In Spain, it is estimated that at the end of 2022 approximately 90 000 people were HIV positive.1 In developed countries, 30% of people living with HIV (PLWH) are currently 50 years of age or older. There are already studies that estimate that by 2030 approximately 75% of global HIV patients will be 50 years old and around 40% will be 65 years or older.2

These findings stem from advancements in antiretroviral (ARV) medications, which have enhanced efficacy and safety. Consequently, there has been a decrease in mortality among PLWH, with their life expectancy now comparable to that of seronegative population.3,4 The aging of PLWH reveals new challenges in their healthcare, which is becoming more complex due to the increase in the prevalence of age-related comorbidities and even the appearance of multimorbidity patterns.5,6

This scenario implies an increase in the prevalence of polypharmacy, aimed at the management of these concurrent comorbidities.7 The negative effects that polypharmacy can have on patients, such as lack of adherence, interactions, or toxicity associated with drugs, are known.8

Traditionally, the approach to measuring polypharmacy has been quantitative, that is, based on the number of prescribed medications. However, in recent years, a more qualitative approach has been used, as the same number of drugs can differ in aspects such as administration instructions, preparation or administration methods, pharmaceutical formulation, or dosage regimen. George et al. introduced the ‘Medication Regimen Complexity Index’ (MRCI), which allows the consolidation and weighting of all these factors.9 In this sense, a threshold value of 11.25 has been established for this index as a benchmark for classifying a polypharmacy patient as having high complexity in the context of PLWH.10

Static measures, like cross-sectional studies examining the proportion of polymedicated patients or those with MRCI values exceeding 11.25, are commonly used to analyzed polypharmacy and pharmacotherapeutic complexity. However, these approaches do not capture the dynamic nature of these phenomena over time, influenced by various changing factors. Group-based trajectory modeling (GBTM) is a data-driven statistical method used to analyzed developmental trajectories, representing the evolution of an outcome with age or time.11 This methodology has been used to analyze drug adherence and polypharmacy in both the general population and HIV patients. Identifying individuals with high MRCI trajectories is crucial for developing tailored interventions and strategies.12,13 Multiple characteristics influencing patients with elevated pharmacotherapeutic complexity may be more relevant for detection than singular risk factors. Compared to conventional methods that rely on dichotomized cut-off points and specific time points to determine MRCI levels, GBTM can predict treatment outcomes with temporal trends from onset to patient care, offering more comprehensive insights into patient management.11 To our knowledge, only one study has utilized this methodology to analyzed polypharmacy and adherence.

The research objective is to utilize GBTM analysis to identify distinct trajectories of pharmacotherapeutic complexity and polypharmacy among PLWH. By examining the associations between identified trajectories and levels of pharmacotherapeutic complexity within the context of polypharmacy, the research aims to provide valuable insights for personalized healthcare approaches and targeted interventions to optimize medication management for PLWH.

Our analysis revealed 4 polypharmacy trajectories, with the ‘low increasing’ group representing the largest proportion of PLWH (50.8%). Considering patients with a high level of pharmacotherapeutic complexity (MRCI value ≥11.25), GBTM identified 2 trajectories, with 80.2% in the ‘low increasing’ polypharmacy trajectory group. The joint model did not reveal any apparent relationship between the polypharmacy trajectories and the high pharmacotherapeutic complexity.

In the current analysis, GBTM provided a unique perspective on the changing nature of polypharmacy behavior in PLWH, differentiating 4 categories. There are different previous studies that analyzed the polypharmacy trajectories in a specific sex. One of them analyzed both adherence and polypharmacy in women with HIV in the United States. As in our case, 4 polypharmacy trajectories were distinguished. As in our study, the largest proportion of these HIV-infected women were in the ‘low increasing’ polypharmacy trajectory group. There is a higher percentage of men with HIV infection in the ‘consistently high’ and ‘consistently low’ polypharmacy category groups. Furthermore, the study by Ware et al., who used GBTM to examine polypharmacy among HIV-infected men in the MACS cohort, was significant. Four groups were identified, with 48% of participants in a non-polypharmacy group.17 This study finds that PLWH and over 50 years of age are associated with polypharmacy trajectory groups, as in our study where the polypharmacy trajectory ‘consistently high’ presents a higher percentage of PLWH over 50 years of age.

A population-based study in Spain analyzed 23 000 individuals with HIV, revealing a higher prevalence of polypharmacy among women (45%) compared to men (30%).18 Contrarily, our study demonstrates a higher prevalence of polypharmacy in men compared to women across all polypharmacy trajectories, as shown in Table 1. This discrepancy may be due to higher rates of chronic diseases in women.19 Notably, our study finds cardiovascular disease-related comorbidities more prevalent among males, aligning with findings by Camps-Vilaró et al.20 The ‘moderately decreasing’ group under 50 years experiencing reduced drug load is often associated with psychiatric–neurological conditions, particularly anxiety and depression.21 This phenomenon in PLWH under 50 years of age can be explained by the development of neuropsychiatric disorders that may be related to the diagnosis of HIV infection and the associated stigma.

