‘Too old to test?’: A life course approach to HIV-related risk and self-testing among midlife-older adults in Malawi

Background Despite the aging HIV epidemic, increasing age can be associated with hesitancy to test. Addressing this gap is a critical policy concern and highlights the urgent need to identify the underlying factors, to improve knowledge of HIV-related risks as well as uptake of HIV testing and prevention services, in midlife-older adults. Methods We conducted five focus group discussions and 12 in-depth interviews between April 2013 and November 2016 among rural and urban Malawian midlife-older (≥30 years) men and women. Using a life-course theoretical framework we explored how age is enacted socially and its implications on HIV testing and sexual risk behaviours. We also explore the potential for HIV self-testing (HIVST) to be part of a broader strategy for engaging midlife-older adults in HIV testing, prevention and care. Thematic analysis was used to identify recurrent themes and variations. Results Midlife-older adults (30–74 years of age) associated their age with respectability and identified HIV as “a disease of youth” that would not affect them, with age protecting them against infidelity and sexual risk-taking. HIV testing was felt to be stigmatizing, challenging age norms, threatening social status, and implying “lack of wisdom”. These norms drove self-testing preferences at home or other locations deemed age and gender appropriate. Awareness of the potential for long-standing undiagnosed HIV to be carried forward from past relationships was minimal, as was understanding of treatment-as-prevention. These norms led to HIV testing being perceived as a threat to status by older adults, contributing to low levels of recent HIV testing compared to younger adults. Conclusions Characteristics associated with age-gender norms and social position encourage self-testing but drive poor HIV-risk perception and unacceptability of conventional HIV testing in midlife-older adults. There is an urgent need to provide targeted messages and services more appropriate to midlife-older adults in sub-Saharan Africa. HIVST which has often been highlighted as a tool for reaching young people, may be a valuable tool for engaging midlife-older age groups who may not otherwise test. Supplementary Information The online version contains supplementary material available at 10.1186/s12889-021-10573-7.

Low levels of recent HIV testing in midlife-older populations, generally ≥30 years of age, stand in contrast with the changing epidemic in southern Africa. In east and southern Africawhere 85% of all PLHIV are now aware of their status [1] many midlife-older adults, especially men, have still never tested or not tested recently [5]. Late diagnosis and initiation of treatment is common among midlife-older men, who prescribe to traditional views of masculinity and age which view accessing HIV services, including testing, as a threat to their position in community and family life [6,7]. This trend among midlife-older men may also contribute to new infections in younger women in southern Africa where intergenerational relationships have been considered a key driver of the HIV epidemic [8]. For midlifeolder women, some of whom no longer need antenatal or family planning services, access to and uptake of HIV testing declines considerably [9]. Declining concerns about unwanted pregnancy, together with minimal public awareness of changes in the demographics of the HIV epidemic [10,11], may lead to greater willingness to engage in condomless sex with existing or new partners of unknown HIV status [12][13][14].
Despite increasing HIV prevalence, risk perception and perceived susceptibility to HIV is believed to be low in midlife-older age groups [9,15,16]. Age-targeted health education and behaviour change campaigns in southern Africa remain focused on adolescents and young people (15-24 years) and have yet to be adapted for other age groups. As a result, in urban Malawi, only 42.8 and 66.7% of over-40-year-old men and women participated in the first year of community-based HIV self-testing (HIVST) distribution in 2012-13 and had an HIV prevalence of 22.5%. In contrast, during this same period, 89.3 and 100% of adolescent boys and girls aged 16-19 years participated with a lower HIV prevalence of 2.5% [17]. Since then, Malawi has scaled-up HIVST and youth continue to be a priority [18].
Poor knowledge of their HIV risk and hesitancy to access HIV testing in the context of high prevalence and incidence among midlife-older men and women undermine global efforts to achieve and maintain low HIV incidence by 2030. Addressing this gap is a critical policy concern and highlights the urgent need to identify the underlying factors, to improve knowledge of HIVrelated risks as well as uptake of HIV testing and prevention services, in midlife-older adults.

