High prevalence of multi-drug resistant organisms in the gut flora of healthy rural individuals in northern India

Material & Methods

This study was conducted at the department of Gastroenterology in collaboration with the departments of Community Medicine and Laboratory Medicine, All India Institute of Medical Sciences, New Delhi, after obtaining approval from the Institute Ethics Committee. Access to the village households was conducted through an existing tripartite MoU with the State Health Society, Haryana. Before undertaking the actual survey a formal permission was obtained from the head of the requisite village. A written informed consent was obtained from all the study participants before administering study procedures such as questionnaire and collection of stool samples.

Sample collection

Samples were collected from healthy individuals in Nuh district, Haryana, in northern India. The district has a population of 1.1 million, with predominant rural inhabitation (89%). The biggest block Nuh with 70 villages was identified of which 15 villages were randomly selected (Fig. 1). The median population size (range) of the selected villages was 2,203 (661-6,597), with median (range) households per villages as 414 (117-1,024), respectively. Random lanes from center of village was approached for sample, the household members were briefed about the study.

Close

Fig. 1.

Number of samples collected from different villages.

Export to PPT

Healthy adults aged 18-45 yr with no history of concurrent acute medical illness or hospitalization in past one year, symptoms pertaining to acute gastrointestinal (GI) disease, history of chronic illness, antibiotics/antifungals/antivirals/painkillers intake over the past three months, and other risk factors which could influence the gut microbiome were recruited (Supplementary Table I). A detailed questionnaire was administered (Supplementary Material). Stool samples were collected in sterile containers containing Copan E-swabs and were transported at 4℃ to the laboratory. The sample was processed for pathogen isolation, antimicrobial susceptibility testing (AST) and genetic profile of AMR genes.

Supplementary Table 1

Download

Supplementary Material

Download

Results

A total of 153 healthy volunteers were enrolled from 15 villages, the sample distribution is detailed in Figure 1. Mean age of participants was 32.5+8.6 yr with 58.2 per cent being females. Mean body mass index (BMI) was 22.2+4.8 kg/m². Among these 105 (68.6%) participants were vegetarian and 48 (31.4%) members consumed both vegetarian and non-vegetarian (mixed) diet. About 62.1 per cent of the participants were exposed to cattle (Table I) and the mean diet score was 5.2+4.3.

A total of 189 organisms were isolated from 153 participants, of these 123 (80.3%) participants were positive for only one organism, while the remaining were positive for >2 organisms (Table I). Village wise distribution of organisms is shown in Figure 2. Among 153 individuals, 137 had their culture positive for E. coli (89.5%). Among all organisms (n=189), the most frequent organism isolated was E. coli followed by K. pneumoniae and Enterobacter species Three participants who were positive for Shigella group bacteria were completely asymptomatic.

Close

Fig. 2.

Organisms isolated from different villages.

Export to PPT

Discussion

We report a high proportion of healthy individuals with fecal carriage of MDR organisms in this study. Healthy individuals from Nuh district were selected through an extensive questionnaire to rule out possible factors which could affect the AMR carriage, and only 31 per cent of the study participants had no fecal carriage for drug resistant organisms. Approximately 42 per cent individuals carried isolates with resistance against single/two drug classes, and 27 per cent carried MDRO. As per a Niti Aayog report¹⁸, Nuh is the most backward district in India with poor health and education infrastructure. We had previously shown that, among healthy individuals, the microbiome of rural individuals was healthier than urban¹². Though we did not have a comparative group of urban healthy adults, the prevalence of MDR in rural individuals was more than expected. Use of antibiotics in deprived areas like Nuh is increasing, possibly due to the increasing rate of irrational antibiotics prescription by informal practitioners in these villages that might be contacted first for simple illnesses. Although, human antibiotic use is the most important factor, other factors such as animal to human transfer (due to indiscriminate antibiotic use in animals and food industry), effect of diet, presence of AMR genes in waste and effluents could also be responsible for increasing MDR burden in healthy individuals¹⁹.

In a similar study from Chandigarh, India, overall prevalence of antibiotic resistance in healthy gut flora was found to be 70.5 per cent, similar to the phenotypic resistance in this study (69.3%). However, the prevalence of MDR was significantly high in this study, 27 per cent, as compared to only 2.4 per cent in Chandigarh⁸, which could be ascribed to differences in population (rural vs. semi-urban), exposure of residents to livestock (62% in this study), and antibiotic use patterns. In other rural-based studies from central India, 70 per cent of the MDR was seen in fecal flora in children of 1-3 yr age¹⁰. Similar heterogeneity has also been observed in other studies, and prevalence of MDR positivity varied from 51 per cent to as high as 93 per cent^9,10,20-23. Possible reasons for this heterogeneity could be the nature of participants, definition of MDR, method of detection, and similar reasons as described above. As an extension of the healthy gut microbiome study described above¹², the AMR profiling of five dominant commensals revealed that all the dominant gut bacteria were MDR, being resistant to at least seven different antimicrobials²⁰. Therefore, multi-drug resistance is also being reported in commensals, possibly due to their inadvertent exposure to antibiotics and toxic compounds.

