Factors Associated with Acute Respiratory Infection among Children Under 5 Years in India: Results from 2019 to 2021 National Family Health Survey

Abstract

Introduction:

Acute respiratory infections (ARIs) remain the leading global cause of death in children under-five. Targeted initiatives are needed to address healthcare inequities and reduce under-five mortality, particularly in disproportionately impacted low- and middle- income countries. To inform initiatives and identify high-risk groups, this study explored regional risk factors for ARIs among Indian children.

Material and Methods:

Our retrospective, observational study utilized India’s National Family Health Survey (NFHS-5). Bivariate and multivariable models were employed to investigate associations between respiratory infections and explanatory variables, including environmental factors, child characteristics, maternal characteristics, enabling factors, and household characteristics.

Results:

Of the 201,133 children under-five included in our sample, 2.85% [2.78-2.92%] experienced a recent respiratory infection. In multivariate analysis, children from northern and central regions had the highest odds of infection, while those from the southern region had the lowest. Healthcare accessibility, maternal smoking, caste, age (child), and birthweight were among additional variables associated with infections. Our study revealed significant regional differences in prevalence of acute respiratory infection symptoms. Notably, inability to access healthcare increased a child’s risk of infection. Several states in southern India, which typically had lower ARI symptom rates, have adopted initiatives to strengthen public health infrastructure, including the WHO’s Integrated Management of Neonatal and Childhood Illnesses program. Such initiatives could serve as models for broader health improvement efforts across regions. Furthermore, observed variability in disease burden suggests that with detailed and deliberate implementation of programs, advancements in under-five mortality due to ARI can be achieved.

INTRODUCTION

The world has seen considerable progress in reducing mortality of children under the age of 5 years. Over the past three decades, the global under-five mortality rate (U5MR) has declined form 93 deaths per 1,000 livebirths in 1990 to 38 deaths per 1,000 livebirths in 2021, signifying a nearly 60% reduction.[1,2] Despite advancements, approximately five million children under the age of 5 years died in 2019, largely due to treatable and preventable ailments such as acute respiratory infections (ARIs).[1]

The World Health Organization (WHO) defines severe ARIs as infections “with a history of fever of ≥38 C and cough with onset within the last 10 days and requires hospitalization”.[3,4] While under-five mortality from respiratory infections has declined in recent years, it remains the leading cause of mortality in children globally, accounting for approximately 10% of all deaths.[5] Along with diarrheal illnesses, studies indicate that ARIs are also the most common reason for hospital admissions among children in low- and middle-income countries (LMICs).[6]

Ensuring child health and reducing U5MR to less than 25 cases per 1,000 livebirths by 2030 is a target of the United Nations sustainable development goals (SDGs).[7] To achieve SDGs, initiatives to address healthcare inequities in 81 disproportionately impacted countries were established by government, multilateral, private sector, and civil society stakeholders.[6,7] In India, these initiatives have contributed to significant progress with U5MR declining from 91.6 cases per 1,000 live births in 2000 to 34.3 cases per 1,000 livebirths in 2019 – largely aligning with global developments.[8]

Despite progress, India will likely need to accelerate current trends to achieve SDGs and research suggests that respiratory infections deserve a central focus in ongoing initiatives.[9,10] Such infections – specifically pneumonia – remain the leading cause of under-five mortality in India. In 2015, the country accounted for 32% of under-five global pneumonia cases – more than any other nation.[6,11] Furthermore, while the incidence of global pneumonia cases declined by 22% from 2000 to 2015, India saw a meager 3% reduction in cases – far less than the 69% reduction observed in China, which led all countries.[6]

While each state in India saw a decline in pneumonia cases from 2000 to 2015, there was significant variability in observed trends. In 2015, pneumonia incidence remained greater than 500 cases per 1,000 children in two states: Uttar Pradesh and Madhya Pradesh. In contrast, pneumonia incidence was dramatically lower in top-performing states: Kerala (137 cases per 1,000 children) and Tamil Nadu (169 cases per 1,000 children).[6] Furthermore, from 2000 to 2017, India observed notable interstate variability in under-five mortality rate, ranging from 10.4 per 1,000 livebirths in Kerala to 59.7 per 1,000 livebirths in Uttar Pradesh.[9,10]

