CD4 cell count and viral load count association and its joint risk factors among adult TB/HIV co-infected patients: a retrospective follow-up study

Abstract

Objective

The aim of this study was to determine CD4 cell count and viral load count association and its joint clinical risk factors among adult TB/HIV co-infected patients. The purpose of this research was leads to contribute improved planning and execution of screening programs, regular follow-ups, and prevention measures.

Result

The baseline variability for viral load () was higher than CD4 cell (819.797). Furthermore, the correlation matrix (R) shows that the baseline subject-specific CD4 cell and viral load were negatively correlated (− 0.). There was also a negative relationship between the subject-specific change CD4 cell and baseline viral load (-), subject-specific baseline CD4 cell, and the patient-specific change of viral load (-0.6513), longitudinal trajectory (association of evolution) for the two random slopes -0.9413). Then, there was very strong negative relationship between subject-specific change CD4 cell and viral load over time. Visit time, white blood cell, good and fair) adherence, hemoglobin ≥ 11 g/dl, baseline CD4 cell ≥ 200 cells/mm³, and baseline viral load < 10,000 copies/mLhematocritand monocytes of patients were a joint determinates for viral load and CD4 cell, respectively.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13104-025-07428-4.

Background

Human Immunodeficiency Virus (HIV) attacks the immune system and reduces people’s defenses against diseases, increasing the risk of tuberculosis (TB). TB screening should be available to People Living with HIV (PLHIV) at the time of diagnosis and at all follow-up visits, and HIV testing should be provided to all TB patients on a regular basis. PLHIV with active tuberculosis require both TB treatment and antiretroviral medication (ART). PLHIV are up to 20 times likely to acquire active tuberculosis than individuals without HIV. Tuberculosis is the primary cause of death among PLHIV worldwide [1].

In 2019, 6300 people died from tuberculosis and HIV across the globe. In 2020, an expected 37 000 HIV/TB co-infections and 5 900 deaths. 61% of notified TB cases recognized their HIV status, and 82% of new and relapsed TB-HIV cases are receiving treatment. Among newly enrolled PLHIV, 39% received preventative therapy [1, 2].

WHO recommends that ART be initiated within two weeks and that patients begin TB treatment as soon as possible, regardless of CD4 count or viral load count. PLHIV without active tuberculosis must get TB preventive medication to lower the risk of developing active TB [2] Mycobacterium tuberculosis co-infection appears to speed up HIV, and active TB has been linked to significant increases in HIV Ribonucleic acid (RNA) levels [3, 4], and a decrease in CD4 cells [5, 6, 7,8]. HIV affects patients CD4 cells [9, 10] due to viral load progression, and patients become infected with several infectious diseases, including tuberculosis [10].

Viral load is a measure HIV RNA copies per milliliter of plasma amount [11, 12]. It is also a more powerful predictor of the death of immune cells in peripheral blood and lymphoid tissues as well as the advancement of the disease [13]. It is employed in determining the likelihood that a disease will worsen, offering direction for starting treatment and tracking the effectiveness of antiretroviral therapy (ART) [14, 15]. Then, decline in CD4 cells and a rise in viral load count correspond to AIDS-related death [10].

High mortality among co-infected patients [16] is a result of the inverse relationship between viral load and CD4 T helper cells [17–19]. I.e., low CD4 cells and high viral load levels were linked to an increased risk of death among individual co-infected with TB/HIV [20].

A gradual decrease in the quantity of circulating CD4 cells indicates high viral load concentrations. These increases the risk of developing HIV/AIDS-related wasting syndrome [21, 22], and opportunistic infections [23]. The problem is related with patients poor treatment follow-up, and poor adherence to ART [21].

However, good adherence to ART lowers the risk of several AIDS-related infections, including tuberculosis (TB), by increasing CD4 cells and enabling PLWHIV to achieve and sustain viral suppression [20, 24]. These biomedical markers remain major measures of TB/HIV disease progression and the well-being of patients on treatment and better health-related quality of life [23]. Thus, the greatest measure of a patient’s immunological and clinical condition is still their CD4 cell and viral load in co-infected with HIV/TB [9, 15].