Theoretical framework
The life-course comprises the set of socially defined events and roles individuals enact over time [19]. Within the life-course, in addition to chronological age, aging is socially and historically constructed by different life events, experiences, and social expectations which shape individual behaviours, perceptions and attitudes [20,21]. These age constructs inform how individuals enact and cultures express "age" through different life stages, such as adulthood, midlife and elderhood, and their meaning in society [20,[22][23][24]. Within these constructs culturallydominant ideals of achievement, respectability and status contribute to shared age-related identities [25]. Diversity of experiences informed by gender, power, class, place and time also influence and define multiple and continuous expressions of age within and across cultures [26][27][28].
Studies in southern Africa have begun to apply lifecourse approaches to understanding HIV among older adults [9,16,[29][30][31][32][33][34][35]. Findings have highlighted how age norms can make accessing HIV prevention, testing and care challenging. This is often due to concerns about loss of respect and age-related stigma and discrimination reflecting experiences from earlier times when treatment was not widely available. Age norms can also be gendered by traditional masculinities reducing HIV testing and ART initiation among older men [7], or increased household and community responsibilities becoming a barrier to health services, including HIV testing, among older women [12,22,29].
The vast majority of studies continue to focus on aging as it relates to trends in HIV epidemiology, risk factors and issues following an HIV-positive diagnosis [10,11,15,29,36]. Framing the discussion around the lifecourse [19], while also considering age and gender [20,25], contributes to the literature by examining the gendered construction of 'respectable' midlife-older age adult life in southern Africa, and how it influences perceptions of HIV-related risk and both conventional HIV testing and self-testing in urban and rural Malawi.
Here we explore perceptions of HIV risk and HIV testing among midlife-older men and women living in rural and urban Malawi following the introduction of HIVST. Drawing on theories of life-course [20,21] and performance of self in society [24], we use qualitative methods to understand how social age is enacted and implications for HIV testing and sexual risk behaviours among midlife-older Malawians. We also explore the potential for HIVST to be part of a broader strategy for engaging midlife-older adults in HIV testing, prevention and care.

