A total of 656 men living with HIV participated in this study. They were recruited from five specialised clinics belonging to two institutions that provide care for people living with HIV, located in three cities in Mexico (see Table 1). The sample consisted predominantly of adult men, with a small proportion of participants identifying with gender identities other than male. In terms of sexual orientation, most participants identified as homosexual, followed by smaller groups identifying as heterosexual or bisexual. Most of the sample reported being single, while a smaller proportion were married or in a civil partnership. With respect to occupational status, the majority of participants were engaged in non-professional activities, although professional employment, student status, homemaking, and unemployment were also represented. Educational attainment varied across the sample, ranging from basic education to postgraduate studies, with most participants reporting upper secondary or higher education. Regarding clinical characteristics, participants had been living with HIV for several years on average and had been receiving antiretroviral therapy for a substantial period. A considerable proportion reported a lifetime history of sexually transmitted infections, and a smaller group had received an AIDS diagnosis. Most participants reported having maintained an undetectable viral load in recent months. In relation to mental health care, a minority of participants reported current or past psychiatric treatment, and a slightly larger proportion indicated receiving psychotherapeutic care at the time of the study.

The version validated by Tafur-Valderrama et al. (2008) was used, which uses 20 Likert-type items to assess treatment compliance, history of non-adherence, interaction with the doctor, patient beliefs, and strategies to improve adherence. Overall, the authors found an acceptable level of reliability (α = .70), similar to the original version validated in Spain (α = .73).

The version validated in the Latin American population by Daza et al. (2002) was used, which comprises three dimensions: depression (7 items), anxiety (7 items), and stress (7 items). The authors reported acceptable reliability for each of the factors (α < .83). The items are in Likert format.

Five specialised clinics belonging to two institutions that provide care for people living with HIV were visited: CEC Condesa and CEC Iztapalapa in Mexico City, CAPASITS Neza in the State of Mexico, and CAPASITS Xalapa and CAPASITS Veracruz in Veracruz. At each of the institutions, users were invited to participate in the research. Those who signed the informed consent form were given a booklet with the instruments. This consent form included information on the anonymity and confidentiality of the information, explicitly stating that participation or non-participation in the research was independent of the care they received at the clinic. On average, it took them half an hour to respond.

Approval of the project by the Research Ethics Committee of the Institute of Health Sciences of the Veracruzana University, which is registered with the National Bioethics Commission (registration number: D/169/2022).

For the first research question, a multiple linear regression analysis was performed using a stepwise method to identify predictors of treatment adherence (dependent variable). In each analysis, the assumptions of normality, independence, and homoscedasticity of the residuals were verified; multicolinearity was also evaluated using tolerance indices and the variance inflation factor (VIF). For the second research question, binary logistic regression analyses were performed; five models were estimated, each with a different dependent variable: (a) lifetime presence of a sexually transmitted infection (STI), (b) previous diagnosis of AIDS, (c) remaining undetectable in the last six months, (d) psychiatric diagnosis, and (e) currently receiving psychotherapeutic treatment. The estimation method used was stepwise forward logistic regression (likelihood ratio), with a significance level set at p < .05. Beta coefficients (β), standard errors, Wald values, significance (p) and odds ratios (OR) were calculated with their 95% confidence intervals. Finally, for the third research question, a multiple linear regression analysis was performed using a stepwise method to identify predictors of the time elapsed since the start of antiretroviral treatment and the time of HIV diagnosis (dependent variables). In each analysis, the indicators already mentioned in the first objective were taken into account.

The results of this study show that anxiety and depression are negatively associated with adherence to antiretroviral therapy (ART), while condom use with casual partners predicts a slight increase in adherence. These findings are consistent with the international literature, which has consistently documented that depressive and anxiety symptoms affect therapeutic compliance and reduce the likelihood of achieving viral suppression (Haas et al., 2023). These effects can be explained by the fact that negative affectivity interferes with essential cognitive functions, such as memory and attention, which are necessary for maintaining adherence (A. Wu & Zhang, 2023). In addition, depression and anxiety often coexist with contextual risk factors, such as stigma and lack of social support, which exacerbate emotional vulnerability and compromise self-care (Guzmán-Mendoza et al., 2025).

