Stress and resilience during pregnancy: a comparative study between pregnant and non-pregnant women in Ethiopia

Introduction

The stress response, an adaptive component of physiology, represents a survival strategy during exposure to threats, adverse experiences or stressors in life (1). When activated, the stress response prepares the body for ‘fight or flight’ to promote safety and protection (1). However, chronic activation of the stress response affects the body negatively and impairs health, wellbeing and performance (2). The extent to which exposure to threats drives the perception of stress varies, however, depending on an individual’s ability to cope or adapt with stressors and successfully bounce back to the normal homeostasis from the effect of adversities (3). The impact of stressors on the stress response is mediated by coping and adaptive strategies, forms of resilience that buffer the adverse effects on health and wellbeing (4). The development of psychological resilience is a dynamic process across the life span formed as a product of the interaction between biological, psychological and socio-environmental factors (5,6). People with different levels of resilience therefore respond differently to a similar set of stressors, such that those with low resilience are more prone to the adverse consequences of stress.

Women in low-income countries experience disproportionate levels of stressors related to household responsibilities, as well as gender inequalities such as an elevated risk of malnutrition, dietary inadequacy and violence (in particular intimate partner violence) (7). Moreover, emerging evidence indicates that the state of pregnancy itself induce additional stress to pregnant women (8). The transition in social role associated with becoming a mother (9) may be accompanied by new physiological sources of stress, and potentially greater sensitivity to stress. This issue has major implications for public health, as experiencing stress during pregnancy has consequences not only for the mother, but also for the offspring who may be exposed to the physiological signals of stress through the placenta (10–12).

Stress in pregnancy can affect maternal health/wellbeing and quality of life by triggering maladaptive emotional and physiological states. Because of its negative impact on maternal and fetal health and nutrition during pregnancy, maternal stress may lead to adverse pregnancy and birth outcomes, such as shorter gestational age, prolonged labor, abortion, stillbirth, low birth weight, congenital anomalies, maternal perinatal infections, preeclampsia and hemorrhage (2,13,14). These associations tend to be of dose-response nature, whereby the greater the maternal stress, the greater the likelihood or magnitude of adverse outcomes. Perinatal complications and prolonged labor were also associated with antenatal common mental disorders in Ethiopian settings (15,16).

In turn, children born to stressed mothers have increased risk of morbidity, growth restriction, and cognitive disability, and may have an elevated risk of mental and behavioral problems during childhood such as anxiety, depression and attention deficit disorder (2,14,17–21). This scenario contributes to the persistent burden of growth stunting and development among children in LMICs (22). Finally, adults who survived stress in prenatal and postnatal life may have higher risk of non-communicable diseases (NCDs) and reduced human capital (23). In turn, this has implications for society, through reduced economic productivity (21,24,25). In this way, stress during pregnancy may contribute to an inter-generational cycle of disadvantage (26,27).

Research on prenatal stress is especially important in countries where poverty and adverse life circumstances are abundant. A recent WHO report indicated that maternal mental health is the missed component of maternal health in LMICs (28,29). To date, for example, there are no published data on stress and resilience during pregnancy in Ethiopia, a low-income country with a high burden of maternal mental health problems, child stunting and other adverse environments undermining optimal child development (30,31). The present study therefore aimed to investigate whether pregnancy is associated with higher levels of stress and lower resilience, by comparing these outcomes between pregnant and non-pregnant women in the city of Jimma, Ethiopia.

Method and participants

Results

Correlation between stress and resilience

Stress and resilience showed a negative correlation to each other (Figure 1). Pregnant women showed higher stress and lower resilience while non-pregnant women contrastingly showed lower stress and higher resilience.

• Download figure
• Open in new tab

Figure 1: Linear prediction for stress on resilience by pregnancy status

Comparison of stress and resilience between pregnant and non-pregnant women Unadjusted analysis

Both stress and resilience scores were normally distributed in the sample population. In unadjusted analyses, except PSS item 1 and 6, all PSS item scores were significantly higher in pregnant compared to non-pregnant women (Figure 2a). Numerical values are given in Supplementary Table 2a. Linear regression analysis showed that in an unadjusted model pregnancy was positively associated with higher PSS stress score (β=4.3; 95% CI: 3.4, 5.3). Similarly, except BRS item 2, all BRS item scores were significantly lower in pregnant compared to non-pregnant women (Figure 2b). Numerical values are given in Supplementary Table 2b. Linear regression analysis showed that pregnancy was associated with lower resilience score in an unadjusted model (β=-3.6; 95% CI: -4.5, -2.6).