The predictors of MDR organism carriage include higher socio-economic status, exposure to antibiotics or hospitalization over the past one year, and recent proton pump inhibitor and non-steroidal anti-inflammatory drug (NSAID) use^21-23. We selected individuals from the same geography having a similar lifestyle to control for confounders which can affect the fecal resistome. Among the possible predictors which were measured in this study age, gender, diet, and BMI did not have any influence. Diet is one of the major determinants of gut microbiome²⁴, although the effect of diet on gut resistome and AMR gene profile has been less explored^25,26. Possible reasons for lack of association between diet and AMR profile could be rather uniform diet in the same geography, and bias associated with recall method of diet assessment.

Approximately 44 per cent isolates were ESBL gene positive, and 14 per cent were positive for Carbapenems gene, similar proportions being observed among E. coli. In a recent meta-analysis, globally, 21.1 per cent of inpatients in healthcare setting, and 17.6 per cent healthy individuals had the presence of ESBL E. coli in fecal samples⁷. However, the prevalence in healthy Southeast Asians in this meta-analysis was 35.3 per cent, similar to our study. In the study from Chandigarh, prevalence of ESBL genes among cephalosporin resistant isolates was 1.9-25 per cent (different genes)⁸. In the present study, among 18 cephalosporin resistant isolates, 10 were ESBL positive. In another study of 115 healthy volunteers, ESBL-producing E. coli had a prevalence of 19 per cent, with CTX-M being the predominant type followed by TEM, unlike our study, where TEM was most common followed by CTX-M²⁷.

Among 40 MDR isolates, about 70 per cent were ESBL positive, while 1/3^rd of phenotypically sensitive isolates also carried ESBL genes. This phenotypic-genotypic discordance could be explained by the presence of silent/cryptic or proto resistance genes, which can encode AMR but may not be expressed or may undergo modifications after expression¹⁹.

Similar to other prevalence studies, E. coli was the most common organism isolated followed by Klebsiella species. Overall E. coli carriage among individual participants was 89.5 per cent (137/153), similar to that reported by Martinson et al²⁸, (⁓90%). Carriage rate for K. pneumoniae (12.4%), was also similar to the review by Podschum R et al²⁹ (K. pneumonaie fecal carriage in Indian population: 5-38%). Surprisingly, Shigella was detected in three individuals, though all were asymptomatic. Asymptomatic Shigella carriage has also been reported in other studies³⁰. Highest antibiotic resistance was seen for fluroquinolones followed by cotrimoxazole. Least resistant antibiotics were colistin, tigecycline and carbapenems. Like our study, 41.5 per cent fluoroquinolones resistance was seen in Chandigarh study, although unlike ours, maximum resistance was seen for cephalosporins. Carbapenem resistance was also similar between the two studies. Heterogeneity on this aspect has also been reported in other studies. In a study from rural Thailand, maximum resistance was seen against tetracycline, and fluoroquinolones resistance was seen in only 19.7 per cent²¹. Studies from Vietnam have reported high frequency of colistin resistance and mcr gene carriage, possibly due to significant colistin use in livestock^31,32. In the present study, only two isolates were colistin resistant and none carried mcr gene.

Though the study was strengthened by detailed dietary evaluation and estimation of phenotypic and genotypic resistance profile, but reported predictors of MDR and ESBL carriage in stool like history of antibiotic use and hospitalization beyond three months was not obtained. However, the study sample was well characterized with the questionnaire and we ruled out maximum factors that could affect the gut microbiome in short-term. The sample size was reasonable enough to characterize the AMR profile but a multicentric large cohort study should be conducted in order to evaluate the effect of diet and lifestyle on gut microbiome. Also, with advancement in the techniques, gut resistome through metagenomics approach should also be evaluated in further studies. A more detailed demographic data on socioeconomic status would also help us to understand the spread of resistance in healthy population.

To conclude we report that 27 per cent individuals from rural northern India are positive for MDR organisms in their fecal flora, with an ESBL carriage rate of ∼40 per cent. These findings provide important insights on burden of MDR in the community, and a roadmap for urgent antibiotic stewardship measures to control the imminent threat of community associated MDR infections.