Latent Markov Modeling of HIV Vulnerability Transitions Among Adolescent Girls and Young Women in Zimbabwe

Fungai Hamilton Mudzengerere; Oliver Bodhlyera; Henry Mwambi

doi:10.37905/jjom.v8i2.40071

Latent Markov Modeling of HIV Vulnerability Transitions Among Adolescent Girls and Young Women in Zimbabwe

Fungai Hamilton Mudzengerere, Oliver Bodhlyera, Henry Mwambi

Abstract

This study uses applied Latent Markov modeling (LMM) to assess longitudinal transitions in HIV vulnerability among adolescent girls and young women (AGYW) enrolled in the Determined, Resilient, Empowered, AIDS-free, Mentored, and Safe (DREAMS) program across nine districts in Zimbabwe, from October 2023 to September 2024. Using routinely collected program data for 4,341 AGYW aged 10--19 years, the analysis identified distinct latent classes of HIV vulnerability and tracked their transitions over a 12-month period. For AGYW aged 10--14 years, four latent classes, that is, very high, high, medium, and low risk were identified with vulnerability factors of in-school dropout risk, orphanhood, alcohol use, and exposure to violence. The probability of maintaining low-risk status was high (0.704), while the probability of transitioning from very high to high risk was comparatively modest (0.252), suggesting only limited movement out of the highest-risk category among AGYW. Among the 15-19-year-old AGYW, vulnerability to HIV was mainly driven by sexual behavior, condom use, pregnancy history, and transactional sex, with the highest probability of remaining in the same class observed in high-risk (0.819) and low-risk (0.714) states. Covariates such as area of residence and source of income significantly influenced transitions, where AGYW with family or informal income were less likely to shift to higher-risk states. Overall, the results demonstrate transitions consistent with reductions in HIV risk levels among AGYW during the period of DREAMS participation, alongside the program's social, educational, and economic empowerment components. In this study, Latent Markov modelling (LMM) was applied as a robust analytical framework for monitoring longitudinal changes in vulnerability and evaluating program impact in low-resource settings.

Keywords

Latent markov modeling; Longitudinal data; HIV; Vulnerability; Adolescent girls and young women

Full Text:

PDF

References

G. J. Melendez-Torres, E. Allen, R. Viner, and C. Bonell, “Effects of a wholeschool health intervention on clustered adolescent health risks: Latent transition analysis of data from the inclusive trial,” Prev. Sci., vol. 23, no. 1, pp. 1–9, 2022, doi: 10.1007/s11121-021-01237-4.

E. A. Kim, H. Chung, and S. Jeon, “Joint latent class analysis for longitudinal data: An application on adolescent emotional well-being,” Commun. Stat. Appl. Methods (CSAM), vol. 27, pp. 241–254, 2020, doi: 10.29220/CSAM.2020.27.2.241.

R. Turner, hmm.discnp: Hidden Markov Models with Discrete Non-Parametric Observation Distributions, 2016, R package version 0.2-4; url: https://CRAN.R-project.org/package=hmm.discnp.

C. Bonell, E. Beaumont, M. Dodd, D. Elbourne, L. Bevilacqua, A. Mathiot, and E. Allen, “Effects of school environments on student risk-behaviours: Evidence from a longitudinal study of secondary schools in England,” J. Epidemiol. Community Health, 2019.

S. W. Noor, T. A. Hart, C. N. Okafor et al., “Staying or moving: Results of a latent transition analysis examining intra-individual stability of recreational substance use among MSM in the multicenter AIDS cohort study from 2004 to 2016,” Drug Alcohol Depend., vol. 220, p. 108516, 2021, doi: 10.1016/j.drugalcdep.2021.108516.

E. O. A. Wambiya, A. J. Gourlay, S. Mulwa, F. Magut, N. Mthiyane, B. Orindi, N. Chimbindi, D. Kwaro, M. Shahmanesh, S. Floyd, I. Birdthistle, and A. Ziraba, “Impact of DREAMS interventions on experiences of violence among adolescent girls and young women: Findings from population-based cohort studies in Kenya and South Africa,” PLOS Glob. Public Health, vol. 3, no. 5, p. e0001818, 2023, doi: 10.1371/journal.pgph.0001818.

