Prevalence and determinants of metabolic abnormalities among adult tuberculosis patients at Gondar Town, Northwest Ethiopia: an institution-based comparative cross-sectional study

Abstract

Background

Metabolic abnormalities, such as diabetes mellitus, hypertension, dyslipidemia, and central obesity, are significant contributors to the global burden of non-communicable diseases. Tuberculosis (TB) patients may be at increased risk of these conditions due to chronic inflammation, nutritional depletion, and anti-TB drug effects; however, locally grounded comparative evidence from resource-limited settings such as northwest Ethiopia remains scarce. This study aimed to assess the prevalence and determinants of metabolic abnormalities among TB patients in Gondar Town, Ethiopia.

Methods

An Institution -based comparative cross-sectional study was conducted from March to May 2024 at three health institutions in Gondar Town. A total of 260 participants 130 confirmed TB patients and 130 individually age- and sex-matched TB-negative controls confirmed by sputum smear microscopy were enrolled using simple random sampling with proportional allocation across facilities. Study Participants data were collected through structured questionnaire, anthropometric and clinical examinations, and laboratory analysis. Normality of continuous variables was assessed using the Kolmogorov-Smirnov test, and supported by histograms and Q-Q plots. Independent t-test and one-way ANOVA with Bonferroni post-hoc test was used for continuous variable comparisons across TB-positive with HIV, TB-positive without HIV, and TB-negative control groups. Chi-square and Fisher’s exact tests were used for categorical between-group comparisons Logistic regression analysis was performed on the pooled sample, to identify independent determinants of metabolic abnormalities; statistical significance was set at p < 0.05 in multivariable analysis.

Results

The overall prevalence of metabolic abnormalities was higher among TB patients compared to TB-negative controls. Nearly half of the TB patients 48.46% (95% CI 39.92–57.08) had at least one metabolic abnormality compared to TB-negative controls 43.85% (95% CI 35.50–52.54). Among TB patients, the prevalence of diabetes mellitus, hypertension, dyslipidemia, and central obesity were 18.46% (95% CI 12.62–26.14), 6.92% (95% CI 3.62–12.83), 29.23% (95% CI 22.00–37.68), and 6.15% (95% CI 3.00–11.88), respectively. In contrast, the magnitudes were 10.00% (95% CI 5.86–16.53), 3.85% (1.59–8.96), 20.00% (13.94–27.83), and 9.23% (95% CI 5.29–15.61) among TB-negative controls. On multivariable logistic regression analysis, advanced age, sex, alcohol drinking habit were the predictor variables significantly associated (p < 0.05) with metabolic abnormalities among study participants.

Conclusion and recommendation

This study provides the first institution-based comparative estimate of metabolic abnormality burden among TB patients in Gondar Town, northwest Ethiopia. The overall prevalence of composite metabolic abnormality was 48.46% among TB patients and 43.85% among controls a difference that did not reach statistical significance (p = 0.455). Similarly, between-group differences in the categorical prevalence of individual conditions including hypertension, hyperglycemia, overall dyslipidemia, and central obesity did not reach statistical significance. However, TB patients exhibited statistically significant differences in key continuous biochemical parameters compared to controls including significantly higher mean triglycerides, LDL-cholesterol, and fasting blood glucose, and significantly lower mean HDL-cholesterol and BMI. Patients aged 45 years and above, female patients, and heavy alcohol consumers represent high-risk subgroups who should be prioritized for targeted metabolic screening. These findings provide locally grounded evidence to support the integration of lipid profile and fasting blood glucose testing into routine TB treatment follow-up within the Ethiopian national TB program, particularly at referral-level facilities in northwest Ethiopia.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12879-026-14086-9.

Introduction

Tuberculosis (TB) is an infectious disease caused by the bacterial Mycobacterium species. It primarily affects lungs, causing pulmonary tuberculosis (PTB); but it can also involve extrapulmonary sites including the colon, meninges, bones, and other bodily parts, resulting in extra-pulmonary form tuberculosis (EPTB) [1]. Despite significant advances in diagnosis, treatment, and prevention efforts, tuberculosis remains a significant burden, with an estimated 10.6 million new cases and 1.6 million deaths in 2022, making it one of the leading causes of death in low- and middle-income countries [2, 3]. Tuberculosis remains a major public health problem in Africa and Ethiopia. The WHO reported that the African region still carries a large share of the global TB burden [4]. Ethiopia ranks among the 30 high TB burden countries globally, with an estimated incidence rate of 140 per 100,000 population and approximately 157,000 new TB cases annually, placing a substantial burden on an already strained healthcare system [5]. Conversely, Ethiopia bears one of the highest TB burdens globally, where socioeconomic deprivation, delayed diagnosis, malnutrition, and limited access to healthcare services continue to amplify the disease’s impact on individuals and communities [6, 7].

Metabolic abnormalities encompassing diabetes mellitus (DM), hypertension, dyslipidemia, and central obesity represent an escalating public health crisis, increasingly recognized as major contributors to the global burden of non-communicable diseases (NCDs) [8]. There are growing evidence highlights a strong link between TB and metabolic diseases, such as impaired glucose metabolism, dyslipidemia, and systemic inflammation [9, 10]. These metabolic derangements may result from the chronic inflammatory state induced by tuberculosis infection, as well as the adverse effects of anti-TB medications [11, 12]. The convergence of infectious disease burdens particularly TB with rising metabolic disease prevalence creates a uniquely challenging dual burden in settings like Ethiopia, where both epidemics coexist within already strained health systems [13].

A growing body of evidence highlights a strong and biologically plausible link between TB and metabolic abnormalities [9, 14]. Metabolic abnormalities, usually linked to behavioral and life style changes have surged alongside endemic TB [8]. This intersection is not merely coincidental; however, there are growing evidence indicates bidirectional relationships between TB and specific metabolic abnormalities [10, 13, 15]. Once considered diseases of affluence, metabolic conditions are now rapidly rising in low- and middle-income countries, driven by urbanization, dietary transitions, physical inactivity, and demographic aging [8].

The relationship between TB and metabolic abnormalities, including diabetes, hypertension, obesity, and dyslipidemia, is multifactorial and bidirectional [16]. On one hand, chronic inflammation is a central mechanism linking TB to these conditions. During TB infection, pro-inflammatory cytokines such as tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), and interferon-gamma (IFN-γ) are elevated, disrupting normal glucose and lipid metabolism [17]. This immune response can impair insulin signaling, leading to hyperglycemia and, in some cases, the development or exacerbation of diabetes mellitus (DM) [18]. On the other hand, Tuberculosis also contributes to metabolic abnormalities through its systemic effects and treatment. Anti-TB drugs like rifampicin and isoniazid further complicate metabolic health by inducing hepatotoxicity, altering lipid profiles, and impairing vascular endothelial function [15].

The co-occurrence of TB and metabolic abnormalities is particularly concerning in LMICs, where both conditions are prevalent, healthcare resources are limited, and integrated care models remain underdeveloped [13, 19] Metabolic complications in TB patients can impair treatment response, increase the risk of adverse drug reactions, prolong infectiousness, worsen treatment outcomes, and elevate long-term cardiovascular risk even after TB cure [14, 20]. Understanding the magnitude and determinants of metabolic abnormalities among TB patients is therefore not merely an academic exercise it is a prerequisite for developing evidence-based, integrated care models that address both the infectious and non-communicable dimensions of patient health [20].

Despite growing evidence about the association between tuberculosis and metabolic abnormalities, data from low-income countries, such as Ethiopia, remain scares and previous results are inconsistent [13, 19]. The mechanisms underlying the bidirectional relationships between TB and specific metabolic conditions require further elucidation [9, 10, 18]. Understanding the magnitude and determinants of these conditions among TB patients is essential for developing comprehensive care models. By addressing the dual burden of TB and metabolic disorders, we can improve health outcomes, reduce disparities, and advance progress toward global TB elimination goals [14, 20]. Most previous studies conducted on this theme have focused on single metabolic disorders specifically on diabetes mellitus, while comprehensive assessment of multiple metabolic abnormalities among TB patients remains underexplored. In addition, many earlier studies lacked appropriate TB-negative comparison groups, making it difficult to clearly determine the relative burden of metabolic abnormalities among TB patients. Furthermore, limited evidence exists regarding the determinants of these abnormalities among adult TB patients in Northwest Ethiopia. Understanding the burden and associated factors of metabolic abnormalities among TB patients is essential for improving integrated patient management and reducing long-term complications. Therefore, this study aimed to assess the prevalence and determinants of metabolic abnormalities among adult tuberculosis patients compared with TB-negative controls in Gondar Town, Northwest Ethiopia.

