Role of risk factors and socio-economic status in pulmonary tuberculosis: a search for the root cause in patients in a tertiary care hospital, South India
Corresponding Author Chiranjay Mukhopadhyay, Department of Microbiology, Kasturba Medical College, Manipal University, Manipal 576104, India. Tel.: +91 820 2571201 Exn. 22322; Fax: +91 820 2571927; E-mail: email@example.com
Objective To determine the frequency of underlying risk factors and the socio-economic impact based on occupation in the development of tuberculosis.
Method Retrospective analysis of 207 clinically and microbiologically diagnosed patients with pulmonary tuberculosis (PTB) admitted to Kasturba Hospital in 2005 and 2006. Demographic details and underlying risk factors were statistically evaluated.
Results Diabetes mellitus (DM) (30.9%) was the most prevalent condition and significantly more common than other risk factors like smoking (16.9%), alcoholism (12.6%), HIV (10.6%), malignancy (5.8%), chronic liver diseases (3.9%), history of contact with TB (3.4%), chronic corticosteroid therapy (2.9%), chronic kidney diseases and malnourishment (1.5%). There were 82 patients (39.6%) with no underlying risk factor. Men (M:F = 3.7:1) and patients older than 40 years had a higher incidence of co-existing conditions. PTB was significantly more common in blue-collar (44%) and white-collar (27.1%) workers than household workers (12.1%), students (10.6%) and retired/unemployed people (6.3%).
Conclusion Pulmonary tuberculosis had a significant impact and predominated in male patients co-existing with DM. Patients with DM and suggestive pulmonary symptoms should be screened for tuberculosis. More stringent health education and awareness programme should be implemented at the grass root level.
Objectif: Déterminer la fréquence des facteurs de risque sous-jacents et l’impact socio-économique sur base de l’occupation lors du développement de la tuberculose.
Méthode: Analyse rétrospective de 207 cas de tuberculose pulmonaire (PTB) diagnostiqués cliniquement et microbiologiquement et admis à l’Hôpital de Kasturba en 2005 et 2006. Les données démographiques et les facteurs de risque sous-jacents ont étéévalués statistiquement.
Résultats: Le diabète sucré (30,9%) était la condition la plus prévalente et significativement plus fréquente que d’autres facteurs de risque comme le tabagisme (16,9%), l’alcoolisme (12,6%), le VIH (10,6%), le cancer (5,8%), les maladies chroniques du foie (3,9%), l’histoire d’un contact avec la tuberculose (3,4%), la corticothérapie chronique (2,9%) et les maladies rénales chroniques et la malnutrition (1,5%). Il y avait 82 patients (39,6%) ne présentant aucun facteur de risque sous-jacents. Les hommes (M:F = 3,7:1) et les patients âgés de plus de 40 ans avaient une incidence plus élevée pour des conditions coexistantes. La PTB était significativement plus fréquente chez les ouvriers (44%) et les agents de bureaux (27,1%) que chez les travailleurs à domicile (12,1%), les étudiants (10,6%) et les retraités/chômeurs (6,3%).
Conclusion: La PTB a un impact significatif et prédominait chez les hommes, coexistant avec le diabète sucré. Les patients atteints de diabète sucré et suggestifs des symptômes pulmonaires doivent être testés pour la tuberculose. Une éducation plus stricte à la santé et des programmes de sensibilisation devraient être implémentés dès le plus bas niveau.
Objetivo: Determinar la frecuencia de los factores de riesgo subyacentes y del impacto socio-económico, según la ocupación, en el desarrollo de la tuberculosis.
Método: Análisis retrospectivo de 207 casos de tuberculosis pulmonar (TBP), con diagnóstico clínico y microbiológico, admitidos en el Hospital de Kasturba entre el 2005–2006. Los detalles demográficos y los factores de riesgo subyacentes se evaluaron estadísticamente.