Methods
This qualitative study was nested in two clusterrandomised trials of community-based HIVST, one (ISCTN02004005) in three high-density townships of Ndirande, Likhubula and Chilomoni in urban Blantyre and one (NCT02718274) in Manyenje and Nkoka villages in rural Blantyre, Malawi. Results, including uptake of HIVST by age and sex have been described elsewhere [17,18,37]. Briefly, volunteers from intervention communities were trained as community-based distributors (CBDs) responsible for providing HIVST to their neighbours [17,38]. Distributors provided residents with pretest information, including a demonstration on how to use an oral HIVST kit (OraQuick HIV Self-Test, Ora-Sure Technologies, LLC, Bethlehem, PA, USA). Community members could self-test with a CBD or take a kit home to test later. All self-testers were informed by CBDs that they needed to confirm reactive results and where to access treatment. Disclosure of self-test results was not required, though many shared results with CBDs [17,18,37].
Five focus group discussions (FGDs) (n = 48) and 12 in-depth interviews (IDIs) were conducted between April 2013 and November 2016 in five HIVST intervention areas. Community residents were eligible for FGDs if 35 years or older, and IDIs if over 30 years. FGD and IDI participants were recruited prospectively, by liaising with CBDs. Participants were only engaged once for approximately 1 h and were provided transport and refreshments. FGDs included: two with both men and women cluster residents, two with women only, and one with community distributors only. FGDs ranged from 8 to 12 participants, including 48 participants in total. IDIs were limited to community members who had self-tested (Table 1).
Community FGDs and IDIs explored personal experiences and community perceptions, including aspects of the life-course, such as: age and gender norms, social positions, as well as knowledge of HIV and treatment, risk perception, stigma, social harm, self-testing, relationships with distributors, and perceptions of barriers to testing. The distributor FGD explored their experiences offering HIVST to middle-aged and older community members, and perceived barriers to testing. Topic and interview guides are publicly available on the London School of Hygiene and Tropical Medicine (LSHTM) website [39]. All participants in IDIs and FGDs were assured that information shared was confidential and identifiable information would not be shared. FGD participants were requested to maintain confidentiality related to discussion. All participants were anonymized and identifying information was de-linked using a unique study code prior to review and analysis. Unique participant study codes were only made accessible to the research team.
FGDs and IDIs were conducted in the local language (Chichewa) by three male field workers experienced in qualitative methods (MK, MP, and LK) and audio recorded. One field worker led each FGD and IDI. All field workers were trained, participated in development of interview guides and had extensive experience working in the community as qualitative researchers and were native speakers. All recordings were transcribed into Chichewa and then translated into English by trained personnel. Electronic data was stored in password protected servers (RedCap) and password protected computers. Paper-based data was kept in a locked cabinet which was only accessible to the research team.
Thematic content and framework analysis were conducted on FGD and IDI transcripts using a common coding framework developed under the STAR Initiative for data sorting and indexing. Transcripts were first read (CJ, JM, MK) and grouped by themes and then triangulated across participants by age, gender, HIV status and setting (urban or rural). Transcripts were read and reread in English and Chichewa, then coded themes were jointly re-reviewed to ensure consistency (CJ, MK). Following which, themes were extracted and used to refine the framework iteratively and inform the analysis that was software assisted by NVIVO 12 (QSR International Pty Ltd.). Study findings, from Self-Testing Africa (STAR) and HitTB, were then disseminated to local communities and the ministry of health. Individual participants, however, did not review results.
Ethical approval was obtained from LSHTM, and Malawi College of Medicine and Research and Ethics Committee (COMREC) (P011/10/1020). According to COMREC approved procedures, all participants were informed about the study and goals to evaluate HIVST. All literate participants provided written informed consent and all illiterate participants provided verbal witnessed informed consent and a thumb print. The study and presentation of results followed COREQ guidelines.

Results
Community residents participating in the study were aged 30-74 years. Additional characteristics and key FGDs and IDIs quotes are summarised in Tables 1 and 2 in addition to select quotes which highlight key themes. Data presented, and findings were consistent. Supplementary information includes full summary of participant quotes (S1).
The main themes, using a life-course framework, identified that there were stages of elderhood, with midlifeolder age beginning around 30 to 35 years but mostly defined by respectable behaviour (e.g. low sexual activity, fidelity), attributes (e.g. wisdom) and life events (e.g. marriage, number of children) (see Fig. 1). Exploration of perceptions on HIV risk and HIV testing revealed views on respectable behaviour and age drove midlife-older adults to associate HIV with youth, but this reflected lack of knowledge and awareness of their own age-specific risks, including that HIV could have been acquired earlier and in previous relationships. The risk of sexual transmission at older ages was ignored: instead, there was a strong focus on non-sexual modes of transmission, implicitly considered more socially acceptable.
Because of their age, participants felt HIV testing could undermine their respectability, their roles and relationships in family and community life, and that health workers may stigmatize them if diagnosed with HIV. Learning one's HIV-positive status later in life was considered stressful and deemed unhelpful, particularly as knowledge of the full benefits of ART and its role in preventing further transmission among those virally suppressed was very limited. As a result, midlife-older adults often considered conventional HIV testing unacceptable. Conversely, HIVST was highly preferable, for its convenience and privacy,  The one who is looked at as an old person, is the one who follows the advice that he has been given, because when an old person is told something he follows the rules. The one whom we consider as a young person, is the one who doesn't follow the advice that people give him.  especially through door-to-door distribution, though there were differences by age and sex. Figure 1 illustrates emerging themes within a life course approach with variation by age and gender among 'respectable midlife-older adults' in urban and rural Malawi.

