Diabetes and tuberculosis syndemic in India: A narrative review of facts, gaps in care and challenges

3.1 Epidemiology of diabetes and TB globally and in India

3.2 TB-DM comorbidity

There is a bidirectional relationship between DM and infections, and it is true for TB as well.⁵ It is estimated that 15% of adult TB cases are attributable to DM, which is nearly the same for the HIV-TB association.^{13, 19, 20} According to a World Health Organization (WHO) report, in 2020, an estimated 370 000 new cases of TB globally were related to DM. In 2019, just over 15% of people with TB were estimated to have diabetes globally, compared with 9.3% among the general adult population (aged 20–79 years). This equates to about 1.5 million people with TB and diabetes who required coordinated care and follow-up to optimize the management of both conditions.¹⁸ No authentic data on the overall number of patients with TB-DM comorbidity are available from India. TB cases in India were attributable to the following risk factors: undernutrition, alcohol use disorders, tobacco smoking, diabetes, and HIV.²¹

Furthermore, the International Diabetes Federation estimated that the number of cases of DM will increase by about 50% globally between 2019 and 2045, with a median increase of 99% (interquartile range [IQR]: 69%–151%) in high TB burden countries.^{22, 23}

3.3 Prediabetes and T2DM among patients with tuberculosis

In a systematic review, Gautam et al²³ showed that the prevalence of DM in cases with TB was variable in South Asia (Afghanistan, Bangladesh, Bhutan, Maldives, Nepal, India, Pakistan, and Sri Lanka). They reported a pooled prevalence of DM in TB cases of 21%, varying from 11% in Bangladesh to 24% in Sri Lanka (with India at 22%). New-onset hyperglycemia with TB was evaluated in a meta-analysis. Eleven studies were included yielding a total of 677 (27.3%). The pooled unresolved (not reverting back to normal glucose regulation, see section 3.5, “Transient Hyperglycemia in Patients with TB”) new cases of hyperglycemia at the end of follow-up were 50% (95% confidence interval [CI]: 36%–64%) and the total pooled burden of hyperglycemia at 3–6 months of follow-up was 11% (95% CI: 7%–16%), with both estimates displaying a high heterogeneity.²⁴ In a study from Kerala, routine screening of TB cases for DM using glycated hemoglobin (HbA1c) yielded a large number (44%) of DM cases.²⁵

There are several factors associated with TB in India. A household community-based tuberculosis disease survey conducted targeting 69 054 population from 43 villages of five blocks in a district of South India during 2015–2018 revealed that older age, male sex, low body mass index, diabetes, earlier history of TB, and alcohol use were significantly associated with TB.²⁶ Mave et al²⁷ screened 1793 participants and found 890 cases of microbiologically confirmed TB. The median age of the participants was 32 years; of these 66% were male. The prevalence of pre-DM and DM was 33% and 18%, respectively. Of the 162 with DM, 59% were known patients and 41% were recently diagnosed. Prevalence of DM among TB cases aged <25 years was 2%, in 25–40 years, 12%, and in patients >40 years, 45%. In this study, each percentage of increase in HbA1c (odds ratio [OR] 1.42, 95% CI 1.01–2.01) was associated with >1 + smear grade or ⩽9 days to TB detection.²⁸ There are regional differences in prevalence of DM in TB cases. In a cross-sectional analysis of the cohort data under Regional Prospective Observational Research for Tuberculosis-India Consortium, newly diagnosed TB patients (n = 1188) were recruited between 2014 and 2018. In this cohort from south India, the prevalence of DM among TB patients was very high, 39% (95% CI: 36.2% to 41.8%).²⁹ Interestingly, household contacts (HHCs) of TB patients have a high prevalence of DM. Of 652 adult HHCs, 175 (27%) had prediabetes and 64 (10%) had DM. Forty (64%) HHCs were newly diagnosed with DM and 48 (75%) had poor glycemic control (HbA1c > 7.0%).²⁶ This study emphasizes that all HHCs of TB patients should be screened for diabetes.

