Water, Air, & Soil Pollution

, Volume 212, Issue 1, pp 89–100

Air Pollution and its Impact on Lung Function of Children in Delhi, the Capital City of India


    • Air Contaminants Lab, Exposure and Biomonitoring Division, Environmental Health, Science & Research BureauHealth Canada
  • Madhuchanda Banerjee
    • IIT
  • Manas Ranjan Ray
    • Dept. of Experimental HematologyChittaranjan National Cancer Institute
  • Twisha Lahiri
    • Nature, Environment & Wildlife Society (NEWS)

DOI: 10.1007/s11270-010-0324-1

Cite this article as:
Siddique, S., Banerjee, M., Ray, M.R. et al. Water Air Soil Pollut (2010) 212: 89. doi:10.1007/s11270-010-0324-1


Air pollution is a major contributor to several respiratory problems, it affects the whole population in general but children are more susceptible. Exposure to automobile exhaust is associated with increased respiratory symptoms and may impair lung function in children. In view of this, the study was conducted among the children of Delhi, the capital city of India, where ambient air quality was much above the National Ambient Air Quality Standards. The study was conducted in children aged 9–17 years. Pulmonary function test was carried out following the guideline of American Thoracic Society using a portable, electronic spirometer. Air quality data was collected from Central and State Pollution Control Boards. In addition, the level of particulate matter in indoor air was measured by portable laser photometer. Lung function was reduced in 43.5% schoolchildren of the urban area compared with 25.7% of control group. The urban children had increased prevalence of restrictive, obstructive, as well as combined type of lung functions deficits. Besides higher prevalence, the magnitude of lung function deficits was also much more in them. After controlling potential confounders like season, socioeconomic conditions and ETS, PM10 level in ambient air was found to be associated with restrictive (OR = 1.35, 95% CI 1.07–1.58), obstructive (OR = 1.45, 95% CI 1.16–1.82), and combined type of lung function deficits (OR = 1.74, 95% CI 1.37–2.71) in children. Spearman's rank correlation test reaffirmed the association. The study confirms that the level of air pollution is affecting the children.


Air pollutionChildrenLung functionPM10

1 Introduction

Air pollution is recognized as a major contributor to several respiratory problems, it affects the whole population in general but factors like lower breathing zone, more time spent outdoors, immature immunity, and developing organs makes a child more susceptible to the ill effects of air pollutants (Gilliland et al. 1999).

Lung function is an important measure of respiratory health and a predictor of cardiorespiratory morbidity and mortality (Gotschi et al. 2008). Over the past two decades, researchers worldwide have investigated long-term effects of ambient air pollution on lung function with most finding adverse effects. Several studies have also suggested effects from traffic-related air pollution with strong support for effects on the development of lung function in children and adolescents (Gotschi et al. 2008).

The functional organization of the lungs requires a coordinated ontogeny of critical developmental processes that include branching morphogenesis, cellular differentiation and proliferation, alveolarization, and maturation of the pulmonary immune, vasculature, and neural systems. Therefore, exposure to environmental pollutants during crucial periods of prenatal and/or postnatal development may determine the course of lung morphogenesis and maturation (Kajekar 2007).

Epidemiological studies of air pollution and children’s lung function (Dockery et al. 2005) reveals that: (i) living in areas of high air pollution is associated with lower lung function, (2) chronic exposure to elevated level of air pollution is associated with lower rates of lung function growth, (3) improvement in air quality leads to improvement in lung function level and/or growth rate, and (4) children who spend a significant amount of time outdoors in polluted environments or those with poor nutrition may be more strongly affected by air pollution.

Acute exposure to automobile exhaust is associated with increased respiratory symptoms and may decrease and impair lung function in children. In view of this, we felt it necessary to examine the level of lung function among the children of Delhi, the capital city of India, where ambient air quality was much above the NAAQS.