Defining age and midlife-older adulthood
Participants drew from a range of norms and attributes to define elderhood and the start of midlife-older age in Malawi. While some focused on chronological age, with midlife-older age starting between 30 and 35 years, most defined this period by attributes and experiences [20],  (2) Social positions of respect and being leaders in the community relates to midlife-older adults perceptions that testing is inappropriate for their age this HIV risks resulting in the "othering" of HIV-related risk behaviours as those which are contrary to characteristics of midlife-older adults; and (3) Social positions among older women as faithful and trustworthy relates to men's views that they are low risk, and because of their lack of knowledge about HIV serodiscordancy among couples, they do not think they need HIV testing including the start of declining health or loss of strength, increased wisdom, respectable behaviour, major life events and increased responsibilities (e.g. marriage, parenthood). Because of this, many acknowledged that youth could be treated as "older", based on marriage and through being compliant and respectable. Whereas those engaged in 'bad' or 'unwise' behaviours were considered "childish" at any age ( Table 2, quotes (Q) 1-3).
The one who is looked at as an old person, is the one who follows the advice that he has been given, because when an old person is told something he follows the rules. The one whom we consider as a young person, is the one who doesn't follow the advice that people give him. -Community resident, 35-49 years, urban, FGD Participants voiced disapproval however when life events challenged life-course norms, such as older men leaving families and children, or pregnancy among young girls and older women.

Midlife social positions and sexuality
Gender roles and responsibilities continued to follow traditional heteronormative dichotomies of men as head-of-household and women caring for the household and community. With age, both midlife-older adults, especially women, were expected to become more responsible, with diminishing sexual activity and infidelity. Being faithful and trustworthy was considered important for both genders, although most commonly described as a 'mature' female trait. Older participants described themselves as 'less sexually active', more likely to be married, and less likely to have many partners. Indeed, having a 'highly active sex drive' in later life was considered socially unacceptable (Q4).
How could an elderly person like this be found with a disease like this? It should have happened to the youth because they are the ones who 'run faster' (are more active sexually). -Community resident, 35-49 years, urban, IDI

Risk perceptions in relation to modes of HIV transmission
Men and women felt that only youth were affected by HIV, and midlife-older adults were at low risk (Q7). Sexual risk was discussed purely in the context of current behaviour, without acknowledging that a recent HIV diagnosis could reflect infection acquired years earlier.
When an old person is looking at a young person, he thinks that a young person has [HIV] in his body. But when a young person is looking at an old person he is 100% sure that this old person does not have any [HIV] in his body. -Community resident, 35-49 years, urban, FGD Infidelity by oneself or one's partner was an acknowledged risk for acquiring HIV within both urban and rural communities. Few participants however were willing to acknowledge this as a risk within their own relationship. Men spoke more frequently and openly about infidelity, including hinting of doubts about their partner, but still rated themselves as being at low risk. The exception was when a partner was known to be HIVpositive or was known to have had an affair with someone known to be HIV-positive (Q10).
[Interviewer: You had any perception of risk of HIV then, before found positive with self-testing?] Yes, because my husband had a relationship with a woman who was HIV positive and she was on ARVs … My husband doesn't stop his immoral behaviour. -Community resident, 35-49 years, rural, IDI While sexual transmission of HIV was acknowledged, midlife-older adults strongly emphasised non-sexual modes of transmission as a reason for older adults to worry about HIV and to consider HIV-testing. This reflected stated age norms of sexual inactivity, marriage, fidelity and respectability assigned to those considered older. Women in both rural and urban areas expressed concern about acquiring HIV through caring for the sick and bathing the dead. Routinely sharing items was another concern cited by both men and women, and including beard shavers, razors, soap and needles used for removing thorns. For CBDs, emphasising non-sexual routes of transmission provided a socially acceptable way of promoting HIVST amongst midlife-older adults, avoiding detailed discussions about sex which made participants less comfortable (Q11).
We explain to them that one can contract the virus through different ways. It might be that you helped a certain person, or maybe you used something sharp, and from nowhere you can easily contract the virus. Because of that, they say 'I think that you are explaining well' and you will find that they get tested. -CBD, < 35 years, urban, FGD