Resultados: La diabetes mellitus (30.9%) era la condición más prevalente y significativamente más común, comparada con otros factores de riesgo tales como el fumar (16.9%), el alcoholismo (12.6%), el VIH (10.6%), los tumores malignos (5.8%), las enfermedades hepáticas crónicas (3.9%), una historia de contacto con TB (3.4%), la terapia crónica con corticoesteroides (2.9%) y la enfermedad renal crónica y la desnutrición (1.5%). Se hallaron 82 pacientes (39.6%) que no presentaban ningún factor de riesgo subyacente. Los hombres (H: M = 3.7:1) y los pacientes mayores de 40 años tenían una mayor incidencia de condiciones coexistentes. La TBP era significativamente más común entre obreros (44%) y empleados (27.1%) que en trabajadores del hogar (12.1%), estudiantes (10.6%) y personas retiradas/sin empleo (6.3%).
Conclusión: La TBP tiene un impacto significativo y predominante en pacientes masculinos con diabetes mellitus. A los pacientes con diabetes mellitus y síntomas pulmonares sugerentes se les debería realizar las pruebas para descartar una tuberculosis. A nivel local deberían implementarse una educación sanitaria más rigurosa y programas de concienciación.
Tuberculosis (TB) is a chronic pulmonary disease causing high morbidity and mortality since ancient times. According to WHO (2007), one-third of world’s population is currently infected with TB and every second someone is newly infected. In India, there are around 1.8 million new cases of TB annually (Steinbrook 2007). With the emergence of Human Immunodeficiency Virus (HIV) pandemic, there has been a resurgence of TB infection. TB is the most common HIV-related opportunistic infection (Decker & Lazarus 2000; Devi et al. 2005; Steinbrook 2007), and HIV is considered the most powerful risk factor for the progression of TB infection to disease as it weakens the cell-mediated immunity and macrophage function (Decker & Lazarus 2000).
Although more focus is being given on TB–HIV co-infection as a major concern for public health, there are other underlying risk factors which compromise the immune status. Diabetes mellitus (DM) widely impairs neutrophil and macrophage functions (Ljubic et al. 2004) and thus can be a major aggravating risk factor for TB. There are several studies indicating a higher prevalence of TB in patients with chronic corticosteroid therapy, malignancies (Pal et al. 2002; Kamboj & Sepkowitz 2006; Golsha et al. 2009), chronic kidney disease (CKD) (Al-Shoaib 2000; Venkata et al. 2007) and chronic liver disease (CLD) (Kim et al. 2009). Despite the evolution of new risk factors, close contact to smear-positive pulmonary tuberculosis (PTB) patients, malnourishment and poverty are the more likely predisposing factors in developing countries for acquiring PTB (Hernandez-Garduno & Perez-Guzman 2007). Alcoholism and smoking are also closely associated with the development of PTB (Khan 2006; Lönnroth et al. 2008).
Very few studies have been carried out in India to determine the major predisposing factors and the impact of socio-economic status. This retrospective study was planned to document the probable underlying risk factors in the development of TB and correlate them with the occupation of the patients and to assess the socio-economic impact of microbiologically confirmed cases of pulmonary TB.
Materials and methods
The medical records of 207 clinically and microbiologically diagnosed (based on smear microscopy and/or culture) cases of PTB admitted to Kasturba Hospital, Manipal, South India, in 2005 and 2006 were studied to evaluate the frequency distribution of common underlying risk factors such as HIV infection, DM, malignancy, history of contact to a smear-positive PTB patient, chronic corticosteroid therapy, malnourishment, smoking, alcoholism, CKD and CLD along with other demographic details like age, sex and the occupation of the patients. The patients were categorized into four age groups: <20, 21–40, 41–60 and >60 years. We considered five occupational groups namely blue-collar workers (agricultural, non-agricultural and skilled labourers); white-collar workers (salaried service personnel or professionals, self-employed businessmen, drivers and hotel workers); retired and unemployed people; household workers and students.
Statistical analysis of underlying risk factors was carried out by chi-square test and Fisher’s exact test using GraphPad InStat 3.0 software. Occupational impact was analysed by one-tail hypothesis testing using t-statistics.
