Exploring the impact of pregnancy on cognitive function: a comparative study in a low-income setting

Abstract

Background

Cognitive dysfunction is a significant contributor to mental health complexities during pregnancy, potentially leading to heightened rates of pregnancy-related mortality and inadequate prenatal care. However, limited research has been conducted to explore the relationship between pregnancy and cognitive decline, especially in low-income settings such as Pakistan. Therefore, this study aimed to establish a clear link between cognitive function and pregnancy.

Methods

A cross-sectional comparative study was conducted at a tertiary care hospital in Karachi, Pakistan with a sample size of 160 participants, divided into two groups of 83 pregnant (aged 25.63 ± 4.22) and 77 nonpregnant women (aged 27.79 ± 3.89). First, the participants were interviewed to collect demographic information and pregnancy status. Then, the Montreal Cognitive Assessment (MoCA) scale, which evaluates cognitive function across multiple domains, including visuospatial/executive function, naming, attention, language, abstraction, delayed recall, orientation, and memory was used on each group separately. The analysis investigated the relationship between cognitive function and pregnancy, considering the influence of low-income status and gestational age. The statistical analyses included Spearman Rho (for non-normal data), t-tests, and linear regression models. T-tests were used to compare the means of MoCA scores between different groups and to analyze the effect of pregnancy status on the specific domains of MoCA. Multiple linear regression models were employed to examine the relationships between MoCA scores and various predictors, such as pregnancy status, education level, gestational age, and active complaints.

Results

The study found a significant difference in MoCA scores between pregnant and nonpregnant women (B=-1.55, t=-2.37, p = 0.019), indicating a decline in cognitive function during pregnancy. Education level (B = 2.34, t = 8.38, p = 0.000) and gestational age (B=-1.61, t=-2.51, p = 0.014) were identified as significant factors influencing cognitive function. Higher education was associated with better cognitive function while increasing gestational age correlated with a decline in cognitive function. In addition, active complaints (B=-1.86, t=-2.25, p = 0.028) during pregnancy were associated with lower MoCA scores.

Conclusion

Our preliminary analyses suggest that there is notable cognitive impairment associated with pregnancy. More attention and research in this aspect can contribute to better prenatal care and promote the well-being of pregnant women.

Background

The significance of adequate prenatal care cannot be overstated, both in developing and developed countries. Insufficient care during pregnancy can lead to adverse outcomes such as premature births, low birth weight, and an increased risk of miscarriage [1]. Shockingly, global statistics from Maternal Mortality Review Committees indicate that approximately 1 in 5 women (22%) died during pregnancy in 2020, with more than 84% of these deaths potentially preventable through appropriate measures [2]. Emerging research suggests that cognitive function may be influenced in pregnant women by various physiological and environmental factors leading to cognitive impairment during pregnancy which may contribute to maternal morbidity by affecting self-care, adherence to prenatal care, and the ability to recognize and respond to health complications [3].

Cognitive impairment is defined as a range of difficulties, from mild to severe, in memory, learning, concentration, decision-making, and other aspects of daily life [4]. Mild cognitive impairment has been shown to have a negative effect on the overall well-being of individuals, including their physical and emotional health [5, 6]. Additionally, physical discomfort, metabolic changes, and pre-existing health conditions (e.g., thyroid dysfunction, and gestational hypertension) may also contribute to cognitive difficulties in pregnant women. Emotional and psychological adjustments, including stress and anxiety, further interact with these physical changes, potentially exacerbating cognitive challenges during pregnancy [7] and thus decreased well-being [8].

Cognitive impairment during pregnancy can influence both maternal and fetal health, potentially leading to negative pregnancy outcomes [9]. This impairment may be influenced by steroid hormones, such as estradiol and progesterone, which are normally high in pregnancy, crossing the blood‒brain barrier and potentially affecting different regions of the neurological system [7]. Furthermore cognitive changes in pregnancy have itself been linked to hormonal fluctuations, including cortisol alterations which may impact fetal brain development and emotional regulation [10]. Imaging studies suggest that maternal cognitive changes can contribute to increased activity in the fetal amygdala, which has been associated with emotional conditions such as anxiety and depression later in life [11]. Furthermore, maternal cognitive instability during pregnancy may affect fetal hippocampal development and brain chemistry, potentially influencing learning abilities and emotional expression [12]. Physical discomfort, hormonal fluctuations, and emotional adjustments can also collectively contribute to cognitive impairment [7].

