Chronic exposure to emissions from solid fuels such as wood, charcoal, crop residues and cow dung is a significant cause of ill health conditions such as chronic obstructive lung diseases (chronic bronchitis and asthma), lung cancer, acute respiratory infections (ARI) in children and pregnancy-related outcomes [1]. In sub-Saharan Africa and many other low income countries where the main energy source for cooking/heating houses is biomass, women are chronically exposed to emissions which contain a lot of pollutants harmful to their health. Many other studies in developing countries have demonstrated a number of health effects linked to household biomass combustion and these include chronic obstructive pulmonary disease (COPD), asthma, respiratory infection, heart diseases, pulmonary diseases, cataract and adverse pregnancy outcomes [2,3]. In all these health burdens, women and children carry the heaviest burden compared to men; they are more susceptible to these health problems from household air pollutants because they spend longer times in the home environment [4,5]. A large body of evidence exists of biomass fuel being a significant contributor to increased respiratory symptoms and impaired lung functions [4,5].

Many studies have explored the relationship between pregnancy outcomes and exposure to air pollution due to indoor biomass combustion [6], second hand smoking [7] or smoking in pregnancy [8] but very few have tried to relate maternal respiratory symptoms and lung function to pregnancy outcomes. It is clear now that combustion products of unprocessed biomass fuels used by the poor urban and rural people for cooking and heating the house are the most important causes of indoor air pollution in developing countries. Indoor biomass fuel smoke has been shown to adversely affect the host defense against the respiratory infections and causes inflammation of airways and alveoli and it is thus plausible that it increases the incidence of respiratory ill health among the exposed [9,10]. Since air pollution has been associated to a number of adverse pregnancy-related outcomes, and then maternal respiratory health, influenced by indoor air pollution should have an influence on pregnancy-related outcomes. However, there has been a paucity of epidemiologic evidence regarding the relationship between maternal respiratory health and pregnancyrelated outcomes. Although there is a lack of evidence investigating how maternal respiratory health is associated with adverse pregnancy outcomes, it appears that respiratory health of the woman is linked to the pregnancy outcome through the impact of chronic indoor air pollution exposure, a situation that is common in many developing.

Other available associations of maternal respiratory health and lung functions are those that demonstrate that adverse environment in utero may affect fetal growth and have lasting effects on lung function in adulthood and old age [11]. For instance, previous studies indicate that low birth weight predicts lower lung function later in adult life after adjusting for maternal and adult factors [12]. However, this is not the focus for our study; the focus for our study is to relate maternal respiratory health with selected birth outcomes such as low birth weight, preterm birth and small for gestational age and available research has shown that poor maternal respiratory health is associated with intrauterine factors that affect fetal growth [13]. Studies elsewhere have also shown that suboptimal lung function in pregnancy is associated with adverse birth outcome [14]. The study of the relationship between asthma during pregnancy and selected infant and maternal outcomes, after significant confounding variables are adjusted for, gives important information regarding the association between respiratory ill health during pregnancy and several adverse infant outcomes, including preterm birth, small for gestational age and maternal outcomes like idiopathic preterm labor [15]. In addition this, several other studies previously have evaluated the association between birth weight and lung function with some, [11,16] though not all obtaining positive association. Therefore, this study aims at investigating the relationship between maternal respiratory ill health and pregnancy-related outcomes.

Questionnaire

A structured questionnaire was used to obtain information on background characteristics and symptoms of respiratory conditions.

Spirometry

Lung function was determined using a spirometer, MIR Spirobank-G (Italy) with an attached printout of forced vital capacity (FVC), forced expiratory volume in 1 second (FEV₁), and FEV₁/FVC% ratio.

For each assessment a research nurse demonstrated the technique to the participant. The participants then performed some practice efforts. They were then required to perform a minimum of three reproducible FVC measures (within 5% of maximum FVC produced). The output that produced the highest sum of FVC and FEV₁ were used in the analyses. Women who could not perform three reproducible measures or who were unable to attempt the lung function assessment were excluded. All assessments were carried out by one of nine research nurses who were trained to carry out the assessment in a similar manner.

Pregnancy outcomes

Pregnancy outcomes were recorded by the research nurse and the main outcomes of interest in the study were low birth weight (LBW), preterm birth (PTB) and small for gestational age (SGA). Neonatal length, weight and gestation age on delivery was recorded by the study nurse immediately after birth. General maternal health status, sex of the neonate and birth order was recorded together with any other birth outcomes such as stillbirth and congenital abnormalities.