Of 207 patients studied in 2 years (2005–2006), 163 were men (78.74%) and 44 women (M:F ratio = 3.7:1). PTB was most prevalent in the age group 21–40 years (n = 86), followed by 41–60 years (n = 79), >60 years (n = 24) and <20 years (n = 18). Eighty-two patients (39.61%) presented with no significant history of any underlying conditions, and 125 patients (60.39%) had at least one underlying risk factor with a relative risk of 1.52 (95% CI = 1.25–1.86) and P-value < 0.0001 (χ2 = 17.043). DM (30.92%) was the most common and significant co-morbid condition, followed by HIV infection (10.63%) (χ2 = 24.672; P-value < 0.0001), smoking (16.9%) (χ2 = 10.408; P-value < 0.0013), alcoholism (12.6%) (χ2 = 19.436; P-value < 0.0001), malignancy (5.8%) (χ2 = 41.919; P-value < 0.0001), CLD (3.9%) (χ2 = 50.859; P-value < 0.0001), history of recent close contact with a smear-positive TB patient (3.4%) (χ2 = 53.312; P-value < 0.0001), chronic corticosteroid therapy (2.9%) (χ2 = 55.859; P-value < 0.0001), malnourishment and CKD (1.5%) (χ2 = 64.106; P-value < 0.0001).
Of the 125 patients presenting with potential risk factors, 85 (68%) patients presented with a single risk factor, while 40 patients (32%) in the age group above 20 years had two or more associated risk factors.
HIV infection as a predisposing factor was more common in the age group 21–40 years (48.3%) than 41–60 years (13.6%), whereas no HIV-infected patients were found in the other two age groups. DM was found more common with increasing age above 40 years (54.4%), as were smoking (27.2%), alcoholism (18.4%) and malignancy (9.7%).
Patients older than 40 years had a higher incidence of co-existing conditions (81.6%): DM (54.4%), smoking (27.2%), alcoholism (18.4%), malignancy (9.7%), HIV (7.8%), CLD (4.9%), CKD (2.9%) and chronic corticosteroid therapy (0.09%). Although among women, the percentage of diabetic patients was greater in the age range above 40 years (18.2%), men with (56.0%) or without (34.8%) underlying predisposing factors were more prone to have PTB than women. Table 1 shows the frequency distribution of patients with PTB based on underlying risk factors.
Table 1. Underlying risk factors in patients with pulmonary tuberculosis along with age and sex distribution
|H/O contact||2||1||3||1||0||0||0||0||7 (3.38)|
|Allergy/asthma on corticosteroids||1||0||0||1||3||1||0||0||6 (2.9)|
Of the 207 patients with pulmonary TB, 44% were labourers, 27.1% were white-collar workers, 12.1% were household workers, 10.6% were students and 6.3% were retired or unemployed. Prevalence of PTB was significantly more common among labourers (t-value = 17.86; P-value < 0.0001) and white-collar workers (t-value = 17.58; P-value < 0.0001). Statistically, labourers >61 years are most vulnerable to pulmonary TB. Table 2 shows the occupational distribution of the patients with PTB. We observed a 36.4% prevalence of DM among labourers (n = 33), which was not significantly different to 61.5% among white-collar workers (n = 8), 28% among household workers (n = 7) and 26.8% among retired and unemployed patients (n = 15) with PTB.
Table 2. Occupation chart of patients with pulmonary tuberculosis along with age and sex distribution
In this descriptive retrospective study, we have studied the frequency distribution of disease and underlying risk factors followed by TB in southern Karnataka. We found that DM was the most common co-morbid condition followed by smoking, alcoholism and then HIV, although a rising trend of HIV in TB has been observed by Rajasekaran et al. (2000). The global burden of DM is increasing, and it is possible that areas with higher prevalence of DM could have a greater impact on TB than HIV. There were other conditions like smoking, alcoholism, malignancy, CLD, CKD, chronic corticosteroid therapy (signifies immunosuppressed conditions), history of close contact to patients with TB (signifies crowding and lack of health education) and malnourishment (signifies poverty as well as lack of health education).
Yamagishi et al. (1996) stated that the onset of tuberculosis is more frequent in DM patients than in individuals with any other underlying disease. Even a systematic review of 13 observational studies from various parts of the globe suggests that DM was associated with an increased risk of tuberculosis regardless of study design and population (Jeon & Murray 2008). From India, Suryakirani et al. (1998) reported 30% smear-positive cases among DM patients. In South India, a matched case–control study found DM to be a significant risk factor (Shetty et al. 2006). In our study, 30.92% patients with PTB had DM as a co-morbid condition, which was very similar to the observation by Golsha et al. (2009) in Iran (23.05%). Moreover, our finding that tuberculosis has a significant impact in men co-existing with DM was supported by a study conducted by Nissapatorn et al. (2005) in Malaysia.
