Cointegration and Causality Analysis of Human Immunodeficiency Virus and Tuberculosis Coinfection in Nasarawa State, Nigeria
Augustine Akyenyi Attah *
Department of Statistics, Faculty of Physical Sciences, Federal University of Lafia, Nasarawa State, Nigeria.
Ayeni Omini Abam
Department of Statistics, Faculty of Physical Sciences, Federal University of Lafia, Nasarawa State, Nigeria.
Ibrahim Musa Saleh
Department of Statistics, Faculty of Physical Sciences, Federal University of Lafia, Nasarawa State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Aim: This study investigated the long-run and short-run dynamics of Human Immunodeficiency Virus (HIV) and tuberculosis (TB) coinfection in Nasarawa State, Nigeria, to provide empirical evidence for integrated disease prevention and control strategies.
Study Design: The study employed a retrospective longitudinal time-series research design using advanced econometric techniques, including descriptive statistics, unit root tests, Johansen cointegration, and Fully Modified Ordinary Least Squares (FMOLS), Vector Error Correction Model (VECM), and Pairwise Granger causality analysis.
Place and Duration of Study: The study was conducted in Nasarawa State, Nigeria, using quarterly surveillance data obtained from the National Tuberculosis and Leprosy Control Programme (NTBLCP) Quarterly Reports on Case Finding. The dataset covered the period from the fourth quarter (Q4) of 2015 to the fourth quarter (Q4) of 2024.
Methodology: The study analysed 37 quarterly observations on HIV infections, TB infections, and HIV/TB coinfections. Descriptive statistics were used to summarise the data, while Augmented Dickey-Fuller and Phillips-Perron unit root tests examined stationarity. Johansen cointegration analysis assessed long-run relationships among the variables. Long-run effects were estimated using FMOLS, short-run dynamics were examined with the VECM, and pairwise Granger causality tests determined the direction of causal relationships.
Results: Descriptive analysis showed that the mean number of HIV infections was approximately 133 cases, TB infections averaged 1,043 cases, and HIV/TB coinfection averaged 115 cases, with TB exhibiting the highest variability (SD = 536 cases). Unit root tests confirmed that all variables were integrated of order one, I(1), supporting cointegration analysis. Johansen cointegration identified three long-run equilibrium relationships among HIV, TB, and HIV/TB coinfection. FMOLS estimates revealed that a one-unit increase in HIV infection increased HIV/TB coinfection by 0.981 units, while a one-unit increase in TB infection increased HIV/TB coinfection by 0.512 units, both statistically significant at the 1% level. The VECM produced a significant error correction coefficient of −0.579, indicating that approximately 58% of deviations from long-run equilibrium were corrected within one quarter. Short-run results showed that changes in HIV infection significantly increased HIV/TB coinfection, whereas TB infection exhibited negative short-run effects. Pairwise Granger causality analysis established unidirectional causality from HIV infection to TB infection (F = 8.746, p = 0.0012), from HIV infection to HIV/TB coinfection (F = 6.826, p = 0.0074), and from TB infection to HIV/TB coinfection (F = 5.423, p = 0.0098).
Conclusion: HIV and TB infections exhibit strong long-run equilibrium relationships and significant short-run causal interactions in Nasarawa State, with HIV serving as the primary driver of TB infection and HIV/TB coinfection. These findings underscore the need for strengthened integrated HIV-TB surveillance, prevention, early diagnosis, treatment programmes, and sustained investment in healthcare systems to reduce the burden of coinfection.
Keywords: HIV infection, Tuberculosis, HIV/TB coinfection, cointegration analysis, granger causality