Given regional disparities in health outcomes, state-specific strategies should be adopted to supplement India’s country-wide programs to reduce child mortality secondary to respiratory infections.[10] While well-established risk factors for ARIs have been identified for LMICs, country-specific research highlighting pertinent regional factors and high-risk children is lacking.[12] The aim of this study was to identify such factors and inform program developers in designing targeted, regional initiatives to lower the burden of respiratory infections among Indian children.

MATERIALS AND METHODS

Data source

Our study utilized data from the fifth and most recent iteration of India’s National Family Health Survey (NFHS-5), which was carried out over two phases between June 2019 and April 2021.[13] Coordinated by the International Institute for Population Sciences (IIPS), the NFHS-5 was a multiround cross-sectional survey conducted in a nationally representative sample of 639,699 Indian households.[13] Using a standardized questionnaire, 724,115 eligible women aged 15–49 years were interviewed. Information regarding basic sociodemographic characteristics, maternal health, pregnancy and postnatal care, and child health indicators was collected. Additional survey design details, including the two-stage stratified random sampling technique utilized, can be found in the official NFHS-5 report.[13]

Study population and sample size

Responses from women with children born in the 5 years preceding the survey were reformatted using common variable names to generate a Children’s Recode dataset, which was made publicly available through the Demographic Health Survey (DHS) data distribution system. The Children’s Recode was utilized in our analysis as it contained information relevant to our target population of children under the age of 5 years – including data on recent illnesses and various health parameters. Data were weighted before our analysis, and surveys with missing information were omitted, yielding a final sample size of 201,133 children [Table 1].

Table 1.

Distribution of variables

Variable	ARI- N	ARI + N (%)a
Environmental
Place of residence
Urban	51,386	1,259 (2.39)
Rural	144,015	4,473 (3.01)
Region
North	26,366	824 (3.03)
Central	53,240	1,775 (3.23)
East	51,599	1,696 (3.18)
Northeast	7,413	193 (2.53)
West	24,342	636 (2.55)
South	32,441	608 (1.84)
Seasonb
Winter	58,886	1,935 (3.18)
Pre-Monsoon (Summer)	29,834	785 (2.56)
Southwest Monsoon	59,624	1,663 (2.71)
Post-Monsoon	47,058	1,349 (2.79)
Child characteristics
Age (years)
<1	37,667	1,332 (3.41)
1	38,102	1,315 (3.34)
2	39,294	1,134 (2.80)
3	39,585	1,038 (2.56)
4	40,753	913 (2.19)
Sex
Female	94,462	2,507 (2.58)
Male	100,939	3,225 (3.10)
Birthweight
Low (<2.5 kg)	31,242	1,039 (3.22)
Normal (2.5–4.0 kg)	145,264	4,081 (2.73)
High (>4.0 kg)	2,273	69 (2.93)
Not weighted	13,379	461 (3.33)
Don’t know	3,242	82 (2.46)
Weightb
Severely underweight	19,798	636 (3.11)
Underweight	42,206	1,264 (2.91)
Normal weight	131,791	3,773 (2.78)
Overweight	1,605	58 (3.51)
Stuntedb
Not stunted	127,816	3,679 (2.80)
Stunted	38,312	1,078 (2.74)
Severely stunted	29,273	975 (3.22)
Maternal characteristics
Age (years)
15-19	4,989	210 (4.21)
20-24	59,002	1,866 (3.07)
25-29	78,922	2,210 (2.72)
30-34	36,371	1,009 (2.70)
35-39	12,752	347 (2.65)
40-44	2,771	62 (2.17)
45-49	594	28 (4.54)
Education level
No education	40,994	1,234 (2.92)
Primary	23,873	845 (3.42)
Secondary	100,030	2,924 (2.84)
Higher	30,504	729 (2.33)
Marital status
Married	193,249	5,663 (2.85)
Not married	2,152	69 (3.09)
Smokes cigarettes or tobacco
Yes	5,810	248 (4.10)
No	189,591	5,483 (2.81)
Enabling factors
Household wealth index
Poorest	47,368	1,587 (3.24)
Poorer	42,594	1,374 (3.13)
Middle	38,611	1,069 (2.69)
Richer	36,462	909 (2.43)
Richest	30,365	792 (2.54)
Health insurance coverage
Covered	46,340	1,242 (2.61)
Not covered	149,061	4,489 (2.92)
Distance to health facility
No problem [ref]	78,068	1,961 (2.45)
Not a big problem	68,866	2,045 (2.88)
Big problem	48,467	1,726 (3.44)
Getting money for health services
No problem [ref]	90,474	2,390 (2.57)
Not a big problem	60,472	1,745 (2.80)
Big problem	44,454	1,597 (3.47)
Household characteristics
Religion
Hindu	155,530	4,567 (2.85)
Muslim	31,275	912 (2.83)
Christian	4,194	115 (2.68)
Sikh	2,385	75 (3.04)
Other	2,017	62 (2.99)
Caste
Scheduled Caste	45,512	1,438 (3.06)
Scheduled Tribe	19,617	479 (2.38)
Other Backward Class	84,958	2,479 (2.84)
None	43,534	1,286 (2.87)
Don’t know	1,779	49 (2.67)
Household size
1-2 members	903	28 (3.03)
3-5 members	87,079	2,648 (2.95)
>5 members	107,419	3,056 (2.77)