Consequences of HIV testing and diagnosis in later life
Midlife-older adults considered themselves to be more subject to HIV stigma and at greater risk of losing social standing than younger people if diagnosed with HIV, or even if seen to be testing. They anticipated being considered 'childish', mocked and laughed at if diagnosed HIV-positive, and that their diagnosis would be interpreted as a 'lack of wisdom' and sexual impropriety (Q14-16).
We look at those old people who contract HIV as if they lack wisdom. -Community resident, 35-49 years, urban, FGD The extent to which these concerns were justifiable, and from what age, however, was unclear: for instance, neither of the two married women in their 30s who disclosed that they were diagnosed with HIV through selftesting experienced any negative reactions (Q24-25).
Awareness of HIV was considered psychologically stressful, with some older adults considering themselves to be "already finished" with little to gain from learning their HIV status. ART was considered beneficial to health by all, although sometimes difficult to access. And there was little evidence of awareness of treatmentas-prevention, with newly diagnosed participants instead stating intent to use condoms or practice sexual abstinence with their spouse (Q17; Q19).
Some older people say 'I have already grown upwhat is remaining here is just dying. Why should I go to test? Even if they will mend [treat] me, what will that do for me?' -Community resident, 50+ years, urban, FGD Since that incident happened [both diagnosed with HIV], the community health worker came and gave us condoms. That's what we are using now. We are using condoms, apart from that we usually having sex once per week or two weeks. -Community resident, 35-49 years, rural, IDI In this context, testing for HIV was considered stigmatizing for older adults as there was widespread belief that wanting to test would be interpreted as evidence of recent infidelity or sexual risk-taking and that testing HIVpositive would only be harmful. The need for complete confidentiality was stressed for the act of testing, as well as the results, with caution expressed even for homebased or community programmes visible to family and neighbours even though participants recognised that those testing HIV-positive would inevitably lose all confidentiality as soon as they were seen to be attending their local ART clinic.

Experiences and concerns relating to self-testing
Community-based HIVST, with support and guidance from a CBD and the option to give a kit to a partner, was considered to have many advantages for midlifeolder adults, addressing their concerns by providing confidentiality and stigma, as well as convenience. Older participants desired more support compared to younger participants while self-testing, which was confirmed by CBDs (Q26-28).
Old people prefer different things. Those who have reached 45 to 70 years are the ones who test in our presence so that we should help them in reading the results, and so you can explain the instructions to them properly. -CBD, 35-49 years, urban, FGD Being able to give an HIVST kit to a partner or self-test with a spouse, having decided and received information and counselling together, was considered advantageous by midlife-older adults (Q29-32).
I found myself to be HIV positive together with my husband.
[Before] we had plans to go for testing, so we took self-testing together as an advantage to us, [and] we accepted the results … There is benefit because it [self-testing] will bring trust and love to each other. -Community resident, 35-49 years, rural, IDI Neither gender, however, liked the idea of having a selftest kit imposed on them by their partner via "secondary distribution", reflecting themes of HIV testing undermining social position, as well as questioning one's elderhood by doubting their fidelity.
Concerns about risks posed to the community by HIVST were negligible for all age-groups, with anticipated benefits considered to outweigh harms. No social harm was reported by participants who all previously self-tested, including two women who disclosed that they were diagnosed with HIV through self-testing.

Future service delivery preferences for self-testing
Many urban participants considered younger CBDs inappropriate for older community members, being unable to discuss personal issues and unlikely to be persuasive (Q33; Q35). However, older participants in rural settings prioritized trustworthiness over age or sex of distributors (Q34).
[Interviewer: Who should distribute self-test kits in terms of age and sex?] Anyone, as long as the person is trustworthy. -Community resident, 35-49 years, urban, IDI This reinforced views that chronological age alone is not important, but that respectable behaviours and experience can also define what is considered "older".
Acceptable alternatives to door-to-door distribution varied by age and gender, with women in their 30s suggesting outreach linked to antenatal and family-planning clinics, older women suggesting health facilities, and older men preferring fixed community collection points, workplaces or bus depots. These preferences aligned closely with perceptions of what was deemed age-gender appropriate.
Views on linkage post-HIVST did not appear to vary by age or gender, but some participants strongly preferred face-to-face post-test support. Following HIVST, having accompaniment from a relative or health worker or a referral slip (as in the study) was considered useful. Few other tools and approaches to support linkage were suggested.