USAID, “DREAMS: Partnership to reduce HIV/AIDS in adolescent girls and young women,” url: https://www.usaid.gov/global-health/health-areas/hiv-and-aids/technical-areas/dreams; accessed: May 21, 2024.

F. Bartolucci, S. Pandolfi, and F. Pennoni, “LMest: An R package for latent Markov models for longitudinal categorical data,” J. Stat. Softw., vol. 81, no. 4, pp. 1–38, 2017, doi: 10.18637/jss.v081.i04.

S. Bonhomme, K. Jochmans, and J.-M. Robin, “Estimating multivariate latent structure models,” Ann. Statist., vol. 44, pp. 540–563, 2016.

C. Kam, A. J. Morin, J. P. Meyer, and L. Topolnytsky, “Are commitment profiles stable and predictable? A latent transition analysis,” J. Manage., vol. 42, no. 6, pp. 1462–1490, 2016.

K. Saengtrakul, S. Kanjanawasee, and N. Wiratchai, “Student factors affecting latent transition of mathematics achievement measuring from latent transition analysis with a mixture item response theory measurement model,” Procedia Soc. Behav. Sci., vol. 217, pp. 729–737, 2016.

J. Zyberaj, C. Bakac, and S. Seibel, “Latent transition analysis in organizational psychology: A simplified ‘how to’ guide by using an applied example,” Front. Psychol., vol. 13, p. 977378, 2022, doi: 10.3389/fpsyg.2022.977378.

K. G. Card, H. L. Armstrong, A. Carter, Z. Cui, L. Wang, J. Zhu, N. J. Lachowsky, D. M. Moore, R. S. Hogg, and E. A. Roth, “Assessing the longitudinal stability of latent classes of substance use among gay, bisexual, and other men who have sex with men,” Drug Alcohol Depend., vol. 188, pp. 348–355, 2018, doi: 10.1016/j.drugalcdep.2018.04.019.

I. Zych, J. Rodríguez-Ruiz, I. Marín-López, and V. J. Llorent, “Longitudinal stability and change in adolescent substance use: A latent transition analysis,” Children Youth Serv. Rev., vol. 112, 2020.

A. H. Weinberger, J. Platt, and R. D. Goodwin, “Is cannabis use associated with an increased risk of onset and persistence of alcohol use disorders? A three-year prospective study among adults in the United States,” Drug Alcohol Depend., vol. 161, pp. 363–367, 2016, doi: 10.1016/j.drugalcdep.2016.01.014.

P. M. M. Okuda, W. Swardfager, P. S. Lucio, G. B. Ploubidis, T. Liu, M. Pangelinan, and H. Cogo-Moreira, “The trajectory of balance skill development from childhood to adolescence was influenced by birthweight: A latent transition analysis in a British birth cohort,” J. Clin. Epidemiol., vol. 109, pp. 12–19, 2019.

Y. Ni, J. Y. Tein, M. Zhang, Y. Yang, and G. Wu, “Changes in depression among older adults in China: A latent transition analysis,” J. Affect. Disord., vol. 209, pp. 3–9, 2017.

B. A. Hultgren, R. Turrisi, M. J. Cleveland, K. A. Mallett, R. Reavy, M. E. Larimer, I. M. Geisner, and M. M. Hospital, “Transitions in drinking behaviors across the college years: A latent transition analysis,” Addict. Behav., vol. 92, pp. 108–114, 2019, doi: 10.1016/j.addbeh.2018.12.021.

F. Liao, E. Molin, H. Timmermans, and B. van Wee, “The impact of business models on electric vehicle adoption: A latent transition analysis approach,” Transp. Res. Part A Policy Pract., vol. 116, pp. 531–546, 2018.

R. Di Mari, F. Dotto, A. Farcomeni, and A. Punzo, “Assessing measurement invariance for longitudinal data through latent Markov models,” Struct. Equ. Modeling, vol. 29, no. 3, pp. 381–393, 2021, doi: 10.1080/10705511.2021.1993857.