Regarding condom use with casual partners, its positive relationship with adherence suggests a consistent pattern of self-care, where preventive behaviors reinforce each other. It is possible that in contexts of less intimacy or knowledge of the other person, the perception of risk activates behavioral control mechanisms that promote both sexual protection and therapeutic compliance (Bowleg et al., 2020). This interpretation coincides with the idea that consistent condom use is not an isolated practice, but part of a repertoire of behaviors regulated by risk awareness and commitment to health (Silva et al., 2024). Thus, adherence to ART and condom use in casual contexts seem to be part of the same motivational framework of self-care and responsibility.

Finally, stress was not a significant predictor of treatment adherence, possibly because its effect is more indirect and operates through other psychological variables such as anxiety and depression. When these factors are included simultaneously in the model, the specific contribution of stress may be reduced due to shared variance.

Analysis of clinical indicators revealed that stress is associated with a higher probability of having had a sexually transmitted infection (STI), while reduced condom use with stable partners and sex workers also increases this risk. These results support previous studies showing significant links between psychological distress and higher STI incidence, particularly when perceived stress is present (Singh & Singh, 2021). Recent evidence suggests a bidirectional relationship between stress and sexual risk, framed within a syndemic phenomenon where mental health problems and STIs reinforce each other (Bretz et al., 2023). Furthermore, the experience of an STI is often accompanied by feelings of guilt and shame (Scheinfeld, 2021), which could intensify emotional distress and create a cycle of vulnerability.

The finding that condom use with sex workers is associated with an AIDS diagnosis may reflect a post-diagnosis adaptive response, where increased risk perception promotes the adoption of protective measures in contexts considered to be more dangerous (Luz et al., 2023). In this sense, diagnosis could function as a turning point that reorients sexual behaviors towards safer strategies, at least in interactions characterized by lower trust or greater uncertainty.

Furthermore, the fact that viral undetectability is associated with greater adherence to ART and condom use with casual partners reinforces the notion of an integrated self-care profile. Both behaviors involve discipline, emotional regulation and a sense of responsibility towards one’s own health and that of others. This finding coincides with research that emphasizes the need for holistic approaches to HIV care, where clinical and psychological dimensions are addressed jointly (Brown & Adeagbo, 2022).

On the other hand, the association between higher stress and psychiatric treatment indicates that those with more severe emotional distress are more likely to receive specialized care, suggesting adequate detection of psychological needs within the health system, although probably after the development of clinically significant symptoms.

The time elapsed since diagnosis and since the start of ART showed different associations with emotional and behavioral variables. A longer time since diagnosis was associated with more frequent condom use with a stable partner and with sex workers , while depression was associated with a more recent diagnosis. Similarly, those who had been in treatment for longer reported more consistent condom use with stable partners and lower levels of depression. These findings suggest a process of progressive adaptation to living with HIV and treatment, in which preventive practices and emotional regulation tend to stabilize over time.

The higher levels of depression observed in the early stages of diagnosis coincide with the findings of Xia et al. (2023), who point out that the first few months after diagnosis represent a critical phase characterized by intense emotional distress, uncertainty and the need for psychological support. In contrast, people who have been living with HIV for longer seem to have developed more effective coping strategies and a greater internalization of self-care (Thurman et al., 2023). Although some authors suggest that condom use decreases over time due to preventative fatigue (Malekinejad et al., 2021), the results of this study point in the opposite direction, suggesting that experience and accumulated learning can strengthen protective behaviors in certain contexts.

This study has several theoretical and methodological limitations that should be considered when interpreting the results. First, from a theoretical perspective, the findings show that psychological, behavioral, and clinical factors are more complexly interrelated than anticipated, posing challenges for linear or unidirectional explanatory models. The associations observed between emotional variables, adherence and condom use suggest that self-care processes do not necessarily respond to fixed causal sequences, but rather to reciprocal dynamics that depend on context, time of diagnosis and living conditions. This highlights the need for more integrative conceptual frameworks that consider the simultaneous interaction of individual, clinical and social levels.

