Assessing Biological Vulnerability of Acute Respiratory Tract Infection Among Children: Evidence from Bangladesh Demographic and Health Survey 2017-18

Background: Acute respiratory tract infections (ARIs) are the leading infectious disease worldwide and continues to be the single largest morbidity contributor in children. One of the most densely populated countries, Bangladesh also threatens by alarming under-ve childhood morbidity, which has aggravated in past years with the COVID-19 pandemic. This study attempts to understand the biological factors affecting the pre-existing respiratory tract infections in under 5 children of Bangladesh. Methods: The present study uses data from 8398 children aged below 5 years during the survey from the Demographic and Health Survey of Bangladesh (BDHS 2017-18). Both bivariate and multivariate analyses were performed to understand the biological vulnerability factors of pre-existing acute respiratory tract infections (ARIs) in under ve Bangladeshi children and relate them with the potential impact of the COVID-19 pandemic. Further, to show effectively the effect of different risk factors on child morbidity status, we have summarized all the results into prediction graphs at various levels of one variable as the other variable changes Results: Children aged one year were 1.40 [95% CI: 1.16, 1.67] and 2.01 [95% CI: 1.70, 2.36] times more likely to experience single morbidity and comorbidity respectively compared to children aged four years. We observe that male children were 1.18 [95% CI: 1.07, 1.31] times more likely to experience comorbidity compared to their female counterparts. Prediction graphs conrm the multivariate analysis as the probability of comorbidity remains higher in the monsoon season among children, with little change in the summer and winter seasons. Further, Rajshahi administrative division followed by Barisal and Rangpur shows the highest probability of comorbid condition in Bangladesh. Conclusion: Biological factors emerged as the prominent in child ARIs condition. More care is required pandemic not only isolated the people from also Insightful