View this table:

• View inline
• View popup
• Download powerpoint

Table 1: Background characteristics of study participant, n=320

View this table:

• View inline
• View popup
• Download powerpoint

Table 2: Stress and resilience in pregnant and non-pregnant women

View this table:

• View inline
• View popup
• Download powerpoint

Table 3a: Association of pregnancy with stress score

View this table:

• View inline
• View popup
• Download powerpoint

Table 3b: Association between pregnancy statuses with resilience score

• Download figure
• Open in new tab

Figure 2:

Mean difference in (a) Perceived Stress Scale (PSS) items and (b) Brief Resilience Scale (BRS) items between pregnant and non-pregnant women in a simple model

Adjusted analyses

In a fully adjusted regression model, all PSS items except item 1 and 6 were significantly greater in pregnant compared to non-pregnant women (Figure 2a). Numerical values are given in Supplementary Table 2a. For the overall PSS score, after adjusting for marriage, social support, religion, family size, income, education, occupation, household food insecurity status, distress, and current physical activity level, pregnancy was associated with greater stress (β=4.0; 95% CI: 2.9, 5.1).

Adjusted analysis

In adjusted analyses, except PSS item 1, all PSS item scores were significantly higher in pregnant compared to non-pregnant women (Figure 3a). Numerical values are given in Supplementary Table 2a. Linear regression analysis showed that in a fully adjusted model pregnancy was associated with higher stress score (β=4.1; 95% CI: 3.0, 5.2). Similarly, except BRS item 2, all BRS item scores were significantly lower in pregnant compared to non-pregnant women (Figure 3b). Numerical values are given in Supplementary Table 2b. Linear regression analysis showed that pregnancy was associated with lower resilience score in the adjusted model (β=-3.3; 95% CI: -4.5, -2.1).

• Download figure
• Open in new tab

Figure 3:

Mean difference in (a) PSS-10 item scores and (b) BRS-6 item scores between pregnant and non-pregnant women, in the fully adjusted model.

Lastly, in models that adjusted stress for resilience, and vice versa, pregnancy was associated independently with higher stress (β=2.8; 95% CI: 1.7, 3.9), and with lower resilience (β=-1.6; 95% CI: -2.8, -0.5) (Figure 4).

• Download figure
• Open in new tab

Figure 4:

Association of pregnancy with stress and resilience in the mutually adjusted models

Other factors associated with higher stress in the overall sample include household food insecurity score (β=0.21; 95% CI: 0.03, 0.38) in model 4, increased level of physical activity (β=0.24; 95% CI: 0.90, 3.58), decreased level of physical activity (β=0.25; 95% CI: 0.13, 0.37), distress (β=0.25; 95% CI: 0.03, 0.38), while resilience was associated with lower stress (β=-0.4; 95% CI: -0.48, -0.28) in model 5. Similarly, household food insecurity (β=-1.2; 95% CI: -0.18, - 2.2) and stress were associated with lower resilience (β=-0.4; 95% CI: -0.53, -0.31), while compared to no change in physical activity increased physical activity (β=2.1; 95% CI: 0.7, 3.5), decreased physical activity (β=1.2; 95% CI: 0.18, 2.2), and ever use of substance (β=1.4; 95% CI: 0.4, 2.4) were associated with higher resilience in model 5.

The mutually adjusted model explained 46% of the variance in PSS score and 38% of the variance in the BRS score. The associations between pregnancy with stress and resilience were stable across the different regression models indicating the robustness of the finding.

Discussion

In this study we compared stress and resilience between pregnant and non-pregnant women. Pregnant women had higher stress and lower resilience than non-pregnant women. The proportion of moderate to severe level of stress was higher in pregnant than non-pregnant women. Similarly, one-third of pregnant women compared to two-thirds of non-pregnant women had normal levels of resilience, suggesting that their coping strategies were hampered during pregnancy. Independent of potential confounders, pregnancy status increased stress by 4 points and decreased resilience by 3 points, and the associations with stress and resilience were also largely independent of each other.

Global data showed that the mean stress score in a sample of 1406 women recruited from different countries was 13.7±6.6 (32) which is lower than our finding in a sample of mixed pregnant and non-pregnant women (16.7±4.8) (ref). In the current study, the pregnant women had a much higher mean stress score (18.7±4.3) while the non-pregnant women had only slightly higher mean stress score (14.4 ±4.3) compared to the global data. Similar to our findings, previous studies have documented higher proportions of moderate to severe levels of stress in pregnant compared to non-pregnant women in Thailand (38), Saudi Arabia (39) and Iran (40). Unlike previous studies, we used a comparative study design to measure the association of pregnancy with stress and resilience so that we are able to investigate the independent association of pregnancy with both outcomes. As such, the current study provides more robust findings from a low income setting compared to previous work.