2 Materials and Methods

2.1 Participants

The study was conducted in different areas of Delhi, West Bengal, and Uttaranchal. The exposed group consisted of 5,671 children (3,708 boys and 1,963 girls). The control group comprised of 2,245 children (1,438 boys and 807 girls) from rural areas of West Bengal and Uttaranchal. The children were in the age group of 9–17 years. Children suffering from asthma were not included in this study. Only normal healthy children were recruited. After obtaining consent from the school principal, the questionnaires were distributed among the school children a day before the lung function tests were conducted. The children were instructed through class teachers to take home the questionnaire and ask their parents or guardians to fill these up. Filled-up questionnaire forms signed by the parent or guardian and countersigned by the class teacher were collected from the respective schools and the lung function tests were performed.

2.2 Determination of Socioeconomic Status

Socioeconomic status (SES) of the child’s family was ascertained following the procedure of Srivastava (1978) and Tiwari et al. (2005).

2.3 Assessment of Lung Function

PFT was carried out following the guideline of American Thoracic Society (ATS 1995) using a portable, electronic spirometer (Spirovit SP-1) with disposable filters (SP-150), designed for ambulatory pulmonary function measurements. The device measures actual respiratory flow at a precision of 2%, in addition to predicted values according to age, sex, height, weight, and race. Before performing the PFT, the height and weight of the child was measured with shoes removed. Each child performed forced expiratory maneuvers while sitting with free mobility and nose closed with a nose clip to prevent passage of air through the nose. Each spirometric test was repeated three times to allow the choice of the best values, and two values of FEV1 should not differ by more than 5% according to the ATS criteria. Using a computer-assisted quantitative assessment the best maneuver for acceptance was determined.

The data were compared with predictive values based on age, sex, height, and ethnic group. Flow was plotted against volume to display a continuous loop from inspiration to expiration, as the overall shape of the flow volume loop is important for interpreting spirometric results. The specific spirometric parameters (absolute and relative values such as ratio of actual and predicted values) such as forced vital capacity (FVC), forced expiratory volume at 1 s (FEV1), ratio of FEV1 to FVC (FEV1/FVC), expressed as percentage, forced expiratory flow at 25–75% (FEF25–75%) and peak expiratory flow rate (PEFR) during expiration were recorded for analysis.

Decrement of lung function detected by spirometry could be generally of two types: obstructive type and restrictive types of impairment. In some cases combined (both obstructive and restrictive) type of lung function impairment could be encountered.

2.3.1 Obstructive Type of Lung Function Impairment

In obstructive type of lung function deficits such as emphysema or chronic bronchitis the FEV1 is reduced disproportionately more than the FVC, resulting in an FEV1/FVC ratio less than 70%. Thus, FEV1/FVC <70% diagnoses airway obstruction. Subjects with obstructive lung had a rapid peak expiratory flow but the curve descends more quickly than normal and takes on a concave shape, reflected by a marked decrease in the FEF25–75%. With more severe obstruction, the peak becomes sharper and the expiratory flow rate drops precipitously.

2.3.2 Restrictive Type of Lung Function Impairment

In restrictive lung type of lung function decrement, the FVC is reduced below 80% of predicted value. The shape of the flow volume loop is relatively unaffected in restrictive disease, but the overall size of the curve appears smaller when compared to normal on the same scale.

2.3.3 Combined Type of Lung Function Impairment

In combined type of lung function impairment, both FVC and FEV1/FVC ratio are appreciably decreased. Subjects having this problem had FVC less than 80% of predicted value and FEV1/FVC ratio < 70%.

2.3.4 Calibration and Quality Control

In epidemiological studies in which spirometry results are primary outcome measurements, the results depend not only on the true lung function of the subject population, but also on quality of their test performance. Factors related to age, gender, size, ethnicity, and subject-technician affinity can influence the performance. But their overall effect is small with well-trained, experienced technicians/researchers. We calibrated the spirometers using a 2.0 l syringe each morning before the tests. The instrument was calibrated again after measurements in 50 children.

2.4 Assessment of BMI

Body mass index of a child was calculated by dividing the body weight in kilogram by the square of the standing height in meter. The body weight was categorized following the procedure of BMI-for-age separate growth charts for girls or boys formulated by the Center for Disease Control, USA.

