Alcohol use is a significant risk factor for tuberculosis (TB), contributing to approximately 233,000 new TB cases in India in 2020 (WHO, 2020), representing 8.9% of all incident cases. Alcohol impairs the immune system, increases susceptibility to TB, and can negatively impact treatment outcomes due to altered pharmacokinetics of TB medications (Lonnrot et al., 2008; Rehm et al., 2009). Social mixing in congregate settings, particularly where alcohol is shared, significantly increases TB transmission risk (Classen et al., 1999; Diel et al., 2002). This study aimed to quantitatively analyze the alcohol-driven social mixing patterns of TB patients using a social network approach to better understand and address this understudied aspect of TB transmission in high-burden settings like Chennai, India.

Previous research has highlighted the association between alcohol consumption and increased risk of tuberculosis. Studies have indicated that alcohol use disorders increase the likelihood of developing active TB and re-infection, primarily due to its immunosuppressive effects and its contribution to alcohol-related diseases (Lonnrot et al., 2008). Furthermore, the impact of alcohol on treatment outcomes is also well documented, as alcohol can interfere with the pharmacokinetics of TB medications (Rehm et al., 2009). The significance of social mixing patterns, especially in settings involving alcohol sharing, has also been established, suggesting that these patterns may significantly increase the risk of transmission. However, there is a limited understanding of this issue from a quantitative perspective. Previous studies have alluded to the importance of alcohol sharing groups and drinking venues in amplifying TB transmission, but a systematic and quantitative analysis using social network methods has been lacking. This study addresses this gap by leveraging social network analysis to provide quantitative insights into the social mixing patterns of TB patients who consume alcohol.

This study was conducted in Chennai, India, a high TB burden area, between February 2018 and June 2019. 300 newly diagnosed adult pulmonary (drug-sensitive) TB patients who had resided in the study area for at least a year were recruited from 24 Designated Microscopy Centers (DMCs). An ego-centric social network survey was used to collect data on first-degree social contacts with whom the patients had consistent social interaction (at least three days a week for two to four hours) during the six months prior to TB diagnosis. The survey used a name generator method to identify contacts and alcohol sharing venues. Patients with regular alcohol use (RAU) were defined as those sharing alcohol for at least three days a week with their contacts in a neighborhood venue. Contacts were categorized as those with RAU or without RAU. Data collected included age, sex, occupation, TB status, and relationship type. Two networks were created: a 'Patient-contact' network and a 'Patient-contact-venue' network. Network analysis, including degree centrality, betweenness centrality, network density, and network components, was performed using Gephi software (Version 0.9.2). Descriptive statistics and confidence intervals were used to assess differences in TB proportions between contacts with and without RAU. Ethical approval was obtained from the Institutional Ethics Committee of NIRT, ICMR.

Of the 300 TB patients, 52 (17%) were identified as patients with regular alcohol use (RAU), who regularly shared alcohol with 106 (4%) of their first-degree contacts. Alcohol sharing occurred in 16 overcrowded and poorly ventilated venues. A significantly higher proportion of contacts with RAU were diagnosed with TB (12.3%; 95% CI: 6.6–20.00) compared to contacts without RAU (3.5%; 95% CI: 2.8–4.3). The 'Patient-contact' network (excluding venues) was less dense (density ratio 0.009) and less connected (mean degree centrality 1.3, mean betweenness centrality 0.5), indicating weaker transmission potential. In contrast, the 'Patient-contact-venue' network was denser (density ratio 0.018), more connected (mean degree centrality 3.1, mean betweenness centrality 154.2), and had a giant component linking 67.3% of patients with RAU, 38.4% of contacts with RAU and TB, and 72.3% of contacts with RAU but without TB. This indicates that the inclusion of alcohol venues as nodes significantly increases the apparent transmission potential of the network. The pooled TB transmission exposure for contacts with RAU increased tenfold when considered from a network perspective. Most of those with RAU were male friends, neighbors, or occupational contacts.

This study provides novel evidence, using social network analysis, on the role of alcohol sharing in amplifying TB transmission. The significantly higher TB prevalence among contacts with RAU compared to those without highlights the crucial role of alcohol-driven social mixing in TB spread. The network analysis reveals that incorporating venues into the network dramatically increases the observed transmission potential, indicating that venues serve as critical hubs for TB transmission among alcohol-consuming individuals. This has significant implications for TB contact tracing, which currently primarily focuses on household contacts in India. Our findings suggest a need to prioritize contact tracing among individuals who regularly share alcohol with TB patients and to consider alcohol sharing venues as high-risk locations for TB intervention strategies. The increased transmission exposure of contacts with RAU, as shown by the network analysis, underscores the potential effectiveness of focusing interventions on these high-risk venues.

This study demonstrates the utility of a social network approach in understanding the complex interplay between alcohol use, social mixing, and TB transmission. The findings emphasize the importance of considering alcohol sharing venues and contacts with RAU during TB contact tracing and intervention programs. Future research could explore interventions targeting these venues to reduce transmission, as well as investigate social mixing patterns in non-alcohol settings to provide a more comprehensive understanding of TB spread among this vulnerable population.

This study focused solely on alcohol-related social mixing and did not assess social mixing in non-alcohol settings. The limited ability to analyze social mixing in non-alcohol settings, which was sparse and mostly occurred in open public spaces, prevents a comparison with the closed and crowded settings where alcohol was shared. Further research is needed to comprehensively analyze the transmission dynamics of TB considering both alcohol and non-alcohol social mixing.