INTRODUCTION

Tuberculosis (TB) affects 8.9 to 11.0 million people and is the cause of death of 1.2 to 1.4 million people globally ¹. Multidrug-resistant TB (MDR-TB) has a higher rate of treatment failure compared to sensitive TB ². MDR-TB is present in 3.5% of new TB cases and 18% of previously treated cases, which causes 230,000 deaths worldwide ³.

Peru is one of the countries from the Latin America and Caribbean (LAC) region with the highest rates of MDR-TB, with an incidence of 9.6 per 100,000 inhabitants, and about 1,205 incident cases per year ⁴. The incidence of MDR-TB has been unusually high in Peru for decades and the factors involved in the higher number of cases in Peru (1.5 times more MDR-TB cases than in other LAC countries) are unknown ⁵.

The emergence of MDR-TB is associated with the amplification of antibiotic resistance patterns through incomplete treatment and transmission in the most vulnerable populations ⁶. The factors that increase the risk of MDR-TB are family contact with MDR-TB diagnosis ⁷, co-infection with human immunodeficiency virus (HIV), heavy tobacco smoking, previous incarceration, history of TB ⁸, and type 2 diabetes mellitus (T2DM) ⁹.

On the other hand, the epidemic of overweight and obesity has doubled since 1980. In 2018, about 2 to 5 billion adults were overweight, and 30% of this population was obese ¹⁰. Like MDR-TB, Peru is one of the LAC countries with the highest overweight rate (38.8%), which is above the regional average (34.4%). In Peru, the prevalence of obesity has increased from 9% in 1975 to 13.8% in men and 23.3% in women in 2010; and up to 23.8% in people aged 30 to 59 years from 2013 to 2014 ¹¹.

The immune and the metabolic systems work in synergy because they evolved from common structures, with adipose tissue maintaining the same lineage as the immune cells ¹². Malnutrition, defined as a body mass index (BMI) under 16, is associated with sputum culture non-conversion within the first four months of MDR-TB treatment ¹³. In addition, a BMI under 18 is a risk factor for developing MDR-TB compared to patients with sensitive TB who have a BMI greater than or equal to 18 ¹⁴. However, evidence is scarce on whether BMI higher than 25 (overweight/obese) would alter the homeostasis of the metabolic immune axis ¹⁶.

The main factors affecting the tissue distribution of drugs are BMI, regional blood flow and the affinity of the drug for plasma proteins ¹⁷. The biomedical literature states that tissue blood flow decreases in obese individuals, and cytochrome P450-2E1 activity increases; however, the effect of overweight/obesity on glomerular filtration is unknown, which motivates the application of weight-standardized maintenance doses to correct for differences in drug clearance ¹⁸. This is relevant in drugs with a low threshold of pharmacokinetic variation such as rifampicin, one of the main anti-TB drugs where, as a rule of thumb, the dose increases in overweight/obese patients ¹⁹. Early findings showed that T2DM affected the pharmacokinetics of anti-TB drugs, that blood concentrations of isoniazid and rifampicin were lower in patients with TB and T2DM and that T2DM had an effect on the absorption rate and volume of distribution of rifampicin in patients with TB and T2DM ¹⁹; however, these statements were later questioned, as the real cause of all these effects was the increased weight of patients with T2DM and TB ²⁰.

Low plasma concentrations of isoniazid and rifampicin are associated with increased therapeutic failure and relapse; overweight/obesity affects the concentration-time curve for isoniazid ²¹. Evidence suggests that low plasma concentrations of drugs with high pharmacokinetic variability, such as rifampicin, are involved in the development of MDR-TB ²². Weight gain could alter metabolic rates, as occurs in the systemic clearance of other antimicrobials ²³, and could play a role in modifying plasma concentrations of low pharmacokinetic threshold drugs and be a key determinant for the development of MDR-TB²⁴. Therefore, this study aims to evaluate the association between overweight/obesity and multidrug resistance in patients with and without a history of TB treatment.

MATERIALS AND METHODS

RESULTS

We evaluated 4500 cases, of which 1149 had negative solid-media cultures for Mycobacterium tuberculosis at their baseline visit. The final study population consisted of 3351 culture-positive cases. The median age was 28 years with a range of 15 to 94 years, 62.3% were male, 5.2% had previous incarceration, 2.7% reported heavy social smoking, and 44.2% reported heavy social drinking.

Of the 3351 cases, 15.1% had a BMI greater than or equal to 25, 17.1% had a history of TB treatment, 3.5% had HIV coinfection, 5.6% reported having T2DM, and 33.6% had some type of resistance: monoresistance, polyresistance and resistance to isoniazid and rifampicin (MDR) (Table 1).