2.5 Air Quality Data

Air quality data with respect to particulate matter with a diameter of less than 10 µm (PM10), oxides of sulfur and nitrogen (SOx and NOx), in ambient air of study areas was collected from Central and State Pollution Control Boards from their fixed-site monitoring stations. In addition, the level of PM10 and PM2.5 (particulate matter with diameter <2.5 µm) in indoor air of the households of participating children was measured by portable, battery-operated laser photometer (DustTrak™ Aerosol monitor, model 8520, TSI Inc., MN, USA). The indoor air quality was measured randomly and was averaged over several days.

2.6 Statistical Analysis

The collected data were processed and analyzed in EPI info 6.0 and Statistical Package for Social Sciences software. Chi-square test was done for dichotomous or multinomial qualitative variables, and the Student’s t test for quantitative variables of normal distribution and homogeneous variances. A descendant stepwise logistic regression adjusted over potential confounding variables was carried out for multivariate analysis.

3 Results

3.1 Demography

The characteristics of both urban and rural children are presented in Table 1. The urban and rural children were comparable (p > 0.05) with respect to age, gender, BMI, parental smoking, religion, and food habit. But, the control group had lower parental education and family income than that of exposed (p < 0.05). Moreover, a substantially increased number of households of control children (p < 0.05) used biomass fuel such as dung, firewood, and agricultural refuse (dried leaves, hay, jute stick etc.) for domestic cooking.
Table 1

Demographic characteristics of the children


Control (n = 2,245)

Delhi (n = 5,671)

p value

Median age in years




Boys: girls




Height in cm ± SD

149.5 ± 10.5

151.5 ± 11.3


Mean body weight in kg ± SD

38.0 ± 10.3

42.5 ± 12.9


Mean BMI in kg/m2 ± SD

16.8 ± 3.8

18.2 ± 3.9


Parental smoking (%)




Parents’ education (%)

Up to 5 years of schooling




10 years of schooling












Religion (%)





Muslim, Sikh, Jain, and others




Food habit (%)









Cooking fuel use at home (%)













Average family income/month (Rs.)




NS statistically not significant

3.2 Air Pollution Level

The annual average concentrations of PM10 was 161.3 ± 4.9 µg/m3. In contrast, the concentrations of these pollutants were significantly lower in control areas—74.6 ± 3.3 µg/m3. Mean concentrations of sulfur dioxide (SO2) and nitrogen dioxide (NO2) in the urban air during this period were 9.6 ± 1.0 and 50.1 ± 7.1 µg/m3, respectively, and it was within the permissible limit. In the control areas the concentrations were 5.6 ± 2.2 and 30.3 ± 5.2 µg/m3, respectively (Table 2).
Table 2

Comparison of air quality of the residential areas of Delhi and the control area

Pollutant (µg/m3)

Standard for residential areas

Control areas in West Bengal




167.6 ± 28.3

341.8 ± 38.3*



74.6 ± 17.3

136.8 ± 16.5*



30.3 ± 5.2

43.4 ± 5.9*



5.6 ± 2.2

9.4 ± 1.3*

The data (mean ± SD) are annual average values

*p < 0.05 compared with control

3.3 Decreased Lung Function

Lung function was reduced in 43.5% schoolchildren of the urban area (Delhi) compared with 25.7% of control group. The urban children had increased prevalence of restrictive (20.3% v. 14.3% in control), obstructive (13.6% vs. 8%), as well as combined (both restrictive and obstructive) type of lung functions deficits (9.6% vs. 3.5%; Table. 3; Fig. 1). Besides higher prevalence, the magnitude of lung function deficits was much more in them. For example, 7.3% schoolchildren of the city had severe lung function deficits compared with 2.2% children in control group (Table 4). Reduced lung function was more prevalent in girls than in the boys both in rural and urban settings. In the urban group, 51% of the girls enrolled in this study had reduced lung function compared with 39.8% of age-matched boys (Fig. 2). In control group, 28.1% of the girls had lung function deficits compared with 24.4% of the boys (Table 3). It was seen that in the exposed group the prevalence of reduced lung function was high in the 9–11-year-old group and 15–17-year-old group whereas in the control group the highest lung function reduction was seen in the age group of 9–11 years (Table 5).
Table 3

Prevalence of lung function deficits

Type of deficit



Total (n = 2,245)

Boys (n = 1,438)

Girls (n = 807)