Data processing and analysis

Lung function parameters of interest were categorized as follows: FEV₁/FVC%: > 70 is Normal, FEV₁% and FVC%: >80 is Normal. In the logistic regression, categories mild to severe were classified as impaired lung function. Preliminary analysis involving bivariate analysis was conducted in Epi Info. The independent predictors of outcome were obtained using the stepwise logistic regression in SPSS. Magnitudes of association were estimated using odds ratios and their 95% confidence intervals. Statistical significance was set at 5%. Ethical Consideration: The project was approved by the tropical Diseases Research Centre (TDRC) Ethics Committee. Further approval was obtained from the Zambia Health Research Authority of the Ministry of Zambia.

Background characteristics

Out of a total of 1,210 pregnant women recruited from the nine antenatal clinics (ANC), forty (3.3%) households were excluded from the analysis due to incompleteness of data and failure to perform three reproducible Spirometry measures or who were unable to attempt the lung function assessment, leaving a total of 1,170 participants for analysis. Majority (80%) of the participants enrolled in the study were 35 years and below. Close to three thirds (91.2%) of the participants were married and over half (54.6%) had attained secondary and a third (30.0%) primary education.

Above two thirds of the participants (69.6%) were unemployed housewives. Over a quarter participants (87.8%) were taking nutritional supplements (ferrous sulphate and folic acid) and about the same proportion (84.6%) were under malaria chemo prophylaxis-Fansida. Close to one fifth (19.7%) of participants took alcohol and only 3.1% cigarette smoking in pregnancy. Hypertension was reported in 3.7% and diabetes in 1.8%.

Respiratory symptoms of the pregnant women in the study population

The summary of the maternal respiratory symptoms in the study population has been shown in Table 1. Majority (71.5%)of the pregnant women reported prolonged cough, while not having flu, during the last 12 months and close to half (45.9%) of them had recurrent secretion of phlegm from lungs during the last 12 months. Close to one third (30.4%) of the pregnant women reported wheezing during the last 12 months (wheezing caused by bronchi, not nose) while 16.3% of the pregnant women presented with uncomfortable feeling of breathlessness (tight chest and complicated breathing), during the last 12 months.

Table 1: Showing proportions of maternal respiratory symptoms according to region.

Prolonged nasal symptoms when not having flu during the last 12 months were reported by 45.6% of the pregnant women and about one third (32.9%) presented throat symptoms when not having a flu during the last 12 months.

Lung function test results of the pregnant women in the study population

Almost all the pregnant women (99.3%) had an FEV₁/FVC higher than 70% and close to two thirds (60.5%) of pregnant women had a normal FEV₁ while more than one third (38.4%) recorded mild obstruction. Normal FVC was obtained in 35.6% of pregnant women while mild reduction and moderate reduction was recorded in 37.4% and 26.9% of pregnant women respectively. Only a very small proportion (0.4%) of the population presented with COPD (Table 2).

Table 2: Showing proportions of lung function of participants according to region.

Pregnancy outcomes in the study population according to region (Table 3)

Table 3: Showing proportions of adverse birth outcomes with their respective confidence intervals for rural and urban areas.

Mean differences of selected birth outcomes between rural and urban (Table 4)

Table 4: Showing mean differences of selected birth outcomes between rural and urban.

There was a statistically significant difference in gestational age and neonatal weight between rural and urban areas.

Mean differences of selected lung function parameters between rural and urban area

The mean FEV₁/FVC were 96.69% (SD 0.64) and 104.2% (SD 0.43) respectively observed in rural and urban area. The two means varied significantly with a p-value of <0.0001. The mean FVC between rural and urban did not vary significantly while the mean FEV₁ in the rural (84.75% SD 0.44) and urban (83.07% SD 0.39) varied significantly with the p-value 0.003. Table 5 below presents the mean differences of the lung functions together with the standard error, standard deviations, and p-values at 95% confidence interval between rural and urban area.

Table 5: Showing mean differences of Spirometry together with the standard error, standard deviations, and p-values at 95% confidence interval between rural and urban area.

There were statistically significant mean differences in FEV₁/FVC (p value<0.0001) and FEV₁ (p value=0.003) between rural and urban areas.