^aRow-wise percentages are included in parenthesis. ^bWinter: January – February; Pre-Monsoon: March – May; Southwest Monsoon: June – September; Post-Monsoon: October – December

Location data for each survey were also collected by DHS and utilized to create a map illustrating prevalence of ARI by state or union territory [Figure 1]. The DHS grouped responses into one of ~30,000 geographic clusters throughout India. Of note, to protect respondent privacy, the clusters were displaced from their actual locations by up to 2 kilometers for urban clusters and 10 kilometers for rural clusters, ensuring clusters do not change administrative units during the displacement procedure.[13]

Figure 1.

Map of India with southern states highlighted [created with historicalmapchart.net]

Outcome variable

Our outcome variable was defined as the occurrence of recent acute respiratory infection symptoms in children under the age of 5 years. Children were considered to have had recent symptoms if, during the two weeks before the survey, they had a cough accompanied by 1) short, rapid breathing that was chest-related, and/or 2) difficulty breathing that was chest-related. For our analysis, we categorized responses to produce a dichotomous-dependent variable with the following outcomes: ARI positive (presence of recent symptoms) and ARI negative (absence of recent symptoms).

Explanatory variables

After reviewing the relevant literature, 19 explanatory variables [Table 2] belonging to five categories – environmental factors, child characteristics, maternal characteristics, enabling factors, and household characteristics – were identified from survey data.[12,14,15] Region was adopted from categories of states/union territories provided in the NFHS-5 report.[13] To derive seasons, authors applied definitions provided by the Indian Meteorological Department.[16]

Table 2.

Binary Logistic Regression (Univariate & Multivariable)