Background
With the quick spread of the COVID-19 pandemic, health authorities had also prioritized the well-being of individuals especially the vulnerable section of society. While data from Italy shows that the elderly were more susceptible to fatality from this disease [1], studies also show that children were most likely to be spared from the direct mortality status through COVID-19 involvement [2]. However, the indirect effect of COVID-19 on children due to interruption in health care facilities and services was yet to be discovered. A study from 118 low-and middle-income countries by the Johns Hopkins Bloomberg School of Public Health shows that the indirect effect of reduction in health coverage services during a pandemic can be the reason for 2 million under-ve deaths which in turn would be responsible for bringing back the decades of progress across the world [3]. One of the most densely populated countries worldwide, Bangladesh is also threatened by the indirect effect of COVID-19 on children's health as they still face an alarming situation in under ve childhood morbidity [4].
Diarrhoea and respiratory tract infections including fever, cough, and breathing problems are the common infectious diseases that are prevalent across Bangladeshi children [5]. In 2015, a systematic analysis projected the death of 4.4 million under 5 children due to infectious diseases like diarrhoea and respiratory tract infections in 2030 [6]. Acute Respiratory Tract Infection, caused by virus or bacteria, is classi ed as upper respiratory tract infections (URIs) (airways from nostrils to vocal cords including paranasal sinuses and middle ear) and lower respiratory tract infections (LRIs) (airways from trachea and bronchi to the bronchioles and the alveoli) [7]. In developing countries, ARIs remains to be the single largest illness as it is responsible for 70% of under-ve morbidity condition [8]. ARIs are the common cause of illness and mortality among under 5 age children who suffer from an average of three to six episodes regardless of their living condition [9]. The situation can be worsened when the child experiences other conditions like fever and cough along with acute respiratory infection. As such comorbidity can lead to further extension of ARIs to infection, in ammation and reduced lung function [7]. Comorbidity is de ned as the occurrence of more than one morbid condition in the same individual either at the same time or in some causal sequence [10]. Today, it is common to experience more than one morbid condition for an individual as the comorbid condition is not limited to any age group. The situation can be more precisely explained by introducing the term biological vulnerability as it de nes a condition when an individual who is vulnerable to something is more likely to get affected by it [11]. For instance, children under 5 years of age are more biologically vulnerable to infectious diseases as they are more likely to get affected by such diseases. Further, to understand such a condition it is essential to consider the situation of comorbidity which can be the causal sequence of a disease.
Studies have associated ARIs with the exposure, environment, and background of children [12]. Evidence also shows that the risk of acute respiratory infection was common in younger ages [13,14]. A study from Bangladesh had shown that smoking habits among family members, location of the kitchen, and type of cooking fuels play an important role in acute respiratory infection incidence [15]. Studies have also shown that urban and male children were lesser likely to experience these diseases due to a preference for better food and health care facilities [15]. Numerous factors like parental education, household income and living condition were found to be associated with ARIs [13,[16][17][18]. According to UNICEF, approximately 33 million Bangladeshi children are residing in poor living conditions with deprivation from basic human needs like food, health, education and sanitation [19]. Studies have shown that the poor living condition of children is highly associated with their deteriorated health which can be further aggravated due to lack of access to improved water and sanitation sources [20,21]. In Bangladesh, water and sanitation insecurity had continued to be the largest challenge throughout the years as still, 60% of the population are lacking the proper accessibility which had further increased the burden of infectious disease like acute respiratory tract infection (ARIs) [22].
Despite numerous efforts to reduce the mortality condition in under-ve children, Bangladesh continues to experience high ARI-related mortality of one in every ve deaths [23,24]. Moreover, with COVID-19, immense pressure is expected on Bangladeshi children who are already facing health-related challenges.
In this regard, it is reasonable to understand the current situation of pre-existing morbidity conditions (in terms of acute respiratory tract infections) of under-ve Bangladeshi children. Recently released data of Demographic and Health Survey of Bangladesh (BDHS 2017-18) provides an opportunity to understand the morbidity condition of under-ve children. To the best of our knowledge, this is the most recent data which provides national-level information from Bangladesh, before the outbreak of the COVID-19 pandemic. This study can provide insights for the planning of mitigation strategies to take care of underve children during and after the COVID-19 pandemic. The present study aims to provide the biological and availability and accessibility of health and family planning services at the community level. The survey follows a two-stage strati ed sample design. Further details regarding sample design, survey instruments, eldwork and training of staff, data collection and processing, and response rates are available in the BDHS 2017-18 reports [24].
We used the data for 8759 children aged under-ve years born to 7562 mothers aged 15-49 years in Bangladesh. However, we dropped the records of 361 children who were not alive during the survey period and had no information regarding their morbidity status. Therefore, the analytical sample for this study is 8398 children under ve years of age.

Outcome variables
The outcome variable of morbidity status comes from the mother's responses regarding the knowledge of their children's morbidity. BDHS 2017-18 collected information regarding whether the children had suffered from fever, cough, and acute respiratory infections (ARI) within two weeks before the survey. We combined these three variables into a single variable of morbidity status with contained three categories which are -children who did not suffer from any of the three morbidities ("no condition"); children who suffered from "single condition"; and children experiencing two and more conditions ("comorbidity"). The advantage of this approach is that it aloows us to take into account the severity of the children's in rmity [5].

