The respiratory system is affected by anatomical and physiological changes associated with pregnancy. This was the base of our hypothesis that spirometry profiles of pregnant women are different from those predicted by their age and height if they were not pregnant. Therefore this study was designed to examine the profiles of lung function of pregnant African women using spirometry and compare them to profiles of non-pregnant women. In this study, FVC, FEV1, and PEF values of pregnant and non-pregnant women were lower than the values predicted by their age and height. Also, FVC, FEV1, and PEF values of pregnant women were lower than values of non-pregnant women.
We report mean spirometry test values in pregnant women which are lower than the means reported in Brazilians (20). The non-pregnant means of FVC, FEV1, PEF and their % predicted in this study are comparable to values reported from other studies done in Tanzania (29), Rwanda (30) and Mozambique (31). Our values were slightly higher than the other studies because of the age difference as the mean age of non-pregnant women was less than 30 years in this while it was more than 35 in the other studies. Also, the values of non-pregnant women in this study were lower than values recorded in Europeans and Australians (45), Asians (46), and Scandinavians (47) but FEV1/FVC ratio was higher. Like ours, other studies have reported lower spirometry profiles in African decency which could not be explained by anthropometric and skin color differences alone (5, 48). A portion of this could be explained by lower seating height and socio-economic status which were reported to relate to lower values. Even so, the values are considered normal since the prognosis has not been different (49–52). In addition to ethnic differences, we found lower values probably because we did not administer a bronchodilator prior to spirometry unlike in the previous studies.
Interestingly, a phasic relationship between age and spirometry test values of non-pregnant women was noted. There was an increase to a peak, followed by a decrease in spirometric values. A similar pattern was observed in pregnant women for FEV1/FVC, PEF, and PEF%. However, the peak age for FVC, FEV1, and PEF was earlier with lower values in pregnant women. The spirometry test values have been known to increase with age then peak around 25 years before starting to decline (5, 47, 53, 54). This is thought to occur as a part of the aging process. After peak age, pulmonary elastic recoil decrease with age due to progressive loss of lung tissue elasticity and increase of chest wall stiffness resulting in the decline of lung function (55–59). Also, it could be partly due to a decrease in spirometry performance with aging. Hence, age has been an important factor in spirometry test values predicting equations. Pregnancy factors could have influenced the pattern observed among pregnant women in our study.
Similar to previous studies, FVC, FEV1, and PEF of both pregnant and non-pregnant women increased with height (29, 30, 47). Height also has been an important factor in spirometry prediction equations together with age (3, 29, 53, 60–65). However, FVC% and FEV1% decreased as height increased. This could mean that as height increased participants were more likely to have lower than expected FVC and FEV1 values but also it could be a reference equation over predicting expected values. Reference values have been reported to over predict spirometry test values in different populations (60, 61, 64). Other studies have found a difference in prediction even when references were derived from a closely related population such that abnormal findings in one reference were deemed normal by the other (65). There was no significant relationship between height and FEV1/FVC in pregnant and non-pregnant women. This was in line with other studies (47) and it could be due to the equal effect of height on FEV1 and FVC.
FVC, FEV1, and PEF of pregnant women increased with weight, peaked at 61-70Kg then decreased. Non-pregnant women's values decreased when women were becoming overweight and obese. Despite such a pattern, neither weight nor BMI appeared to statistically significantly affect FVC, FEV1, or PEF in neither pregnant nor non-pregnant women after adjusting for age and height. This has been found by several other studies (29, 30, 66, 67). However other studies have demonstrated a negative effect of the increasing waist to hip ratio (WHR) and weight gain on FEV1 and FVC (68, 69). This could be for the reason that quantification of body mass and its index is not specific to the distribution of body composition while fats in hips, thighs, gluteal regions, and breasts are less likely to affect lungs, diaphragm, and chest wall mechanics (46). While this study was limited to FVC, FEV1, and PEF, other studies have found an inverse relationship between increasing BMI and vital capacity, total lung capacity, and functional residual capacity (70, 71).
The mean FVC, FVC% FEV1%, PEF, and PEF% were higher in parous than nulliparous and first birth showed the greatest effect on the pattern in both pregnant and non-pregnant. Despite that, only FVC% and FEV1% were statistically significantly related to parity in non-pregnant women and the relationship disappeared after adjusting for age, height, and weight. Similar results were found in a longitudinal study that involved pregnant women (20). However, other studies found a significant adjusted positive effect of parity on spirometry test values (54, 67). It has been postulated that the hormonal effects of pregnancy to compensate for mechanical changes and maintain lung function persists even after the uterus have returned to its small size (67, 72). The median FEV1/FVC ratio was lower in parous than nulliparous in both pregnant and non-pregnant women but was statistically significant only in pregnant women after adjusting for age, height, and weight. Similar findings have been presented by other studies (20). This could be due to disproportionate changes between FVC and FEV1.
Spirometry test values decreased as gestation age advanced. This is in line with other studies conducted previously (20, 73, 74). This decline has been attributed to the limited maternal effort as gestation advances due to an increase of maternal weight, uterine enlargement, and a degree of pulmonary edema (Brancazio, Laifer, and Schwartz, 1997). However, spirometry test values have been observed to remain within normal limits (18, 19). Other studies have reported values that increased during pregnancy and persisted to the postpartum period (33, 66, 67). But those studies concentrated on whether spirometry test values were normal as compared to a known range or not. In our study, we compared absolute values and their % of predicted values of pregnant women at different gestational periods.
FVC, FEV1, and PEF values of pregnant women were significantly lower than values predicted by age and height if they were not pregnant. The observation was similar for non-pregnant women. The other study that was done in Tanzania also reported a similar finding (76). This suggests that the reference equation derived from non-African settings could have over-predicted expected values. Also, the other study done on young men in Tanzania concluded that spirometry reference equations developed from non-African populations tended to overpredict measurements of black Africans (76). Likewise, it has been noticed by other studies in which reference values over-predicted expected values (60, 61, 64).
Nevertheless, when compared to non-pregnant women; FVC, FEV1, and PEF of pregnant women were significantly lower even after adjusting for age, weight, and parity. This could be explained by ribcage and volume displacement long known to take place during pregnancy (13, 15, 19, 77). However, Le Merre et al discussed that changes during pregnancy do not cause significant respiratory functional changes since mechanical effects are balanced by hormonal factors (78). Unlike other studies which compared pregnant values against the established normal range, this study compared values of pregnant women against values of non-pregnant women.
To our knowledge, this is the first study on lung function conducted among pregnant women using spirometry in African settings. Spirometry test values of pregnant women were compared against values of non-pregnant women rather than comparing against the established normal range of values. This study was able to adhere to standard operating procedures and infection prevention protocol
This study was not without limitations. Non-pregnant healthy women were likely to hesitate to participate in the study as they would feel a lack of need for tests. Only women who booked their first visit in their first trimester were included in a study. Pregnant women were obtained by random sampling while non-pregnant controls were obtained consecutively. Also, many potential participants hesitated to participate worrying that they were tested for the Coronavirus. These could have influenced the nature of women who participated in this study and limited our ability to match the characteristics of pregnant women. Our study was limited to spirometry, therefore, could not explain other observations which would be well explained by other lung function testing methods such as measuring static lung volumes. Also, we did not quantify hormonal effects on the spirometry profile by hormonal assay.