In methodological terms, the cross-sectional design limits the possibility of establishing causal relationships between the variables analyzed, so the results should be interpreted as associations rather than direct effects. Likewise, the use of self-report instruments may have introduced memory or social desirability biases, especially in questions related to sexual behavior and treatment adherence, where social norms and expectations could influence responses.

Another limitation relates to convenience sampling. While this strategy facilitated access to a population that is difficult to contact, such as men living with HIV, it may have generated selection bias. The participants come from three specialized urban clinics, so the results may reflect particular characteristics associated with the type of care, the availability of institutional resources, or the quality of medical and psychological support, which limits the generalization of the findings to other contexts or regions of the country.

Similarly, the absence of a complementary qualitative approach prevented the exploration of subjective meanings and personal experiences related to diagnosis, treatment, and sexual practices. The incorporation of interviews or focus groups would have allowed for a deeper understanding of the motivations, emotions, and barriers experienced by the participants, enriching the interpretation of the quantitative results. Furthermore, this study is subject to limitations regarding the contextual variables of sexual behaviour, as data were not collected concerning sexual positioning (insertive or receptive roles), the use of pre-exposure prophylaxis (PrEP) by partners, or their serological status. Given that these factors significantly influence risk perception and condom negotiation among MSM, it is essential for future studies to incorporate these variables into their design to provide a more comprehensive understanding of preventive practices.

Finally, relevant structural factors (e.g., perceived stigma, discrimination, or differential access to health services), which have a decisive influence on adherence and preventive behaviors, were not included. The absence of these variables limits our understanding of the role that social inequalities and contextual determinants play in the health of people living with HIV.

These limitations suggest the need for longitudinal, multi-method studies that integrate psychological, clinical, and social dimensions, allowing for clarification of self-care trajectories and adaptation processes over time.

The practical implications of these results indicate the necessity of establishing a clinical care pathway where mental health screening functions as a trigger for preventive actions. Given that depression and anxiety predict lower adherence, healthcare centres should implement early warning protocols in which high scores in these areas automatically activate behavioural reinforcement interventions. This approach is essential for operationalising the Undetectable = Untransmittable (U = U) principle, ensuring that psychological support serves as the foundation for maintaining viral suppression. Likewise, counselling must be differentiated by relationship type, prioritising condom negotiation within stable partnerships where trust often reduces protective measures. By integrating stress management as a tool to decrease STI prevalence, care programmes transition towards a combined prevention model that proactively manages both biological and psychological risks.

This study possesses several notable strengths, starting with the large and diverse sample of 656 men living with HIV, which provides significant statistical power and robustness to the findings. The multicentric nature of the research, involving five specialised clinics, enhances the external validity and representativeness of the results within the region. By integrating emotional, behavioural, and clinical variables into a single analysis, this research transcends traditional biomedical assessments to offer a holistic understanding of treatment success. This integrated approach constitutes a significant scientific advance by providing empirical evidence for a shared self-regulation framework between sexual health and therapeutic compliance.

The results of this study show a critical interaction between emotional symptoms, preventive behaviours, and treatment adherence, validating the need for comprehensive care models that unify mental health and sexual prevention within a single clinical protocol. The evidence underscores that the early detection of anxiety, depression, and stress should be considered a key indicator of quality of care, given its direct impact on therapeutic outcomes and the optimisation of healthcare resources.

Furthermore, the positive relationship between condom use and adherence suggests that both behaviours share a common self-regulation framework, supporting the development of integrated psychoeducational programmes that simultaneously address sexual health and therapeutic compliance. The association observed between stress, STIs, and psychiatric care highlights the importance of strengthening emotional well-being monitoring to prevent costly clinical complications. The variations found according to the time since diagnosis indicate that care must be adapted to the different stages of the HIV adaptation process. This study provides a significant scientific advance by integrating emotional, behavioural, and clinical dimensions into a single analysis, contributing to a more efficient, humane, and evidence-based model of care.