Explanatory variables
Guided by extant research, we identi ed relevant factors that are associated with the occurrence of morbidity among children [5,12,13]. Accordingly, we included relevant explanatory variables, conditional upon their availability in BDHS 2017-18. The child-related characteristics are -age in years (less than one, one, two, three, four) and gender (male, female). The parent-related characteristics are -mother's level of education (no formal education, upto primary, secondary and above), father's level of education (no formal education, upto primary, secondary and above). The household-level factors are -household sanitation condition (poor, average, good), household members drink treated water (no, yes), type of handwashing place (Private space, Public place, No handwashing place), shares toilet with other households (Not shared, Shared by two households (HH), Shared by three HH, Shared by four and more HH), wealth quintile (poorest, poor, middle, rich, richest), the religion of household head (Islam, Hinduism, others). Further, the season during the interview (Summer, Monsoon, Winter), place of residence (City Corporation, urban areas other than City Corporation, Rural areas), and administrative division (Dhaka, Chittagong, Barisal, Khulna, Mymensingh, Rajshahi, Rangpur, Sylhet) were also included.
Taking a cue from extant research, the household sanitation condition variable was constructed from three variables -type of source of drinking water, type of sanitation facility and the number of members per room in the household [25]. Respondents were asked about the source of household drinking water and as per prevalent standards, we recoded the source of household drinking water into two categories -"unimproved" coded as 0 (consisting of "dug, open well", "river", "pond", "truck" and "bottled" categories from the original variable) and "improved" coded as 1 (consisting of "piped", "tube well", "hand pump", "covered well" and "rainwater" categories from the original variable) [26]. Similarly, we recoded the type of household toilet facility into -"unimproved" coded as 0 (consisting of "defecation in open elds" and "traditional pit latrine" categories from the original variable) and "improved" coded as 1 (consisting of "ventilated improved pit latrine" and " ush toilet" categories from the original variable) [26]. Similarly, households with less than 3 members per room were coded as "1" and those with 3 or more members were coded as "0". After this, we added the three variables to obtain a household sanitation condition score. Households with a score of 3, score of 2 and score less than 2 were categorized as having "good", "average" and "poor" sanitation condition respectively.
To avoid multicollinearity, we coded a new wealth quintile variable after excluding information on household water source and toilet facility. The wealth quintile variable was prepared using standard procedures that are documented elsewhere [27].

Statistical methods
We performed bivariate and multivariate analysis to realize the study objectives. Owing to the categorical nature of the outcome variable, the bivariate association was examined using the chi-square test for association. Equivalently, multivariable analysis was performed by estimating multinomial regression models. The multivariate association of morbidity status of children with the explanatory variables was shown using relative risk ratios. Relative risk ratio gives the risk (multiple times) of having comorbidity (or single morbidity) compared to having no morbidity among those children belonging to a particular category of an explanatory variable given the effect of all the other explanatory variables remain constant [28]. To show effectively the effect of different risk factors on child morbidity status, we have summarized the regression output into graphs of predicted probability [29].
We checked for multicollinearity in the regression model and found the mean value of the variance in ation factor (VIF) to be less than 1.25. Therefore, multicollinearity is negligible [30]. Further, the Hausman-McFadden test revealed that our estimated model did not violate the independence from irrelevant alternatives (IIA) assumption [31]. All statistical estimations were performed using the STATA software version 13.0 [32]. 15% of children had a mother and father with no schooling education. One in every ten children comes from a household with poor sanitation conditions and 89% of children are from households where drinking water is untreated. Public space handwashing was common (64%) and most of the children did not share toilets with their family members (67%). Nearly, 27% of the population belongs to the poorest wealth quintile households and 65% reside in rural areas. In terms of population numeric, Chittagong is the largest division (17%) followed by the Dhaka division (15%) which includes the country's capital city Dhaka.       Note -(a) RRR: relative risk ratio; (b) 95% Con dence Interval (CI) is given in brackets; (c) Statistical signi cance is denoted by asterisks where * denotes p-value < 0.05; (d) ® denotes reference category; (e) Morbidity status of children categorized into: no condition, single condition, comorbidity

Sample description
Results of the multivariate analysis were further con rmed by showing the predicted probability graphs for the factors having a statistically signi cant association with morbidity status. Figure-1 shows that the probability of acquiring single or comorbid (or multiple morbid) conditions increases with the decreasing age. Here, the highest single or comorbid condition is seen in the child's rst year of life. The probability of acquiring comorbidity condition among male children is higher, while little change in single morbidity condition is noticed across both genders (Figure-2). Figure-3 shows that the probability of comorbidity increases in the monsoon season among children, with little change in the summer and winter seasons. Children residing in other urban areas show a higher probability of comorbid conditions ( gure-4). While observing the administrative division of Bangladesh, a scattered picture is noticed. The probability of comorbidity condition was higher at the Rajshahi division ( gure-5).