Mean differences of lung functions between mothers of LBW and of normal weight babies

There was a significant variation with a p-value of 0.023 at 95% C.I of 100.37%-101.86% between the mean FEV₁/FVC (101.60%, SD 12.31) of mothers to normal weight babies and mothers to babies who had a low birth weight (99.92%, SD 14.55). The mean difference was 1.68% with a standard error of 0.84. The mean FVC for mothers to normal weight babies (75.49%, SD 7.80) and that of low birth weight babies (76.55%, SD 7.78) also varied significantly at a p-value of 0.0176 at 95% C.I of 75.72%-77.38% with a mean difference of -1.06% and a standard error of 0.50. The mean FEV₁ of 83.53% (SD 10.30) and 84.32% (SD 9.62) respectively observed in mothers to normal birth weight and low birth weight, did not vary significantly. However, the mean difference was -0.79% with a standard error of 0.651. Table 6 below shows the summary of the mean differences of the lung functions together with the standard error, standard deviations, and p-values at 95% confidence interval between normal birth weight mothers and low birth weight mothers. The mean differences were statistically significant between.

Table 6: Showing mean differences of Spirometry between mothers of low birth weight babies and normal weight.

The mean differences were statistically significant between LBW and normal weight for FEV₁/FVC (p value 0.023) and FVC (p value=0.0176).

Mean differences of lung functions between mothers of SGA and of normal for gestational age babies

The mean FEV₁/FVC of mothers to normal for gestational age babies (102.50%, SD 12.10) and mothers to SGA babies (98.60%, SD 14.27) varied significantly with a p-value of < 0.0001 at 95% C.I of 100.37% - 101.86%. The mean difference was 3.88% with a standard error of 0.79. The mean FVC for mothers of normal for gestational age babies (75.64%, SD 7.83) and that of SGA babies (76.10%, SD 7.75) did not vary significantly. However, there was a mean difference of -0.46% at a standard error of 0.48. The mean FEV₁ of 83.28% (SD 10.22) and 84.65% (SD 9.84) respectively observed in mothers of normal for gestational age babies and SGA babies, varied significantly with a p-value of 0.0134 at 95% CI of 83.18% - 84.37. The mean difference was -1.37% at a standard error of 0.62. Table 7 below shows the summary of the mean differences of the lung functions together with the standard error, standard deviations, and p-values at 95% confidence interval between mothers for normal gestational age babies and SGA babies.

Table 7: Showing mean differences of Spirometry between mothers to SGA and mothers to normal babies.

The mean differences were statistically significant between SGA and normal for gestational age for FEV₁/FVC (p value<0.0001) and FEV₁ (p value=0.0134).

Mean differences of lung functions between mothers of preterm and mothers of full term babies

The mean FEV₁/FVC of 101.01% (SD 12.85) and 102.72% (SD 15.2) respectively observed in mothers of full term babies and mothers of preterm babies did not vary significant (p-value 0.1368). However, there was a mean difference of -1.712% at a standard error of 1.56. The FVC had a mean of 75.75% (SD 7.77) among mothers of full term babies and 76.60% (SD 0.960) among mothers of preterm babies. The two means did not vary significantly (p-value 0.183). However, there was a mean difference of -0.85% at a standard error of 0.94. The mean FEV₁ of 83.80% (SD 10.13) and 83.14% (SD 9.70) respectively observed in mothers of full term babies and mothers of preterm babies did not vary significant (p-value 0.295) at 95% CI of 83.18% - 84.37%. There was however, a mean difference of 0.653% at a standard error of 1.21. Table 8 below shows the summary of the mean differences of the lung functions together with the standard error, standard deviations, and p-values at 95% confidence interval between mothers of full term babies and mothers of preterm babies.

Table 8: Showing mean difference of Spirometry between mothers to preterm and mothers to full term babies.

No statistically significant difference in lung functions between mothers with preterm delivery and term delivery.

Bivariate analysis

Association between maternal respiratory symptoms and low birth weight (Tables 9-11)

Table 9: Association between low birth weight and maternal respiratory symptoms in rural and urban areas.

Table 10: Association between maternal lung function and low birth weight.

Table 11: Showing association between preterm birth and maternal lung functions.

There was no significant association between maternal respiratory health and preterm delivery. (Tables 12 and 13).

Table 12: Showing association between Small for Gestational Age and maternal respiratory health.

Table 13: Showing association between small for gestational age and maternal lung functions.

There was no significant association between maternal respiratory health and small for gestational age.

Assessment of influence of gestational age on the relationship between lungs functions and birth outcomes (Table 14)

Table 14: Showing association between maternal lung functions in different trimesters and birth outcomes for rural and urban areas.