Based on multiple studies, some clinical risk factors for repeated measure CD4 cells were duration of ART [24], weight loss [27], follow-up time, adherence level, functional status, type of TB [26], baseline clinical stage IV [29], lower baseline CD4 cell count, opportunistic infection, lower hemoglobin [28], and types of ART [31]. Similarly, at the 5% level of significance, clinical risk factors for repeated measure viral load were ART regimen with (AZT + 3TC + NVP) [32], baseline viral load < 10,000 copies/mL, hemoglobin level ≥ 11 g/dl, CD4 cell count ≥ 200 per mm³, weight ≥ 50 kg, BMI between 18.5 and 24.9 kg/m², fair and good treatment adherence, advanced WHO clinical stages, patients with OCC [33], duration of ART [34], and opportunistic infections [35], underweight [36], and visit time [37].

Based on the above previous literatures, the research gap of this study is empirical gaps (lack of data or empirical studies on TB/HIV patients), methodological gaps (nonexistence of joint model to study CD4 cell and viral load simultaneously), and population gaps (nonexistence of prior study among co-infected patients in terms of CD4 cell and viral load). The novelty of this study is new perspective or approach of joint model, emerging areas of joint risk factors and association to CD4 cell and viral load count. On the other hand, novelty of this study is an introduction of original ideas of co-infected patients CD4 cell and viral load count, and methods or perspectives that significantly advance to CD4 cell and viral load count.

Besides, there is limitations of a study that done simultaneously between CD4 cell count and viral load count. Then, the quantitative relationship between CD4 cells and viral load among adult TB/HIV co-infected patients and their risk factors is not well-documented. Drawing from the aforementioned context, this study aimed to determine CD4 cell count and viral load count association and its joint clinical risk factors among adult TB/HIV co-infected patients.

Thus, based on a cooperative model of joint risk factors among adult co-infected patients and their association of CD4 cell count and viral load count leads to contribute improved planning and execution of screening programs, regular follow-ups, and prevention measures. In addition, the study’s findings also contribute to how CD4 cells, and viral load count increase/decrease, which indicate HIV replication linked to tuberculosis.

Materials and methods

Study area

This study was conducted at the University of Gondar Comprehensive Specialized Hospital (Fig. 1).

Fig. 1.

Study area geographical location

Study design

A retrospective follow-up study (repeated measure) was performed to retrieve necessary information. In these TB/HIV co-infected patients, each patient’s CD4 and viral load follow-up measurements on average was 12 months. Repeated measure was taken based on the number of times for both outcome variables (CD4 cells, and viral load count) was approximately measured within every 12 months starting from the baseline months, 12 months, 24 months, 36 months, 48 months, and 60 months.

Study period

Patients treated for co-infection with HIV/TB between March 2017 and March 2022.

Inclusion and exclusion criteria

All adult co-infected patients with TB and HIV who had at least two visit time for repeat measurements of their CD4 cell and viral load, patients who were at least 15 years old, and patients who started treatment during the treatment follow-up period (March 2015 to March 2022) were included in this study. All adult co-infected patients with TB and HIV who had only one visit time for repeat measurements CD4 cell and viral load, patients whose age less than 15 years old, patients with incomplete data, lost to follow-up before two measurements, or with other co-infections, and patients who started treatment outside study period were excluded.

Study population and sample size determination

Adult TB/HIV co-infected patients were considered as the study population. Then, based on inclusion criteria, the current study included 148 adult TB/HIV co-infected patients (Fig. 2).

Fig. 2.

Sample size determination conceptual framework

Data source

Secondary data sources that were accessed through the patient’s chart served as the data source.

Procedure of data collection and data quality control

The medical registration number (MRN) can be used in this study to select co-infected patient charts. Subsequently, two trained TB/HIV data clerks examined co-infected patient charts to gather the necessary data. The quality of the data was checked by data collectors/health professionals and hospital data management.

Data confidentiality protection

Confidentiality relates to the obligation of healthcare providers and hospitals to protect patient information from unauthorized access and disclosure. UGCSH requires that explicit consent be obtained from ours before their health data can be collected and processed. In this study, data confidentiality protection through patient consent, training and policies, and physical and technical safeguards.

Variables included in the study

Response variables

Viral load count in copies/mL and CD4 cell count in cells/mm³ were considered as response variable.

Independent variable

TB/HIV co-infected patients’ clinical independent factors are displayed in Tables 1 and 2.

Table 1.

