Author ORCID Identifier

Date of Award

Summer 8-10-2021

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Public Health

First Advisor

Richard B. Rothenberg, MD, MPH, FACP

Second Advisor

Matthew J. Magee, PhD, MPH

Third Advisor

Russell R. Kempker, MD, MSc

Fourth Advisor

Katherine E. Masyn, PhD, MA


The global strategy to end tuberculosis (TB) includes interim goals to reduce TB incidence by 80% and TB mortality by 90% by the end of 2030. However, global TB control primarily focuses on achieving favorable TB treatment outcomes (i.e., microbial cure or treatment completion). Yet clinical and public health understanding of the long-term impacts of TB on post-TB health is severely limited. Emerging evidence suggests that TB patients may have increased post-TB risk of non-communicable diseases such as chronic obstructive pulmonary disease, type-2 diabetes mellitus (T2DM), cardiovascular disease, or stroke. Compared to general population mortality rates, patients previously treated for TB have approximately double the rate of all-cause mortality. However, guidance on how to improve patients’ post-TB health is unavailable. Furthermore, the impact of comorbidity factors such as pre-existing T2DM, human immunodeficiency virus (HIV), and hepatitis C co-infections on post-TB health has not been assessed. To evaluate the relationship between comorbidities and post-TB treatment health, we conducted three observational studies (two retrospective and one prospective) in the country of Georgia.

In Study 1, we estimated the association between common comorbidities (T2DM, HIV, and hepatitis C co-infections) and post-TB mortality among patients with drug-resistant TB (DRTB) between 2009-2017 (n=1,032). Competing risks models were used to estimate the hazard rate of all-cause mortality post-TB treatment comparing DRTB patients with and without comorbidities. We reported a strong association between HIV co-infection and post-TB treatment mortality (adjusted hazard ratio [aHR] 4.40, 95% confidence interval [CI] 2.17 – 8.93). The hazard rate of post-TB mortality was non-significantly higher among patients with hyperglycemia (aHR 1.19, 95%CI 0.73 – 1.96) or hepatitis C co-infection (aHR 1.25, 95%CI 0.71 – 1.96) compared to those without hyperglycemia or hepatitis C.

In Study 2, we estimated the association between the use blood glucose-lowering agents with TB treatment outcomes among DRTB patients with hyperglycemia (n=128). Hyperglycemia was determined according to fasting blood glucose level at TB treatment initiation or a record of self-reported previous T2DM diagnosis from a health care provider. We evaluated TB treatment outcomes using three metrics, including 1) time to sputum culture conversion, 2) final TB treatment outcome, and c) mortality post-TB treatment. We used log-binomial and Cox proportional hazard regression models to determine the association between T2DM characteristics and the study outcomes. Among DRTB patients with T2DM with a record of blood glucose-lowering agents use (n=60), metformin use was associated with a reduced risk of poor TB treatment outcomes (adjusted risk ratio [aRR] 0.23, 95%CI 0.05 – 0.97).

In Study 3, we prospectively followed patients who were successfully treated for TB to evaluate cardio-metabolic risks post-TB treatment. Newly diagnosed adult pulmonary TB patients with a favorable outcome at the completion of TB treatment were eligible (n=105). For this study, interim data from study baseline and 6-month follow-up were used. We used log-binomial regression to estimate the risk ratio of metabolic syndrome at the end of TB treatment, comparing patients treated for drug-susceptible (DSTB) vs. DRTB. We also used general linear models to estimate the association between drug-resistant TB and visceral adipose index (VAI) at the end of TB treatment. Among a subset of individuals with 6-month follow-up information available (n=62), mixed models with random intercepts were used to estimate the association between drug-resistance and change in VAI. Compared to patients treated for DSTB, patients treated for DRTB had a higher prevalence of metabolic syndrome (aRR 2.29, 95%CI 0.50 – 10.37) and elevated VAI levels (adjusted mean difference 0.37, 95%CI -0.13 – 0.86) at the end of TB treatment. Although non-significant, the result from the mixed model suggested that the mean VAI among patients formerly treated for DRTB was higher by 0.60 (95%CI -0.20 – 1.40) points compared to those treated for DSTB.

Our preliminary findings suggest a clinical need for post-TB continued care even when patients have a successful TB treatment outcome. Importantly, patients with pre-existing comorbidities and DRTB, who may have increased risks of metabolic diseases or mortality post-TB treatment, should receive targeted follow-up care after TB treatment completion.


File Upload Confirmation


Available for download on Wednesday, July 31, 2024