Methods and materials

Study design, area and period

A facility based comparative cross-sectional study was conducted between March 11th and May 17th 2024. Three health facilities (University of Gondar comprehensive specialized hospital, Maraki health center and Azezo health center) were used for data collection. These three facilities were selected purposively based on high TB patient flow and their representativeness of different levels of the healthcare system a tertiary referral hospital and two primary-level health centers ensuring a diverse study population.

Gondar town is found in Amhara regional state, and is located about 627 km from Addis Ababa, the capital city of Ethiopia. According to the Gondar City Administration Bureau (2022/2023), the town has an estimated population of approximately 432,000. Gondar city has eight health centers, 14 health posts, 32 private clinics, and one comprehensive specialized hospital serving as the main referral center for the northwest Ethiopia catchment area [21].

Population

Source population

TB-patients group - All adult TB confirmed patients on treatment at the selected health facilities of Gondar town in 2024.

TB-negative controls - Age and Sex matched TB- negative population at Gondar town.

Study population

TB-patients group - All adult TB confirmed patients who visited the selected health facilities during data collection period (March 11th - May 17th 2024) and were accessible for data collection.

TB-negative controls – Sex and Age matched TB- negative adults who attended the selected health facilities during the data collection period and met the eligibility criteria.

Eligibility criteria

Inclusion criteria

TB-patients group - Adults aged 18 years and above, confirmed with tuberculosis with GeneXpert or both GeneXpert and smear microscopy and on anti-TB treatment; and willing to provide informed consent.

TB-negative controls - Adults aged 18 years and above with: no prior history of tuberculosis diagnosis or treatment; no clinical signs or symptoms suggestive of active TB at the time of enrollment specifically, absence of chronic cough ≥ 2 weeks, fever, night sweats, or significant weight loss; a negative sputum smear microscopy result (Ziehl-Neelsen staining) performed at the time of enrollment; and willingness to provide written informed consent.

Exclusion criteria

Participants were excluded if they were on medications that may independently affect metabolic profiles (corticosteroids, or any lipid-lowering drugs); Pregnant or lactating women (due to altered metabolism during pregnancy/lactation); with history of recent hospitalization for any other acute illness within the preceding four weeks. Individuals with mental health disorders, hearing impairments, or other significant medical conditions that could potentially affect their ability to provide accurate data were excluded. Furthermore, individuals receiving steroid or protein therapy were excluded from the study.

Variables

Dependent variables

• The primary outcome was composite metabolic abnormality, defined as the presence of at least one of the following: diabetes mellitus, hypertension, dyslipidemia, or central obesity. Individual metabolic outcomes were analyzed as secondary outcomes.

Independent variables

• Socio-demographic factors - Age, Sex, Educational status, Marital status, Average Income, Residence and Occupation.

• Anthropometric characteristics - Body Mass Index (BMI), Mid Upper Arm Circumference (MUAC), Waist Circumference (WC), Hip Circumference (HC) and Waist to Hip Ratio (WHR).

• Clinical parameters – Co-morbidity, Medication use, family history of Diabetes, and hypertension.

• Behavioral and life style factors – coffee drinking Habit, Alcohol consumption, Smoking, Khat chewing, Physical activity, dietary diversity score, usual Food source.

• Tuberculosis related Factors – Microscopic smear status, Previous history of Tuberculosis diagnosis, treatment stage, Tuberculosis class, Tuberculosis related signs and symptoms.

Sample size and sampling technique

The sample size was calculated using a double population proportion formula Open-epi software. The values of the outcome variable were taken from the previous study conducted in South-Eastern Amhara Region, Ethiopia [22]; Taking the population proportion to be 8.3% and Assuming 95% confidence interval and 80% power, the minimal sample size in each group is expected to be 117. Considering 10% non- response rate a total of 260 confirmed tuberculosis patients and apparently healthy TB-negative controls were recruited on the present study.

The three health facilities were selected based on high TB patient volume and representativeness of facility type. Sample size was allocated proportionally across the three facilities based on their respective TB patient registers., the University of Gondar Comprehensive specialized hospital, “Maraki” Health Center and “Azezo”; and simple random sampling technique was applied to recruit each study participants.

TB patient group: TB patients were recruited using simple random sampling from the TB clinic registers at the three selected health facilities, with proportional allocation across facilities. The sampling frame consisted of all confirmed TB patients enrolled in the TB treatment registers during the data collection period. Randomness was achieved using a lottery method. For each enrolled TB patient, one TB-negative control was recruited through individual matching on sex and age (± 5 years) from outpatient department attendees, hospital staff, and medical interns at the same facilities. Controls were therefore not independently randomly sampled but were deliberately recruited to achieve 1:1 individual matching with enrolled TB patients. Matching: TB-negative controls were recruited from health facility workings staffs, Students and outpatient department attendees at the same three health facilities who were working/attaching at the facilities or a presented for routine checkups or minor non-infectious conditions. Individual matching was applied on a 1:1 ratio. Each TB patient was matched to one TB-negative control of the same sex and within ± 5 years of age. Matching was performed at the time of enrollment by trained data collectors who verified age and sex prior to enrolling each control participant.

Data collection

The data collection process involved the use of pre-prepared structured questionnaires (Supplementary File 1) which were translated into Amharic for convenience. The translated questionnaires were pre-tested on 5% of the target population through face-to-face interviews to ensure understanding based on the study objectives. Trained clinical nurses were selected as data collectors who screened participants for eligibility; participants who met the inclusion criteria were invited to participate voluntarily after being provided detailed information about the study and provided written and signed consent.

Anthropometric measurements: Weight was recorded using a digital weighing scale (SECA 813 digital flat scale (SECA GmbH & Co. KG, Hamburg, Germany) with an accuracy of ± 0.1 kg, ensuring participants wore light clothing and removed their shoes. Height was measured using a stadiometer (SECA 213 portable stadiometer (SECA GmbH & Co. KG, Hamburg, Germany), accuracy ± 0.1 cm, with participants standing barefoot on a flat surface, their heels, buttocks, shoulders, and the back of the head touching the stadiometer. The head was positioned in horizontal plane, and the height was recorded to the nearest 0.1 cm. Body mass index (BMI) was calculated as weight (kg) divided by the square of height (m²) and categorized using WHO classifications [23].

Waist circumference (WC) was measured using a non-stretchable measuring tape positioned at the midpoint between the lower margin of the last palpable rib and the top of the iliac crest. Hip circumference (HC) was measured at the widest part of the buttocks using the same tape, ensuring it remained parallel to the ground. Both WC and HC measurements were taken while participants stood upright with feet together, arms relaxed, and breathing normally and measurements were recorded to the nearest 0.1 cm. Mid-upper arm circumference (MUAC) was measured using a flexible, non-stretchable measuring tape on the participant’s arm. The measurement was taken at the midpoint between the acromion process of the scapula and the olecranon process of the ulna, with the arm relaxed and hanging freely.

Blood pressure (BP): Blood pressure was measured using an analog sphygmomanometer and stethoscope from Omron, Japan, with readings taken from the upper arm of participants who had been seated for at least 5 min. The BP-cuff was placed securely on the upper arm of the non-dominant side, with the arm supported at heart level. Two readings were taken five minutes apart, and the average of the two was used. In cases where the two readings differed significantly (≥ 10 mmHg), a third reading was taken, and the closest two readings were averaged.

Clinical data: clinical data were extracted from participants’ medical charts and records to ensure comprehensive data collection. For TB patients, this included TB diagnosis type (pulmonary or extrapulmonary), treatment regimen, treatment phase, treatment duration, sputum smear status, HIV status, and any documented comorbidities; and for TB-negative controls clinical data were obtained from self-reports and OPD medical records, including any current medications, and any documented comorbidities.

Laboratory procedures and analysis

Following face-to-face interviews with structured questions, fasting venous blood samples were collected from each participant after a minimum of 8-hour fast, processed according to standard procedures of the hospital’s laboratory. Five milliliter (ml) fasting venous blood sample of was collected from the medial cubital vein of the left arm with a 19-gauge syringe then transferred to a jell coated serum separator tube (SST); the tube labeled with a unique ID number.

The blood was left to form a blood clot at room temperature for 30 min, then centrifuged with 3000 RPM for 5 min to separate cells from serum. Serum fasting blood glucose and lipid profile (total cholesterol, triglycerides, HDL-cholesterol, and LDL-cholesterol) were analyzed on the same day of collection using the Beckman Coulter DxC 700 AU automated clinical chemistry analyzer (Beckman Coulter Inc., Brea, California, USA). All results were reported in mg/dL.