Baseline clinical characteristics of patients

Variables	Categories	Censored (%)	Death (%)	Total (%)
Types of TB	Pulmonary	47(74.6)	16(25.4)	63(42.6)
	Extra-pulmonary	62(72.9)	23(27.1)	85(57.4)
Hemoglobin	< 11 g/dl	50(69.4)	22(30.6)	72(48.7)
	≥ 11 g/dl	59(77.6)	17(22.4)	76(51.3)
Weight	<50 kg	57(70.4)	24(29.6)	81(54.7)
	≥ 50 kg	52(77.6)	15(22.4)	67(45.3)
Baseline viral load	<10,000 copis/mL	81(79.4)	21(20.6)	102(68.9)
	≥ 10,000 copis/mL	28(60.9)	18(39.1)	46(31.1)
Treatment adherences	Poor	34(69.4)	15(30.6)	49(33.1)
	Fair	43(68.3)	20(31.7)	63(42.6)
	Good	32(88.9)	4(11.1)	36(24.3)
WHO clinical stage	Stage-I	40(85.1)	7(14.9)	47(31.8)
	Stage-II	36(87.8)	5(12.2)	41(27.7)
	Stage-III	23(62.2)	14(37.8)	37(25.0)
	Stage-IV	10(43.5)	13(56.5)	23(15.5)
Baseline CD4 cell	< 200cells/mm3	42(60.0)	28(40.0)	70(47.3)
	≥ 200cells/mm3	67(85.9)	11(14.1)	78(52.7)
OIs	No	94(88.7)	12(11.3)	106(71.6)
	Yes	15(35.7)	27(64.3)	42(28.4)
INH	No	70(89.7)	8(10.3)	78(52.7)
	Yes	39(55.7)	31(44.3)	70(47.3)
CPT	No	74(89.2)	9(10.8)	83(56.1)
	Yes	35(53.8)	30(46.2)	65(43.9)
ART regiment	1d	30(66.7)	15(33.3)	45(30.4)
	1c	39(79.6)	10(20.4)	49(33.1)
	1e	15(68.2)	7(31.8)	22(14.9)
	Others	25(78.1)	7(21.9)	32(21.6)
Functional status	Working	73(86.9)	11(13.1)	84(56.8)
	Ambulatory	20(52.6)	18(47.4)	38(25.7)
	Bedridden	16(61.5)	10(38.5)	26(17.5)
BMI	< 18.5 kg/m2	56(77.8)	16(22.2)	72(48.6)
	18.5–24.9 kg/m2	32(72.7)	12(27.3)	44(29.8)
	≥ 25 kg/m2	21(65.6)	11(34.4)	32(21.6)
OCC	No	97(85.1)	17(14.9)	114(77.0)
	Yes	12(35.3)	22(64.7)	34(23.0)
Total	-	109(73.6)	39(26.4)	148(100.0)

Key: 1d refers to ART treatment of AZT-3TC-EFV,1c refers to ART treatment of AZT-3TC-NVP,1e refers to TDF-3TC-EFV and other means other ART treatments, like 1j (TDF + 3TC + DTG), and 1f (TDF + 3TC + NVP)

Table 2.

Continuous clinical predictor’s descriptive statistics

Continuous Variables	Minimum	Maximum	Mean	Standard Deviation
Hematocrit in %	27.90	53.60	38.1595	4.73104
WBC in 10^3/	2.60	10.90	5.9074	1.70816
RBC in 10^6/	2.00	8.10	4.0813	1.12154
Platelet in 10^3/	20	566	265.09	99.187
Lymphocyte in %	21.0	69.1	45.196	12.3547
Monocyte in %	2.3	55.4	6.603	12.3547

Method of data analysis

In this study, data analysis were done by using statistical package for social science (SPSS) and R statistical software.

Statistical model

The bivariate generalized linear mixed model (GLMM) of viral load and CD4 cells served as the foundation for this investigation. The least Akaike Information Criterion (AIC) and BIC: Bayesian Information Criterion (BIC) values of the random intercept and slope models were also incorporated.

Method for handling missing data

Missing or incomplete data can result in incorrect insights and poor decision-making [39, 40]. We then utilized multiple imputation methods to accommodate missing or incomplete data by through SPSS software.