Total (n = 5,671)

Boys (n = 3,708)

Girls (n = 1,963)


320 (14.3)

202 (14.0)

118 (14.6)

1,152 (20.3)*

669 (18.0)*

483 (24.6)*


179 (8.0)

104 (7.2)

75 (9.3)

772 (13.6)*

462 (12.5)*

310 (15.8)*


79 (3.5)

45 (3.1)

34 (4.2)

544 (9.6)*

344 (9.3)*

200 (10.2)*


578 (25.7)

351 (24.4)

227 (28.1)

2,468 (43.5)*

1,475 (39.8)*

993 (51.0)*

Results are expressed as number of children with percentage in parenthesis

*p < 0.05, compared with respective control in Chi-square test

Fig. 1

Comparison of lung function tests in control (a) and Delhi’s children (b)

Table 4

Magnitude of lung function reduction

Magnitude of deficits in PFT































Results are expressed as percentage of total children

*p < 0.05, compared with respective control in Chi-square test

Fig. 2

Comparison of prevalence (%) of lung function decrement between girls and boys of Delhi

Table 5

Prevalence of reduced lung function in children in relation to age


% children with lung function deficits



9–11 year



12–14 year



15–17 year



*p < 0.001 compared with control in Chi-square test

Restrictive type of lung function decrement (FVC < 80% of predicted value) was recorded in 20.3% of urban children, in contrast to only 14.3% of children in control area (p < 0.05, Table 3). Girls had greater prevalence of restrictive type of lung function deficits than the boys in both the groups (Table 3), and the gender difference in this regard was significant (p < 0.05). The difference in the prevalence of mild, moderate and severe restriction between control and children from the urban area was significant (p < 0.05). The prevalence of severe restriction was more in girls in both the areas. The hallmark of obstructive type of lung function decrement is reduction in FEV1/FVC ratio to less than 70%. We found the presence of this type of lung function deficits in 13.6% schoolchildren of the urban area and 8.0% children of control group (Table 3). Girls in general had greater prevalence of obstructive lung than the boys in both the groups (Table 3). Children having both restrictive and obstructive lung function reduction are grouped under ‘combined’ type. The number of such children was 3.5% in control and 9.6% in the urban area; combined type of lung function deficits was also more common in girls (Table 3).

Overall, the school children examined in Delhi had 2.19 ± 0.45 (SD) liters FVC that was 83.6 % of the predicted value based on age, weight, height, and ethnicity. In contrast, school children in control group had 2.67 ± 0.55 l FVC, which was 104.7% of the predicted value (Table 6). Thus, Delhi’s children showed a significant 480 ml reduction (p < 0.001) in mean FVC. Compared with controls, boys in Delhi illustrated 520 ml reduction (−18%) in mean FVC, while in case of girls the deficit was 390 ml (−17%). Mean measured forced expiratory volume in one second (FEV1) was 1.93 ± 0.25 l in Delhi compared with 2.19 ± 0.36 in control group, thereby showing a reduction of 12% in mean FEV1 in Delhi’s children (p < 0.05, Table 7). Comparison of FEF25–75% in relation to age between control and Delhi boys and girls revealed a lower value in Delhi in all the age groups (Fig. 3). Overall, school children of Delhi demonstrated 21% reduction in FEF25–75% (p < 0.001) than the controls, suggesting greater prevalence of underlying small airway obstruction. Compared with respective controls, the reduction of FEF25–75% was 24.4% in boys, and 21.6% in girls of Delhi. It was observed that the measured value of FEF25–75% was much lower in both boys and girls of Delhi than the control group (Table 8). A total of 70.3% children of Delhi had small airway obstruction as evidenced by reduction of FEF25–75% below 80% of predicted value. In contrast, 32.6% control children had FEF25–75% < 80% predicted (p < 0.001). In 17.8% of children of Delhi, the reduction in FEF25–75% was <40% of predicted value, implying severe obstruction of the small airways (Table 9). Severe small airway obstruction was found in 9.2% control children, suggesting a twofold rise in the prevalence of this problem among the children of Delhi.
Table 6

Comparison of FVC

FVC (liter)




Control (n = 1,438)