Variable	cOR	P	95% CI	aOR	P	95% CI
External environment
Place of residence
Urban [ref]	1	--	--	1	--	--
Rural	1.27	<0.001*	1.13-1.42	1.13	0.074	0.99-1.29
Region
North	1.67	<0.001*	1.44-1.94	1.69	<0.001*	1.43-1.99
Central	1.78	<0.001*	1.54-2.05	1.73	<0.001*	1.47-2.04
East	1.75	<0.001*	1.52-2.03	1.54	<0.001*	1.31-1.80
Northeast	1.39	<0.001*	1.17-1.65	1.19	0.071	0.99-1.44
West	1.40	0.001*	1.14-1.71	1.36	0.003*	1.11-1.67
South [ref]	1	--	--	1	--	--
Season
Winter	1	--	--	1	--	--
Pre-Monsoon (Summer)	0.80	0.001*	0.70-0.92	0.79	0.001*	0.69-0.91
Southwest Monsoon	0.85	0.003*	0.76-0.95	0.99	0.939	0.87-1.14
Post-Monsoon	0.87	0.018*	0.78-0.98	1.04	0.547	0.92-1.17
Child characteristics
Age (years)
	Unadjusted β: -0.0032 [95% CI: -0.004 to -0.0027] P<0.001*			Adjusted β: -0.0031 [95% CI: -0.0037 to -0.0026] P<0.001*
Sex
Female [ref]	1	--	--	1	--	--
Male	1.20	<0.001*	1.13-1.29	1.21	<0.001*	1.13-1.30
Birthweight
Low (<2.5 kg)	1.18	0.001*	1.07-1.31	1.14	0.010*	1.03-1.26
Normal (2.5 – 4.0 kg) [ref]	1	--	--	1	--	--
High (>4.0 kg)	1.07	0.636	0.80-1.43	1.07	0.631	0.80-1.44
Not weighted	1.23	0.004*	1.07-1.41	1.10	0.191	0.95-1.27
Don’t know	0.90	0.454	0.68-1.19	0.83	0.193	0.63-1.10
Weight
Severely underweight	1.12	0.070	0.99-1.27	1.02	0.607	0.94-1.12
Underweight	1.05	0.297	0.96-1.14	1.03	0.686	0.91-1.16
Normal weight [ref]	1	--	--	1	--	--
Overweight	1.27	0.150	0.92-1.76	1.30	0.118	0.94-1.79
Stunted
Not stunted [ref]	1	--	--	1	--	--
Stunted	0.97	0.627	0.89-1.07	0.97	0.475	0.88-1.06
Severely stunted	1.16	0.006*	1.04-1.29	1.07	0.241	0.96-1.19
Maternal characteristics
Age (years)
	Unadjusted β: -0.0018 [95% CI: -0.0026 to -0.0011] P<0.001*			Adjusted β: -0.0004 [95% CI: -0.0012 to 0.0003] P=0.240
Education level
No education	1.26	<0.001*	1.10-1.43	1.03	0.702	0.89-1.19
Primary	1.48	<0.001*	1.29-1.70	1.25	0.003*	1.08-1.46
Secondary	1.22	<0.001*	1.09-1.37	1.12	0.056	1.00-1.27
Higher [ref]	1	--	--	1	--	--
Marital statusa
Married [ref]	1	--	--
Not married	1.09	0.621	0.78-1.51
Smokes cigarettes or tobacco
No [ref]	1	--	--	1	--	--
Yes	1.48	<0.001*	1.26-1.73	1.49	<0.001*	1.27-1.76
Enabling factors
Household wealth index
Poorest	1.28	<0.001*	1.13-1.46	0.97	0.747	0.81-1.16
Poorer	1.24	0.002*	1.08-1.41	0.98	0.845	0.83-1.16
Middle	1.06	0.403	0.92-1.22	0.92	0.329	0.79-1.08
Richer	0.96	0.536	0.83-1.10	0.89	0.152	0.77-1.04
Richest [ref]	1	--	--	1	--	--
Health insurance coverage
Covered [ref]	1	--	--	1	--	--
Not covered	1.12	0.008*	1.03-1.22	1.05	0.287	0.96-1.15
Distance to health facility
No problem [ref]	1	--	--	1	--	--
Not a big problem	1.18	<0.001*	1.08-1.30	1.11	0.072	0.99-1.24
Big problem	1.42	<0.001*	1.29-1.56	1.25	<0.001*	1.11-1.42
Getting money for health services
No problem [ref]	1	--	--	1	--	--
Not a big problem	1.09	0.068	0.99-1.20	0.98	0.764	0.88-1.10
Big problem	1.36	<0.001*	1.24-1.50	1.13	0.045*	1.00-1.27
Household characteristics
Religiona
Hindu [ref]	1	--	--
Muslim	0.99	0.903	0.88-1.12
Christian	0.94	0.608	0.73-1.20
Sikh	1.07	0.556	0.86-1.33
Other	1.05	0.810	0.70-1.57
Caste
Scheduled Caste	1.29	0.001*	1.10-1.45	1.37	<0.001*	1.18-1.58
Scheduled Tribe [ref]	1	--	--	1	--	--
Other Backward Class	1.19	0.024*	1.02-1.31	1.34	0.007*	1.17-1.53
None	1.21	0.039*	1.01-1.33	1.34	0.009*	1.15-1.57
Don’t know	1.12	0.504	0.77-1.71	1.14	0.595	0.74-1.77
Household sizea
1-2 members [ref]	1	--	--
3-5 members	0.97	0.908	0.61-1.55
>5 members	0.91	0.692	0.57-1.45