Results

Clinical characteristics of adult TB/HIV co-infected patients

Less than half of the 148 co-infected individuals (47.3%) had CD4 cell counts < 200 cells/mm³, with 40.0% of those deaths being related to HIV/TB. Comparably, of the subjects, less than one-third (31.1%) had a viral load level of more than 10,000 copies/mL, and 39.1% of them died from the illness. Similarly, among co-infected patients, less than one-fourth (24.3%) adhered to their treatment plan, with 11.1% of them dying from the illness. More than half of the patients (56.1% and 52.7%) of study participants had not used Cotrimoxazole Preventive Therapy (CPT) and Isoniazid (INH); of them, 10.8% and 10.3%, respectively, had death. 56.5% of deaths in WHO Clinical Stage IV were related to TB/HIV. The overall death rate among co-infected patients was 26.4%. The remaining co-variants can be interpreted similarly as above (Table 1).

Risk factors for CD4 cell and viral load count

Visit time, WBC, hematocrit, monocyte count, adherence, hemoglobin level, baseline CD4 cell count, and baseline viral load count were the joint clinical risk factors that affected repeated measure CD4 cell, and viral load (Table 3).

Table 3.

GLMM parameter estimate results for repeated measure viral load and CD4 cell count

Variables	Categories	viral load count			CD4 cell count
		Values	Standard error	p-values	Values	Standard error	p-values
Intercept	-	4.6103	0.0236	0.000*	1.4306	0.0161	0.000*
Visit time	-	-0.1694	0.0254	0.000*	0.0714	0.0013	0.000*
WBC	-	-0.0645	0.0431	0.004*	0.7430	0.3046	0.016*
RBC	-	0.1365	0.1290	0.104	-0.6145	0.0306	0.018*
Platelet cell count	-	0.0349	0.0338	0.061	0.1840	0.1477	0.146
Hematocrit	-	0.4875	0.0610	0.000*	-0.3649	0.0023	0.000*
Lymphocyte	-	-0.0906	0.0864	0.140	-0.6947	0.0048	0.004*
Monocyte	-	-0.0274	0.0046	0.002*	1.0064	0.0106	0.030*
Weight(Ref = < 50 kg)	≥ 50 kg	-0.6410	0.0346	0.000*	0.6320	0.6103	0.720
BMI(Ref = Underweight)	Normal	-0.1364	0.1317	0.060	0.6102	0.5921	0.203
	Obesity	0.9236	0.2469	0.003*	0.9236	0.0127	0.091
Adherence (poor)	Fair	-0.0236	0.0217	0.024	0.4109	0.0106	0.006*
	Good	-1.2491	0.2784	0.007*	0.6940	0.0391	0.002*
WHO(Ref = Stage-I)	Stage-II	1.8059	0.0183	0.103	-0.0326	0.198	0.230
	Stage-III	0.1631	0.0108	0.002*	0.1930	0.1870	0.060
	Stage-IV	0.2810	0.0327	0.010*	-0.3980	0.3206	0.261
ART regiment (Ref = 1d)	1c	0.7841	0.7081	0.880	0.7841	0.7081	0.880
	1e	0.1704	0.1208	0.102	0.1704	0.1208	0.102
	Other	1.3078	0.191	0.310	-2.2627	0.2321	0.003*
OIs(Ref = No)	Yes	0.0736	0.0641	0.136	-0.2954	0.0124	0.006*
OCC(Ref = No)	Yes	0.2641	0.0341	0.003*	-0.1426	0.1322	0.063
Hemoglobin level(> 11)	≥ 11 g/dl	-2.3610	0.3784	0.000*	0.4870	0.0291	0.000*
CD4cell (Ref = < 200)	≥ 200cells/mm3	-1.3694	0.0243	0.001*	0.6694	0.0236	0.002*
INH(Ref = No)	Yes	-1.0404	0.8413	0.261	-1.6321	0.0242	0.005*
CPT(Ref = No)	Yes	-0.0329	0.0316	0.988	0.0329	0.0316	0.988
Viral load (Ref=>=10,000)	< 10,000 copies/mL	-1.4683	0.3069	0.001*	0.9541	0.0367	0.000*
Types of TB (Ref = pulmonary)	Extra-pulmonary	0.06142	0.06003	0.060	-0.2637	0.06003	0.006*

Key: * indicates statistically significance at 5% level of significance, - refers to continuous risk factors without any categories, and Ref is the reference category

Based on Table 4, the estimated variance-covariance matrix and correlation matrix (R) are as follows;

Table 4.

