Dietary practices among antenatal women in Sri Lanka: prevalence and associated factors

Abstract

Objective

To assess the prevalence and associated factors of dietary practices among antenatal women in Colombo district, Sri Lanka.

Design

This descriptive cross-sectional study examined dietary practices among antenatal mothers in four Medical Officer of Health areas in Colombo, Sri Lanka. A total of 422 participants were selected using stratified random sampling. Data were collected via a validated Food Frequency Questionnaire and analysed using SPSS V.26. Dietary diversity, food variety and animal-source food consumption were assessed. Poisson regression identified predictors of dietary practices, adjusting for socio-economic and pregnancy-related factors. The statistical significance was set at p<0.05.

Results

Of the 380 antenatal mothers (mean age: 30.72±3.96 years), most were married (98.2%) with 73.7% living in urban areas. Regarding dietary practices, 64.7% had high dietary diversity, while 35.3% had low diversity. Of the sample, 52.1% had a high food variety score and 64.7% had a high animal-source food score. More than half (64.7%) had appropriate dietary practices. Fruits, vitamin A-rich vegetables and rice were the most consumed foods. Key factors influencing dietary practices included age, religion, education, employment and geographical location.

Conclusions

This study highlights the prevalence and factors influencing dietary practices among antenatal mothers. Although the predominant mothers had fair dietary diversities, a considerable number were found to have poor dietary practices. Better dietary practices were associated with major educational attainment, formal employment status and selected residential areas, while younger age, low educational qualification and housewife status were associated with poorer nutrition. The findings indicate that there is an urgent need for interventions related to nutrition for specific vulnerable groups so that they can improve their maternal nutrition and produce better pregnancy outcomes through education and support programmes.

STRENGTHS AND LIMITATIONS OF THIS STUDY

• This study used stratified random sampling across multiple Medical Officer of Health (MOH) areas with a high response rate, improving internal validity.

• Dietary practices were assessed using a locally validated Food Frequency Questionnaire and multiple diet quality indicators.

• Use of Poisson regression with robust variance enabled appropriate estimation of prevalence ratios.

• The cross-sectional design limits causal inference between associated factors and dietary practices.

• Findings may have limited generalisability beyond the selected MOH areas and the first trimester of pregnancy.

Introduction

Maternal nutrition is a key determinant of pregnancy outcomes in terms of fetal growth, birth weight and maternal well-being.¹ Proper nutrition during pregnancy allows for optimal fetal development and reduces the risk of pregnancy-related complications such as preterm birth, low birth weight and neonatal mortality.¹ Although maternal nutrition is now gaining concern as an important aspect of pregnancy health, eating habits still vary widely across regions and socio-economic strata.² These differences are the outcome of a plethora of factors, including cultural beliefs, economic status, education level, access to healthcare services and food preferences.¹

Globally, pregnancy-related malnutrition is still a public health concern,³ while overnutrition and excessive weight gain during pregnancy have caused an increase in gestational diabetes and hypertensive disorders.⁴ Furthermore, many low- and middle-income countries have a significant number of pregnant women suffering from iron, folic acid, calcium and vitamin D deficiencies, all of which are required for proper fetal development.⁵ In high-income countries, poor eating habits such as high intake of processed foods, excessive sugar consumption and non-compliance with dietary guidelines are linked with adverse maternal and neonatal outcomes.⁶

Dietary diversity, defined as the number of different food groups consumed over a specific reference period, is a key indicator of diet quality and nutrient adequacy.^{7 8} Previous studies in Ethiopia have shown that higher dietary diversity is associated with improved nutritional status and reduced adverse pregnancy outcomes.^{9 10} Limited dietary diversity during pregnancy has long-term effects on the health of both mother and fetus, which may predispose to impaired child development and non-communicable diseases.¹¹ Therefore, investigating the dietary practices of pregnant women yields important information on what nutritional challenges they encounter and will be very useful to shape public health strategies that aim at enhancing maternal nutrition.

Several studies have demonstrated the association of socio-demographic factors with dietary diversity, including higher socio-economic status, which has been associated with improved dietary diversity, as well-off women typically have more access to diverse foods.^{12 13} The educational level also has a significantly positive effect, which means that by increasing the education of women, we can provide more knowledge about nutrition and health.^{14 15}

Moreover, dietary diversity is highly dependent on food availability and access to healthcare services. Women who attend antenatal care are more likely to receive nutritional counselling and support which can lead them to positively change their dietary practices.^{16 17} In contrast, in rural settings, food diversity scores are often lower due to limited access to various types of foods due to infrastructural barriers such as financial¹⁸ and physical accessibility limitations along the production chain.¹⁹

In this light, the current study aimed to assess the prevalence and associated factors of dietary practices of antenatal mothers attending clinics in selected Medical Officer of Health (MOH) areas in the Colombo district, Sri Lanka.

Materials and methods

Study design and setting

This descriptive cross-sectional study was conducted in antenatal clinics within four purposively selected MOH areas in the Colombo district, from June 2024 to February 2025. In Sri Lanka, the MOH and its respective teams are responsible for delivering preventive healthcare services, including antenatal care. The Colombo district comprises 12 MOH areas, all of which operate antenatal clinics. For this study, the Homagama, Maharagama, Nugegoda and Dehiwala MOH areas were selected due to all four areas having well-established antenatal clinics with high client turnout, ensuring adequate sample sizes and data quality. Further, these MOH areas represent a diverse cross-section of the Colombo district in terms of socioeconomic status and urbanisation. For instance, Dehiwala and Nugegoda are highly urbanised with a more heterogeneous population, while Homagama and Maharagama represent semi-urban settings, allowing the study to capture a broader range of antenatal care experiences and practices.