H. L. Nguena Nguefack, M. G. Pagé, J. Katz, M. Choinière, A. Vanasse, M. Dorais, O. M. Samb, and A. Lacasse, “Trajectory modelling techniques useful to epidemiological research: A comparative narrative review of approaches,” Clin. Epidemiol., vol. 12, pp. 1205–1222, 2020, doi: 10.2147/CLEP.S265287.

J. H. Ryoo, C. Wang, S. M. Swearer, M. Hull, and D. Shi, “Longitudinal model building using latent transition analysis: An example using school bullying data,” Front. Psychol., vol. 9, p. 675, 2018, doi: 10.3389/fpsyg.2018.00675.

B. Muthén and T. Asparouhov, “Latent transition analysis with random intercepts (RI-LTA),” Psychol. Methods, vol. 27, no. 1, pp. 1–16, 2022, doi: 10.1037/met0000370.

L. V. D. E. Vogelsmeier, J. K. Vermunt, L. Keijsers, and K. De Roover, “Latent Markov latent trait analysis for exploring measurement model changes in intensive longitudinal data,” Eval. Health Prof., vol. 44, no. 1, pp. 61–76, 2021, doi: 10.1177/0163278720976762.

K. Nylund-Gibson, A. C. Garber, D. B. Carter et al., “Ten frequently asked questions about latent transition analysis,” Psychol. Methods, vol. 28, no. 2, pp. 284–300, 2023, doi: 10.1037/met0000486.

O. Perra, “Latent transition analysis,” in SAGE Research Methods: Foundations. SAGE Publications, 2020, doi: 10.4135/9781526421036878157.

F. Li, A. Cohen, B. Bottge, and J. Templin, “A latent transition analysis model for assessing change in cognitive skills,” Educ. Psychol. Meas., vol. 76, no. 2, pp. 181–204, 2016, doi: 10.1177/0013164415588946.

N. S. Ahanchi, F. Hadaegh, A. Alipour, A. Ghanbarian, F. Azizi, and D. Khalili, “Application of latent class analysis to identify metabolic syndrome components patterns in adults: Tehran lipid and glucose study,” Sci. Rep., vol. 9, no. 1, 2019.

A. Abarda, M. Dakkon, M. Azhari, A. Zaaloul, and M. Khabouze, “Latent transition analysis (LTA): A method for identifying differences in longitudinal change among unobserved groups,” Procedia Comput. Sci., vol. 170, pp. 1116–1121, 2020, doi: 10.1016/j.procs.2020.03.059.

R. E. Turpin, T. V. Dyer, D. T. I. Dangerfield, H. Liu, and K. H. Mayer, “Syndemic latent transition analysis in the HPTN 061 cohort: Prospective interactions between trauma, mental health, social support, and substance use,” Drug Alcohol Depend., vol. 214, p. 108106, 2020, doi: 10.1016/j.drugalcdep.2020.108106.

A. Gourlay, S. Floyd, F. Magut, S. Mulwa, N. Mthiyane, E. Wambiya et al., “Impact of the DREAMS partnership on social support and general self-efficacy among adolescent girls and young women: Causal analysis of population-based cohorts in Kenya and South Africa,” BMJ Glob. Health, vol. 7, p. e006965, 2022, doi: 10.1136/bmjgh-2021-006965.

X. Song, Y. Xia, and H. Zhu, “Hidden Markov latent variable models with multivariate longitudinal data,” Biometrics, vol. 73, no. 1, pp. 313–323, 2017, doi: 10.1111/biom.12536.

X. Zhou, K. Kang, and X. Song, “Two-part hidden Markov models for semicontinuous longitudinal data with nonignorable missing covariates,” Stat. Med., vol. 39, pp. 1801–1816, 2020.

F. Pennoni, M. Barbato, and S. Del Zoppo, “A latent Markov model with covariates to study unobserved heterogeneity among fertility patterns of couples employing natural family planning methods,” Front. Public Health, vol. 5, p. 186, 2017, doi: 10.3389/fpubh.2017.00186.

DOI: https://doi.org/10.37905/jjom.v8i2.40071