Random effect covariance matrix for CD4 cell and viral load

Coefficient	Std. Deviation	Correlation
Intercept1 ( )	28.6321	( )
Visit time1 ()	2.6478	-0.7436	()
Intercept 2 ()	33.6145	-0.5624	-0.8649	()
Visit time2 ()	3.8456	-0.6513	-0.9413	-0.4684
Residual Std. error (Sigma)	26.4720
	30.3984

From the estimated variance-covariance matrix (), it can be seen that the baseline variability for viral load was higher (=) than CD4 cell (= 819.797). Then, the estimated baseline variation of random intercepts of viral load and CD4 cell was larger than the estimated variation of the random slope viral load (=) and CD4 T helper cell (=). This indicates the subject’s trajectory at baseline is larger than longitudinal trajectories.

Furthermore, the correlation matrix (R) shows that the baseline subject-specific CD4 cell measurement and viral load were negatively correlated ( = −0.). There was a negative relationship between the subject-specific change CD4 cell measurement and viral load (=). Similarly, there was a negative relationship between subject-specific baseline CD4 cell measurement and the patient-specific change of viral load over time (= -0.6513).

Similarly, the correlation between CD4 cell and viral load of the longitudinal trajectory (association of evolution) for the two random slopes 𝑎𝑛 was negative . On the other hand, there was a negative relationship between subject-specific change CD4 cell and viral load over time (.

Discussions

The association between CD4 cell and viral load of the longitudinal trajectory (association of evolution) for the two random slopes was negative . This can be concluded that there was a strong negative association between CD4 cells and viral load through visit time. The result of this study is similar with previous studies [20, 40–44]. The crucial similarity of this study findings with the above former studies were may be similar methodological approach. However, this study is difference from previous studies done in South Africa [25]. The result said there was no association between CD4 cell and current viral loads. The crucial difference of this study findings with former studies were may be different in study population, variables included under study, size of study participants, study area, and way of methods to do different results. Then, both CD4 cell count and viral load count can be measured in order to assess TB/HIV therapy outcomes. However, due to the negative relationship between CD4 cell count and viral load count, CD4 cell count monitoring occasionally fails to forecast virological failure, resulting in wasteful switching of treatment lines, medication resistance, and treatment limits.

When the patient’s visit times of co-infected patients increased by one unit, the log of expected CD4 cell count and viral load count were increased by 0. 0714 cells/mm³ (p-value = 0.000) and decreased by -0.1694 copies/mL (p-value = 0.000), respectively. The potential reason for this might be the proper follow-up of patients, which leads for being good treatment adherent and this patients additionally leads to good health progressions. The result of this study is similar with previous studies [38, 40, 45–47].The potential similarities of this study were similar in adult study population. According to this study, the goal of antiretroviral therapy (ART) for co-infected patients is to increase their CD4 count, decrease their viral load, and reduce their TB infection. In another hand, patients continuously used ART had a decreased in viral load and an increased CD4 cell. This study recommended that patients do not default form treatment in any reasons in order to overcome better immune system and minimum amount of HIV RNA.

Patients WBC increased by a 10^³/, the log of expected CD4 cell count and viral load count were increased by 0.74 cells/mm³ (p-value = 0.016) and decreased by -0.06 copies/mL (p-value = 0.004), respectively while all other clinical risk factors remained constant. This study indicated an increased CD4 cells and decreased viral load, allowing co-infected patients to quickly reduce high WBC infections. Then, patient high WBC, a clinician might order additional blood tests to determine the cause. The idea of this study is similar with previous literatures [45, 46, 48, 49]. The potential reasons for this similarities were similar risk factors that included in a study. However, the result of this study is different from former studies [50–52]. The potential reasons for this difference may be different in methodology, study areas, study participants, and study periods.

