Health literacy and its determinants among pregnant women in Portugal | BMC Public Health

Sample and items description

The study included a sample of 886 pregnant women with a mean age of 31.09 years (± 5.60 years). To facilitate participation and ensure convenience for pregnant women, the survey, which included the HLS₁₉-Q12 instrument, was offered in three formats to achieve a homogeneous sample. Participants could respond digitally via a QR code or link distributed during pregnancy appointments (28.8%), through an interview conducted in primary healthcare or hospital settings (45.4%), or by self-completing a paper-based version (25.8%). Among those who received the paper version, the response rate was 76%.

The participants ranged in age from 18 to 51 years. More than half of the sample (53.3%) were aged between 30 and 39 years, 39.7% were between 18 and 29 years, and 7.0% were 40 years or older. Most participants were Portuguese nationals (81.2%), with 18.8% born in other countries. Educational attainment was relatively high, as 42.6% had completed university education, and 40.7% had finished high school. Regarding marital status, 73.4% were married or in a common-law relationship, while 25.2% were single. In terms of employment, 77.8% of the participants were professionally active. Among these, 52.5% were employed in technical occupations, 27.3% held intellectual or scientific roles, and 20.2% worked in undifferentiated professions. Additionally, 69.2% worked for others, 8.1% were self-employed, and 1.2% were family workers. However, 21.4% of participants were not engaged in professional work, including students, domestic, unemployed individuals, or retirees. Regarding housing, 46.4% owned their homes, 32.5% rented, and 20.8% lived with relatives, while 0.3% resided in social housing. Notably, 15.2% of participants reported having training in a healthcare profession, while 84.8% did not. The analysis of financial stress among participants reveals that a majority faced challenges in meeting their expenses. Specifically, 61.7% of participants reported finding it “difficult” or “very difficult” to pay their expenses, indicating a significant level of financial strain. Conversely, 38.3% of participants found paying expenses to be “very easy” or “easy,” suggesting a minority experienced financial ease. Further details can be found in Table 1.

Health behaviors and lifestyles and health status were examined to better understand factors influencing maternal health. Regarding smoking behavior, 74.0% of participants reported never smoking, while 13.9% smoked before pregnancy, 4.9% quit during pregnancy, and 7.2% smoked occasionally. Tobacco smoke exposure was reported by 32.3% of participants. Alcohol or psychoactive substance use were reported by 33.2% of pregnant women. Among these, 16.8% consumed substances before pregnancy, 7.6% quit during pregnancy, and 8.8% used them occasionally. Physical activity levels varied, with 34.9% engaging in light activity, 30.2% in heavy activity, and 25.4% reporting no physical activity. Fruit and vegetable consumption were notably high, with 89.4% reporting heavy intake. Pre-pregnancy BMI classifications revealed that 55.9% had a normal BMI, 24.5% were overweight, and 15.0% were classified as obese. Regarding self-perceived health, 69.8% rated their previous health as good or very good, while 29.0% considered it fair, and 1.2% described it as bad or very bad. During pregnancy, 11.2% felt their health had worsened, while 81.7% reported no change, and 7.1% indicated improvement. Chronic diseases or disabilities were reported by 19.2% of participants, with 30.6% finding it difficult to manage these conditions. Additionally, 12.8% experienced no activity limitations, 5.5% reported some limitations, and 0.9% faced significant limitations. Detailed results are provided in Table 2.