Data quality control

Prior to data collection, the data collectors underwent comprehensive training on the study objectives, data confidentiality, participant rights, consent procedures, interview techniques, and laboratory test protocols. A pre-test of the prepared questionnaire was conducted on a 5% sample of the target population to validate its accuracy and consistency. Sample analysis was carried out using a Beckman-Coulter DXC 700 chemistry analyzer, following quality control measures derived from feedback reports of the Amhara regional Laboratory. The research team closely monitored the data collection and laboratory analysis process to maintain the data integrity and consistency.

Laboratory quality control: A comprehensive quality management system was maintained throughout the study period. Daily two-level internal quality control (IQC) was performed on the Beckman Coulter DxC 700 AU analyzer before any patient sample was analyzed, using commercially supplied normal-range and pathological-range control materials. QC runs were evaluated using Westgard rules (1–2 s warning rule; 1–3 s rejection rule) to detect random and systematic analytical errors. Analytical runs were accepted only when IQC results fell within ± 2 standard deviations of the assigned target mean; runs failing QC were repeated after troubleshooting and recalibration before patient results were reported. The coefficient of variation (CV%) for all analytes remained within the internationally accepted threshold of ≤ 5% throughout the study period, confirming acceptable analytical precision [24, 25]. Calibration was performed using manufacturer-supplied Beckman Coulter reagent-specific calibrators and was repeated whenever QC results fell outside acceptable control limits, when reagent lots were changed, or following significant maintenance procedures. External quality assurance (EQA) was maintained through participation in the Amhara Regional Health Bureau Laboratory proficiency testing program; the laboratory received satisfactory performance feedback for all analytes measured in this study during the study period.

Statistical analysis

All participant data was first entered to Epi-data software, followed by transfer to STATA Version 14 for statistical analysis. Descriptive statistics including frequencies, percentages, means, and standard deviations were used to characterize participant sociodemographic, clinical, and biochemical profiles.

The normality of continuous variables was assessed using the Kolmogorov-Smirnov test. Normality assessment was further supported by graphical methods histograms and Q-Q (quantile-quantile) plots for all continuous variables. Variables satisfying normality assumptions by both K-S test (p > 0.05) and Q-Q plot visual inspection were analyzed using parametric tests.

Independent samples t-tests were used to compare continuous biochemical and anthropometric parameters between TB patients and TB-negative controls. To assess whether HIV co-infection independently contributed to the observed metabolic differences, a one-way ANOVA with Bonferroni post-hoc test was performed comparing key continuous metabolic parameters across three groups: TB-positive with HIV co-infection, TB-positive without HIV, and TB-negative controls.

Logistic regression model was fitted to identify associated factors. Categorical independent variables were assessed using the chi-square (X²) or Fisher’s exact tests, with those meeting assumptions subjected to bivariate logistic regression analysis. Finally, the logistic regression model was fitted. All variables meeting the threshold of p < 0.25 in bivariable analysis were advanced to the multivariable logistic regression model. TB status and HIV status did not meet this threshold and were therefore excluded from the multivariable model per the pre-specified variable selection criteria this was a statistically justified exclusion, not an intentional omission. Although individual 1:1 matching was performed on age and sex, unconditional rather than conditional logistic regression was used for the multivariable analysis. This approach is valid when matching variables are included as covariates in the model as was done in the current study and when the number of matched pairs is large relative to model complexity. Both age and sex were included as covariates in all regression models, thereby controlling for their effects. Variables exhibiting multicollinearity were identified through variance inflation factor (VIF) assessment and excluded. The final multivariable logistic regression model was fitted on the pooled sample (n = 260), with residence and occupation included as covariates to adjust for sociodemographic differences between groups. Model fit was assessed using the Hosmer-Lemeshow goodness-of-fit test (p = 0.50), indicating adequate fit. Results are reported as crude odds ratios (COR) and adjusted odds ratios (AOR) with 95% confidence intervals, with statistical significance set at p < 0.05 in the multivariable model. The Hosmer-Lemeshow test was analyzed to test the goodness of fit of the logistic regression model, yielding a p-value of 0.50.

Operational definitions

• Tuberculosis (TB) Cases - Study participants were considered TB positive, if their sputum samples are positive for Mycobacterium species by GeneXpert or both GeneXpert and smear microscopy [2].

• TB negative controls - were defined as adults with no prior history of tuberculosis diagnosis or treatment; absence of any clinical signs or symptoms suggestive of active TB (chronic cough ≥ 2 weeks, fever, night sweats, or significant weight loss) at the time of enrollment; and a negative sputum smear microscopy result (Ziehl-Neelsen staining) performed at the time of enrollment.

• Metabolic abnormality- was defined as the presence of at least one of the following conditions: hyperglycemia (Fasting blood glucose (FBG) ≥ 126 mg/dL (7.0 mmol/L) after an 8–10 h fasting [26], hypertension (Systolic blood pressure (SBP) ≥ 140 mmHg, or Diastolic blood pressure (DBP) ≥ 90 mmHg [27], dyslipidemia (Total cholesterol ≥ 200 mg/dL (5.2 mmol/L), or LDL cholesterol ≥ 130 mg/dL (3.4 mmol/L), or Triglycerides ≥ 150 mg/dL (1.7 mmol/L), or HDL cholesterol < 40 mg/dL (1.0 mmol/L) for men or < 50 mg/dL (1.3 mmol/L) for women [28], or central obesity (was considered WC > 90 cm (35.4 inches) for Men and WC > 80 cm (31.5 inches) for Women [29, 30]).

• BMI- was considered < 18.5 kg/m² (Underweight), 18.5–24.9 kg/m² (Normal Weight), 25.0–29.9 kg/m²(Overweight), and ≥ 30 kg/m² (Obese) [23].

• MUAC – was considered (Severe malnutrition) < 21 cm, (Moderate Malnutrition) 21–22.9 cm; and (Normal) ≥ 23 cm [31].

• Waist circumference (WC) - was considered Normal: <90 cm (Men) or < 80 cm (Women); and Higher ≥ 90 cm or ≥ 80 cm (Women) [29, 30].

• Waist to Hip Ratio (WHR)- was considered Normal: < 0.9 (Men) or < 0.85 (Women); and Increased risk ≥ 0.9 or ≥ 0.85 (Women) [29, 30].

• Previous History of TB - refers to whether an individual has been diagnosed with and treated for tuberculosis (TB) in the past before the current regime.

• Current Microscopic Smear Status - refers to the results of a sputum smear examination for acid-fast bacilli (AFB).

• TB-Related Signs and Symptoms - refers to clinical manifestations associated with active TB infection based on participants response about the presence of specific symptoms within the past 2–4 weeks; such as persistent cough lasting ≥ 2 weeks, Fever, Night sweats, Weight loss (≥ 5% of body weight), Fatigue, nausea, vomit, loss of appetite and/or Chest pain.

• TB Type – “PTB” Evidence of TB in the lungs, including positive sputum smear, culture, or radiographic findings; and “EPTB” TB being diagnosed at any extrapulmonary sites.

Ethical considerations

The study protocol was approved by the Institutional Review Board of School of Biomedical and Laboratory Sciences, college of medicine and health sciences University of Gondar; with reference number SBLS/763/16. Permission was also gained from the medical director of University of Gondar comprehensive specialized hospital and each of the health centers. Informed consent was obtained from each participant after appropriate information was given about the aim of the project and objective of the study, and confidentiality of data was maintained throughout the study. The collected data was not used for another purpose other than the present study. The study was conducted in accordance with the Declaration of Helsinki guidelines and regulations.

Results

Socio-demographic characteristics

A total of 260 participants were enrolled, evenly divided between TB patients (n = 130) and TB-negative controls (n = 130). The mean age of for TB group was 33.52 (± 13.26) while the mean age for controls was 32.63 (± 11.70). The majority of participants 180 (69.23%) had had at-least attained secondary education, and 198 (76.15%) were Urban dwellers. The study population was predominantly younger (18–34 years: 68.46%) and predominantly urban (198, 76.15% overall) (Table 1).

Table 1.

