Effects of violence on adherence to antiretroviral therapy among adolescents living with HIV in Lilongwe, Malawi

¹Amref International University, Nairobi, Kenya, ²United States Government, State Department, U.S. President´s Emergency Plan for AIDS Relief (PEPFAR), Lilongwe, Malawi, ³Kamuzu University of Health Sciences, Blantyre, Malawi, ⁴University of Malawi, Zomba, Malawi

^&Corresponding author
Virginia Maria Thonyiwa, Amref International University, Nairobi, Kenya

Introduction    

Significant strides have been made in Human Immunodeficiency Virus (HIV) epidemic control in the 21^st century. Global statistics on HIV/AIDS show approximately 1.3 million individuals were diagnosed with HIV in the year 2024; this is a 40% decline in new HIV infections since 2010 and a decline of 61% since the peak in 1995 [1]. Additionally, there has been a 70% reduction in AIDS-related deaths since the peak in 2004, these strides are in line with the Sustainable Development Goals (SDG) number 3.3, aiming at ending HIV and AIDS as a public health threat by the year 2030 [2]. The World Health Organization (WHO) test-and-treat strategy has been key to the reduction of HIV incidence and AIDS-related mortality as it removes barriers to eligibility for antiretroviral therapy (ART) initiation among people living with HIV. This means that all people living with HIV, including children and adolescents, are eligible for antiretroviral therapy, regardless of their immunological status and clinical staging [3,4].

Notwithstanding the worldwide decline in new HIV infections and AIDS-related fatalities, the UNAIDS 2025 report reveals that sub-Saharan Africa continues to be the focal point of the HIV epidemic, accounting for approximately 67% of the population of individuals living with HIV/AIDS globally. Adolescents represent a big number of people living with HIV; approximately 1.7 million adolescents aged 10-19 years are living with HIV worldwide [1]. Sub-Saharan Africa hosts about 86% of all ALWHIV globally. The extraordinary scale-up of access to antiretroviral therapy (ART) for pediatrics around the world has contributed to more HIV-positive infants surviving into adolescence following perinatal infection over the past decades [5]. However, the success of an ART program is highly dependent on adherence to antiretroviral therapy.

Optimal adherence to antiretroviral therapy (ART) is central to achieving viral suppression and positive health outcomes in people living with HIV. People with suppressed viral loads can also reduce the risk of HIV transmission to uninfected individuals. Thus, ART adherence has become both an HIV treatment and an HIV prevention strategy [6]. Considering that the benefits of ART are tied to optimal adherence, and that the gains can be reversed if adherence is not maintained, leading to the risk for development of opportunistic infection and death, the World Health Organization has recommended a strict lifelong treatment adherence of 95% and above to fully benefit from ART and avoid poor health outcomes [6,7].

Despite efforts to improve and maintain adherence, suboptimal adherence to ART has attracted global concern, with adolescents exhibiting lower levels, only 62.3% adherence against the WHO-recommended target of 95% and above [7-10]. Data from Eswatini, Uganda, and Malawi indicate that adolescence is the only age group with the worst outcomes at all HIV stages, and rising HIV-associated mortality, largely driven by poor ART adherence [11].

Studies in sub-Saharan Africa have found that exposure to violence increases the risk of ART non-adherence among ALWHIV [12-14]. Specifically, emotional violence, such as verbal abuse, threats, and humiliation, reduces self-esteem, increases depressive symptoms, and decreases motivation to take ART consistently. Physical violence, including hitting, beating, or other bodily harm, can induce fear, trauma, and disruption of daily routines, negatively impacting adherence. Sexual violence, including harassment or assault, is associated with shame, psychological distress, and avoidance behaviors, which may result in missed doses or disengagement from care [15-17].

Globally, violence has been increasing, with sub-Saharan Africa being among the most affected regions [18]. In Malawi, ALWHIV already face challenges such as poverty, stigma, and social marginalization, and experiences of violence may further compromise ART adherence [19,20].

Although few studies have examined the association between violence and ART adherence, most have focused on women and girls, excluding boys who may also experience violence, and most studies were conducted outside Malawi [21-23]. This creates a significant knowledge gap in understanding how different forms of violence, physical, sexual, and emotional, affect ART adherence among both male and female ALWHIV in Malawi. This study assessed the effect of these three forms of violence on ART adherence among ALWHIV attending ART teen clubs in Lilongwe, Malawi.

