [43], Chinese, especially elderly people, do not talk about sex [44], elderly men involving in commercial heterosexual behaviors may be too shy to admit it. It can only become more difficult to access the HIV education for them than in other age groups.
Worldwide, there is huge regional variation of highrisk sexual behaviors [45], mainly determined by structural factors, such as social and economic context [45,46]. The greatest risk factors for commercial heterosexual behavior in this study were experiences of work out of town. Since the 1980s, rural-to-urban migration have been part of their lives for men from rural areas in China. In this study, more than half of the respondents had rural-to-urban migration experience. They were sexually active migrating alone [47,48], easy to be involved in high-risk sexual behaviors [27,31]. Their first visit to FCSW mostly happened during their migration. Visiting FCSW was cheap and easy to get access. No one used condoms consistently, mainly because of lack of skills and knowledge. Condom in China has been used mainly for contraception, meanwhile the main contraception method among married couples were intrauterine device (IUD) for women [49]. When they were young, respondents did not take major responsibility for contraception, neither did they have too much chances to learn how to use condom, someone even never used condom in their lifetime. Moreover, carrying condom in brothels is taken as an evidence for visiting FCSW in China where sex work is illegal [49], which hold elderly men from initiating condom use when visiting FCSW.
In this study, receiving HRHE through mass media,which cover all groups of population [39] and has been part of the systematic HIV prevention [5], helped elderly men avoid visiting FCSW. Health facilities have played an important role in improving elderly men's HIV knowledge, but been far from success. In terms of "90-90-90 targets", even though it is not impossible to achieve the goals of 90% of HIV diagnosed population receiving sustained ART, as well as 90% of those receiving ART be viral suppressed by 2020, it is very hard to achieve 90% of diagnosis rate. Not only because free HIV testing is a voluntary action in China [5], but also one of five elderly men have not even heard the term of HIV/AIDS, let alone utilizing the free test. Current HIV prevention strategies in County A succeed in reducing blood transmission route, and most elderly men are aware of blood transmission route and none use illicit drugs through injections. However, HIV related knowledge level among elderly men is very low (31.41%). There are misunderstandings about transmission routes and stigma toward PLHIV which probably hold them back from utilizing the free test [43]. When they were detected as HIV positive, it is highly possible that they had been infected for a long time. During their undetected HIV status, they would infect their sexual partners unintentionally. FCSW infected by elderly men will also infect other age group men [25,[31][32][33]50]. That is why elderly men have been listed as key groups in HIV prevention in 2017 [19].

Limitations
The current study should be noted including a crosssectional study design and self-report behavior information. Face-to-face interviews may heighten socially desirable responses such as conservative attitudes toward high-risk behaviors, and low report of high-risk sexual behaviors. In order to confront the problems, our interviewers were well trained, interviews were conducted in separate room, and local slangs were used.
Migration experience is a risk factor for high-risk sexual behavior. At the time of investigation, we missed around 6% elderly men who migrated outside Chengdu. Fortunately, the prevalence of commercial heterosexual behavior in this study was only underestimated by 1.33%.
Due to limited sample size for bisexual respondents, we cannot draw the conclusions that bisexual elderly men were more likely to be involved in high-risk sexual behaviors than heterosexual ones, but the fact that both two self-identified bisexual respondents were involved in commercial heterosexual behaviors gave a clue for further study.

Conclusions
Comparing with illicit drug use, high-risk sexual behaviors are more common among elderly men from rural areas in Chengdu. HIV related knowledge level among elderly men is very low (31.41%). Receiving HRHE from mass media and undecided and objection attitude toward commercial sex prevent elderly from being involving in commercial heterosexual behavior. According to the results, health intervention strategies for elderly men could be the following: Health facilities continue to conduct systematic interventions, paying more attention to 50-59 years old group. Sex and condom use need to be talked in public. Working with mass media, health facilities give elderly men health education not only focusing on transmission route of HIV/AIDS, but also on knowledge and skills of condom use.