In Pakistan, an estimated 30,000 women die each year from pregnancy-related causes, with a maternal mortality rate of 276 per 100,000 births annually [13]. Lack of prenatal care, limited awareness of maternal health issues, and socioeconomic barriers significantly contribute to this alarming scenario. Despite existing research on cognitive function changes in pregnancy, little is known about the specific cognitive domains affected and their relationship with gestational age, particularly in low-income settings such as Pakistan. Therefore, further research is warranted to explore the relationship between cognitive impairment and pregnancy, especially in this region to which our research contributes.

The objective of this study was to investigate the relationship between cognitive function and pregnancy while considering the influence of low-income status and gestational age. Additionally, we aimed to identify the most and least affected cognitive domains during pregnancy. By addressing these objectives, our research will provide valuable insights to recognize the level of cognitive impairment in pregnancy, contributing to an improved understanding of maternal health. Establishing a clear link between the two factors is vital for understanding the root causes and developing effective interventions. Therefore, the findings of this study will have important clinical implications, ultimately leading to better outcomes for pregnant women and promoting their overall well-being.

Method and materials

A cross-sectional comparative study was conducted at the obstetrics outpatient department (S) of Jinnah Postgraduate Medical Center, the largest public tertiary care hospital in Karachi. The sample size was calculated using a study by Davies et al. [14], which reported a standardized mean difference of 0.52 between cognitive function in pregnant and nonpregnant women. This was converted to an odds ratio of 2.568 according to the Cochrane handbook guidelines [15]. Using a power of 80% and a 95% confidence interval, a sample size of 148 participants (74 pregnant and 74 nonpregnant) was determined using Open Epi 3.01 [16]. To account for errors and increase precision, a total of 160 participants were recruited using nonprobability convenience sampling.

Inclusion criteria for the study included confirmed pregnancy by ultrasound for pregnant participants, age between 18 and 35, and the ability to speak and understand Urdu or English. Exclusion criteria comprised a history of psychiatric illness, dementia or delirium, multiple strokes, brain tumor, or brain damage resulting from other medical, neurological, or autoimmune diseases that could affect cognitive function.

The questionnaire used in the study was an interview-administered questionnaire consisting of two parts. The first part collected demographic information such as age, education, active complaints, comorbidities, pregnancy status, and socioeconomic status. The second part included the Urdu version of the Montreal Cognitive Assessment (MoCA) scale [17], which assessed participants’ cognitive status based on visuospatial/executive function, naming, attention, language, abstraction, delayed recall, orientation, memory, serial subtraction, and trail making. The MoCA scale was chosen for its comprehensive assessment of multiple cognitive domains, enabling a thorough evaluation of cognitive status. Its extensive validation and widespread use further strengthened its suitability for research [18].

We further assessed comorbid conditions that could influence cognitive function, including diabetes, thyroid disorders, hypertension, and asthma, as these may be exacerbated by pregnancy. Patients with cognitive-impairing conditions such as dementia, depression, and anxiety disorders were excluded to minimize bias. Additionally, we inquired about active pregnancy-related symptoms like nausea, vomiting, fatigue, and vaginal discharge. For nonpregnant participants, we considered the primary reason for their visit and excluded those with conditions potentially affecting cognition, such as postpartum depression, infertility, or UTIs, to ensure a fair comparison between groups. The study obtained ethical approval from the Institutional Review Board at Jinnah Sindh Medical University.