Methods     

Study design: a cross-sectional mixed-methods study was conducted to determine the effect of violence on ART adherence among adolescents living with HIV (ALWHIV) in Lilongwe, Malawi. A cross-sectional design was appropriate for this study because it allows for the simultaneous assessment of exposure (violence) and outcome (ART adherence) within a defined population at a specific point in time.

Study setting and population: the study was conducted in Lilongwe, the capital city of Malawi. Although Lilongwe has more than 30 ART clinics offering services to adolescents living with HIV, the study purposively selected three sites (Baylor, Mtenthera, and Area 25) because they host some of the largest and longest-running teen clubs in the district. These sites serve a relatively high number of adolescents compared to smaller facilities, thereby providing a diverse and representative sample of ALWHIV in Lilongwe. While not all ART clinics were included, the cohorts from these facilities provided sufficient power for meaningful analysis and reasonable generalization of the findings to the broader Lilongwe context. The participants in each selected site were chosen through simple random sampling to avoid bias in selection. The study included six ART staff members (mentors, a nurse, and a clinician), and these cadres are primarily responsible for adolescent care within teen clubs. Additionally, two focus group discussions (FGDs) were conducted with 22 adolescents not included in the quantitative survey (12 and 10 participants per FGD). In these facilities, ART is dispensed and monitored at the HIV clinics (teen clubs), and drug reactions are managed within the same clinical setting by the clinician or nurse.

Variables

Outcome variable: ART adherence, categorized as: optimal adherence: ≥95% of prescribed doses taken in the previous month; suboptimal adherence: <95% of prescribed doses taken.

Clients are generally considered adherent to their ART medication if their pill adherence percentage, defined as the number of pills taken in a given time period (“X”) divided by the number of pills prescribed by the physician in that same time period, is at least 95% [24,25]. The adherence rate was calculated as:

The number of pills taken was the subtraction of the number in the bottle from the number prescribed at the last visit. One limitation to calculating adherence using this pill count method is that it assumes that the number of pills absent from the bottle was actually taken by the client. In addition, this method may not be representative of long-term adherence patterns because patients may exhibit white-coat adherence, or improved medication-taking behavior in the 5 days before and 5 days after a health care encounter [24].

For clients who didn´t bring their ART bottles, they were asked to recall how many pills they missed over the past month. In order to mitigate the effect of recall bias or social desirability of respondents on self-report, adherence reports were cross-checked against medical records, including missed clinic visits and viral load results over the past year. Viral load suppression was used as a proxy validation measure, consistent with WHO recommendations [19,26].

Independent variables: sociodemographic characteristics and experience of physical, emotional, and sexual violence.

Data management

Data collection tool: quantitative data were collected using a semi-structured questionnaire adapted from standardized instruments validated in adolescent HIV research. Emotional, physical, and sexual violence assessment used questions adapted from Violence against Children Surveys [17,21,27]. The questionnaire included information on ART adherence and experiences of violence. Medical records were reviewed for viral load and clinic attendance over the previous year. Qualitative data were collected using FGD guides and key informant interview guides. FGDs provided collective insights from adolescents not included in the survey, while key informants were purposively selected based on their roles at the teen clubs and willingness to participate.

Validity and reliability: tools were translated into the local language, Chichewa, and back-translated to ensure linguistic validity. The questionnaire was pretested among a small sample of adolescents outside the study sites to ensure clarity and appropriateness. Cronbach´s alpha coefficients were computed to assess internal consistency. Training of data collectors emphasized consistent administration of questions.

Sample size determination: a minimum sample size of 190 adolescents was computed using the formula for prevalence studies:

Where: z=1.96 (95% confidence level); p=0.86 (expected adherence prevalence from previous studies [8-10]); d=0.05 (margin of error). To account for potential non-response, the sample size was increased by 10%.

Data analysis: quantitative data were exported from KoboToolbox to IBM SPSS v22 for cleaning and analysis. Descriptive statistics summarized demographic characteristics. Associations between adherence and forms of violence were examined using chi-square tests and multivariate logistic regression, controlling for potential confounders (age, sex, education, and orphan status). Significance was set at p ≤ 0.05. Qualitative data from FGDs and key informants were analyzed thematically using NVivo, triangulating findings with quantitative results to enrich the understanding of violence and ART adherence. Clinic attendance, availability of adolescent-friendly services, and support from teen club staff were considered as contextual factors potentially influencing adherence, explored through both quantitative data (missed clinic visits) and qualitative interviews.