Delhi (n = 3,708)

Control (n = 807)

Delhi (n = 1,963)

Control (n = 2,245)

Delhi (n = 5,671)


2.90 ± 0.73

2.38 ± 0.48*

2.24 ± 0.48

1.85 ± 0.35*

2.67 ± 0.55

2.19 ± 0.45*


2.76 ± 0.49

2.81 ± 0.69

2.19 ± 0.24

2.26 ± 0.35

2.55 ± 0.42

2.62 ± 0.39

% predicted







*p < 0.05 compared with respective control in Student’s t test; **p < 0.001 compared with control in Chi-square test

Table 7

Comparison of FEV1 between control and Delhi’s children

FEV1 (liter)




Control (n = 1,438)

Delhi (n = 3,708)

Control (n = 807)

Delhi (n = 1,963)

Control (n = 2,245)

Delhi (n = 5,671)


2.31 ± 0.58

2.05 ± 0.44*

2.01 ± 0.41

1.72 ± 0.37*

2.19 ± 0.36

1.93 ± 0.25*


2.36 ± 0.38

2.38 ± 0.45

2.01 ± 0.22

2.03 ± 0.40

2.23 ± 0.38

2.25 ± 0.33

% predicted







*p < 0.05 compared with respective control in Student’s t test; **p < 0.001 compared with control in Chi-square test

Fig. 3

Changes in FEF25–75% in boys and girls in relation to age in control and exposed group

Table 8

Mean FEF25–75% in control and Delhi’s children





2.99 ± 0.65

2.26 ± 0.79*


2.45 ± 0.54

1.92 ± 0.80*


2.66 ± 0.39

2.10 ± 0.59*

*p < 0.05 compared with respective control in Student’s t test

Table 9

Percentage of children with reduced FEF25–75%

FEF25–75% (% predicted)































*p < 0.001 compared with respective control in Student’s t test

Lung function in schoolchildren varied considerably with season. In control group, lung function reduction was highest during winter (in 32.9% children) when the particulate pollution level in ambient air was highest. Conversely, lowest prevalence of lung function deficits in schoolchildren was recorded during monsoon (19.9%) when the breathing air is cleanest. In the urban area especially in Delhi, high prevalence of lung function deficits was observed both in winter and summer (52.7% in both seasons); while a much lower prevalence (39.9%) was observed in monsoon (Table 10). Frequent dust storms coming from the Rajasthan deserts and hitting Delhi during summer elevate the background particulate level in breathing air. This could be a contributory factor to higher prevalence of lung function deficits in Delhi’s children during summer. The difference in the prevalence of lung function deficits between winter/summer and monsoon in Delhi was significant (p < 0.05).
Table 10

Seasonal variation in the prevalence (%) of lung function decrement













*p < 0.05 compared with control in Chi-square test

An inverse relationship exists between SES and the prevalence of lung function deficits. Conditional logistic regression analysis showed that the correlation was significant (p < 0.05; Table 11).
Table 11

Conditional logistic regression analysis for association between lung function deficits and socioeconomic status


Lung function deficits







1.34 (1.12–1.89)*

1.47 (1.29–1.74)*


1.98 (1.25–2.67)*

1.80* (1.56–2.54)*

The results are expressed as odds ratio with 95% CI in parentheses

*p < 0.05 compared with high SES

About 28% of control and 27% of Delhi’s children were exposed to environmental tobacco smoke (ETS) at home due to smoking habit of their fathers or some other male member of the family. These children had lower lung function than those unexposed to ETS.

School children of Delhi, 5.4%, enrolled in this study were overweight on the basis of their BMI data. In contrast, 2.4% children of the control group were overweight, and the difference between these two groups in this regard was highly significant (p < 0.001). Moreover, 9% of Delhi’s school children were at risk of being overweight compared with 4.4% children of control group (p < 0.001). On the other hand, the prevalence of underweight children was greater in control group (Table 12). BMI of the children had profound influence on their lung function. It was observed that overweight and underweight children had poor lung function than children with normal weight (Table 13).
Table 12

Body mass index of boys and girls of Delhi




Boys (n = 3,708)

Girls (n = 1,963)

Total (n = 5,671)