Sociodemographic characteristics of TB patients and TB-negative controls at Gondar Town, Northwest Ethiopia, 2024 (n = 260)

Variables	Categories	TB+	Controls	Total	P-Values
		n (%)	n (%)	n (%)
Age Group	18 − 34	88 (67.69%)	90 (69.23%)	178 (68.46%)	0.99
	35–44	21 (16.15%)	20 (15.38%)	41 (15.77%)
	45–54	7 (5.38%)	7 (5.38%)	14 (5.38%)
	55/+	14 (10.77%)	13 (10.00%)	27 (10.38%)
Sex	Male	70 (53.85%)	70 (53.85%)	140 (53.85%)	1
	Female	60 (46.15%)	60 (46.15%)	120 (46.15%)
Residence	Urban	75 (57.69%)	123 (94.62%)	198 (76.15%)	< 0.001*
	Rural	55 (42.31%)	7 (5.38%)	62 (23.85%)
Educational Attainment	No formal Education	10 (7.69%)	5 (3.85)	15 (5.77%)	0.22
	Primary	30 (23.08%)	35 (26.92%)	65 (25.00%)
	Secondary	48 (36.92%)	58 (44.62%)	106 (40.77%)
	Higher Education	42 (32.31%)	32 (24.62%)	74 (28.46%)
Marital Status	Single	66 (50.77%)	43 (33.08%)	109 (41.92%)	0.01*
	Married	54 (41.54%)	73 (56.15%)	127 (48.85%)
	Widowed/ Divorced	10 (7.69%)	14 (10.77%)	24 (9.23%)
Average Monthly Income	< 5000 ETB	35 (26.92%)	24 (18.46%)	59 (22.69%)	0.10
	≥ 5000 ETB	95 (73.08%)	106 (81.54%)	201 (77.31%)
Occupation	Governmental Job	1 (0.77%)	24 (18.46%)	25 (9.62%)	< 0.001*
	Farmer	58 (44.62%)	4 (3.08%)	62 (23.85%)
	Merchants	24 (18.46%)	40 (30.77%)	64 (24.62%)
	Unemployed/Retired	31 (23.85%)	20 (15.38%)	51 (19.62%)
	Daily Labor	3 (2.31%)	31 (23.85%)	34 (13.08%)
	Private employee	13 (10.00%)	11 (8.46%)	24 (9.23%)

* Statistically significant on chi² test, ETB- Ethiopian birr (1 ETB = 0.0079 USD, December, 2024)

Anthropometric and clinical characteristics

The mean body mass index (BMI) was 20.75 (± 3.09) kg/m² among TB patients’ group and was 22.77 (± 3.96) kg/m² TB-negative controls. The mean MUAC of the TB group was 21.27 (± 2.71), with 143 (63.85%) were undernourished based on MUAC measurement. On the other hand, mean MUAC measurement of the non-TB controls were 22.5 (± 2.17). Thirty-five of TB- patients observed to have HIV/AIDS- TB comorbidity and the prevalence of HIV/AIDS-TB co-morbidity found to be 26.92% (Table 2).

Table 2.

Anthropometric and Clinical characteristics of TB patients and TB-negative controls at Gondar Town, Northwest Ethiopia, 2024 (n = 260)

Variables	Categories		TB+	Controls	Total	P-values
			n (%)	n (%)	n (%)
BMI	Under weight		28 (21.54%)	17 (13.08%)	45 (17.31%)	< 0.001*
	Normal		88 (67.69%)	74 (56.92%)	162 (62.31%)
	Over weight		14 (10.77%)	31 (23.85%)	45 (17.31%)
	Obese		-	8 (6.15%)	8 (3.08%)
MUAC	Sever MN (< 21 cm)		57 (43.85%)	33 (25.38%)	90 (34.62%)	0.004*
	Moderate MN (21–22.9 cm)		26 (20.00%)	27 (20.77%)	53 (20.38%)
	Normal (≥ 23 cm)		47 (36.15%)	70 (53.85%)	117 (45.00%)
WHR	≤ 0.80		3 (2.31%)	4 (3.08%)	7 (2.69%)	0.45
	0.81–0.85		64 (49.23%)	54 (41.54%)	118 (45.38%)
	> 0.85		63 (48.46%)	72 (55.38%)	135 (51.92%)
Blood Pressure	SBP	< 140 mmhg	129 (99.23%)	129 (99.23%)	258 (99.23%)	1
		≥ 140 mmhg	1 (0.77%)	1 (0.77%)	2 (0.77%)
	DBP	< 90 mmhg	121 (93.08%)	125 (96.15%)	246 (94.62%)	0.27
		≥ 90 mmhg	9 (6.92%)	5 (3.85%)	14 (5.38%)
Family History of DM	Yes		7 (5.38%)	9 (6.92%)	16 (6.15%)	0.60
	No		123 (94.62%)	121 (93.08%)	244 (93.85%)
Family History of HTN	Yes		6 (4.62%)	9 (6.92%)	15 (5.77%)	0.42
	No		124 (95.38%)	121 (93.08%)	245 (94.23%)
Comorbidity	None		95 (73.08%)	130 (100%)	225 (86.54%)	< 0.001*
	HIV/AIDS		35 (26.92%)	-	35 (13.46%)

* Statistically significant association on Chi² / Fisher- exact test

Life style, behavioral and diet related factors

Smoking habit and khat chewing differed significantly between groups. All smokers — both current (n = 8, 6.15%) and ex-smokers (n = 14, 10.77%) were TB patients; no controls reported any smoking history (p < 0.001). Regarding alcohol consumption, almost half of the participants 130 (50.00%) reported to be never have habit of consuming alcohol. In terms of physical activity, The majority of participants (158, 60.77%) reported low physical activity (1–2 days per week), with no significant difference between groups (p = 0.27). A low dietary diversity score was common in both groups, with 92 (70.77%) of TB patients and 76 (58.46%) of controls falling in the low dietary diversity category (Table 3).

Table 3.

Life style, behavioral and diet related characteristics of TB patients and TB-negative controls at Gondar Town, Northwest Ethiopia, 2024 (n = 260)

Variables	Categories	TB+	Controls	Total	P-values
		n (%)	n (%)	n (%)
Coffee Drinking	Rare/Never	11 (8.46%)	4 (3.08%)	15 (5.77%)	0.05
	Weekly drinker	96 (73.86%)	91 (70.00%)	187 (71.92%)
	Daily drinker	23 (17.69%)	35 (26.92%)	58 (22.31%)
Smoking habit	Non -smoker	108 (83.08%)	130 (100%)	238 (91.54%)	< 0.001*
	Smoker	8 (6.15%)	-	8 (3.08%)
	Ex-smoker	14 (10.77%)	-	14 (5.38%)
Alcohol Drinking Habit	Non-drinker	58 (44.62%)	72 (55.38%)	130 (50.00%)	0.05
	Occasional Drinker	37 (28.62%)	40 (30.77%)	77 (29.62%)
	Moderate Drinker	18 (13.85%)	7 (5.38%)	25 (9.62%)
	Heavy Drinker	17 (1308%)	11 (8.46%)	28 (10.77%)
Khat Chewing Habit	Chewer	9 (6.92%)	1 (0.77%)	10 (3.85%)	0.01*
	Non-Chewer	121 (93.08%)	129 (99.23%)	250 (96.15%)
Physical Activity Exercise	Inactive	14 (10.77%)	21 (16.15%)	35 (13.46%)	0.27
	Low (1–2 days/weeks)	86 (66.15%)	72 (55.38%)	158 (60.77%)
	Moderate (3–5 days/weeks)	23 (17.69%)	31 (23.85%)	54 (20.77%)
	High (≥ 6 days/weeks)	7 (5.38%)	6 (4.62%)	13 (5.00%)
Dietary diversity score	Low	92 (70.77%)	76 (58.46%)	168 (64.62%)	0.10
	Moderate	29 (22.31%)	39 (30.00%)	68 (26.15%)
	High	9 (6.92%)	15 (11.54%)	24 (9.23%)
Usual Food Source	Plant /Vegetables sources	112 (86.15%)	121 (93.08%)	233 (89.62%)	0.06
	Animals and dairy products	18 (13.85%)	9 (6.92%)	27 (10.38%)

* Statistically significant association on Chi² / Fisher- exact test

Clinical profile and tuberculosis-related variables among tuberculosis patients

Among the 130 TB patients, the majority (119, 91.54%) were diagnosed with TB for the first time and were on their first treatment regimen. Pulmonary TB was the predominant form, accounting for 119 (91.54%) of cases, while extrapulmonary TB accounted for 11 (8.46%). Regarding treatment phase, 78 (60.00%) were in the intensive phase (2–8 weeks) and 52 (40.00%) were in the continuation phase (9–24 weeks). Smear microscopy was positive in 45 (34.62%) and negative in 85 (65.38%) of TB patients Regarding TB-related symptoms, loss of appetite was the most frequently reported symptom (44, 33.85%), followed by vomiting (31, 23.85%), nausea (31, 23.85%), and fever (28, 21.54%). Cough was reported by 26 (20.00%) of TB patients (Table 4).

Table 4.