The study also examined gynecological and obstetric variables to provide insights into participants’ health profiles and pregnancy care (Table 3). Regarding the use of contraceptive methods, 89.7% of participants reported using contraception, while 10.3% did not. Cervical cytology results showed that 12.8% had never undergone the procedure. Among those who had, 36.6% underwent it within the last 12 months, 24.4% between one and two years ago, 13.3% between two and three years ago, and 13.0% more than three years ago. The mean gestational age was 29.60 weeks (± 10.61). Most participants were in their third trimester (28–41 weeks; 68.3%), followed by the first trimester (10–13 weeks; 19.1%) and the second trimester (14–27 weeks; 12.6%). Preconception care was reported by 57.3% of participants, while 42.7% did not attend such consultations. Planned pregnancies accounted for 66.6% of cases, while 33.4% of pregnancies were unplanned. In terms of pregnancy surveillance, the majority (88.8%) initiated care between 1 and 11 weeks of gestation, 10.7% began care between 12 and 27 weeks, and 0.5% started care at 28 weeks or later. Pregnancy risk were identified in 25.8% of participants, while 74.2% report a low risk pregnancy. Regarding parity, 57.4% of participants were experiencing their first pregnancy, while 42.6% had one or more children, with a mean of 0.57 children (± 4.21). Antenatal care sites varied, with 43.3% receiving care at both health centers and private clinics, 32.5% exclusively at health centers, 14.1% at health centers and hospitals, and 10.0% across health centers, hospitals, and private clinics. When asked about breastfeeding intentions, 92.6% of participants intended to breastfeed, 5.8% were undecided, 1.2% did not intend to breastfeed, and 0.5% reported being unable to breastfeed. Regarding childbirth preparation programs, 39.4% of participants did not attend, 32.1% were already attending, 17.6% intended to attend, and 10.9% had not yet decided or could not attend, probably because they are at risk of premature birth.

Distribution of limited health literacy across sample subgroups

The distribution of limited health literacy across sample subgroups were calculated using data from participants, ranging from 856 for the “Access” dimension of health information processing to 885 for “Health Promotion,” a domain of health literacy. Specifically, the general health literacy score was based on 875 participants, while the domain scores included 875 for “Healthcare” and 879 for “Disease Prevention.” For the dimensions of health information processing, the number of participants was 880 for “Understand,” 883 for “Appraise,” and 885 for “Apply.” The variation in sample size reflects the number of pregnant women who provided at least 80% of valid responses required for the calculation of each specific score.

A mean general health literacy score of 68.31 (± 10.92) was observed, which was lower than the mean scores for each of the six sub-indexes. These ranged from 68.32 (± 16.77) in the “Access” dimension of health information processing to 72.50 (± 11.30) in the “Apply” domain of health information processing. An exception was noted in the “Disease Prevention” domain, which exhibited a slightly lower mean score of 67.48 (± 12.83).

The distribution of general health literacy (HL) among pregnant women, as presented in Fig. 1, shows a substantial proportion of participants with intermediate levels of HL. Specifically, 4.2% were classified as having inadequate HL and 42.5% as problematic, indicating that nearly half of the sample experienced difficulties in accessing, understanding, or using health information. An equal proportion of participants (42.5%) demonstrated sufficient HL, suggesting functional skills to manage health-related tasks and decision-making during pregnancy. Only 10.7% of participants were classified in the excellent category, reflecting a relatively small subgroup with the highest levels of autonomy and confidence in dealing with health information.

Levels (%) of general health literacy

The analysis of health literacy across the three health information domains (Healthcare, Disease Prevention, and Health Promotion) reveals domain-specific patterns in the distribution of HL levels among pregnant women. In the Healthcare domain, 6.2% of participants were classified as having inadequate HL and 24.8% as problematic, while 55.1% reported sufficient HL and 13.9% excellent. These findings suggest that most women felt confident navigating the healthcare system and interacting with medical professionals, possibly due to their regular contact with antenatal care services. In the Disease Prevention domain, HL levels shifted slightly, with 7.1% of participants classified as inadequate and 33.1% as problematic. Meanwhile, 47.1% reported sufficient HL and 12.7% excellent. This domain presented the highest proportion of participants in the problematic category, indicating potential difficulties in understanding or applying preventive information, such as vaccinations, screenings, or behavioral risk avoidance strategies. The most favorable distribution was observed in the Health Promotion domain, where only 3.2% of participants had inadequate HL and 15.7% problematic, while 66.1% were classified as sufficient and 15.0% as excellent. These results suggest that participants were more confident in engaging with information aimed at maintaining or improving general well-being, such as nutrition, physical activity, or stress management during pregnancy. Overall, the data indicate that while healthcare navigation appears to be relatively well-managed, there are important gaps in understanding and acting upon preventive strategies. Strengthening communication related to disease prevention may help reduce risk behaviors and improve maternal and fetal health outcomes (Fig. 2).