Ethical considerations: the study got ethical approval from the Malawi National Health Sciences Research Committee (NHSRC), Protocol approval number 25/03/4599. Administrative approval was also obtained from the Lilongwe District Health Office and the hospital authorities at each study site. Additionally, parental or guardian consent was obtained from all parents or caregivers of adolescents below the age of 18, as they were recognized as minors. Adolescents aged 18 and above provided their own consents. While all minors below 18 provided an assent in addition to the parental/guardian consent.

Confidentiality was maintained at all stages of the study, as it was a critical aspect of study ethics. Real-time submission of each questionnaire on the KoboToolbox prevented data collectors from having access to the completed questionnaire after submission. Only the investigator and statistician had access to the soft copy of completed questionnaires on the Kobo toolkit. The Kobo platform auto-generated a unique ID number, and respondents' names or codes to trace them were not used in the analysis or report writing. All collected information was kept confidential in a password-protected electronic file, with access limited to the study investigator and statistician only.

Participation in the study was voluntary, and all respondents were informed about the study's goal and nature before their involvement. Respondents' right to decline participation was respected, and they were not coerced or persuaded to take part in the study.

Results     

Socio-demographic characteristics: a total of 190 ALWHIV aged 10-19 were enrolled. The mean age was 17 years ±2.316 [95% CI]. Above half of the respondents, 56.3% (n = 107) were females and 43.7% (n = 83) were males. The proportion of orphans was slightly above half the respondents (53.7%, n=102), orphanhood meant they had lost one or both parents.

Adherence to antiretroviral therapy: after computing individual adherence rate, the average adherence for the study was 93.7%. Of the respondents, 29% had missed clinic visits. Additionally, of all the study respondents, only 110 (57%) had their viral load results in the past year. Almost all respondents from the Area 25 health center did not have viral load results. Overall, viral suppression was at 78%.

Prevalence of violence among adolescents living with HIV: adolescents were asked if they had experienced different forms of violence, namely emotional, physical, and sexual. A response of ’yes’ to any experience of the three types of violence was categorized as ’ever experienced violence’, and a response of ’no’ to all was translated as ’no experience of violence’. The study found that 86.3% of adolescents (n=164) living with HIV had experienced some form of violence in their lives, while 13.7% reported no exposure to violence. Different forms of violence were analyzed separately, and emotional violence was more prevalent at 76.8%, followed by sexual violence at 53.7%, and physical violence was at 30.5% (Figure 1).

Bivariate analysis: using the chi-square test, we explored the relationships between social demographic variables and ART adherence. All demographic characteristics were not statistically significant. For the age category, the proportion of optimal adherence was higher in younger adolescents (56.2%) than in older adolescents, 53.2%. However, the association between age group and adherence was not statistically significant (p=0.750). Sex was also not significant (p=0.648), education, orphanhood, and disclosure were all not statistically significant (p≥0.05) (Table 1). On the other hand, the chi-square test results indicated a strong association between ART adherence and both clinic visit attendance (p=0.01) and viral suppression (p=0.001). All forms of violence exposures showed a statistically significant association with ART adherence (Table 2), the relationship between exposure to sexual violence and ART adherence (p=0.011), emotional violence and ART adherence (p=0.011), and finally, physical violence and ART adherence (p=0.001).

Multivariate logistic regression analysis: a multivariate logistic regression was conducted to examine the association between various forms of violence and ART adherence among ALWHIV. The model controlled for education level, sex, and orphanhood as socio-demographic confounders. To avoid potential multicollinearity, each form of violence was entered into the model separately, rather than simultaneously, as adolescents who experience one form of violence are often exposed to other forms of violence, leading to high correlation between variables. Emotional violence was statistically significant, associated with lower odds of adherence. Adolescents with exposure to emotional violence were about 58% less likely to adhere to ART, compared to those who had not experienced emotional violence (p=0.040, OR=0.421, 95% CI: 0.185-0.961). Physical violence was also statistically significantly associated with lower odds of adherence (p=0.021, OR=0.432, 95% CI: 0.212-0.879), indicating that adolescents who were exposed to physical violence were 57% less likely to adhere to treatment. Sexual violence was not statistically significantly associated with adherence in the model (p=0.128, OR=0.580, 95% CI: 0.287-1.170).

Education level was not significantly associated with adherence (p=0.557, OR=1.183, 95% CI: 0.674-2.077), suggesting that educational attainment did not independently influence adherence in this model. Orphanhood status approached significance (p=0.086, OR=1.281, 95% CI: 0.966-1.698), implying that having parents alive may be modestly protective, though the association was not statistically significant. Respondent sex (p=0.433, OR=1.302) and age (p=0.657, OR=1.038) were not significant predictors of ART adherence (Table 3).