Clinical profile and tuberculosis-related variables among tuberculosis patients at Gondar Town, Northwest Ethiopia, 2024 (n = 130)

Variables	Categories	Frequency (n)	Percentage (%)
Previous History of TB	Yes	11	8.46%
	No	119	91.54%
Microscopic Smear Status	Smear Positive	45	34.62%
	Smear Negative	85	65.38%
Treatment group	Intensive phase (2–8 weeks)	78	60.00%
	Continuation phase (9–24 weeks)	52	40.00%
TB Type	Pulmonary (PTB)	119	91.54%
	Extrapulmonary (EPTB)	11	8.46%
Tuberculosis related sign and symptoms
Cough	Yes	26	20.00%
	No	104	80.00%
Fever	Yes	28	21.54%
	No	102	78.46%
Vomit	Yes	31	23.85%
	No	99	76.15%
Nausea	Yes	31	23.85%
	No	99	76.15%
Loss of appetite	Yes	44	33.85%
	No	86	66.15%

Biochemical and anthropometric parameters

TB patients exhibited a statistically significant (p < 0.05) higher mean value of serum triglyceride (TG), Low density cholesterol (LDL-C) and Fating blood sugar (FBS) than TB-negative controls. However, the mean High-density cholesterol (HDL-C) and mean BMI of TB patients shown to be lower than that of controls. Conversely, no statistically significant difference was observed on the mean serum value of total cholesterol (TC), Wiest to hip ratio (WHR) and systolic and diastolic blood pressures, with (p > 0.05) (Table 5).

Table 5.

Biochemical and anthropometric parameters among TB patients and TB-negative controls at Gondar Town, Northwest Ethiopia, 2024 (n = 260)

Parameters	TB+	Controls	Mean difference (%)	P-values
	Mean (SD)	Mean (SD)
TC	159.56 (± 26.02)	154.68 (± 27.90)	3.15%	0.14
TG	93.70 (± 21.86)	78.23 (± 22.89)	19.77%	< 0.001 *
HDL-C	45.33 (± 10.13)	50.36 (± 7.26)	11.09%	< 0.001 *
LDL-C	96.69 (± 27.48)	85.13 (± 28.04)	13.57%	0.0009*
FBS	100.76 (± 33.64)	91.17 (± 22.43)	10.51%	0.0076*
WHR	0.8572 (± 0.036)	0.8653 (± 0.039)	0.94%	0.08
BMI	20.75 (± 3.09)	22.77 (± 3.96)	9.73%	< 0.001 *
SBP	105.92 (± 9.70)	105.42 (± 9.00)	0.47%	0.667
DBP	70.88 (± 8.59)	70.42 (± 7.60)	0.65%	0.646

* Statistically significant based on p-values were obtained by independent t-test

To assess whether the observed metabolic differences were attributable to TB itself or to HIV co-infection given that 26.92% of TB patients were HIV-positive a one-way ANOVA was performed comparing key continuous metabolic parameters across three groups: One-way ANOVA demonstrated statistically significant differences in TG, HDL-C, LDL-C, FBS, and body mass BMI among TB-negative controls, TB+/HIV−, and TB+/HIV+ study groups (p < 0.05). Bonferroni post hoc analysis revealed significant differences in TG, HDL-C, and BMI between TB+/HIV− patients and TB-negative controls, as well as between TB+/HIV+ patients and TB-negative controls. In contrast, a statistically significant difference in LDL-C was observed only between TB+/HIV− patients and TB-negative controls. Although FBS showed an overall statistically significant difference in the ANOVA analysis, pairwise group comparisons using Bonferroni post hoc testing did not demonstrate statistically significant differences between individual groups. On the other hand, no statistically significant differences were observed between TB+/HIV + and TB+/HIV− groups for any of the assessed metabolic parameters (p > 0.05) (Table 6).

Table 6.

Biochemical and anthropometric parameters among TB+/HIV+, TB+/HIV − and TB-negative controls at Gondar town, northwest Ethiopia, 2024 (n = 260)

Parameter	TB + HIV+ (n = 35)	TB + HIV- (n = 95)	TB- Controls (n = 130)	P -Values
	Mean (SD)	Mean (SD)	Mean (SD)
TC	160.68 (± 31.50)	159.15 (± 23.87)	154.68 (± 27.90)	0.33
TG	91.68 (± 28.11)	94.45 (± 19.18)	78.23 (± 22.89) ab	< 0.001*
HDL- C	46.17 (± 12.66)	45.02 (± 9.08)	50.36 (± 7.26) ab	< 0.001*
LDL-C	92.28 (± 25.27)	98.31 (± 28.20)	85.13 (± 28.04) a	0.002*
FBS	101.94 (± 37.62)	100.33 (± 32.26)	91.17 (± 22.43)	0.026*
WHR	0.85 (± 0.039)	0.85 (± 0.034)	0.8653 (± 0.039)	0.137
BMI	20.98 (± 2.91)	20.67 (± 3.16)	22.77 (± 3.96) ab	< 0.001*
SBP	105.14 (± 8.53)	106.21 (± 10.12)	105.42 (± 9.00)	0.772
DBP	70.85 (± 8.53)	70.89 (± 8.65)	70.42 (± 7.60)	0.9

*Statistically significant in One way ANOVA

a-b Superscript letters represent statistical significance in post hoc Bonferroni post hoc test

a- p < 0.05 when TB + HIV- group compared to Controls

b- p < 0.05 when TB + HIV+ group compared to Controls

Magnitude of metabolic abnormalities

The prevalence of individual and composite metabolic abnormalities among TB patients and TB-negative controls is presented in Table 7. The prevalence of hypertension was 6.92% among TB patients and 3.85% among controls. Hyperglycemia was present in 18.46% of TB patients compared to 10.00% of controls. Dyslipidemia was the most prevalent metabolic abnormality in both groups, identified in 29.23% of TB patients and 20.00% of controls; low HDL-cholesterol was the most common lipid abnormality, present in 26.15% of TB patients and 14.62% of controls. Central obesity was less common among TB patients than controls (6.15% vs. 9.23%), consistent with the lower BMI and catabolic state associated with active TB. Overall, 63 (48.46%) TB patients had at least one metabolic abnormality compared to 57 (43.85%) of controls. (Table 7; Fig. 1).

Table 7.

Magnitude of metabolic abnormalities among TB patients and TB-negative controls at Gondar Town, Northwest Ethiopia, 2024 (n = 260)

Parameters	TB+ (n = 130)		Controls (n = 130)		Overall (n = 260)		P – Values
	n (%)	CI	n (%)	CI	n (%)	CI
Hypertension	9 (6.92%)	3.62–12.83	5 (3.85%)	1.59–8.96	14 (5.38%)	3.20–8.91	0.272
Hyperglycemia	24 (18.46%)	12.62–26.14	13 (10.00%)	5.86–16.53	37 (14.23%)	10.46–19.06	0.051
Higher -WC	8 (6.15%)	3.00 -11.88	12 (9.23%)	5.29–15.61	20 (7.69%)	5.00–11.65	0.352
Increased TC	6 (4.62%)	2.07–9.95	4 (3.08%)	1.14–7.97	10 (3.85%)	2.07–7.02	0.519
Increased TG	3 (2.31%)	0.73–9.92	-	-	3 (1.15%)	0.36–3.54	NA
Decreased HDL	34 (26.15%)	19.27–34.44	19 (14.62%)	9.48–21.85	53 (20.38%)	15.89–25.75	0.021*
Increased LDL	24 (18.46%)	12.65–26.14	10 (7.69%)	4.16–13.77	34 (13.08%)	9.47–17.78	0.010*
Dyslipidemias	38 (29.23%)	22.00–37.68	26 (20.00%)	13.94–27.83	64 (24.62%)	19.72–30.25	0.084
Met+	63 (48.46%)	39.92–57.08	57 (43.85%)	35.50–52.54	120 (46.15%)	40.14–52.28	0.455

“CI” 95% confidence interval “Met+” - having at least one metabolic abnormality

*Statistically significant (p < 0.05) on chi-square / Fisher exact

Fig. 1.

Magnitude of Metabolic Abnormalities among TB patients and TB-negative controls at Gondar town, northwest Ethiopia, 2024 (n = 260)

Chi-square analysis demonstrated statistically significant differences in the prevalence of decreased HDL-cholesterol and increased LDL-cholesterol between TB patients and TB-negative controls (p < 0.05). The prevalence of decreased HDL-cholesterol and increased LDL-cholesterol was higher among TB patients compared with TB-negative controls. In contrast, although the prevalence of hypertension, hyperglycemia, dyslipidemia, and central obesity was numerically higher among TB patients, these differences did not reach statistical significance (p > 0.05) (Table 7).