Health Literacy domains (%)

Health literacy performance in each of the four dimensions of information processing (access, understand, appraise, and apply) demonstrates distinct patterns. The access dimension presented the highest proportion of participants with inadequate HL (13.0%) and also the highest problematic HL level (25.5%), indicating that locating and obtaining health-related information may be the most challenging task for many pregnant women. In contrast, the apply dimension showed the lowest percentage of inadequate HL (4.6%) and problematic HL (9.9%), while presenting the highest proportion of sufficient HL (72.8%) and a relatively elevated excellent level (12.7%). This suggests that once information is understood, most participants felt capable of integrating it into their health decisions. For the understand dimension, 11.1% of participants were classified as inadequate, 15.8% as problematic, 63.3% as sufficient, and 9.3% as excellent. Regarding the appraise dimension, 8.2% were inadequate, 16.9% problematic, 63.6% sufficient, and 11.3% excellent. These results suggest that while comprehension and judgement of health information are generally adequate, they remain more complex than direct application. Overall, the data indicate that challenges are most prominent in the early stages of processing health information, particularly in access and appraisal, whereas applying information is reported more confidently. This may reflect the structured support provided in antenatal care but also highlights potential weaknesses in information navigation and evaluation skills during pregnancy (Fig. 3).

Dimensions of health information processing (%)

Prevalence of limited health literacy across participant subgroups

Table 4 presents the descriptive findings based on a recoded variable in which the “Inadequate” and “Problematic” categories of general health literacy were combined into a new construct termed “Limited Health Literacy.” Among pregnant women with at least 80% valid responses to the 12 items of the HLS₁₉-Q12 instrument (n = 875), 46.7% were classified as having limited health literacy. To enhance clarity and focus, only variables that showed statistically significant associations with limited health literacy (p < 0.05) are included in the table. The full version of this analysis, including non-significant comparisons, is available in Supplementary Material (Additional File 1).

Limited health literacy was more pronounced among specific subgroups, reflecting significant disparities across sociodemographic, behavioral, and health-related variables. Younger pregnant women, particularly those aged 18 to 29 years, had the highest prevalence of limited health literacy at 56.2%. Women born outside of Portugal exhibited a markedly higher prevalence (74.7%) compared to Portuguese nationals. Education level also strongly correlated with health literacy, as 90.9% of those with only primary education (2nd cycle) showed limited health literacy. Similarly, marital status revealed disparities, with divorced, separated, or widowed women reporting the highest prevalence at 92.3%. Occupational status highlighted further inequalities, with 80.8% of undifferentiated workers and 76.2% of those not professionally active demonstrating limited health literacy. Regarding housing conditions, 57.6% of women living in homes owned by relatives had limited health literacy, compared to lower rates among those renting or owning their homes. The absence of healthcare training played a critical role, as 51.3% of those without such training exhibited limited health literacy. Financial strain was a key determinant, with 61.3% of women who found it difficult to meet expenses reporting limited health literacy.

Behavioral and lifestyle factors also played a role. Women who smoked occasionally had the highest prevalence of limited health literacy (71.9%), followed by those who quit smoking during pregnancy (60.5%). Exposure to tobacco smoke was associated with higher rates of limited health literacy (58.5%) compared to those not exposed. Occasional alcohol or psychoactive substance users had a prevalence of 62.3%. Among physical activity levels, the highest prevalence was observed in women who reported never engaging in physical activity (63.2%). Regarding dietary habits, limited health literacy was most common among those with occasional (75.0%) or light fruit and vegetable consumption (77.6%). Obesity prior to pregnancy was associated with 60.6% limited health literacy, followed by overweight women (46.7%).

Self-perceived health status also revealed disparities. Women who rated their previous health as “bad or very bad” had the highest prevalence of limited health literacy (81.8%), while those who rated their current health as “much worse or worse” had a prevalence of 58.6%. In terms of gynecological history, 73.2% of women who had never undergone cervical cytology exhibited limited health literacy, as did 65.6% of those who did not use contraceptive methods.

Pregnancy-related variables further emphasized disparities. Women in their third trimester (28–41 weeks) exhibited the highest prevalence of limited health literacy (47.4%) compared to those in the first (45.7%) and second trimesters (44.5%). Among women with unplanned pregnancies, 67.0% demonstrated limited health literacy, significantly higher than the 36.5% observed among those with planned pregnancies. Delayed initiation of pregnancy surveillance was strongly associated with limited health literacy, with 72.6% of those starting at 12–27 weeks affected and 100% of those starting surveillance at 28 weeks or later. Women without children had a lower prevalence of limited health literacy (44.0%) compared to those with children (50.4%). Healthcare engagement and intentions regarding breastfeeding and childbirth preparation programs further highlighted disparities. Women attending queries only at health centers had the highest prevalence of limited health literacy (63.6%) compared to 37.5% among those attending multiple types of healthcare facilities. Those undecided about breastfeeding (76.5%) or who did not intend to breastfeed (72.7%) exhibited higher rates of limited health literacy compared to women who intended to breastfeed (44.5%). Among those not attending childbirth preparation programs, 60.9% had limited health literacy, while only 30.9% of those intending to attend such programs were affected.