The qualitative findings were thematically grouped to provide clarity. First, limited access to viral load testing emerged as a major challenge, with participants noting that facilities had stopped conducting viral load tests following the suspension of U.S. government foreign aid in January 2025. As one key informant, a nurse explained, “the facility stopped conducting viral load testing due to a stop work order by the US government”. Secondly, reluctance to report violence was evident, particularly when perpetrators were caregivers or guardians. Adolescents explained that even when they knew where to report, they feared losing shelter or support, with one adolescent boy, an FGD discussant, noting, “even if you know where to report to, you can´t report violence perpetrated by your guardian if you don´t have anywhere else to go”. Third, the consequences of sexual violence were highlighted, with several adolescents linking sexual abuse to sexually transmitted infections (STIs) and a missed opportunity for health workers to assess for violence among adolescents reporting at the STI clinic. An adolescent girl another FGD discussant, shared in vernacular Chichewa that “mtsikana umati ukapita ku STI ndi matenda opatsirana pogonana ma nurse amangopereka mankwala achindoko koma osafunsa kuti wachitenga kuti nde pena munthu umaopa kuti unene zoti unagwililidwa nawenso umangokhala chete” (“when an adolescent girl goes with an STI to the clinic, nurses only treat the STI and never probe for sexual violence; sometimes one is afraid to report, so we just keep quite”).

Finally, violence and ART adherence emerged as a strong theme. Adolescents described how physical violence disrupted adherence, such as running away from beatings, being locked out of the house, or prioritizing personal safety, which often resulted in missed doses. As one 14-year-old boy narrated, “sometimes I am made to sleep outside and can´t take medication”. Another adolescent, an FGD discussant added humorously but seriously, “with physical violence one tends to run away from the beatings, and how can you tell a perpetrator ‘Taimani nditenge kaye botolo la mankhwala?´ (wait, let me get my bottle of ARVs) before you run away?” These narratives illustrate how experiences of violence not only compromise adolescents´ well-being but also undermine their ability to consistently adhere to antiretroviral therapy.

Discussion     

The results from this study show that most adolescents face challenges in attaining optimal adherence to antiretroviral therapy. Overall adherence rate was suboptimal at 93.7%, which means adolescents in Lilongwe, Malawi, are lagging behind the WHO-recommended adherence threshold of ≥95% required to fully benefit from ART and avoid poor health outcomes [19]. A study in South Africa among ALWHIV similarly found suboptimal adherence of 81% [17], while a study in Ethiopia reported an adherence rate of 71% [20]. Results were higher than those of another study in South Africa, where only 62.9% of adolescents demonstrated consistent adherence [21].

The dangers of suboptimal adherence include poor health outcomes, such as failure to attain viral suppression, thereby increasing the risk of new infections, as adolescents may transmit the virus to others. This may also reverse the gains of ART, lead to drug resistance, and increase morbidity and mortality among ALWHIV [17,22]. In this study, only 78% of adolescents achieved viral suppression, aligning with the observed rates of suboptimal adherence. This falls short of the UNAIDS global HIV 95-95-95 targets, particularly the third “95”, which requires that 95% of people on ART should be virally suppressed [23]. A systematic review of adherence data for adolescents aged 10-19 from 2010 to 2024 revealed similarly poor outcomes, with adherence as low as 65% and viral suppression of only 55% in sub-Saharan Africa [9]. In Kenya, a study reported that only 40% of ALWHIV attained viral suppression [28]. These findings underscore persistent challenges in adolescent HIV care in sub-Saharan Africa.

This study also found a high prevalence of violence among adolescents living with HIV in Lilongwe. All forms of violence were significant barriers to ART adherence, as chi-square analysis demonstrated statistically significant associations with suboptimal adherence. A study in South Africa also confirmed that violence among ALWHIV significantly reduced ART adherence [21]. Specifically, our findings indicate that adolescents exposed to emotional violence were about 58% less likely to adhere to ART, while those exposed to physical violence were 57% less likely to maintain adherence. Adolescents who reported no exposure to violence were over three times more likely to achieve optimal adherence. These results illustrate the critical role of psychosocial factors in shaping treatment outcomes and confirm existing evidence that violence is strongly associated with reduced ART adherence [21].