Factors associated with metabolic abnormalities

A logistic regression model was fitted to identify predictor variables associated with metabolic abnormalities. Bivariable logistic regression was performed on the pooled sample (n = 260) to identify candidate variables for multivariable analysis. In bivariable logistic analysis participants age group, sex, educational status, marital status, BMI category, previous history of TB diagnosis, smoking and alcohol drinking habit had p value less than 0.25, Thus was fitted in to multivariable logistic analysis. On multivariable logistic regression, four variables were independently and significantly associated with metabolic abnormalities (p < 0.05).

The odds of having Metabolic abnormalities were 2.53 higher among participants with age between 35 and 44, while participants with age between 45 and 54 had 6.26-fold higher odds for having metabolic abnormalities; participants with age ≥ 55 had 8.84 times higher odds for metabolic abnormalities. Female gender had almost twice higher odds for occurrence of metabolic abnormalities with (AOR 2.01, 95% CI 1.10–3.66). Conversely, heavy alcohol drinkers had five-fold higher odds for metabolic abnormalities (AOR = 5.03; 95% CI 1.84–13.72). Occupation was also significantly associated with metabolic abnormality in the multivariable model: farmers had significantly higher odds compared to government employees (AOR = 3.99, 95% CI: 1.24–12.81, p = 0.02). On the other hand, Current or past smokers appeared to have lower odds of metabolic abnormality compared to non-smokers (AOR = 0.21, 95% CI: 0.06–0.75, p = 0.01). However, this finding must not be interpreted as a genuine biological protective effect of smoking. All 22 smokers in this study were TB patients; no controls reported any smoking history. This complete structural imbalance in smoking distribution between groups means the estimate is entirely confounded with TB status and is affected by severe sparse-data bias, as evidenced by the very wide and unstable confidence interval (0.06–0.75). This finding is considered a methodological artifact and is biologically implausible given the established evidence that smoking increases metabolic risk through oxidative stress, endothelial dysfunction, and dyslipidemia (Table 8).

Table 8.

Factors associated with metabolic abnormalities among TB patients and non-TB controls at Gondar Town, Northwest Ethiopia, 2024 (n = 260)

Variables	Categories	Met Abnormalities		COR (95% CI)	AOR (95% CI)	p-value
		Yes	No
Age	18–34	63 (35.39%)	115 (64.61%)	ref	ref
	35–44	23 (56.10%)	18 (43.90%)	2.33 (1.17–4.64)	2.53 (1.12–5.71)	0.025*
	45–54	10 (71.43%)	4 (28.57%)	4.56 (1.37–15.14)	6.26 (1.49–26.17)	0.012*
	55/+	24 (88.89%)	3 (11.11%)	14.60 (4.23–50.40)	8.84 (2.16–36.17)	0.002*
Sex	Male	57 (40.71%)	83 (59.29%)	ref	ref
	Female	63 (52.50%)	57 (47.50%)	1.60 (0.98–2.63)	2.01 (1.10–3.66)	0.022*
Educational Status	No Formal Edu	14 (93.33%)	1 (6.67%)	18.37 (2.29–147.11)	6.48 (0.48–87.13)	0.15
	Primary	30 (46.15%)	35 (53.85%)	1.12 (0.57–2.19)	0.80 (0.36–1.76)	0.58
	High school	44 (41.51%)	62 (58.49%)	0.93 (0.51–1.69)	0.78 (0.39–1.55)	0.48
	Higher Edu	32 (43.24%)	42 (56.76%)	ref	ref
Marital status	Single	40 (36.70%)	69 (63.30%)	ref	ref
	Married	67 (52.76%)	60 (47.24%)	1.92 (1.14–3.24)	1.22 (0.63–2.36)	0.54
	Divorced/ Widowed	13 (54.17%)	11 (45.83%)	2.03 (0.83–4.97)	1.50 (0.49–4.56)	0.46
Occupation	Gov’t	8 (32.00%)	17 (68.00%)	ref	ref
	Farmer	32 (51.61%)	30 (48.39%)	2.26 (0.85–6.02)	3.99 (1.24–12.81)	0.02*
	Merchant	25 (39.06%)	39 (60.94%)	1.36 (0.51–3.62)	1.43 (0.48–4.30)	0.51
	Unemployed/ Retired	27 (52.94%)	24 (47.06%)	2.39 (0.87–6.52)	2.82 (0.89–8.95)	0.07
	Daily Labor	13 (38.24%)	21 (61.76%)	1.31 (0.44–3.90)	2.11 (0.62–7.19)	0.23
	Private	15 (62.50%)	9 (37.50%	3.54 (1.08–11.50)	3.90 (0.99–15.28)	0.05
Smoking Habit	Non-Smoker	115 (48.32%)	123 (51.68%)	ref	ref
	Smoker/ Quit	5 (22.73%)	17 (77.27%)	0.31 (0.11–0.88)	0.21 (0.06–0.75)	0.01*
Alcohol Drinking Habit	Non- drinker	55 (42.31%)	75 (57.69%)	ref	ref
	Occasional- drinker	35 (45.45%)	42 (54.55%)	1.13 (0.64–2.00)	1.08 (0.56–2.08)	0.81
	Moderate-drinker	10 (40.00%)	15 (60.00%)	0.90 (0.37–2.17)	0.75 (0.25–2.24)	0.61
	Heavy drinker	20 (71.43%)	8 (28.57%)	3.40 (1.39–8.30)	5.03 (1.84–13.72)	0.002*
Previous History of TB	No	113 (45.38%)	136 (54.62%)	ref	ref
	Yes	7 (63.64%)	4 (36.36%)	2.10 (0.60–7.37)	1.50 (0.31–7.29)	0.61

*P < 0.05 Statistically significant, ref= Reference category, COR= crude odd ratio, AOR= adjusted odd ratio

Discussion

This study assessed the magnitude and determinants of metabolic abnormalities among adult TB patients and individually age- and sex-matched TB-negative controls at three health facilities in Gondar Town, northwest Ethiopia. The overall prevalence of at least one metabolic abnormality was 48.46% among TB patients and 43.85% among controls, a difference that was numerically notable but did not reach statistical significance in unadjusted between-group comparison (p = 0.455). However, statistically significant between-group differences were observed for key continuous biochemical parameters including triglycerides, HDL-cholesterol, LDL-cholesterol, and fasting blood glucose providing robust objective evidence of subclinical metabolic dysregulation among TB patients even where categorical threshold-based prevalence differences were not statistically significant. Dyslipidemia, particularly low HDL-cholesterol, was the most prevalent metabolic abnormality in both groups, and was significantly more common among TB patients than controls for both decreased HDL-C (26.15% vs. 14.62%, p = 0.021) and increased LDL-C (18.46% vs. 7.69%, p = 0.010).

The overall prevalence of at least one metabolic abnormality among TB patients 48.46% (95% CI 39.92–57.08) This finding aligns closely with reports from Gabon [13] where 46.0% of TB patients exhibited metabolic abnormalities, and the Philippines [32], where the prevalence was 40.0%. However, this finding is higher than the prevalence reported in Indonesia, which was 35.5% [33]; This variation in prevalence rates across studies could be attributed to several factors; such as the differences in population demographics may influence the observed prevalence of metabolic abnormalities; Notably, our study simultaneously assessed four metabolic domains — diabetes mellitus, hypertension, dyslipidemia, and central obesity using a comparative design with matched controls, which is methodologically more comprehensive than many prior studies that assessed single outcomes without a comparison group.

The prevalence of hyperglycemia among tuberculosis patients in this study was found to be 18.46% (95% CI; 12.62–26.14); This is consistent with prior studies conducted Eastern Ethiopia, 13.5% [34], Addis Ababa, Ethiopia 15.8% [35], Tanzania 16.7% [36], Vietnam 13.7% [37] and Pakistan 14.8% [38]. The between-group difference in hyperglycemia prevalence (18.46% vs. 10.00%) was numerically substantial but did not reach statistical significance on formal chi-square testing (p = 0.051), though it approached the conventional threshold. This borderline finding may reflect limited statistical power for this outcome given the sample size, and warrants confirmation in larger studies. The higher prevalence among TB patients may be attributed to the bidirectional relationship between TB and diabetes, wherein TB induces an inflammatory state that worsens glycemic control [9]. Moreover, the chronic stress and metabolic derangements caused by TB might exacerbate hyperglycemia, particularly in individuals with pre-existing impaired glucose tolerance [9, 18].

