In the present study, overall prevalence of diabetes among the pulmonary positive TB cases was 4.3%. This was consistent to previous study that reported a prevalence of 4.7% in Zambia [18]. Higher prevalence of TB-diabetes comorbidity was also reported from Ethiopia 8.5% [19], Portugal 6% [20], Mexico 29.63% [21], Iran 24.3% [22], and India 25.3% [23]. The prevalence of diabetes among TB patients observed in this study is comparable to the occurrence of the disease in the general population. While the difference among countries could be due to the inherent difference in socio economic and demographic characteristics among their population and the existing burden of both diseases in general population.
Many studies showed that male TB patients were more likely to suffer from TB-diabetes comorbidity than female TB patients [19, 20, 22, 23,24]. The current study had also a consistent finding in which male subjects were proportionally dominant in both TB-diabetes cases and those without diabetes. In contrary to this, a study in Zambia described that majority of the TB-diabetes cases were females [18]. Presence of the comorbidity predominantly among male cases could be related to the relatively higher probabilities of smoking and alcohol drinking among men in Eritrea in which these risk factors could have increased the occurrence of both disease in this particular group.
Among TB-diabetes cases nearly three fourth were 45 and above years old whereas in TB cases without diabetes almost half of them were aged 25–44 years. Similarly, mean age among TB-diabetes cases was 51.04 years which is higher than the mean age of those without diabetes 40.60 years. Previous similar studies reported a consistent finding in which TB-diabetes cases were older than TB cases without diabetes and the difference was significant [20, 22, 25]. This could be described as age related diabetes because Type 2 diabetes frequently develops in older adults. This in turn could lead to the development of TB in infected individuals with the decreasing immunity due to diabetes. Opposing findings were also reported from Zambia where the mean age of participants was 33 and 37 years for TB-diabetes cases and those without diabetes respectively. None of the participants with coexisting TB-diabetes were above 50 years of age [18].
The present study failed to report BMI (body mass index) due to unavailability of height record for all patients. Thus, our result could not be compared to available literatures that reported MBI and should be interpreted carefully. Higher mean weight (53.4 kg) was observed among TB-diabetes cases in comparison to the TB cases without diabetes (46.8 kg). A likely description for effects of weight in relation to TB and diabetes has been documented in which higher risk of TB is detected in undernourished or underweight individuals and this risk decreases as individuals gain weight [12]. According to a study in Zambia, patients with coexisting TB-diabetes were more likely to have a normal or low BMI [19]. Likewise, a study from India reported lower mean BMI of TB-diabetes than those TB cases without diabetes [26]. Another study from UK found comparable BMI among TB-diabetes and those without diabetes [25]. Significant higher BMI in TB-diabetes patients was also reported [24, 27] that describes excessive weight gain predisposes to diabetes which is another independent risk factor for TB [3, 24]. People with diabetes could have been over weight and may remain in over weight status after being affected by TB. On the other hand, there could be weight loss among diabetes cases due to TB that is infection induced hyperglycemia.
Residential difference among study subjects had been related to varying occurrence of these diseases. Consequently, majority of the TB-diabetes cases and almost three fourth of TB cases without diabetes in the current study were residents of an urban areas. Earlier studies showed higher prevalence of diabetes among TB patients who were urban residents [19,26]. In contrast, a study in Northern India reported a higher number of study participants with TB-diabetes were living in rural areas [28]. The reason for higher TB-diabetes comorbidity among urban residents could be linked to the sedentary life style such as physical inactivity and unhealthy dietary intake. On the other hand there could be under diagnosis of these diseases in rural populations in relation to limited healthcare access or delay in healthcare seeking.
Depending on TB treatment history and category, majority of the cases were new case and in first treatment category in both TB-diabetes and without diabetes respectively. Previous TB treatment was reported among few (4.29%) TB-diabetes and in 10.66% of TB cases without diabetes. Among those previously treated cases only relapse were found among TB-diabetes but treatment after failure and lost to follow up were reported only among TB cases without diabetes comorbidity. In agreement to the present study results, a study from UK reported that almost all cases of TB-diabetes and those without diabetes were new TB case and only less than 2% of them were previously treated [25]. Nevertheless, relatively higher number of TB-diabetes in category two with comparable figures among those TB cases without diabetes was reported previously [24, 26]. This could be due to the fact that Eritrea has an excellent TB cases management and follow up reflected by very high TB treatment success rate which could have minimized the possibility of retreatment of TB cases.
Another health-related characteristic investigated by the study was HIV-infection. The overall HIV positivity in this study was 10% in which comparable HIV positivity was found among TB-diabetes and without diabetes. Studies in Zambia and Ethiopia had higher HIV positivity rate among TB-diabetes case [18, 19,22]. A study in Portugal reported discrepant results with absence of HIV infection being a factor significantly associated with diabetes among TB patients [20]. HIV positivity rate found in this study was almost double of the prevalence of HIV in TB and almost 10 times higher than the prevalence in the general population. This could be due related to the study setting (Maekel zone) which may contain the highest number of HIV cases in the country.
In the current study, univariate logistic regression showed that participants with an older age (45–90) were more likely to have TB-diabetes comorbidity (COR: 4.397[CI: 1.675–11.545], p < 0.01) than those aged 10–44 years old. In addition, subjects whose weight was 65 kg and above were more likely (COR: 7.647[CI: 1.965–29.758], p < 0.01) to have TB- diabetes comorbidity than those with lower body weight. It’s expected that at an older age in the presence of immunity compromising infection people could be more prone to develop TB disease leading to comorbidity.
Prior study revealed that TB patients aged 41and above had higher odds of having diabetes than younger patients [23]. Other literature also noted increasing age, family history of diabetes, sedentary occupation, body mass index, type of TB, category of TB, and smoking were significantly associated [24,26].