Interpreted through the lens of the Metatheory of Critical Realism, these findings reveal layered realities shaping adherence. The actual domain is reflected in observed rates of ART adherence (optimal or suboptimal), the empirical domain in observable indicators such as missed appointments, missed doses, and high viral loads, while the real domain highlights the underlying causal mechanisms in this case, exposure to violence that remains largely hidden but directly influences adherence outcomes [24]. Violence often occurs without reporting, particularly when adolescents live with stepparents or non-parental guardians, making it a concealed yet powerful determinant of ART outcomes.

Implications: the implications of these findings are multifaceted. For health systems, there is a pressing need to integrate psychosocial and violence screening into routine HIV care for adolescents to identify and address barriers early. STI clinics offer a great opportunity for tracing sexual violence. Policy makers should integrate violence screening in STI management strategies for law enforcement, strengthening legal protections for adolescents, particularly orphans and those in vulnerable households, is essential to mitigate exposure to violence. For adolescent support programs, interventions must extend beyond clinical adherence counseling to incorporate mental health care, peer support, and safe spaces where adolescents can disclose experiences of violence without fear of stigma or reprisal.

Strengths and limitations of the study: the strengths of this study include its focus on a vulnerable population, adolescents living with HIV within a high-burden setting, and its use of robust quantitative methods to examine the relationship between violence and ART adherence. However, limitations must be acknowledged. The cross-sectional design restricts causal inference, and self-reported adherence may be subject to recall or social desirability bias. Furthermore, the study was conducted in a single urban setting in Lilongwe, which may limit generalizability to rural contexts or other regions.

Recommendations: based on the findings, several recommendations emerge. Healthcare providers should integrate routine violence screening into HIV care and provide trauma-informed counseling to adolescents. Policy makers should strengthen adolescent protection laws and ensure their enforcement, especially in guardianship contexts. Programs supporting adolescents should incorporate psychosocial interventions that address violence, stigma, and mental health alongside adherence support. Policy makers should integrate routine violence screening in STI management, especially targeting adolescents. Finally, future research should employ longitudinal designs to further elucidate causal pathways between violence and ART adherence among adolescents.

Conclusion     

This study assessed the effects of violence on ART adherence among adolescents living with HIV in Lilongwe, Malawi. Adolescents in Lilongwe have suboptimal ART adherence and high rates of unsuppressed viral loads, with violence highly prevalent among them. Violence occurs in both male and female adolescents. There is a strong association between experiences of emotional and physical violence and suboptimal ART adherence. STI clinics provide an opportunity for screening and identifying sexual violence among ALWHIV. Low ART adherence and high rates of unsuppressed viral loads pose an increased risk for HIV transmission, drug resistance, morbidity, and mortality, a challenge to ending HIV/AIDS as a public health threat by the year 2030.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Virginia Maria Thonyiwa: conception and designing of the study, full research protocol development, data collection, analysis and leading the draft and final write up; Tom Marwa: substantial contributions to conception and design, critically reviewing of protocol, tools and the manuscript for important intellectual content and approval of manuscript for publishing; Victor Mwapasa: substantial contributions to conception and design, critical reviewing of protocol, tools and the manuscript for important intellectual content and approval of manuscript for publishing; Uchechi Roxo: reviewing and editing of manuscript; Samuel Salwaco Banda: data analysis, reviewing and editing manuscript. All the authors read and approved the final version of this manuscript.

Acknowledgments     

We would like to appreciate Dr Josephat Nyagero, Helen, and Abdul Malik Abubakar, Dr Ehiagwina Braimah, Alinafe Mbendera, Dr Phillip Omondi, Reuben Phiri, Shiloh 2024, Amanda and Hannah Kayoyo, Mr and Mrs Thonyiwa, Yvonne Thonyiwa, Precious Gunde, Katie Norwood, Sarah Mwale, Dr Priscilla and John Kauye, Jean Manda, Pastor Bright Fiito Olubari, Pastor Joseph Chikwenga, for the tremendous support.

Tables and figure     

Table 1: demographic characteristics versus adherence to antiretroviral therapy of study respondents recruited from select teen clubs in Lilongwe, Malawi, in April, 2025 (N=190)

Table 2: bivariate analyses of different forms of violence and adherence to antiretroviral therapy among adolescents living with HIV recruited from three health facilities in Lilongwe, Malawi (N=190)

Table 3: multivariate logistic regression showing negative associations between all three forms of violence and ART adherence among adolescents living with HIV in Lilongwe, Malawi (N=190)

Figure 1: prevalence of violence among adolescents aged 10-19 years living with HIV, recruited from select facilities in Lilongwe, Malawi

References