However, the current finding was higher than from previous reports from Ethiopia 11.5% and 8.5% [39, 40], Eritrea 9.88% [41], Uganda 8.5% [42], Mozambique 1% [43], Ghana 9.4% [44], China 3.2% [45], and India 8.2% [46]. On the other hand was the reported prevalence was than reports from Taiwan 29.5% [47], Mexico 29.63 [48] and Brunei 33.9% [49]; a wide range of difference in the prevalence of DM in different studies might be due the variation in diagnostic criteria that could influence prevalence estimates; This study specifically employed fasting blood glucose for diagnosis diabetics. Moreover, the inclusion of pre-diabetic individuals could also elevate the reported prevalence in other studies. On the other hand, temporal changes may also explain these differences; the increasing global burden of diabetes, particularly in low- and middle-income countries, has been well-documented in recent years [50].

Hypertension was observed among 6.92% (95% CI: 3.62–12.83) of TB patients’ group and 3.85% (95% CI: 1.59–8.96) of TB-negative controls; The elevated prevalence of hypertension among TB patients could result from chronic inflammation, stress, and corticosteroid use during TB treatment [51, 52]. However the prevalence of hypertension among tuberculosis patients on the current study was lower than previous reports from Gabon 16% [13], Angola 19.4% [11], USA 48.9% [52], India 15.00% and 24.5% [51, 53]; differences in population characteristics, such as age, ethnicity, and comorbidities may account for this discrepancy, In addition, level of healthcare access might result in variation on hypertension prevalence.

Dyslipidemia was identified among 29.23% (95% CI: 22.00–37.68) Tuberculosis patients and 20.0% (95% CI: 13.94–27.83) of tuberculosis negative controls; These findings indicate a notable prevalence of dyslipidemia in both groups, though the rate is higher among TB patients. However, this prevalence is significantly lower than a recent report from Ethiopia [54] and Uganda [55]; The disparity in prevalence between the current study may be attributed to differences in study populations and study methodologies. Both the previous study from Ethiopia and Ugandan study selectively enrolled TB patients with multidrug-resistant tuberculosis (MDR-TB), a subgroup known to have more severe disease progression and metabolic disturbances, including dyslipidemia [56]. Additionally, the Ugandan study did not exclude individuals with comorbid metabolic conditions such as diabetes mellitus and HIV infection. Both diabetes and HIV are independently associated with an increased risk of dyslipidemia [57], which likely contributed to the higher prevalence reported in Uganda.

In the current study, low HDL-c was the most predominant form of dyslipidemia among tuberculosis patients; with 26.15% (95% CI: 19.27–34.33) of tuberculosis patients exhibited lower serum HDL level (< 40 mg/dl for male or < 50 mg/dl for women); This finding mirrors results from other studies conducted in Ethiopia [54, 58], Uganda [55]; HDL plays a crucial role in the reverse cholesterol transport system, shuttling cholesterol from peripheral tissues back to the liver for excretion, thereby preventing the formation of atherosclerotic plaques and further progression to Cardiovascular problems [59, 60]. In Mycobacterium tuberculosis infection, low HDL-c levels could exacerbate systemic inflammation, which high HDL-c has been shown to counter by suppressing tumor necrosis factor production [59, 61].

In the present study, TB patients exhibited and lower rates of central obesity compared to tuberculosis-negative controls (6.15% vs. 9.23%). This finding aligns with the well-documented relationship between TB and significant weight loss [62–65], which is often a hallmark of the disease. Tuberculosis is associated with a hypermetabolic state, systemic inflammation, and anorexia, all of which contribute to reduced body mass and fat distribution [62, 63]. Furthermore, the mean waist circumference (WC) and body mass index (BMI) were significantly lower among TB patients than in the non-TB control group. These results reflect the nutritional and metabolic challenges faced by TB patients, including chronic energy deficiency and muscle wasting. Lower WC measurements suggest a reduction in abdominal fat stores, consistent with the overall catabolic state induced by active TB [62, 65]. In contrast, the TB-negative controls group demonstrated relatively higher BMI and WC values, indicative of better nutritional status and fat reserves [66]. This disparity emphasizes the impact of TB on body composition and highlights the importance of nutritional interventions as a critical component of TB management. Addressing malnutrition in TB patients may not only improve their overall health outcomes but also enhance their response to anti-tuberculosis therapy [66, 67].

A critical analytical consideration in this study is the potential confounding effect of HIV co-infection, present in 26.92% of TB patients and absent in all controls. HIV infection and antiretroviral therapy are well-established drivers of dyslipidemia, insulin resistance, and metabolic syndrome through mechanisms including ART-induced lipodystrophy, mitochondrial toxicity, and direct viral effects on lipid metabolism [68, 69]. To address this, a three-group one-way ANOVA was performed comparing metabolic parameters across TB+/HIV+, TB+/HIV−, and TB-negative control subgroups. This analysis confirmed that TB+/HIV− patients showed significantly higher triglycerides, lower HDL-cholesterol, higher LDL-cholesterol, and lower BMI compared to controls while no significant differences were observed between TB+/HIV + and TB+/HIV patients for any parameter indicating that the observed metabolic dysregulation is attributable to TB itself, independent of HIV co-infection. Similarly, HIV/AIDS patients who are on ART regimen type particularly those regimens containing stavudine, efavirenz, or protease inhibitors is known to independently influence lipid profiles and glucose metabolism beyond the effect of HIV infection itself [70, 71].

In the present study, advancing age was identified as a significant risk factor for metabolic abnormalities among the study participants. Participants aged 35–44 years had 2.53 times higher odds of experiencing at least one metabolic abnormality compared to younger age groups (AOR 2.53, 95% CI: 1.12–5.71, p = 0.025). Furthermore, individuals in the 45–54 age category exhibited 6.26 times higher odds (AOR 6.26, 95% CI: 1.49–26.17, p = 0.012), while those aged 55 years and above had the highest risk, with 8.84 times greater odds (AOR 8.84, 95% CI: 2.16–36.17, p = 0.002). This pattern is consistent with previous studies conducted in Ethiopia [22], Eritrea [41], Angola [11], Gabon [13], and Pakistan [38]. The underlying biological mechanisms for this association may be attributed to age-related physiological changes, such as increased visceral adiposity, progressive insulin resistance, declining sex hormone levels, reduced physical activity, and sarcopenia cumulatively increase susceptibility to metabolic dysregulation [72]. Additionally, aging is often accompanied by cumulative exposure to various lifestyle and environmental risk factors, such as unhealthy diets, sedentary behavior, and chronic stress, which further exacerbate the risk of metabolic dysregulation [72, 73].

Female participants had approximately twice the odds of metabolic abnormality compared to male participants (AOR = 2.01, 95% CI: 1.10–3.66, p = 0.022). This finding aligns with previous research conducted in Ethiopia and the United States, which also reported a higher prevalence of metabolic abnormalities among women compared to men [16, 22]; This association may be attributed to several biological and sociocultural factors. Biologically, hormonal differences, particularly the effects of estrogen, play a critical role in lipid metabolism, glucose regulation, and body fat distribution. Women often experience increased susceptibility to metabolic abnormalities during hormonal transition periods such as menopause, where estrogen levels decline, leading to unfavorable changes in lipid profiles and increased visceral adiposity [74]. But this finding is in contrast to the study done in Nigeria [75]; likely reflecting differences in population age structure, hormonal status distribution, and lifestyle factors between settings.

An unusual finding in the current study was that participants who reported a current or past smoking habit appeared to have a significantly lower risk of metabolic abnormalities compared to non-smokers (AOR 0.21, 95% CI: 0.06–0.75, p = 0.01). However, this finding is biologically implausible and must not be interpreted as a genuine protective effect of smoking on metabolic health. Smoking is well established in the literature to increase metabolic risk through mechanisms including oxidative stress, endothelial dysfunction, dyslipidemia, and insulin resistance [76]. While counterintuitive, this result warrants careful consideration and further investigation to determine potential underlying explanations. The most probable explanation for this spurious finding is sparse data bias combined with complete confounding by TB status. All 22 smokers in this study were TB patients; not a single control reported any smoking history. This complete one-sided distribution of the smoking exposure means the regression model cannot produce a valid, generalizable estimate of the smoking-metabolic abnormality association across both groups. The resulting odds ratio is severely confounded by TB status and is highly unstable, as clearly evidenced by the very wide confidence interval (0.06–0.75), which spans nearly the entire range of plausible effect sizes. Within the TB patient subgroup, smokers may have been younger or at an earlier disease stage characteristics independently associated with lower metabolic risk producing a spurious apparent protective association in the pooled analysis. This finding is therefore a methodological artifact of sparse, imbalanced data and should not be cited as evidence of any protective biological mechanism. Future studies with larger, more balanced samples are needed to validly assess the smoking-metabolic relationship in TB populations.