Statistical analysis

Statistical analysis was conducted using SPSS software v25 (IBM Inc., USA). Continuous normally distributed data are described using the mean and standard deviation, while nonnormally distributed data are presented using the median and interquartile range. Categorical data are reported as frequencies and proportions. The normality of the data was assessed using skewness, kurtosis, and histograms with a normal curve. Independent sample t tests and ANOVA were used for normally distributed continuous data comparisons, while the chi-square test was employed for categorical variables. Spearman’s rho was used to evaluate correlations between nonnormally distributed independent variables and MoCA scores. A significance level of p < 0.05 was considered.

Multiple linear regression analyses were performed to examine the association between study groups and cognition status, satisfying the assumptions of linearity, independence of errors, homoscedasticity, multivariate normality, and no multicollinearity. Two regression models were created using the stepwise method. The first model included pregnancy status, education, age, occupation, comorbidities, active complaints, and socioeconomic status scores as predictors for MoCA scores. In the second model, pregnancy status was replaced by gestational age. This substitution was made to address concerns over collinearity and facilitate the assessment of individual effects. The coefficients were examined to determine the influence of each factor on MoCA scores.

Results

Clinical and demographic characteristics of sample

The study analyzed data from 160 participants, including 83 pregnant and 77 nonpregnant individuals. As summarized in Table 1, nonpregnant respondents had a higher mean age (27.79 ± 3.88) than pregnant respondents (25.63 ± 4.22). The majority of respondents had an education level equal to or below matriculation (70.7%). Most patients were housewives (88.2%), with a greater number of working women among the nonpregnant respondents. The majority of respondents belonged to a lower socioeconomic status (82.5%), and no known comorbidities were present in the majority of respondents (68.2%).

Table 1.

Clinical and demographic characteristics of the sample

	Nonpregnant, N(%)	Pregnant, N(%)	p value
Age (mean ± SD)	27.79 ± 3.88	25.63 ± 4.22	0.001*
Education			0.666
Not formerly educated	6, 3.8 (%)	7, 4.4 (%)
Primary	23, 14.4 (%)	21, 13.1 (%)
Matriculation	24, 15.0 (%)	32, 20.0 (%)
Intermediate	14, 8.8 (%)	17, 10.6 (%)
Graduate or higher	10, 6.3 (%)	6, 3.8 (%)
Occupation			0.006*
Full time house wife	62, 38.8 (%)	79, 49.4 (%)
Working woman	15, 9.4 (%)	4, 2.5 (%)
Socioeconomic Status			0.838
Lower class	63, 39.4 (%)	69, 43.1 (%)
Middle class	14, 8.8 (%)	14, 8.8 (%)
Comorbidities			0.047
Comorbidity Present	32, 20.0 (%)	22, 13.8 (%)
No known comorbidity	45, 28.1 (%)	61, 38.1 (%)
Nature of Visit
Active complaint	37, 23.1 (%)	43, 26.9 (%)	0.752
No active complaint	40, 25.0 (%)	40, 25.0 (%)

*significant at p < 0.05

MoCA scores among pregnant and non-pregnant respondents

As illustrated in Table 2, the average MoCA score for pregnant women was 20.65 (SD = 4.63), while for nonpregnant women, it was 22.27 (SD = 4.68). Language scores were significantly different between pregnant (1.47, SD = 0.85) and nonpregnant (1.88, SD = 0.81) women. Delayed recall scores were also significantly different between pregnant (2.16, SD = 1.67) and nonpregnant (3.22, SD = 1.68) women.

Table 2.

Mean MoCA scores of individual domains

	Nonpregnant (N = 77) Mean (SD)	Pregnant (N = 83) Mean (SD)	p value
Visuospatial/Executive	3.30 ± 1.33	3.13 ± 1.46	0.452
Naming	2.23 ± 0.60	2.29 ± 0.53	0.538
Attention	4.14 ± 1.46	3.76 ± 1.72	0.131
Language	1.88 ± 0.81	1.47 ± 0.85	0.002*
Abstraction	1.14 ± 0.82	0.99 ± 0.79	0.226
Delayed Recall	3.22 ± 1.68	2.61 ± 1.67	0.023*
Orientation	5.73 ± 0.66	5.55 ± 0.98	0.218
Total MoCA score	22.27 ± 4.68	20.65 ± 4.63	0.029*

*significant at p < 0.05

Multivariate regression models evaluating variables affecting cognitive function in pregnancy

Two multivariate regression models were employed to evaluate variables affecting cognitive function. The first model included MoCA scores as the dependent variable and pregnancy status, education, age, occupation, comorbidities, active complaints, and socioeconomic status as predictors. The second model replaced pregnancy status with gestational age in the predictor variable list.