As indicated in previous study (8), various factors could have contributed to the higher burden of stress in this study. Women in general and pregnant women in particular are at a greater disadvantage in LMICs because of a high burden of responsibility with “unpaid care”, combined with very low levels of financial and non-financial reward (10,41–43). While life and role transitions to new parenthood and expectations regarding a new baby are sources of personal and family satisfaction, they can also be stressful experiences given uncertainty and concerns over adverse outcomes for the pregnant women and the offspring, which can lead to a range of negative psychological, physical, and mental consequences, all contributing to an increased state of stress (44,45). In addition, pregnant women require an increased level of psycho-social support to cope with the pregnancy alongside regular daily responsibilities; but in this study around 72% of pregnant compared to 51% non-pregnant women reported that their social support was inadequate. Moreover in a setting like Ethiopia where maternal mortality is high (46,47), all pregnancies are considered to be “between life and death” and this situation increases maternal stress significantly during pregnancy (8). All of these disadvantages that pregnant women encounter on top of their pregnancy are additional life stressors (48–51). Last, but also important, are the physiological and biological changes taking place during pregnancy such as changes in homeostasis, hormonal levels, body weight, and changes in energy metabolism, all of which can induce stress (52).

Pregnancy is associated with stress and resilience in opposing directions in the current study. While perceived stress score was higher, resilience score was lower in pregnant compared to non-pregnant women. At the first encounter, stressors/adversities affect or challenge an individual’s level of resilience or coping strategies and subsequently they progress to induce stress (4). While resilience is the ability to absorb shock/stressors, stress is the product of impaired interactions between stressors and coping mechanisms. During immediate exposure to stressors/adversities, resilience helps the body rapidly initiate acute stress responses through activation of the Hypothalamus-pituitary-adrenal (HPA) axis, leading to the release of cortisol preparing the body for ‘fight or flight’ responses (6). Our data in both groups (Figure 1) show that there is an inverse association between these outcomes, whereby higher resilience is associated with lower perceived stress. However, each group also showed substantial variability in this association where by pregnant women for the same level of resilience as to non-pregnant women showed higher stress scores. (Figure 1)

Normal resilience facilitates a rapid activation of the HPA axis to benefit the body from its activation and release of cortisol followed by a quick culmination of this process (53). In contrast poor resilience fails to control the continued sustained activation of the HPA axis, leading to a chronic increase in cortisol and resulting in uncontrolled stress (53). Resilience therefore helps to protect normal homeostasis in the body by enabling it to bounce back from the psychological, physiological and biological effects of stressful situations (3). Through these mechanisms, normal resilience is mostly inversely associated with pathological stress, anxiety or depression (54). Thus it is crucial to understand what and what levels of stressors overcome an individual’s resilience and what support/intervention types improve resilience so stressors can be overcome.

This is especially important for women during pregnancy to benefit both the mother and her offspring, improving the health and productivity of the next generation by breaking the intergenerational transmissions of stress. The fact that women in general have higher stress score in the current study could indicate the competing responsibility between unpaid household/family responsibility and their aspirations for personal development compounded by an unfavorable educational environment for females, unrelenting household responsibility and lack of social support making it all difficult to balance their personal development and family life (55,56) indicating the need for females friendly environment to improve women empowerment in low income settings. Moreover, higher level of household food insecurity, distress and increased physical activity have contributed to the higher stress and lower resilience score in the current study. Household food insecurity is a form of environmental adversities associated with poor mental health outcome (57) and is a common phenomenon in a LMICs. While it is difficult to explain the association between ever use of substance and higher resilience score, this might be attributed to financial access or freedom those women have compared to others. Consistent to previous studies (58,59) increased level of physical activity is associated with higher resilience but not with lower stress. This could happen because the increase in physically active might be due to the increased household domestic activities making them stressed, or otherwise stressed women might have tried physical activity to reduce their level of stress.

Conclusion and recommendation

Perceived stress is higher and resilience is lower in pregnant women compared to non-pregnant women in Ethiopia. There is a need for more research into the different stress response mechanisms and stress biomarkers during pregnancy. Moreover mechanistic studies and context relevant interventions to improve psychological coping and resilience, and to reduce stress are required so as to improve the health and wellbeing of the mother and her offspring.