Determinants of limited health literacy

Variables showing statistical differences in limited health literacy across categories met the selection threshold for inclusion in the regression analysis (p < 0.10). Table 5 presents the final adjusted odds ratios (ORs) for limited health literacy, highlighting associations with key sociodemographic, economic, health status, and pregnancy-related factors. Full regression outputs, including crude and intermediate estimates, are available in Supplementary Material (Additional File 2).

OR, odds ratio; CI, confidence interval. ¹Binary logistic regression model (1st block: forward, LR method; 2nd bloc: enter method) adjusted for country of birth, educational level, profession/occupation, payment of expenses, planned pregnancy, surveillance pregnancy, previous self-perceived health, current self-perceived health, and BMI. Full model specifications, including crude and intermediate estimates, are available in Supplementary Material (Additional File 2).

In the first regression model (univariate, not adjusted), several variables show a significant relationship with limited health literacy. Regarding age groups, younger individuals demonstrate lower odds of limited literacy compared to those aged 40 and above. Those aged 18–29 years have 47% lower odds of limited literacy (Crude OR = 0.53, 95% CI: 0.40–0.71), while individuals aged 30–39 years have 51% lower odds (Crude OR = 0.49, 95% CI: 0.28–0.85). Country of birth is a significant factor, with individuals born outside Portugal having more than four times the odds of limited health literacy compared to Portuguese-born individuals (Crude OR = 4.39, 95% CI: 3.00–6.42). Education level is also crucial. Compared to university graduates, individuals with high school education have 96% lower odds of limited literacy (Crude OR = 0.04, 95% CI: 0.01–0.14), those with 3rd cycle education have 89% lower odds (Crude OR = 0.11, 95% CI: 0.03–0.37), and those with education up to the 2nd cycle have 78% lower odds (Crude OR = 0.22, 95% CI: 0.06–0.79). Regarding profession/occupation, compared to individuals in intellectual and scientific professions, those in technical professions have 92% lower odds of limited literacy (Crude OR = 0.08, 95% CI: 0.05–0.13), while individuals in undifferentiated professions have 81% lower odds (Crude OR = 0.19, 95% CI: 0.12–0.29). Employment status also shows a significant association with health literacy. Individuals not working professionally have more than five times the odds of limited literacy compared to those employed (Crude OR = 5.12, 95% CI: 3.55–7.38). Financial hardship is another key predictor. Those experiencing difficulty paying monthly expenses have more than five times the odds of limited literacy compared to individuals who find it easy to cover their expenses (Crude OR = 5.34, 95% CI: 3.92–7.27). Regarding pregnancy-related factors, individuals with an unplanned pregnancy have more than three times the odds of limited literacy compared to those with a planned pregnancy (Crude OR = 3.53, 95% CI: 2.62–4.75). Similarly, individuals who initiated pregnancy surveillance at ≥ 12 weeks have more than three times the odds of limited literacy (Crude OR = 3.67, 95% CI: 2.29–5.88). In terms of self-perceived health, individuals who currently rate their health as equal to before have 68% lower odds of limited literacy (Crude OR = 0.32, 95% CI: 0.16–0.62), while those who perceive their health as worse or much worse have 39% lower odds (Crude OR = 0.61, 95% CI: 0.40–0.93). For body mass index (BMI) prior to pregnancy, individuals with normal weight have 52% lower odds of limited literacy compared to those who are underweight (Crude OR = 0.48, 95% CI: 0.23–0.98).