Multivariate regression model 1 results

This model included MoCA scores as the dependent variable and pregnancy status, education, age, occupation, comorbidities, active complaints, and socioeconomic status as predictors. The level of education had a positive relation with MoCA scores (B = 2.34, t = 8.38, p = 0.000), indicating that higher education was associated with better cognitive function. Pregnancy status had a negative relation with MoCA scores (B=-1.55, t=-2.37, p = 0.019), suggesting that being pregnant was associated with lower cognitive function. Other predictor variables did not show a significant impact on MoCA scores. The first model accounted for 33% of the variance in total MoCA scores (see Table 3).

Table 3.

Multivariate regression model 1

Variable	B	SE	t	95% CI	β	p value
Constant	17.51	0.71	24.77	[16.109, 18.902]
Pregnancy Status	-1.55	0.61	2.37	[-2.68, -0.28]	-0.16	0.019
Education	2.34	0.28	8.38	[1.85, 2.95]	0.56	< 0.000

Note. R = 0.574; F(2,157) = 38.60; p*<0.05

Multivariate regression model 2 results

The second model replaced pregnancy status with gestational age in the predictor variable list. Education had a positive relation with MoCA scores (B = 2.20, t=-5.55, p = 0.000), indicating that higher education was associated with better cognitive function. Gestational age had a negative relation with MoCA scores (B=-1.61, t=-2.51, p = 0.014), suggesting that as gestational age progressed, cognitive function declined. The absence of active complaints also had a negative relation with MoCA scores (B=-1.86, t=-2.25, p = 0.028). The second model accounted for 37.3% of the variance in total MoCA scores (see Table 4).

Table 4.

Multivariate regression model 2 results

Variable	B	SE	T	95% CI	β	p value
Constant	21.21	1.86	11.41	[17.54, 24.89]
Education	2.20	0.40	5.55	[1.42, 2.98]	0.50	< 0.000
Gestational Age	-1.61	0.64	-2.51	[-2.88, -0.34]	-0.23	0.014
No Active Complaints	-1.86	0.83	-2.25	[-3.49, -0.23]	-0.20	0.028

Note. R = 0.611; F(3,78) = 15.5; p*<0.05

Variables influencing MoCA scores among participants

Gestational age showed a statistically significant relation with mean MoCA scores (p = 0.019). Mean MoCA scores decreased progressively between the first trimester (mean MoCA score = 24.0), second trimester (mean MoCA score = 21.4), and third trimester (mean MoCA score = 19.4). Negative correlations were found between the number of pregnancies and MoCA scores (Spearman rho=-0.177, p = 0.025). However, no statistically significant correlations were observed between the number of miscarriages or number of children and MoCA scores.

Discussion

The objective of our study was to investigate the differences in cognition between pregnant and nonpregnant women. We utilized the MoCA scale to measure cognition levels, employing various exercises designed to challenge participants’ cognitive abilities. Our findings revealed a statistically significant difference in MoCA scores between pregnant and nonpregnant women, thereby supporting our first hypothesis that pregnant women experience a decline in cognition. Additionally, we found support for our second hypothesis that cognition levels decrease as gestational age increases.

Educational background emerged as a significant factor influencing participants’ performance on the MoCA scale, with higher education correlating with better scores. Moreover, both pregnant and nonpregnant women who reported active complaints displayed lower MoCA scores, indicating a negative relationship between active complaints and cognition. These findings align with previous studies [14, 19, 20] that have investigated cognitive decline during pregnancy. Specifically, our study identified a decline in recall memory among pregnant women, which has been consistently reported by other researchers [14, 19–21]. Furthermore, Jessica F. Henry and Barbara B. Sherwin’s study [19] specifically focused on pregnant women in their third trimester and observed decreased memory compared to nonpregnant women.