3.4 Tuberculosis in patients with T2DM

The prevalence of TB in patients with DM is variable. In a study in Pune, among 630 adults approached for screening for TB in patients with DM, 70.5% were having poorly controlled DM (HbA1c > 7%). Although 18% of participants reported any TB symptoms, none of these patients were diagnosed with culture-confirmed TB.³⁰ Data were analyzed from the National Family Health Survey (2015–2016), including 107 575 men (aged 15–54) and 677 292 women (aged 15–49).²⁸ These authors reported that among those who self-reported having DM, a total of 866 men and 405 women (per 100 000) also had TB. On the contrary, among those who self-reported not having DM the figures were 407 men and 241 women (per 100 000). Furthermore, these authors found that adults from poor families, with low body mass index, low levels of literacy, and who were unemployed had a higher risk of TB-DM comorbidity.³¹ It is not clear if mycobacterium TB mutates to cause more MDR-TB in patients with diabetes. In a nested cohort (n = 1633) studies of TB patients with and without DM in India, the authors found that phylogenetic reconstruction of all 76 paired Mtb strains showed genetically distinct isolates by participant origin, with 84% pairs acquiring at least one single-nucleotide polymorphism (SNP) at treatment failure or recurrence. The rate of SNP acquisition did not significantly differ by DM status at either treatment failure per 1 person-year (PY) for TB-only vs 1.2 per 1 PY for TB-DM. But there was a significant difference for recurrence per 1 PY for TB-only vs 1.27 per 1 PY for TB-DM. The authors concluded that considerable intrahost Mtb mutation rates were present at recurrence among patients with DM in India.³²

3.5 Transient hyperglycemia in patients with TB

It is also possible that some patients with TB have transient hyperglycemia that may revert back to normal glucose regulation. Gupte et al³³ found transient hyperglycemia was common in patients with TB, and HbA1c levels declined significantly during anti-TB treatment, irrespective of their glycemic status at treatment initiation (Figure 1). These authors suggested that a repeat HbA1c testing should be delayed for at least 3 months from TB treatment initiation to confirm a diagnosis of DM. Second, reversion of HbA1c levels to a nondiabetic range was common in participants with non-DM. These authors did not provide details of hyperglycemia treatment used by patients recruited in the study. Kornfeld et al³⁴ recently reported from south India that the incident TB results in transient glucose elevation but was not conclusively shown to cause chronic dysglycemia. Others have also shown a decreasing trend of hyperglycemia during anti-TB treatment.³⁵

3.6 Clinicoradiological presentation of TB in patients with diabetes

Clinicopathological presentation of TB differs in cases with DM vs those without DM (Table 2). Kuruva et al³⁶ showed that hemoptysis, weight loss, atypical lung lesions (lower lobe involvement), and cavitations in the lungs were more in those with TB and DM vs TB alone (Table 2). These authors also found lung consolidation to be more common in cases with pulmonary TB associated with DM (68.3%) compared to pulmonary TB without DM (60.3%). Similar findings were observed by Qazi et al.³⁷

3.7 Treatment outcomes of TB in patients with diabetes

It appears that the overall course of TB-DM patients is worse than those with TB alone but more data are required. Kuruva et al³⁶ showed that sputum conversion and cure rate are lower and treatment failure is higher in TB-DM patients vs TB alone (Table 2). Mave et al³⁸ also showed that early mortality was higher in participants with DM-TB patients TB alone (10% vs 7%).

In another study 124 patients (cases with diabetes, 68) were recruited. Mortality after therapy was 15% in cases and 7% in controls without diabetes; however, the difference was not statistically significant. Equal proportions in each group (9%) had persistent smear positivity at 2 months.³⁹

3.8 Pharmacological treatment of TB and DM in patients with TB-DM comorbidity

Management of DM in patients with TB should be aggressive to prevent failure of anti-TB treatment. The primary line of treatment in florid TB cases is insulin, but oral antihyperglycemic drugs could be continued with insulin, or even alone if the patient is mildly symptomatic.⁴⁰

The DM affects the effectiveness of anti-TB drugs through its metabolic effects by impairing the drug metabolism, altering the duodenal efflux pump P-glycoprotein and pH in the sites of absorption.⁴¹ Alfrasi et al⁴² showed that patients with TB-DM had lower concentrations of isoniazid than patients with TB alone, mostly related to higher weight and slower isoniazid absorption. Further, pyrazinamide concentrations were significantly lower in patients with TB-DM than in those without DM (21.6 vs 32.8 μg/mL), even after adjusting for weight. Finally, these authors also showed that rifampicin concentrations were not affected by DM or HbA1c values, with the C_max for rifampicin similar in cases with and without DM. Similarly, there were no differences in the rifampicin C_max by HbA1c category (<5.6%, 5.7%–6.5%, and >6.5%). These authors also stated that suboptimal blood concentrations of these important antitubercular drugs may be an important reason for lower sputum conversion and for treatment failures in patients with TB-DM vs those with TB without DM. Kumar et al showed a negative correlation between blood glucose and drug concentrations suggesting delayed absorption/faster elimination of isoniazid and pyrazinamide in the presence of elevated glucose.⁴³