The log of expected CD4cell count and viral load count were increased by -0.3649 cells/mm3 (p-value = 0.000) and decreased by 0.4875 copies/mL (p-value = 0.000), respectively for a one unit increased in hematocrit. In the other expression, patients had a low hematocrit level means there are too few red blood cells, which can lead to anemia and different symptoms of fatigue, weakness, shortness of breath, heart palpitations, and lightheadedness. Low hematocrit levels can be caused by various factors such as nutritional deficiencies, chronic diseases, blood loss, or bone marrow disorders. The result of this study is similar with previous studies [37]. The potential reasons for this similarities were similar risk factors that included in a study, and similar study area. This study is different from previous study done at Ethiopia, Gondar [38]. The potential reasons for this difference were similar different statistical methodology, study population, no of participants, and study periods. Lastly, this result may be contradicted with scientific aspects. Because, in some conditions of patients, low hematocrit does not lead to low CD4/high viral load. This may be the nature of patient’s hematocrit with viral load and CD4 cell data. Then, further investigation is needed to fit real scientific aspects.

Similarly, the log of expected CD4cell count and viral load count were increased by 1.0064 cells/mm³ (p-value = 0.030) and decreased by -0.0274 copies/mL (p-value = 0.002), respectively for a one unit increase in monocytes. Then, patient monocyte counts are significantly leads to a low no of viral load and high no of CD4 cell counts. Viruses enter cells by endocytosis, phagocytosis, macro pinocytosis, or membrane fusion. These processes play important roles in the mechanisms that contribute to the pathogenesis of agents, as well as in the establishment of viral genome persistence and latency. Monocytes also key constituents in chronic inflammation and may be a driver in the pathogenesis of inflammation-related diseases such as atherosclerosis. The result of this study is similar with previous studies [49]. The potential reasons for this similarity were similar risk factors that included in the study. This study found that monocyte count in TB/HIV patients is important for understanding the disease’s effect on the immune system and viral load. In addition, monocytes play an important part in the immune response, and their activation and differentiation can provide information about the severity of the infection and the efficacy of treatment. Then, monitoring monocyte levels can help assess the patient’s state and guide clinical decisions about therapy and care.

The result of this study indicated that when comparing co-infected patients who had fair and good to their prescribed treatments to those who had poor, the log of expected CD4 cell count were significantly higher by 0.4109 cells/mm³ (p-value 0.006), and 0.0.694 cells/mm³ (p-value 0.002), respectively. Conversely, the log of expected viral load count for co-infected patients who had good treatment adherence was significantly lower by -1.2491 copies/mL (p-value = 0.007), compared to patients who had poor treatment adherence. On the other hand, high CD4 cell counts during good and fair ART medication were lower no of viral load, and associated with a lower risk of AIDS related morbidity and mortality. As a result, TB/HIV patients must strictly adhere to their prescribed treatment regimen in order to effectively control their diseases and prevent the emergence of drug-resistant strains. This result is consistent with previous studies [29, 33, 37, 38]. The potential reasons for this similarity were similar risk factors that included in the study, and similar study areas. However, the result of this study is different from previous literatures [30, 50–54]. The possible reasons for this different may be different in statistical methodologies that can be applied to analysis patient’s data.

The log of expected CD4 cell count and viral load count for co-infected individuals whose hemoglobin ≥ 11 g/dl were significantly higher by 0.487 cells/mm³ (p-value = 0.000) and lower by -2.361 copies/mL (p-value = 0.000), respectively, compared to patients whose hemoglobin < 11 g/dl. Hemoglobin levels are important in TB/HIV patients because they are closely related to the CD4 cell count and viral load. This study indicated patients with high CD4 cell leads to suppressed viral load and reduction of high hemoglobin-related infections. Inversely, this finding suggested that patients low hemoglobin level leads to hematological abnormality, like, anemia, which can result in a lower CD4 count, greater viral load, and a higher mortality rate. The link between hemoglobin levels, CD4 cell count, and viral load, as well as the effects of anemia on survival, is critical. Therefore, monitoring hemoglobin levels is critical for determining the illness stage and progression in TB/HIV co-infected individuals, as well as for guiding treatment decisions. The result of this study is similar with previous studies [33, 55].This similarities may be similar baseline hemoglobin predictor were used to variables of interest.

The log of expected repeated measure CD4 cell count and viral load count of co-infected patients whose baseline CD4 cell count of ≥ 200 cells/mm³ were significantly increased by 0.6694 cells/mm³ (p-value = 0.002) and decreased by -1.3694 copies/mL (p-value = 0.001), respectively, as compared to CD4 cell count of < 200 cells/mm³. The baseline CD4 cell aids in determining the need for antiretroviral therapy (ART), the urgency of initiating ART, and the requirement for prevention against opportunistic infections. Then, a high baseline CD4 cell of a patient suggested that better clinical condition for an increased in repeated measure CD4 count and readily controlled virus. This result is similar with different previous literatures [44, 46]. The similarity of this study with former study were similar risk factors considered to variable of interests.