Discussion
Although children are not the face of the COVID-19 pandemic, the impact of this universal crisis can be lifelong for them [2]. COVID-19 mitigation strategies have usually enforced social distancing and isolation measures. However, such strategies may sometimes disrupt life-saving health services. Studies have shown that a sudden outbreak of pandemic had affected regular health care facilities and services [33].
In Bangladesh, acute respiratory infections (ARIs) are the most common morbidity condition which needs proper attention throughout the year. So, keeping in view the COVID-19 situation, the present study uses recent national-level data of Bangladesh to highlight the vulnerability factors of under-ve childhood morbidity.
We found a strong effect of age on the incidence of ARIs along with fever and cough and these results are consistent with the previous Bangladesh studies [5]. It has been usually found that children at their younger ages can get exposed to contaminated water, soil, and food easily as at these ages they usually crawl and tries to explore the environment. However, older ages children who have already moved towards this exposure are well-versed with their environment and sometimes build a strong immunity till that age. Incidence of comorbidity condition among under-ve children also varies according to household wealth index in both single and multiple morbidity conditions. Multiple morbidities were found to be signi cantly higher among monsoon seasons which is consistent with a previous study showing the health impact of climate change [34]. However, in contrast to a previous Bangladesh study, the present study shows that comorbidity condition is higher among male children than females [35].
Rajshahi administrative division followed by Barisal and Rangpur shows the highest probability of comorbid condition. This may be due to the higher indigenous population in this area. Also, a WHO report has shown that throughout the decade, poverty in few administrative divisions like Rangpur had increased facing a weak health care system [36]. Results from predicted probability also con rm the preexisting demographic risk factors of under-ve childhood morbidity in Bangladesh. As the age and sex of the child, place of residence and administrative division of Bangladesh emerged as the detrimental factor for under-ve morbidity. Water and sanitation condition (like sharing toilets with more people in a household) doesn't affect signi cantly the morbidity status. These ndings are consistent with a Bangladesh study where improved water and sanitation sources were not found to be signi cantly associated with childhood morbidity [37]. Further, another study had also provided evidence that water, sanitation and handwashing interventions did not affect the linear growth of children in Bangladesh [18]. This might be due to the reason that most of the Bangladeshi population lack proper access to improved water and sanitation sources. And the combined effect of both water and sanitation interventions should be considered for bringing favourable changes [37]. Also, there is the necessity to consider the other unobserved factors which may play role in comorbid conditions.
The current pandemic can even worsen the situation of food insecurity, poverty, hunger, and malnutrition across the world. Previous studies have also shown the impact of the pandemic on the mental health of children [38]. Although the government had taken different measures to protect the well-being of children, the pandemic had increased the existing inequities and burdened the country with the risk of childhood disease or death. So, the unprecedented situation of COVID-19 draws our attention towards strengthening the public care facilities and identify the vulnerability factors which lead to the morbidity condition of children. The present study is also backed with the recent national-level data of under-ve children in Bangladesh which will help us to evaluate the situation just before the pandemic. Our study will also help policymakers to explore different mitigation strategies.
However, our study has some limitations too. First, our study provides only a cross-sectional view of the scenario and therefore does not allow us to examine causality. Second, the morbidity incidence was evaluated from the self-reported information provided by women. However, the short recall period of morbidity (two weeks before the survey) makes the chances of recall bias minimal. Also, there is a need to consider the unobserved factors which affect the association.

Conclusion
The nationwide lockdown due to the COVID-19 pandemic had not only isolated the people from physical communication but also disrupted the health care facilities critical for mitigating the pre-existing morbidity condition among Bangladeshi children. During the pandemic, it was found that the access to regular health care services and continuity of care become worsened. Our study urges a greater investment by the government to mitigate the adverse impact of the pandemic and to enhance the  Adjusted Predicted probability of Morbidity by Place of residence from the multinomial regression model Adjusted Predicted probability of Morbidity by Country Administrative Region from the multinomial regression model