Usually, physicians resort to insulin if a patient with TB-DM is suffering from weight loss and fever. Metformin is often stopped. This may not be appropriate given some data from Indian researchers. Mave et al³⁰ showed that metformin appeared to mitigate risk of failure of TB treatment and mortality and, therefore, may be important in the management of DM-associated TB. They have reported that the TB-DM patients who did not receive metformin in their study had twice the risk of all-cause mortality, and a six-time increased risk of death during TB treatment, compared with patients with TB alone. Moreover, metformin reduced the risk of recurrence among patients with TB-DM.³⁰

Further, metformin may have an effect on matrix metalloproteinases (MMPs), which are capable of degrading the pulmonary extracellular matrix. MMP is significantly higher in patients with TB-DM vs TB. Interestingly, known patients with DM on metformin treatment exhibited significantly lower levels of MMP.^{43, 44} Experience with other antihyperglycemic drugs (dipeptidyl peptidase- 4 inhibitors, sodium-glucose cotransporter-2 inhibitors, and glucagon-like peptide-1 receptor analogs) in patients with TB-DM is limited.

3.9 Cost of management of TB-DM comorbidity

Cost incurred by patients with DM is already substantial.⁴⁵ The research is sparse regarding the cost incurred by the patient if they have DM-TB. In a study done in the Bhavnagar region of western India, among the 304 patients with TB-DM comorbidity, 72% were male and the median (IQR) monthly family income was Indian rupees (INR) 9000 (8000–11 000) [~US$ 132 (118–162)]). The median (IQR) total costs due to combined TB-diabetes were INR 1314 (788–3170 [~US$ 19 (12–47)]), whereas those due to TB were INR 618 (378–1933 [~US$ 9 (6–28)]). The authors concluded that in addition to a marginal increase in the percentage of catastrophic costs, coexistent diabetes nearly doubled the median total costs incurred among patients with TB.⁴⁶

3.10 Integrated TB-DM programs: Gaps and challenges

Although, the Government of India has structured a program for TB control in India, knowledge and practices for diabetes control among primary healthcare workers and physicians may still need to be boosted. The Ministry of Health and Family Welfare in consultation with several strategic experts and stakeholders in TB control has framed a National Strategic Plan for TB 2017–2025.⁴⁷ This comprehensive plan is intended to contribute to the decisions of the policymakers at the national and state levels, development partners, and other stakeholders supporting the Revised National Tuberculosis Control Programme efforts to end TB in India.

Since the WHO initiated an innovative program of “Directly observed treatment, short course” or DOTS in the late 1990s,⁴⁶ there has been a steady increase in the use of this program for TB control services. Since 1995, it is estimated that around 41 million cases have been successfully treated and up to 6 million lives were saved through DOTS. Moreover, 5.8 million TB cases were notified through DOTS programs in 2009.⁴⁸ The WHO and many national TB programs continue to use DOTS as an important strategy for TB delivery for the fear of drug resistance. In 2006 due to several emerging new priority interventions, the WHO then launched the Stop TB Strategy. The new elements of the strategy were represented by targeted measures against MDR-TB and TB/HIV coinfection, with new attention given to the private sector (up to half of the TB cases are managed in the private sector in the Indian subcontinent), as well as patients and communities. The vision radically evolved in the End TB Strategy launched by WHO in 2014 and consisting of three pillars. The first pillar summarized the core concepts of the previous strategies related to diagnosis and treatment of TB, while introducing for the first time the concept of prevention, mainly based on the identification and treatment of individuals with latent TB infection.⁴⁹

The key components of the strategy of the Government of India include surveillance of the joint burden of TB and people with DM, and monitoring and evaluation of collaborative TB and DM activities. Routine screening of TB in patients with DM and DM in patients with TB (“bidirectional screening”) provides a unique opportunity for its early diagnosis and better management. This has been outlined in a document from the Government of India.⁵⁰

According to a recent study, gestational diabetes screening should also be prioritized in tuberculosis-endemic countries, especially in women living with HIV.⁵¹