In a second regression approach (first block forward), country of birth remains a significant determinant, with individuals born outside Portugal exhibiting more than twice the odds of limited health literacy compared to those born in Portugal (Adjusted OR = 2.44, 95% CI: 1.58–3.75). Profession/occupation continues to demonstrate a strong association with health literacy, compared to individuals in intellectual and scientific professions, those in technical professions show 75% lower odds of limited literacy (Adjusted OR = 0.25, 95% CI: 0.14–0.44), while individuals in undifferentiated professions exhibit 66% lower odds (Adjusted OR = 0.34, 95% CI: 0.22–0.54). Financial difficulties remain a key predictor. Individuals who report difficulty covering monthly expenses have more than three times the odds of limited health literacy compared to those without financial strain (Adjusted OR = 3.24, 95% CI: 2.29–4.60). Previous self-perceived health does not show a statistically significant association after adjustment. Current self-perceived health continues to be a relevant factor. Compared to those who perceive their health as better or much better, individuals who rate their health as the same demonstrate 76% lower odds of limited literacy (Adjusted OR = 0.24, 95% CI: 0.11–0.51). Those who perceive their health as worse or much worse do not present a statistically significant association (Adjusted OR = 0.61, 95% CI: 0.38–0.99). BMI also becomes significant in this model. Compared to underweight pregnant women’s, those with normal weight exhibit 64% lower odds of limited literacy (Adjusted OR = 0.36, 95% CI: 0.15–0.84), however, overweight and obesity do not show statistically significant associations.

In the final adjusted binary logistic regression model (1st bloc: forward, LR method; 2nd bloc: enter method) adjusted for country of birth, educational level, profession/occupation, payment of expenses, planned pregnancy, surveillance pregnancy, previous self-perceived health, current self-perceived health, and BMI remain significant determinants of limited health literacy among pregnant women. Country of birth continues to be a strong predictor, with pregnant women born outside Portugal exhibiting more than twice the odds of limited health literacy compared to those born in Portugal (Adjusted OR = 2.43, 95% CI: 1.56–3.80). Educational level remains a key determinant of health literacy. Compared to pregnant women with university education, those with high school education exhibit 80% lower odds of limited literacy (Adjusted OR = 0.20, 95% CI: 0.05–0.77), while those with 3rd cycle education has 77% lower odds (Adjusted OR = 0.23, 95% CI: 0.06–0.86). Education up to the 2nd cycle does not present a statistically significant association in the fully adjusted model. Profession/occupation continues to demonstrate an association with health literacy. Compared to pregnant women in intellectual and scientific professions, those in technical professions have 64% lower odds of limited literacy (Adjusted OR = 0.36, 95% CI: 0.19–0.70), while those in undifferentiated professions show 58% lower odds (Adjusted OR = 0.42, 95% CI: 0.25–0.68). Financial hardship remains a strong predictor. Pregnant women who report difficulty covering monthly expenses have nearly three times the odds of limited health literacy compared to those without financial strain (Adjusted OR = 2.87, 95% CI: 1.99–4.14). Planned pregnancy is no longer statistically significant in the fully adjusted model. Pregnant women with an unplanned pregnancy exhibit 37% higher odds of limited health literacy (Adjusted OR = 1.37, 95% CI: 0.94–2.00), but this association does not reach statistical significance. Surveillance pregnancy, defined as delayed initiation of prenatal care (≥ 12 weeks), also does not present a significant association with limited health literacy in the final model (Adjusted OR = 0.95, 95% CI: 0.53–1.72). Previous self-perceived health does not show a statistically significant association after adjustment, indicating that retrospective health perception may not be a strong determinant of health literacy. Current self-perceived health remains a relevant factor. Compared to those who perceive their health as better or much better, pregnant women who rate their current health as the same exhibit 75% lower odds of limited health literacy (Adjusted OR = 0.25, 95% CI: 0.11–0.54). However, those who consider their health as worse or much worse do not present a statistically significant association (Adjusted OR = 0.65, 95% CI: 0.40–1.05). Body Mass Index (BMI) prior to pregnancy continues to be an influential factor. Compared to underweight women, those with normal weight demonstrate 64% lower odds of limited health literacy (Adjusted OR = 0.36, 95% CI: 0.15–0.84). Overweight and obesity do not show significant associations in the final model.

At last, the area under the ROC curve (AUC) for the final model is 0.782, indicating a good level of discriminatory power and suggests that the model has a 78.2% probability of correctly distinguishing between individuals with and without limited health literacy. This result demonstrates that the model reliably differentiates between the two categories, providing a strong basis for its predictive validity in this context.