Boys (n = 1,438)

Girls (n = 807)

Total (n = 2,245)

Underweight (%)







Normal (%)







At risk (%)







Overweight (%)







Table 13

Lung function deficits in children with abnormal body weight

BMI status

% children with reduced lung function









At risk






*p < 0.05 in Chi-square test compared with control

3.4 Relationship Between PM10 and the Prevalence of Lung Function Deficits

After controlling potential confounders like season, socioeconomic conditions and ETS, PM10 level in ambient air was found to be associated with restrictive (OR = 1.35, 95% CI 1.07–1.58), obstructive (OR = 1.45, 95% CI 1.16–1.82), and combined type of lung function deficits (OR = 1.74, 95% CI 1.37–2.71) in children.

Spearman’s rank correlation test reaffirmed the association. It was found that the decrease in all the lung function measurements was correlated with PM10 level in ambient air. The correlation was strongest for FEV1/FVC ratio (rho value −0.986, p < 0.0005), followed by FEF25–75% (rho value −0.944, p < 0.0005), FVC (rho = −0.912, p < 0.0005), PEFR (rho = −0.542, p < 0.001), and FEV1 (−0.472, p < 0.001).

Similarly, the existence of a direct relationship between PM10 level in ambient air and lung function deficits was observed in conditional logistic regression analysis. Increasing levels of PM10 were found to be associated with increased lung function deficits of obstructive and restrictive type. In nearly all the cases we found that as PM10 increased lung function deficit also increased (Table 14).
Table 14

Conditional logistic regression analysis of the relationship between PM10 level in ambient air and children’s lung function

PM10 (µg/m3)

Reduced lung function

Restrictive type (FVC < 80%)

Obstructive type (FEV1/FVC < 70%)






1.34* (1.06–1.68)

1.48* (1.15–1.91)

0.96 (0.70–1.23)


1.62* (1.40–1.87)

1.53* (1.30–1.81)

1.42* (1.12–1.80)


3.75* (3.50–4.60)

4.82* (3.90–5.97)

1.59* (1.28–1.20)

The results are expressed as odds ratio with 95% CI in parentheses

*p < 0.05

4 Discussion

Compared with rural controls, we found reduced lung function both in boys and girls of the urban area (Delhi). Lung development and lung function are influenced by several factors. Most important among these are birth weight, infections, nutrition, and environmental factors such as air pollution. Air pollution hampers teenagers’ lung development (Khan 2004). Lung is affected by chronic exposures to high level of NO2, SO2, and PM, of which the effect is strongest for PM (Ackermann-Liebrich et al. 1997). Recent study has identified PM2.5 as the most dangerous particulate fraction in this regard (Bernstein and Abelson 2005). Fine particulates (PM2.5) are ubiquitous because they are largely derived from common combustion processes such as engines of motor vehicles, power generation, burning of biomass, and manufacturing, and they are transported over long distances and readily penetrate indoors (Pope 2004). Exposure to fine particulate matter may be an important public health concern (Pope 2004). Such matters that can be breathed deeply into the lungs include sulfates, nitrates, acids, metals, and carbon particles with various chemicals adsorbed onto their surfaces (Pope 2004). Chemical composition and radical-generating capacity of the PM depends on the source of emission. Therefore, chemico-toxicological characteristics of PM vary at locations with high traffic emissions from that of PM at low traffic locations, regardless of the mass concentration of PM (Hogervorst et al. 2006). It has been shown that for every increase of 10 µg/m3 of PM10 and NO2 lung function parameters are reduced in general, by 1% in school children aged 7–10 years (Moshammer et al. 2006).

The association between air pollution and children’s lung function deficits was unveiled by the epidemiologic studies conducted in Europe and the United States. The Second National Health and Nutrition Examination Survey (NHANES II) in the United States demonstrated significant negative correlations between annual concentrations of TSP, NO2, and ozone and FVC and FEV1 in children, adolescents, and young adults aged 6–24 years (Schwartz 1989). Studies in young adults have shown that lung function decrement is associated with long-term elevated levels of particulates, and lung function decrement can be further worsened by concomitant exposure of PM10 with ozone (Abbey et al. 1998).