Odds Ratio (OR): Odds of ARI +/Odds of ARI-; cOR: crude odds ratio; aOR: adjusted OR. *P<0.05. ^aExplanatory variable did not meet inclusion criteria for multivariate model (P>0.10 in univariate analysis)

For variables identified as child, maternal, or household characteristics, definitions provided by the NFHS-5 report were applied.[13] Of note, stunting and weight categories were determined based on height-for-age and weight-for-age distributions, respectively. Pertinent enabling factors associated with healthcare utilization were identified using the widely adopted Andersen’s behavioral framework.[17] Household wealth index reflected assets, access to basic services, and housing characteristics (e.g. source of drinking water, building materials, etc.) with scores divided into quintiles.

Statistical analysis

Our statistical approach was similar to previous published research. To ensure the sample was representative for different domains, we applied sample weights that adjusted for the survey’s disproportionate two-stage sampling technique.[18,19,20,21] Initially, a descriptive analysis was conducted to report overall prevalence of recent ARI symptoms as well as frequencies by study factor – with a focus on geographic region. We then used binary logistic regression to analyze the association between explanatory variables and the presence of recent acute respiratory infection symptoms. Among explanatory variables, child and maternal age were treated as continuous variables, while remaining factors were treated as categorical variables.

Variables with a P value of ≤0.10 in the preliminary univariate analysis were included in the final multivariable model. We considered associations in the multivariable logistic regression analysis as statistically significant if the P value was ≤0.05. Correlation coefficients and associated 95% confidence intervals (CIs) are presented for continuous variables. For categorical variables, the results are presented as crude (cOR) and adjusted odds ratios (aOR) with confidence intervals.

Our analysis was performed using STATA© 17.0 (StataCorp LLC, College Station, TX, USA) and R v 4.2.2. The authors acquired permission to access and analyze data from the DHS program before study commencement. Program guidelines for data use were followed.

Ethics statement

This study was a secondary analysis of deidentified data from the NFHS-5, which is readily available in the public domain. Informed consent was obtained from survey participants, and survey protocol was approved by Institutional Review Board at the International Institute for Population Sciences, Mumbai.

RESULTS

Distribution of ARI symptoms

Of the 201,133 children represented in our sample, 2.85% (95% CI: 2.78-2.92%) had recent symptoms of an acute respiratory infection (ARI+). As demonstrated in Table 1, the prevalence of acute respiratory infection symptoms was highest during winter months and among children residing in rural areas. Regional differences were also observed, with ARI symptoms being less frequently observed in southern states [Figure 1]. Younger children, particularly those aged <2 years, exhibited a higher frequency of ARI symptoms than older children, and males had a higher prevalence compared with females.

Moreover, the occurrence of ARI symptoms was more common in children who were severely stunted or underweight compared with those who were not stunted or had a normal weight-for-age. The study also indicated that children whose mothers smoked cigarettes or tobacco had a higher prevalence of ARI symptoms. Among enabling factors, children living in the poorest households or households reporting issues with access to healthcare facilities or problems with healthcare payment experienced the highest frequency of symptoms [Table 1].

Factors associated with ARI from multivariate analysis

The results from univariate and multivariate analyses are presented in Table 2. Crude and adjusted odds ratios for presence of recent ARI symptoms are included for each explanatory variable with relevant reference groups, significance levels, and confidence intervals. A hierarchical multivariable logistic regression model with explanatory variables added in forward selection manner is illustrated in Table 3.