On the other hand, 15.9% were found to have MDR-TB. When comparing the proportions of covariates and MDR-TB, we found differences in age by category (p=0.007), previous incarceration (p=0.041), and social to heavy tobacco smoking (p=0.010). On the other hand, the difference between MDR-TB and overweight/obesity was not statistically significant (Table 2). We assessed the proportion of MDR-TB in patients with and without a history of TB treatment according to the demographic characteristics and medical history. In this regard, we found differences between not having a history of TB treatment and social/heavy tobacco smoking (p=0.016). We also found differences between having a history of TB treatment and age (p=0.004) and sex (p=0.001) (Table 3).

The multivariate analysis showed that MDR-TB was not associated with being underweight (aPR: 1.05; 95%CI: 0.74-1.49 in patients with treatment history; aPR: 0.87; 95%CI: 0.59-1.28 in patients with no treatment history) nor with overweight/obesity (aPR: 0.96; 95%CI: 0.68-1.38 in patients with treatment history; aPR: 0.88; 95%CI: 0.57-1.38 in patients with no treatment history).

DISCUSSION

Few studies explore the relationship between the TB epidemic and the overweight/obesity epidemic, particularly regarding MDR-TB; with the aim of implementing strategies to help eradicate TB. We assessed the association between overweight/obesity and MDR-TB in patients with and without a history of TB (cases with a history of TB and new TB cases). Our results show that there is no association between overweight/obesity and MDR-TB, i.e., overweight/obesity was not associated with MDR-TB in new patients without a history of TB treatment or in patients with a history of TB treatment.

Song et al. found an association between new MDR-TB cases and being overweight in data collected from 2004 to 2019 ²⁵, which differs from our results. Earlier studies on this subject reported that overweight/obesity was associated with low risk of active TB, although they did not evaluate MDR-TB ^27,28; their findings also are different from ours. The difference between our results and those from other studies could be due to genetic variations related to pharmacogenetics, which was proposed by Pasipanodya et al. ²⁴. This variability could explain the high rates of MDR-TB in the Peruvian territory, being one of the countries with the highest rates in the LAC region for several decades ⁵. Another possible for this difference is the type of overweight/obesity, Peru being the country with the highest percentage of energy derived from carbohydrate consumption (62.9%), unlike other countries in the region, where obesity is due to the high consumption of fats and proteins ²⁹.

Another possible cause may be the fact that overweight/obese women account for 5-6% of the population in the other studies, whereas in our study they represent 19-20% of the total sample. This agrees with the study by Woolcott et al., who conducted a cross-sectional analysis of 31,549 people from the National Center of Food and Nutrition of Peru and reported that the crude prevalence of obesity is 11.3% for men and 20.7% for women and varies according to the altitude of residence, which complicates overweight and obesity ³⁰. Besides, the aforementioned studies only evaluated new cases, whereas our study included old and new TB cases. Considering these subpopulations, we found that overweight/obesity would not modify the association between previous TB treatment and MDR-TB ^8,31.

Variables classically associated with MDR-TB such as HIV coinfection, T2DM and poverty did not have statistically significant differences in this study. It is possible that we did not find differences between HIV infection and MDR-TB because HIV is more frequent in specific populations that may have not been represented in our study. Other studies reported an association between HIV coinfection and MDR-TB, particularly primary MDR-TB, in populations with higher HIV rates, such as Africa or Asia ³². In our study, data regarding T2DM diagnosis was based on self-reports, and given that T2DM is an almost asymptomatic disease in its early stages, it is likely that it was not accurately diagnosed ³³, which may explain why we did not find differences between T2DM and MDR-TB. Additionally, no differences were found between poverty and MDR-TB, which may be due to the type of instrument we used ^34,35, given that the original study recruited participants from poor districts of Lima, it is possible that the number of non-poor people is too small to determine a statistically significant difference.

Our study has several important limitations. The cross-sectional design prevents us from establishing causality between the variables, and we could only determine their association. The BMI categories were based on single cut-off points for all age groups, without considering that for people aged 60 years or older, who made up 9.5% of the sample, the cut-off points could be different. The history of having received treatment for TB was self-reported during the anamnesis. This, together with the fact that there is no information available to confirm whether the patient was cured or not after the treatment, could cause a recall bias. In addition, T2DM diagnosis was also self-reported, which does not allow diagnostic confirmation and prevents establishing absolute glycemic values. Overweight/obesity changes over time, this condition may have changed between diagnosis and treatment. We did not collect information regarding weight before enrollment, nor during follow-up, this is an important limitation, since, this variable could increase with age. On the other hand, we found that only 2.3% of the participants were obese, which is why we added the overweight population (12.8%). This lack of cases with nutritional disorders could limit the statistical power to evaluate their association with MDR-TB. The most appropriate design would have been a large cohort during several years of observation and data collection, not only for BMI but also for other nutritional parameters such as waist-to-hip ratio.

We conclude that there is no association between overweight/obesity and anti-TB drug resistance in patients with a history of TB treatment.