Prospective cohort study in Poland in preadolescent children demonstrated significantly lower mean lung function growth rate adjusted to height in children living in more polluted areas of Krakow (Jedrychowski et al. 1999). In a large prospective cohort study conducted in Los Angeles, California with 3,000 children showed a strong association between exposures to PM10, PM2.5, NO2, and inorganic acid vapor with deficits in growth of lung function, as measured by changes in FVC, FEV1, and maximum mid-expiratory flow (MMEF; Gauderman et al. 2000). Compared with children living in the least polluted community, those living in most polluted community of the above study had a cumulative reduction of 3.4% in FEV1 and 5% in MMEF over the 4-year study period, and the deficits were more in children spending longer time outdoors (Gauderman et al. 2000). The latest report from the Children’s Health Study in the US with 1,759 children aged 10 to 18 years for a long 8-year follow-up period has illustrated deficits in FEV1, FVC, and MMEF in association with exposure to a variety of air pollutants including NO2, acid vapor, PM2.5, and elemental carbon (Gauderman et al. 2004). Collectively, these studies suggest that long-term exposure to elevated levels of air pollution during childhood can produce deficits in lung function growth. Therefore, poor air quality of the city can account for deficits in lung function growth in Delhi’s schoolchildren. The level of lung function reflects cumulative effects of air pollution over a lifetime. The decrement of FVC, FEV1, FEF, and PEFR in Delhi’s children in comparison with that of rural controls may suggest inhibition of expansion of the bronchioles and alveoli in the face of sustained exposure to high particulate pollution of the city.

Significant reductions of FVC in Delhi’s children relative to age- and sex-matched children from rural areas provide clear signs of pulmonary function deficits in the city. Fall in FVC is associated with fall in total lung capacity and development of restrictive type of lung function decrement that we have seen more often than not in Delhi’s children. Lung development is essentially complete in girls by the age of 18 years, whereas in boys it continues till early twenties (Schwartz et al. 1988). It is therefore unlikely that the deficits in lung function at the age of 17 years that we have found in a large number of schoolchildren of Delhi will be reversed as they complete the transition into adulthood.

In addition to restrictive lung, Delhi’s children had greater prevalence of obstructive type of lung function deficits. They had significantly reduced levels of FEV1, FEV1/FVC, and FEF25–75%, suggesting airway obstruction. As high as 70% of Delhi’s children had FEF25–75% value less than 80% of predicted, and in about 18% children, it was reduced below 40% of predicted value. Reduction in FEF25–75% is associated with peripheral (small) airway obstruction. In fact, it is a more sensitive index of airway obstruction than the FEV1, especially for the small airways that are dependent on the elastic and resistant properties of the distal airways (Seaton and Crompton 2000). Thus, a twofold rise in the prevalence of reduced FEF25–75% in Delhi relative to control suggests a remarkable increase of small airway obstruction in Delhi that could be attributed to sustained exposure of thee children to city’s air pollution. COPD is not a well-defined entity in children (Kabra et al. 2001). A child presenting with chronic cough and wheeze should be investigated for asthma, recurrent aspiration airway compressions, chronic infection, cystic fibrosis, and immune deficiency. In absence of these causes, environmental factors such as ETS and air pollution could be implicated (Kabra et al. 2001).

Indoor air quality has a tremendous effect on the lung function and its development. Households in the rural areas using only LPG as the cooking fuel were included in this study; therefore, the indoor air quality in both the groups was comparable. It was, however, observed that children from households using biomass as the cooking fuel had greater lung function deficits as compared to the children from households of the same rural areas using LPG. A parallel study is being conducted on women and children chronically exposed to biomass fuel in the rural areas; however, presenting the data and the results is beyond the scope of this paper.

Delhi possesses 4.2 million motor vehicles, which is more than the combined number of automobiles possessed by country’s other three metros—Mumbai, Kolkata, and Chennai. No wonder, vehicular pollution is the largest contributor to Delhi’s air pollution, and its relative contribution to the city’s air pollution is increasing every year. Road traffic could be causally associated with poor respiratory health of the children in Delhi. A large number of investigations carried out in Europe and elsewhere support this hypothesis. For instance, a positive correlation between exposures to car traffic and the prevalence of cough, recurrent wheeze, recurrent dyspnea, and reduced lung function particularly PEFR and FEF25–75% has been reported in Germany (Wjst et al., 1993). Road traffic also increases the risk of recurrent bronchitis, broncholitis, and pneumonia among children in Italy (Ciccone et al. 1998).