Table 3.

Hierarchical multivariable logistic regression model with explanatory variables included via forward selection

Model with Explanatory Variables	Outcome variable: ARI Symptoms
	R 2	ΔR2
Model 1 (Place of residence)	0.00027
Model 2 (Model 1 + Region)	0.00106	0.00079
Model 3 (Model 2 + Season)	0.00126	0.0002
Model 4 (Model 3 + Age of child)	0.00199	0.00073
Model 5 (Model 4 + Sex of child)	0.00224	0.00024
Model 6 (Model 5 + Birthweight of child)	0.00237	0.00013
Model 7 (Model 6 + Weight of child)	0.00238	0.00001
Model 8 (Model 7 + Stunted status)	0.00254	0.00016
Model 9 (Model 8 + Maternal age)	0.00269	0.00016
Model 10 (Model 9 + Maternal education)	0.00274	0.00004
Model 11 (Model 10 + Maternal tobacco use)	0.00275	0.00002
Model 12 (Model 11 + Household wealth index)	0.00307	0.00032
Model 13 (Model 12 + Health insurance)	0.00315	0.00008
Model 14 (Model 13 + Money for health service)	0.00318	0.00003
Model 15 (Model 14 + Caste)	0.0032	0.00002

In our adjusted analysis, neither season nor place of residence were significantly associated with ARI symptoms, while region was associated. Children living in southern India had the lowest adjusted odds of having recent ARI symptoms, and those residing in central and northern India had the highest. Specifically, children in central and northern regions had 1.73 [CI: 1.47-2.04; P < 0.001] and 1.69 [CI: 1.43-1.99; P < 0.001] times the adjusted odds of experiencing recent ARI symptoms compared with children in southern states or territories. In addition, those living in western and eastern regions had significantly higher adjusted odds of ARI symptoms compared with those living in the south. There was no significant difference in likelihood of ARI symptoms between children living in northeastern and southern regions of India.

Among child characteristics, age, sex, and birthweight were significantly associated with ARI symptoms, while growth parameters, such as weight and stunting status, were not. Adjusted odds of having recent ARI symptoms decreased with age and were significantly higher in males compared with females. Furthermore, children with low birthweight had 1.14 [CI: 1.03-1.26; P = 0.01] times the odds of having ARI symptoms compared with those with a normal birthweight.

Of the maternal characteristics investigated in our study, smoking status was associated with recent ARI symptom presentation. Children with mother’s who smoke cigarettes or tobacco had 1.49 [CI: 1.27-1.76; P < 0.001] times the adjusted odds of experiencing ARI symptoms compared with those with mothers who did not smoke. Marital status and maternal age were not associated with ARI symptoms. Children of a scheduled caste, other backward class [sic], or no caste were more likely to experience ARI symptoms compared with those from a scheduled tribe. Although prior studies have highlighted healthcare inequities along religious lines in India, our analysis did not find such disparities with respect to prevalence of ARI symptoms in children.[22,23,24]

Notably, children living in households that indicated distance to health facility as a “big problem” had 1.25 [CI: 1.11-1.42; P < 0.001] times the adjusted odds of recent ARI symptoms compared with those in households for which distance was not a problem. Similarly, children living in homes that reported attaining money for medical treatment as a “big problem” had 1.13 [CI: 1.00-1.27; P = 0.045] times the odds of ARI symptoms compared with those in households for which attaining money was not a problem.

DISCUSSION

Respiratory infections remain a leading cause of under-five mortality in India, and our study highlights significant regional differences in ARI symptom prevalence. Specifically, children in southern India were less likely to experience ARI symptoms compared with those from most other regions, suggesting that southern initiatives could serve as models for improving health outcomes more broadly.