Similarly, the log of expected repeated measure CD4 cell count and viral load count of co-infected patients whose baseline viral load < 10,000 copies/mL were significantly increased by 0.9541 cells/mm³ (p-value = 0.000) and decreased by -1.4683 (p-value = 0.001), respectively, as compared to ≥ 10,000 copies/mL. This indicated, a low baseline viral load of co-infected patients easily suppression of viral load and higher levels of CD4 cells. A lower baseline viral load correlates with a higher response to ART, which is the primary goal of treatment. The viral load test aids in the diagnosis of acute HIV infection, treatment planning, and monitoring patient response to ART. It is suggested that patients undergo routine viral load testing to ensure that they are stable on ART and have a high CD4 cell count, which is a strong predictor of disease status and immediate danger of death. The result of this study was similar with several previous studies [33, 50, 51]. The similarity of this study with former study were similar risk factors considered to variable of interests. However, the result of this study were different with former study [38, 53–55].

Conclusion and recommendations

The result of this study concluded that, there was a very strong negative association between subject-specific change CD4 cell and viral load over time. Patients with visit time, WBC, good and fair treatment adherence, hemoglobin ≥ 11 g/dl, baseline CD4 cell ≥ 200cells/mm3, and baseline viral load < 10,000 copies/mL significantly lead to high CD4 cell and low viral load. Moreover, hematocrit and monocytes of patients lead to low CD4 cells and high viral load. TB/HIV treatment must be reasonable and available to all co-infected patients in order to reduce viral load, increase CD4 count, and extend the lives of co-infected. In addition, both health professionals and researches should conduct health-related studies among co-infected patients in order to create better health status, minimize the risk of TB/HIV co-infected patients, reduce the progression of TB/HIV, reduce TB/HIV related morbidity, and death. Health professionals should also increasing adherence monitoring, strengthening early detection, and educating TB/HIV patients regarding adherence.

Study limitation

Some important clinical risk factors like eosinophil, neutrophil, and basophil counts, were not available on patient’s chart at the study period. Some patients who had only one visit time for repeat measure CD4 cell and viral load, patients with incomplete data, and unknown follow-up status were considered as limitation of the study.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AIC

Akaike Information Criterion

ART

Antiretroviral therapy

AZT

Azidothymidine

BIC

Bayesian Information Criterion

BMI

Body Mass Index

CD4

Cluster Differentiation 4

CPT

Cotrimoxazole Preventive Therapy

DTG

Dolutegravir

EFV

Efavirenz

GLMM

Generalized Linear Mixed Effect Model

HIV

Human Immunodeficiency Virus

INH

Isoniazid

NVP

Nevirapine

OCC

Other Comorbid Condition

OIs

Opportunistic Infections

PLWHIV

People Living with Human Immunodeficiency Virus

RBC

Red Blood Cell

RNA

Ribonucleic acid

SPSS

Statistical Package for social science

TB

Tuberculosis

TDF

Tenofovir Disoproxil Fumarate

WBC

White Blood Cell

WHO

World Health Organization

UGCSH

University of Gondar Comprehensive Specialized Hospital

3TC

Lamivudine

Author contributions

NSM was involved in this study from data management, data analysis, drafting, and revising the final manuscript, contributed to the conception, design, and interpretation of data, as well as to manuscript, revisions, and approved the final manuscript. AST was involved in revising the final manuscript, contributed to the conception, manuscript revisions, and approved the final manuscript.

Funding

No agents/institutions that funded this research.

Data availability

The data used in the current investigation is available from the corresponding author and can be attached upon request.

Declarations

Ethics approval and consent to participate

A statement to confirm that all methods were performed by the ethical standards as laid down in the Declaration of Helsinki. Hence, informed consent was waived by the Bahir Dar University research technical and ethical review board with Ref.no Stat-S/166/2022 because of the retrospective nature of the study. The study was approved by the Bahir Dar University Research Technical and Ethical Review Board.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.