Perhaps the most important risk factor for reduced lung function in children is the poor air quality of the city. In 1997, the yearly average of CO, NOx, SO2, and TSP at traffic intersection of ITO were 4,810 ± 2,287 µg/m3, 83 ± 35 µg/m3, 20 ± 8 µg/m3, and 409 ± 110 µg/m3, respectively. In the next year, the levels of these pollutants were 5,772 ± 2,116 µg/m3, 64 ± 22 µg/m3, 23 ± 7 µg/m3, and 365 ± 100 µg/m3, respectively (Aneja et al. 2001). In general, the maximum concentration of air pollutants occurred during winter which can be attributed to a combination of meteorological conditions and photochemical activity in the region (Aneja et al. 2001). The ratio of CO/NOx of approximately 50 indicates that mobile sources are the predominant contributors for these two compounds in urban Delhi. The ratio of SO2/NOx (approximately 0.6) indicates point sources are contributing to SO2 in the city. The average background CO concentration in New Delhi was approximately 1,939 µg/m3, which exceeds those for Eastern USA (approximately 500 µg/m3). More importantly, all measured concentrations of airborne pollutants in Delhi exceeded the US NAAQS except for SO2 (Aneja et al. 2001). Even brief exposures to particulate air pollutants have been associated with acute decrease in lung function. In an early meta-analysis, Dockery and Pope (1994) reported a 1.5% decrease in FEV1 and a 0.8% decrease in PEFR for each 100 µg/m3 increase in PM10. Similarly, Zmirou et al. (1997) reported 2.2% decline in FEV1 and 0.7% decline in PEFR for each 100 µg/m3 rise in PM10. Diesel exhausts appear to have acute and chronic effects on lung function (Sydbom et al. 2001). Elemental carbon, a marker of diesel exhaust, has been shown to be associated with reduced lung function in children (Gauderman et al. 2002).

Besides the grossly increased fleet of vehicles on roads, use of adulterated automotive fuels could be responsible to a large extent for the high level of air pollution in Delhi and its adverse effects on children’s health. Adulteration of petrol and diesel is rampant in Delhi according to the report prepared by Center for Science and Environment (CSE 2002).

An important question is whether deficits in lung function growth related to air pollution are permanent or reversible. Cross-sectional prospective studies have shown that lung function may recover if an individual breathes cleaner air, either because of improvement of air quality or because the person moves to an area with cleaner air (Dockery et al. 2005). Reduction of 10 µg/m3 of both NO2 and PM10 improve most lung function parameters by 1% (Moshammer et al. 2006). The beneficial effects of pollutant reduction was observed in Utah Valley following the closure of a steel mill for 14 months in 1987 due to labor strike. Outdoor particulate concentrations and respiratory hospital admissions in both children and adults fell dramatically during the period when the mill was closed, but returned to the pre-closure levels when the mill reopened (Pope, 1989). Lung function of children who had moved to other communities showed reduction in annual lung function growth rates (FEV1, MMEF, and PEFR) among children with areas of more pollution and, interestingly, in children who had moved to less polluted areas (Avol et al. 2001).

Overall, the available reports suggest that reduction in air pollution in a short period improves lung function of the children. Thus, data from current available studies on responses to improving air quality or migration from areas of high concentrations of air pollutants to ones with lower concentrations suggest that recovery of lung function growth is possible. Thus, this study serves as an eye opener for the policy makers and therein lies its importance.


I would like to thank Central Pollution Control Board (CPCB), Delhi, for providing the fund to carry out the study. I would also like to thank all the children, parents, teachers and the principals of the schools without whose enthusiasm and cooperation the study would have been impossible. I would also like to acknowledge the great support of Saswati Chowdhury, for her guidance and mother like attitude throughout the study period.

Copyright information

© Springer Science+Business Media B.V. 2010