Our study found that children with mothers who smoke cigarettes or tobacco had a higher prevalence of ARI symptoms compared with those with nonsmoking mothers. This finding is consistent with previous studies that have demonstrated the harmful effects of secondhand smoke exposure on children’s respiratory health.[12,25,26,27] Generally, northern India has observed the greatest levels of smoking among adults, and studies suggest that the relatively higher smoking prevalence among women contributes to this trend.[28]

The government of India has implemented several strategies to address the curb tobacco use. One such initiative was the enactment of the Cigarettes and Other Tobacco Products Act (COTPA) in 2003, which established a comprehensive legal framework for tobacco control.[29] This legislation, among other provisions, prohibits smoking in public places, bans advertising of tobacco products, and mandates the display of pictorial health warnings on tobacco packaging.

Notably, the effectiveness of antitobacco initiatives and the enforcement of COTPA regulations have been inconsistent across the country, contributing to observed regional variability in tobacco consumption patterns.[30,31,32,33,34] Multiple factors likely drive these disparities, such as differences in political will, availability of resources, and cultural attitudes toward tobacco use. Given the high burden of tobacco-associated morbidity and mortality in India, it is crucial to identify and address barriers to effective tobacco control and COTPA compliance across India. This will not only help in the reduction of tobacco use but also contribute to the improvement of overall public health, particularly among children and other vulnerable populations.

In addition to smoking, children living in households reporting issues with access to healthcare facilities or problems with healthcare payment were more likely to experience ARI symptoms. A number of southern states have adopted unique initiatives to address these barriers. For example, Tamil Nadu, which is the most populous state in southern India and the first to enact a Public Health Act in 1939, has utilized local and central government funding schemes to transform its health infrastructure in recent decades.[35,36] The state has implemented several unique initiatives to develop its well-organized and relatively robust public health sector, including the Universal Immunization Program, Tamil Nadu Medical Services Corporation, Tamil Nadu Health Systems Project, and Tamil Nadu Health System Reform Program.[35]

The findings from our research also revealed a significant association between low birthweight and an increased prevalence of acute respiratory infection (ARI) symptoms. To help mitigate this problem, states in India have adopted strategies proposed by the WHO’s Integrated Management of Neonatal and Childhood Illnesses program (IMNCI).[37] This multifaceted intervention endeavors to enhance the health of newborns and children by promoting improved healthcare quality and access while implementing preventive strategies addressing prevalent illnesses.[38] Notably, southern states of Kerala and Tamil Nadu were the only ones to achieve an infant mortality rate of less than 27 deaths per 1000 live births as established by the 2015 Millennium Development Goals.[37] This success was likely partly attributable to their progress in implementing IMNCI initiatives.[39] The expansion of such comprehensive programs within the country has the potential to ameliorate birthweight outcomes and alleviate the impact of ARI symptoms among children.

Strengths and limitations

Strengths of our study include a large sample size and the use of nationally representative data, which increases the generalizability of our findings. Additionally, we utilized multivariable logistic regression analysis to account for potential confounding factors, such as maternal smoking, healthcare access, and birthweight, which strengthens the internal validity of our findings.

However, there are also several limitations to our study that should be considered. First, our study was cross-sectional, which limits our ability to establish causal relationships. Second, we relied on self-reported symptoms of ARI, which may be subject to recall bias and may not accurately capture the true burden of infections. Additionally, although the NFHS-5 has a large number of participants, approximately 21,200 surveys (<10%) had missing information and were excluded. Finally, our study did not examine the potential effect of other factors, such as air pollution or indoor cooking fuel, which have been shown to be significant risk factors for ARI in other studies.[40,41]

Further research is needed to address these limitations and to evaluate the effectiveness of state-specific initiatives in India. Such research could help identify best practices and areas for improvement. Moreover, there is a need for a more comprehensive approach that involves collaboration between various sectors, including healthcare, education, and policy, to address the underlying determinants of ARI symptoms.

CONCLUSION

Our study highlights the significant regional differences in the prevalence of ARI symptoms among children in India and the potential for state-specific initiatives to improve health outcomes. The adoption of initiatives to reduce exposure to secondhand smoke, improve healthcare access, and address low birthweight could help reduce the burden of ARI symptoms among children in India.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.