Data on concomitant TB and DM are scarce in India, and we need more studies to assess this. Although more importance is being given on TB–HIV co-infection in developing countries, our study emphasizes that other underlying risk factors cannot be ignored as DM was the most common underlying risk factor for the development of TB. WHO has projected that the global figure of people with diabetes will be 300 million in 2025, of which the greatest increase will occur in India, an estimated 57 million cases (Kant 2003).
It has long been evident that there is an association between smoking, alcoholism and risk of developing pulmonary TB. Lönnroth et al.(2008) in a systematic review established that risk of active TB is substantially elevated in alcoholics, which might be because of direct toxic effect of alcohol on immune system or indirectly through macro and micro nutrient deficiency. Similar is in the case of smoking because of toxic effect of nicotine. We observed a 16.9% and 12.6% prevalence of smoking and alcoholism, respectively, in patients with PTB of southern India. Khan (2006) has suggested a risk of TB about 30–50% more among current and former smokers than never smokers. Studies carried out in developed countries observed 10–50% prevalence of alcohol use disorders among patients with TB (Lönnroth et al.2008). On the contrary, a similar study from South India has shown history of smoking and alcohol consumption not to be a significant risk factor (Shetty et al. 2006).
The percentage of patients with chronic corticosteroid therapy as a predisposing factor was similar in another study (Golsha et al. 2009); however, malignancy made patients more susceptible to have TB (5.8%vs. 2%) in our study. Prevalence of CKD (1.5%) was lower in our study than in that of Golsha et al. (2009) (5.8%). CLD has not been studied much as a risk factor; the 3.9% prevalence we observed among patients with PTB was similar to that in a study from Korea (Kim et al. 2009). With one exception (Golsha et al. 2009), men with or without underlying predisposing factors are more prone to have PTB. However, the fact that fewer women than men asked for admission and microbiological check-up at our hospital cannot be ignored; thus, the actual estimation of prevalence needs more extensive investigation, especially at the community level.
TB can be acquired through close household contact with cavities in the lung and smear positivity. Singh et al. (2005) has shown that 33.8% of children below 5 years in close contact with TB patients showed tuberculin positivity. Another study from Turkey detected TB in 5.6% of household contacts of patients with PTB (Talay & Kumbetli 2008). In our study population, 3.38% of our patients were in close contact with known PTB patients. However, it could not be ascertained whether those patients had acquired their infection from their close contacts only.
TB is a disease of the poor, and malnourished people are at risk for developing TB (Shetty et al. 2006). We saw that 1.5% of our patients with PTB were suffering from malnutrition. The true proportion of malnourished people with PTB might be higher because poor patients with malnutrition and TB have limited access to hospital medical services and a high mortality rate.
TB affected mostly the economically productive age group (Rajeswari et al. 1999; Chakroborty 2004). Rajeswari et al. (1999) had documented that the percentage of PTB was greater in waged workers than salaried or self-employed people in India, while another study from Wardha, Maharashtra, India showed that the percentage was greater in white-collar workers closely followed by cultivators, agricultural labourers, students and professionals (Chakroborty 2004). By contrast, we observed a significantly higher prevalence (P-value < 0.0001) of PTB in labourers (43.96%), followed white-collar workers (27.05%), retired and unemployed (6.28%), household workers (12%) and students (11%). This was very similar to the findings by Khan (2006), who documented a higher prevalence of TB in agricultural workers and labourers than in business people and professionals. The increased incidence of PTB among socio-economically lower classes can be attributed to lower education level and poverty (Khan 2006; Shetty et al. 2006).
In conclusion, PTB has a great impact in patients suffering from DM and occurs mostly in socio-economically lower classes, especially labourers. In a country like India, more efficient clinical diagnosis and prompt microbiological work are needed in the first instance. But this is not enough. Major national action needs to be taken to provide proper health education and create awareness among the people, especially the lower socio-economic classes, if we are to effectively combat tuberculosis, DM and HIV infection.