Discussion
Our study has applied the life course approach [20] to understand how midlife-older age (≥30 years) is defined and culturally expressed by Malawians not only as chronological age, but through social and gender norms and attributes which focus on what is considered respectable behaviour. These norms, coupled with limited awareness of the changes in the HIV epidemic, including high HIV prevalence in their age-group and treatment as prevention, result in low HIV-risk perception among older men and women in Malawi. Such findings support previous studies that have highlighted barriers and challenges with reaching older populations with HIV testing, prevention and treatment services in southern Africa [12,22,29].
Age norms defining midlife-older adulthood drove views that conventional HIV testing later in life is disreputable and unacceptable, implying infidelity, sexual risk-taking and a lack of wisdom that, if discovered, would threaten social position in family and community life. Socioemotional selectivity [40] was observed among older adults who highlighted concerns about stigmatising reactions, and that an HIV-positive diagnosis was too stressful and unhelpful. Awareness of the potential for long-standing undiagnosed HIV to be carried forward from past relationships was minimal, as was understanding of treatment-as-prevention. These norms led to HIV testing being perceived as a threat to status by older-adults, and likely drives the low levels of recent HIV testing in midlife-older adults [10,15], as well as lower uptake of early introduction of HIVST in Malawi [17], compared to younger adults.
There is urgent need to communicate the changing epidemiology of HIV, and provide supportive HIV testing and prevention efforts, to midlife-older African adults, given their substantial underappreciation of personal risks, especially among + 35 year old men where HIV prevalence and new infections now exceeds younger age groups [1,3]. Communicating the importance of condom use later in life, as well as how being on ART and maintaining viral suppression prevents transmission to sexual partners should be prioritized.
Older participants expressed marked preference for access to HIVST over other testing modalities, similar to younger age groups [17,18,41]. HIVST delivered at home, with an additional kit for a partner was generally acceptable among midlife-older adults, addressing concerns about being seen testing; though support in the testing process was needed.
Age and gender norms of what is 'respectable' for midlife and older adults emerged as key issues affecting HIV risk perceptions, knowledge and acceptability of HIV testing services. As previously reported [9,16,42,43], the role and responsibility of women increased with age, including expectations of fidelity, being a caregiver and keeping the family healthy and HIV-negative. These preconceptions led many older men and women in established relationships to perceive their HIV risk to be low. Middle-aged and older adults strongly held the view that only young people could be affected because 'mature adults' were less sexually active, less susceptible to infidelity and had fewer partners.
This belief starkly contrasts with current epidemiological data [3,31], but reflects cultural expectations of midlife and older Malawians, as well as gendered views of HIV, and early experiences and messages dating back to when HIV was first recognized in African societies [44] at which time HIV services and messages focused on reaching women and young people because of elevated risk. Such focus may have inadvertently reinforced views of low risk perception among older age groups, especially men, and undermined messages on the importance of testing and HIV prevention more broadly. These views, as well as limited knowledge of the full benefits of ART, emerged as key drivers of the concerns about potential stigma and social consequences of accessing HIV testing among older adults. The evidence on how to deliver effective HIV interventions and messages for older adults is limited [23,45]. Further implementation research is needed to identify and scale-up messages and methods that effectively address sexual risk behaviour and HIV-related risks amongst midlife-older adults.
Our findings support HIVST as a useful tool for reaching midlife-older adults. The very same gender and social norms that drove low-risk perception and poor acceptability of HIV testing among mature adults, also drive preferences for self-testing. The feeling that testing at health facilities was too stigmatizing, costly and time consuming for older adults mirrored their stated preferences for self-testing because of its discretion and ease. While there was some variation, older adults preferred door-to-door HIVST, with an extra kit for a partner. Distribution by older providers, or those considered respectable, was often preferred. Acceptable HIVST kit access points varied by age and gender, defaulting to social positions and settings deemed gender appropriate (i.e. discreet collection points or workplace for older men and clinics for older women). Given the COVID-19 context, offering HIVST for midlife-older adults may also be increasingly advantageous to decongest health facilities and maintain essential services for those with potential risk factors that might lead to more serious disease in older populations.