There was a statistically significant association between FEV₁/FVC and preterm in the urban (p value<0.0001) and small for gestational age (p value<0.0001) in the rural area for all the three trimesters.

Multivariate logistic regression

Impact of maternal respiratory health on birth outcomes (Table 15)

Table 15: Showing association between maternal respiratory health and birth outcomes in the multivariate analysis.

In the urban area, LBW was statistically associated with recurrent nasal symptoms OR [1.69 (95% C.I; 1.0-2.8)] and prolonged secretion of phlegm OR [0.58 (95% C.I; 0.3-1.0)]. Pregnant women presenting with recurrent nasal symptoms were 1.69 times more likely to have a LBW child compared to the pregnant women with no recurrent nasal symptoms. While pregnant women with prolonged secretion of phlegm were 0.58 times less likely to have a low birth weight.

There was a significant association between FVC and LBW in the rural area OR [0.09 (99% C.I; 0.0-0.4)], pregnant women with mild reduction were 0.09 times less likely to have LBW compared to those with severe reduction of FVC. Preterm delivery was statistically significantly associated with FVC OR [0.39 (99% C.I; 0.2-0.8)] in the entire study population, pregnant women with mild reduction in the FVC were 0.39 times less likely to have a preterm delivery compared to pregnant women with a severe reduction in FVC.

This is the first research in Zambia exploring the association between pregnancy- related outcomes and maternal respiratory symptoms and lung functions among both rural and urban pregnant women. The current study found that impaired lung function is linked to biomass fuel smoke and this is supported by results of a study in Ethiopia that found that exposure to indoor air pollution has been linked with a reduction of forced vital capacity and an increased risk of acute respiratory problems [14]. In the urban area, LBW was statistically associated with recurrent nasal symptoms and prolonged secretion of phlegm. Pregnant women presenting with recurrent nasal symptoms were 1.69 times more likely to have a LBW child compared to the pregnant women with no recurrent nasal symptoms. While pregnant women with prolonged secretion of phlegm were 0.58 times less likely to have a low birth weight. Respiratory ill health has been cited in another study as being associated with adverse pregnancy outcomes. This study observed that the prevalence of adverse birth outcome is high among women with poor respiratory health. This finding is in line with the results of other researchers who also found that mothers with acute respiratory infectious diseases (ARID) during pregnancy had a longer gestational age at delivery than mothers without ARID [17]. Other studies elsewhere also observed that women who presented with asthma like symptoms of wheezing and breathlessness had a high prevalence of adverse birth outcomes for instance a study among Australian population showed that asthma in pregnancy is associated with increased poor outcomes for the mother and neonate and these poor outcomes include preterm birth, small for gestational age and preterm labor [18]. Similarly, Kallen and colleagues [19] also observed that the pregnancies of women with asthma are more likely to be complicated by preterm birth, lower birth weight and pre-eclampsia than pregnancies in non-asthmatic women [20].

This study found that there was a statistically significant association between FVC and LBW in the rural area. Pregnant women with mild reduction in FVC were 0.09 times less likely to have LBW compared to those with severe reduction of FVC. This is in accord with the findings in another study where low rates of fetal growth were observed in reduced lung function in adults [21] and a positive linear trend has been recorded between adult lung function and birth weight after adjusting for maternal factors [12]. Other findings elsewhere however, reported association of FEV₁ and LBW and observed that a direct relationship between maternal FEV₁ during pregnancy and infant birth weight [22]. Low FEV₁ has also been reported to be associated with preterm delivery by Odegaard and colleagues [23]. However, in our current study preterm delivery was significantly associated with FVC and pregnant women with mild reduction in the FVC were 0.39 times less likely to have a preterm delivery compared to pregnant women with a severe reduction in FVC. Our findings in this study showed that the gestational age of the pregnant women did not alter lung functions of the pregnant women. This agrees with findings in previous studies of impact of pregnancy on pulmonary function, results obtained by forced Spirometry, showed largely that forced vital capacity (FVC), forced expiratory volume in 1 second (FEV₁), and peak expiratory flow (PEF) remain unchanged during pregnancy [14].

Maternal respiratory health had an influence on the adverse birth outcomes of pregnant women. Therefore, pregnant women with respiratory disease need frequent monitoring of respiratory symptoms and simultaneously conducting spirometry to help optimize their respiratory health throughout pregnancy.