Our study also highlighted the poor language skills of pregnant women in comparison to their nonpregnant counterparts. To the best of our knowledge, only Guilia Barda et al.‘s study [22] had previously examined language skills as a variable in assessing cognition among pregnant and nonpregnant women, and their findings were consistent with ours.

Regarding other cognitive domains, such as visuospatial/executive functions, attention, naming, and abstraction, the MoCA scale did not detect significant differences between pregnant and nonpregnant women in our study. These results support a meta-analysis [14] that encompassed 17 cross-sectional studies on cognitive impairment during pregnancy. The meta-analysis indicated that attention spans and executive functioning, mediated by the frontal cortex, remain unaffected by pregnancy.

Notably, our study discovered an inverse relationship between the presence of active complaints and MoCA scores among participants, regardless of pregnancy status. No other studies examining cognitive impairment in pregnant and nonpregnant women have specifically investigated the impact of active complaints. Thus, future research should incorporate active complaints as a variable to further explore their role in cognitive decline among pregnant women.

Additionally, our study identified a positive correlation between participants’ level of education and cognition, independent of their pregnancy status. This finding aligns with the literature [23, 24], which consistently indicates that education independently influences cognition.

Several limitations may have influenced our study results. First, there is a potential for selection bias, as our sample was derived from a single tertiary-care hospital, primarily representing low-middle class families, thereby limiting the generalizability of our findings to the entire population of pregnant and nonpregnant women. Second, the cross-sectional and nonrandomized design of our study precludes establishing causal relationships between pregnancy and cognitive status, as well as examining changes in cognition over time. A prospective cohort study design would better facilitate the exploration of causal relationships. Last, the stressful environment of a public hospital, characterized by a high patient influx, may have impacted participants’ MoCA scores.

Our study demonstrates a negative linear relationship between cognitive scores and pregnancy status. Additionally, gestational age appears to influence cognition, with better MoCA scores observed in early pregnancy stages compared to later phases. By addressing the limitations of our study and incorporating active complaints as a variable, future research can contribute to a deeper understanding of cognitive differences between pregnant and nonpregnant women.

Conclusion

This study provides valuable insights into the relationship between cognitive function and pregnancy, shedding light on the impact of pregnancy on cognitive abilities. The findings confirm the presence of cognitive decline during pregnancy, particularly in the domains of language and delayed recall. Furthermore, the study highlights the importance of education in mitigating cognitive impairment, suggesting that higher education levels are associated with better cognitive function. Additionally, gestational age and the presence of active complaints were identified as factors influencing cognitive function, with cognitive decline being more pronounced as pregnancy progresses and active complaints are present. These findings underscore the need for targeted interventions and support systems to address cognitive impairment in pregnant women, particularly those from low-income backgrounds. By improving our understanding of the factors influencing cognitive function during pregnancy, healthcare professionals can develop effective strategies to enhance the well-being and overall outcomes of pregnant women.

Abbreviations

OPD

Outpatient Department

MoCA

Montreal Cognitive Assessment

ANOVA

Analysis of Variance

Author contributions

AA and MS conceived the study and designed the research. AA and MS performed the data analysis and interpretation. AA, MS, LH, GR, DZ, MA, and MS conducted the literature search and contributed to the manuscript writing. AA and LH finalized the manuscript. All authors contributed to data collection. AA led the research and publication procedure, assisted by LH. All authors reviewed and approved the final manuscript. SK provided a critical analysis of the manuscript.

Funding

The authors funded this study on their own.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted after receiving ethical approval from the Institutional Review Board at Jinnah Sindh Medical University (Reference No. JSMU/IRB/2022/690). Approval was also granted by JPMC obstetrics and Gynaecology Department to allow data collection on their premises. Informed consent was obtained from all participants prior to their participation in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.