Although the TB control part of the program has been in operation for several years and is widely ingrained within health workers, screening and control of diabetes may not be optimal. Screening for TB among DM patients was not implemented, despite documents indicating that it had been. A mixed-methods study with quantitative (cohort study involving record reviews of patients registered between November 2016 and April 2017) and qualitative (interviews of patients, healthcare providers, and key district-level staff) components was done in a district of a northern Indian state. Of 562 TB patients, only 137 (24%) were screened for DM. These authors concluded that low patient awareness, poor knowledge of guidelines among healthcare providers, lack of staff, and inadequate training were barriers to screening.⁵² Gurukartick et al⁵³ in operational research involving cases with TB-DM from South India found that the fasting blood glucose was not recorded consistently as per the program guidelines of monitoring at baseline, at the end of intensive, and of continuation phase of TB treatment. These authors also opined that the random blood glucose (RBG) estimation, as done in many cases, is not a reliable indicator of a glycemic control status to analyze the effect on TB treatment outcomes. Second, in the same study, it was noted that a substantial number of patients did not have optimal glycemic control at baseline and during TB treatment but parameters indicating control were not optimal and correct. Specifically, these conclusions are also questionable because it is based on RBG data.

Fazludeen et al¹³ used an outlier case study approach and conducted stakeholder interviews and focus group discussions with relevant program personnel including field staff and program managers of TB and DM control programs as well as officials of partners in Indian states, Kerala (south) and Bihar (north). Several problems were identified (Table 3). Healthcare workers were unable to devote much time to a single program (in this case concerning TB-DM) because they were part of multiple national programs. In addition, there is a major labor and resource shortage. The disease control programs need to generate more data for better decision-making and close the skill gaps to improve integrated TB-DM management.¹³ These authors have suggested decentralization of care and strengthening district-level coordination (Table 3).

But carefully programmed integrated management of TB-DM is achievable by non-physician health workers, even in the low resource setting of rural India. In a pilot study (a randomized controlled trial design with mixed-methods evaluation), conducted in the Guntur district of Andhra Pradesh (South India), 120 newly diagnosed patients with TB and screened for DM were evaluated. Nonphysician health workers were trained to encourage adherence to treatment and monitor treatment response including blood glucose levels and provide lifestyle advice and were considered as intervention arm. Although blood glucose levels in both arms of trial were similar, awareness about DM and TB comorbidity and cardiovascular risk increased in the intervention arm of the study.⁵⁴ Overall, this intevention strategy was well accepted by the patients and healthcare providers. These authors concluded that with appropriate training, availability of infrastructure, and planned intervention implementation, it is feasible to co-manage TB-DM within the existing primary healthcare system in India.⁵⁴

Overall, the findings show that although WHO first recommended collaborative activities to address TB and DM in 2011, implementation in India is variable and inadequate in many states and regions. It has been strengthened after a National Framework for Joint TB-Diabetes Collaborative Activities was laid down. This detailed document lists methods of screening, recording, and reporting of TB-DM comorbidity at all levels of medical care. In addition, method of blood glucose testing and forms for proper reporting of diabetes in TB centers and TB in diabetes and general medical facilities have been provided. Administrative set-up and guidance for education and training have been clearly laid down.⁵⁵ However, ground-level research, studies involving large samples, and prospective cohort data are needed. There are some data available regarding bidirectional screening¹³ and management⁵⁶ but more studies are required. Monitoring and evaluation of the joint response will be critical for driving the scale-up and for assessing the impact of TB and diabetes collaborative activities. The challenges lie in strengthening the individual programs and envisioning and implementing a horizontal-vertical integration. The wide network of public health facilities in India offers clear opportunities to improve service delivery.

3.11 Impact of COVID-19 pandemic on DM and TB

During the COVID-19 pandemic, several studies indicated that patients with DM had increased severity of COVID-19 and also post-COVID-19 syndrome.⁵⁶ It is perhaps due to the effect of DM on the viral entry into the cell and inflammatory response to the infection or the effect of corticosteroid therapy.¹³ Patients with known DM had uncontrolled glycemia, and in addition, new-onset DM were also seen.^{4, 57} Such patients with new-onset DM add to the existing large numbers of DM patients in India. Further, uncontrolled glycemia would increase the tendency for contracting TB.

It is reported that the history of TB (active or latent) increases the risk and susceptibility for SARS-CoV-2 infection. Furthermore, it causes more severe clinical manifestations of the disease and its progression and is associated with poor outcomes. Specifically, TB increases 2.1-fold risk of severe COVID-19 disease.⁵⁸ Because diabetes also increases risk of morbidity and mortality in patients with COVID-19,^{59, 60} it is reasonable to suggest a bidirectional TB-COVID-19 screening, in addition to screening for diabetes, should be done, that is, COVID-19 screening (depending on clinical symptoms and signs) for TB cases at diagnosis and TB screening for all COVID-19 cases.