The current study also revealed that heavy alcohol consumption was significantly associated with an increased risk of metabolic abnormalities. Specifically, heavy alcohol drinkers had five times higher odds of developing metabolic abnormalities compared to non-drinkers or moderate drinkers (AOR 5.03, 95% CI: 1.84–13.72, p = 0.002). This finding is consistent with previous reports from Switzerland, which highlighted a positive association between high-risk drinking and the prevalence of metabolic syndrome [77]. Chronic heavy alcohol consumption is known to disrupt normal metabolic processes by affecting the liver, a key organ in lipid and glucose metabolism. Alcohol induces hepatic steatosis, which can impair the liver’s ability to regulate lipid profiles, leading to hypertriglyceridemia and altered cholesterol levels [78, 79]. Additionally, heavy alcohol intake is associated with caloric excess, weight gain, and visceral adiposity, further compounding metabolic risk [80].

Farmers had significantly higher odds of metabolic abnormality compared to government employees in the multivariable model (AOR = 3.99, 95% CI: 1.24–12.81, p = 0.02). While farming is generally associated with high physical activity which would be expected to reduce metabolic risk farmers in this setting may face other metabolic risk factors including dietary monotony, limited healthcare access for early detection and management of metabolic conditions, and higher exposure to food insecurity and seasonal nutritional deficits.

Conclusion and recommendations

This study provides the first institution-based comparative estimate of metabolic abnormality burden among adult TB patients in Gondar Town, northwest Ethiopia, establishing that the overall prevalence of composite metabolic abnormality was 48.46% among TB patients and 43.85% among controls a difference that did not reach statistical significance (p = 0.455). Similarly, between-group differences in the categorical prevalence of individual conditions including hypertension, hyperglycemia, overall dyslipidemia, and central obesity did not reach statistical significance. However, TB patients exhibited statistically significant differences in key continuous biochemical parameters compared to controls including significantly higher mean triglycerides, LDL-cholesterol, and fasting blood glucose, and significantly lower mean HDL-cholesterol and BMI. At the categorical level, decreased HDL-cholesterol (26.15% vs. 14.62%, p = 0.021) and increased LDL-cholesterol (18.46% vs. 7.69%, p = 0.010) were significantly more prevalent among TB patients. These findings indicate subclinical lipid dysregulation in TB patients in this setting that may not be captured by composite prevalence estimates alone. Advancing age (≥ 45 years), female sex, and heavy alcohol consumption were independently associated with metabolic abnormalities, identifying high-risk subgroups for targeted screening. These locally grounded findings provide the quantitative and subgroup-specific evidence base needed to advocate for integration of metabolic monitoring. These findings emphasize the need for an integrated approach to managing both TB and metabolic conditions, especially in regions where both disease burdens are high.

TB patients particularly those aged 45 years and above, female patients, and heavy alcohol consumers should be prioritized for routine metabolic screening including fasting blood glucose and lipid profile assessment at TB diagnosis and at key treatment phase transitions. Screening should not be limited to diagnosing established metabolic conditions but should track continuous biochemical trajectories from treatment initiation, enabling early intervention before clinical thresholds are reached. Health facilities managing TB patients should integrate nutritional assessment and targeted dietary support into routine TB care, given the significantly lower BMI and higher malnutrition rates observed among TB patients. Restoration of nutritional status improves both TB treatment outcomes and metabolic health. Alcohol use screening and brief counseling interventions should be incorporated into TB clinic visits, given the five-fold increase in metabolic risk associated with heavy alcohol consumption identified in this study. The Ethiopian national TB program should consider piloting an integrated TB-NCD screening protocol at referral-level facilities in northwest Ethiopia, incorporating lipid profile and fasting blood glucose testing into the standard intensive and continuation phase monitoring schedule.

Future longitudinal studies are recommended to: elucidate the temporal relationship between TB treatment and metabolic recovery; assess the specific contribution of individual anti-TB drugs to metabolic dysregulation; collect ART regimen data among HIV-TB co-infected patients to disentangle ART-specific metabolic effects; and validate these findings in larger, multi-site samples across Ethiopia with adequate power to detect between-group differences in individual metabolic outcomes.

Strength and limitation of the study

Strengths of the study

This study has several notable strengths. First, it comprehensively assessed multiple metabolic abnormalities, including diabetes mellitus, hypertension, dyslipidemia, and central obesity, thereby providing a broader understanding of the relationship between tuberculosis and metabolic health. Second, the inclusion of both TB patients and apparently healthy controls enabled a comparative evaluation of metabolic abnormalities between the two groups. In addition, the multi-facility study design with proportional allocation of participants enhanced the representativeness and external validity of the findings. Furthermore, the study identified important demographic and behavioral factors associated with metabolic abnormalities, generating evidence that may support targeted preventive and clinical interventions among TB patients. The implementation of robust laboratory quality assurance procedures also strengthened the reliability and accuracy of the biochemical measurements obtained during the study.

Limitations of the study

Despite these strengths, the study has certain limitations that should be considered when interpreting the findings. The cross-sectional study design limits the ability to establish temporal or causal relationships between TB and metabolic abnormalities. Additionally, although the sample size was adequate for the primary outcome, it may have provided limited statistical power for detecting associations related to less common individual metabolic abnormalities. On the other hand, TB – negative controls were recruited from outpatient department attendees, hospital staff, and medical interns at the same health facilities as TB patients. This facility-based control recruitment strategy, while methodologically necessary for feasibility and 1:1 matching, introduces a potential source of selection bias community-based control recruitment with individual matching would provide a more representative comparison group and is recommended for future studies. The diagnosis of diabetes mellitus was based solely on fasting blood glucose measurements, which may have underestimated the true prevalence compared with more sensitive methods such as glycated hemoglobin (HbA1c) testing or oral glucose tolerance tests. Some behavioral and lifestyle-related information was collected through structured interviews, making the study susceptible to recall bias and inaccuracies in self-reported data. Moreover, the smoking variable was completely imbalanced between the study groups, as all participants with a smoking history were TB patients while none of the controls reported smoking, which may have affected the reliability of comparisons involving smoking-related associations.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AFB

Acid-Fast Bacilli

AOR

Adjusted Odd Ratio

BMI

Body Mass Index

cm

Centimeters

X²

Chi-Square

COR

Crude Odd Ratio

DM

Diabetes Mellitus

DBP

Diastolic Blood Pressure

DDS

Dietary Diversity Score

ETB

Ethiopian Birr

EPTB

Extra-Pulmonary Tuberculosis

FBG

Fasting Blood Glucose

HDL-C

High-Density Lipoprotein Cholesterol

HC

Hip Circumference

HIV/AIDS

Human Immunodeficiency Virus/ Acquired Immunodeficiency Syndrome

LDL-C

Low Density Lipoprotein Cholesterol

MUAC

Mid Upper Arm Circumference

ml

Milliliter

mg/dl

Milligram per deciliters

mmHg

Millimeters of Mercury

NCD

Non-Communicable Diseases

PTB

Pulmonary Tuberculosis

SBP

Systolic Blood Pressure

TC

Total Cholesterol

TG

Triglycerides

TB

Tuberculosis

WC

Waist Circumference

WHR

Waist to Hip Ratio

Author contributions

Elias Chane conceptualized and designed the study, developed the methodology, and wrote the initial manuscript draft. Eshet Gebrie and Bisrat Birke Teketlew conducted the statistical analysis, contributed to the interpretation of results, and reviewed the manuscript critically for intellectual content. Dereje Mengesha Berta and Muluken Walle were responsible for data collection, management, and validation, ensuring the integrity of the data. Abiy Ayele Angelo contributed to the literature review, data visualization, and supported data interpretation. Negesse Cherie provided critical inputs on the study design and assisted with drafting specific sections of the manuscript. Mebratu Tamir and Zufan Yiheyis contributed to administrative and technical support during data collection and provided feedback on the manuscript. Amare Mekuanint supervised the study, guided the overall direction, and provided substantial revisions to the manuscript. All authors read and approved the final manuscript and take responsibility for the accuracy and integrity of the work.

Funding

University of Gondar cover the funding resource for materials and laboratory analysis cost on this study.

Data availability

Most of data generated or analyzed during this study are included in this published article; and Addition research data set associated with a paper is available, can be accessed in contact with corresponding author.

Declarations

Ethics approval and consent to participate

The study was approved by the ethical clearance committee of Biomedical and laboratory Science of the University of Gondar. Each participant was informed in detail and his/her consent was obtained before the data collection.

Consent for publication

All participants provided written informed consent to publish this study.

Competing interests

The authors declare no competing interests.