In general, the results revealed that the predictor of belonging to the consistently low polypharmacy path was patients under 50 years of age. All this is widely described at the bibliographic level, since HIV patients older than 50 years suffer a subclinical chronic inflammatory process that is associated with the development of age-related comorbidities and, therefore, the prescription of medications to control these chronic diseases.22,23 However, the protective nature of the male sex stands out in the ‘low increasing’ and ‘consistently low’ polypharmacy trajectories. This is related to the higher prevalence of polypharmacy shown by women, as described in different studies.24,25

In this analysis, GBTM provided a novel perspective on high pharmacotherapeutic complexity previously unexplored. Prior research has highlighted the significance of pharmacotherapeutic complexity in health-related outcomes, linking elevated complexity to reduced patient-perceived quality of life and adverse health outcomes, such as mortality, particularly among elderly patients.26,27 The findings demonstrated the statistical significance of belonging to the ‘slightly decreasing high’ trajectory of high pharmacotherapeutic complexity, especially for PLWH under 50 years with comorbidities. This trajectory is associated with polypharmacy trajectories like ‘consistently high’ and ‘moderately decreasing’, described in existing literature as being due to age-related comorbidities.22 Categorizing PLWH based on their high-pharmacotherapeutic complexity trajectories can offer new insights into patient identification and help develop strategies to enhance health outcomes.

The interplay between polypharmacy and high pharmacotherapeutic complexity may entail an increased risk of adverse effects, drug interactions, and heightened challenges in patient treatment management and adherence. Additionally, it may necessitate closer monitoring and enhanced multidisciplinary collaboration to optimize the safety and efficacy of treatment.

It is already known that polypharmacy has a positive correlation with pharmacotherapeutic complexity. However, contrary to our hypothesis, the primary analyses did not reveal any interrelationship between the polypharmacy groups and the high-medical regimen complexity index trajectory groups. A potential explanation for our findings could be the presence of an effect modifier between polypharmacy and high MRCI, such as the therapeutic group of the patients' concomitant medication to ARV. There is a study of elderly patients with chronic kidney disease that shows the influence of the pharmacotherapeutic group of the treatment on the pharmacotherapeutic complexity, finding that in this population phosphate binders were associated with more-complex medication regimens.28

This methodology offers new research avenues to explore interrelationships among various effects, facilitating the identification of modifying variants. This knowledge can inform targeted intervention strategies aimed at controlling these effects. Specifically, it enables the development of deprescribing strategies to reduce MRCI values, thereby promoting optimal health outcomes, including improved quality of life.

The findings of this analysis show the importance of evaluating PLWH in the ‘consistently high’ and ‘moderately decreasing’ polypharmacy trajectories due to the association of this effect with the prescription of potentially inappropriate drugs and negative effects associated with polypharmacy, such as interactions.29,30 Interventions such as deprescribing should be implemented, making it possible to address both polypharmacy and reduce the pharmacotherapeutic complexity of the patient.

Strengths of this analysis include the large sample size of PLWH who were followed for an extended period, allowing the model to identify the trajectories of both variables.

Despite these results, one of its several limitations must be highlighted. It is a single-center cohort study and hence the possibility that the trajectories cannot be generalized to other HIV populations. This could be explained by the deprescription and pharmacotherapy optimization programs presented by each center, which reduces both the prevalence of polymedicated patients and those with high pharmacotherapeutic complexity. Finally, despite the sample size included, there is the possibility of losing patients to follow-up over time due to changing centers or abandoning treatment, which is a potential limitation that should be considered.

In conclusion, dual GBTM did not identify an interrelationship between polypharmacy and high MRCI. It is crucial to conduct future studies in this line of research, considering potential effect modifiers such as the type of concomitant drug, due to the outcomes induced by both polypharmacy and the value of MRCI in PLWH.

This study utilizes GBTM to examine polypharmacy and medication regimen complexity among PLWH. It identifies distinct polypharmacy trajectories, including a notable ‘low increasing’ pattern, and identifies trajectories characterized by higher MRCI.

These findings underscore the complex relationship between polypharmacy and medication complexity in PLWH. Age and gender were identified as factors influencing polypharmacy trajectories, highlighting the need for tailored therapeutic approaches. Further research into the specific impacts of medications on treatment outcomes is crucial for optimizing HIV management strategies in clinical practice.

This research did not receive any specific grant funding from funding agencies in the public, commercial, or nonprofit sectors.

The study complied with all ethical requirements and was approved by the Ethical Committee of Clinical Research Ethics Committee of Seville-South (C.I. 1341-N-23). This study was conducted in accordance accordance with the guidelines of the Declaration of Helsinki for biomedical research.

The author for the correspondence, on behalf of the other signatories, guarantees the accuracy, transparency, and honesty of the data and information contained in the study; that no relevant information has been omitted; and that all discrepancies between authors have been adequately resolved and described.

Enrique Contreras Macías: Writing – review & editing, Writing – original draft, Visualization, Supervision, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. María de las Aguas Robustillo Cortés: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation. Ramón Morillo Verdugo: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Conceptualization.