Partner self-testing was viewed positively by all participants. Having an established social role and relationship was linked to this view, as well as having few concerns about harm in midlife-older adults. However, while individuals were inclined to give a self-test to a spouse, none stated they wanted to receive a kit from their partner. These findings further reinforce earlier reports showing both preference for giving a partner a kit [42], while receiving a kit is undesirable [46]. Considering continued reports of acceptability and high uptake of partner selftesting among men and women [47,48], the lack of a stated preferences for receiving a kit from a partner is likely associated with socioemotional selectivity [27,40] and avoiding undesirable relationship concerns, such as infidelity.
Future programmes should consider these preferences, and social and cultural norms of midlife-older adults, when planning HIVST, as well as broader HIV testing, implementation. Greater efforts however are still needed to reorient how age norms affect HIV risk and testing. Strategies including use of older providers and community workers who can provide support and instructions on how to self-test should be considered, as well as ambassadors who can align HIV prevention and testing behaviours with increasing respectability, health and responsibility among those who are older, and integrating HIV testing for older age groups as part of testing for non-communicable diseases [9,16]. Information and counselling messages on the full benefits of treatment, particularly treatment as prevention are critical and may reduce some of the fear and stress of an HIV-positive diagnosis which inhibits testing later in life. Additional outreach strategies using traditional media, e.g. newspaper, radios, and village networks may be important as opposed to social media and other technologies being utilised to reach young people.
Our study has several limitations. First, although our study aimed to include a range of adults aged over 30 years, Malawi has a young population, with only 4 and 20%, respectively, aged over 50 and 35 years, as is reflected in our median age of participants of 41 years [49]. This does mean that our findings primarily concern social norms amongst middle-aged rather than olderaged Malawians and may not generalise to countries with markedly different population age-structures.
Second, our findings under-represent the views of those who declined HIVST, many of whom may have been older. Third, we conducted qualitative studies within the context of broader cluster-randomised trials that primarily delivered door-to-door HIVST. Therefore, preferences on HIVST may have included more general preferences for door-to-door and community-based HIV testing. Fourth, because more women were recruited into the study, and that there was not a male only FGD, the views of men may be under-represented. Lastly, it is possible that a young male data collection team (age < 35 years) may have influenced the quality of data collected especially from women and older participants. However, the experience of the team and support from senior qualitative researchers (ND) helped overcome this challenge.

Conclusions
Age and gender norms are important drivers of HIV risk perception and HIV testing uptake during midlife-older adulthood. Using the life-course approach we highlight how age and gender norms contribute to poor uptake of conventional HIV testing by middle-aged or older individuals, as well as preferences for future self-testing in Malawi. With changes in HIV epidemiology, the increasing ease of access to ART and new HIV testing options (such as HIVST), there is an urgent need to provide targeted messages and services more appropriate to this age-group in sub-Saharan Africa. These messages need to include information on HIV risk and the importance of condoms later in life, as well as education on the benefits of ART including that PLHIV on ART who maintain viral suppression will not transmit HIV to their partners.
Despite concerns that HIV testing in facilities would be viewed as disreputable and undermine the current and future social status of midlife-older adults, HIVST appeared to provide a safe and acceptable alternative for mature adults to test, without challenging social age or gender expectations. While door-to-door, with an extra kit for a partner and support for the self-testing, continues to be preferred by middle-aged and older adults, additional service delivery approaches were considered age-gender appropriate (i.e. clinics for women, discreet community collection points and workplaces for men). Future programmes should consider these preferences as they plan HIV testing services, including HIVST scaleup, among midlife and older adults.