Study population

All antenatal mothers who attended antenatal clinics in four selected MOH areas in Colombo District, Sri Lanka were the selected population for the current study. Antenatal mothers in the first trimester in particular clinics were recruited for the study without considering their age. Antenatal mothers with known acute or chronic illnesses, psychiatric disorders or inability to communicate due to hearing and/or speaking difficulties during the data collection period were excluded.

Sample size and sampling technique

A total of 422 participants were recruited for the study, with the sample size determined using Daniel’s formula,²⁰ incorporating a 10% allowance for non-responders. A stratified random sampling technique was employed to ensure equal representation of antenatal mothers across the selected MOH areas. Although the total antenatal population across these MOH areas was approximately 860 pregnant mothers, only 435 women in their first trimester were identified during the study period and considered eligible for inclusion. Proportional stratified sampling was applied based on the distribution of first-trimester mothers within each area—Maharagama (19%), Nugegoda (18%), Homagama (30%) and Dehiwala (33%). Accordingly, the sample was allocated as follows: Maharagama—80, Nugegoda—76, Homagama—127 and Dehiwala—139. Participants were then selected from the clinic registration list using a computer-generated random sampling method.

Data collection and measurements

After obtaining written informed consent, data were collected using a self-administered questionnaire, which consisted of three sections. Section A comprised seven items to collect the socio-demographic information of the participants including age, marital status, religion, ethnic group, residential area, occupation and monthly income. Section B was dedicated to collecting data regarding the dietary patterns of antenatal mothers using the Sri Lankan context-validated ‘Food Frequency Questionnaire’ (FFQ-15).²¹ FFQ contains 15 of the most commonly listed food items and eight food groups consumed by the Sri Lankan population which were used to assess the dietary diversity of the participants. The final section (C) was employed to obtain pregnancy-related data including parity, state of pregnancy (planned or unplanned), having regular follow-ups, etc.

Three measures were used to assess the dietary practices of pregnant women: dietary diversity, food variety and consumption of animal-source foods (ASFs) by using FFQ.²² The food items in the FFQ were initially identified through consultations with key informants familiar with the local culture, language and dietary habits in the same MOH areas. The questionnaire was pretested on 10% antenatal mothers from an MOH area who were not part of the main study, and necessary modifications were made based on the findings before implementation.

The FFQ measured food consumption over the past 3 months. However, considering variations, dietary intake throughout the week was considered.²² Pregnant women were classified as ‘consumers’ of a particular food item if they had consumed it at least once a week. Internal consistency was determined using Cronbach’s alpha, which was 0.74.

The food items in the FFQ were categorised into eight groups: cereals including cooked rice and bread, vegetables including vitamin A-rich vegetables and other vegetables, pulses including cooked dhal and boiled green gram or cowpea, meat including chicken, seafood including fish and dried fish, eggs, fruits and dairy products including fresh milk, curd, yoghurt and cheese. The dietary diversity score (DDS) was calculated based on the number of food groups consumed over a week. DDS was then categorised into tertiles, with the highest tertile indicating a ‘high’ DDS, while the two lower tertiles were grouped as ‘low’ DDS.²²

Food variety score (FVS) was determined by counting the frequency of individual food items consumed within the study’s reference period of over 7 days,²² with a maximum FVS of fifteen. The mean FVS was calculated, and pregnant women with an FVS above the mean were classified as having a ‘high’ FVS, while those below the mean were categorised as having a ‘low’ FVS.²² Furthermore, the consumption of ASFs was assessed by calculating the frequency of each ASF consumed by pregnant women over the designated reference period. ASF scores were divided into tertiles, where the highest tertile indicated ‘high’ ASF consumption, while the lower two tertiles were combined to represent ‘low’ ASF consumption. Adequate dietary practices were identified if the women had at least four meals a day, high FVS, high DDS and high ASF. Otherwise, it was categorised as inappropriate where women had less than four meals a day, a low FVS, low DDS and low ASF consumption.²²

Data analysis

Data were analysed using SPSS V.26 after the dataset was systematically cleaned, coded and examined for missing values and outliers. Descriptive analysis was conducted to obtain frequencies, percentages and means. The outcome variable, dietary practice, was dichotomised as 1 (appropriate) and 0 (inappropriate). To identify predictors of dietary practices among women, a Poisson regression analysis model with a robust variance estimate was employed. In the multivariate analysis, only variables with a p value of less than 0.25 in the bivariate analysis were included in the adjusted model. A backward regression approach was applied, incorporating selected socio-economic and pregnancy-related variables. The results were reported as adjusted prevalence ratios (APRs) with corresponding 95% CIs. Statistical significance was determined at an alpha level of 5%. Before inclusion in the multivariable model, explanatory variables were assessed for multicollinearity using a correlation matrix, standard errors and variance inflation factor values. Additionally, potential interactions between covariates were evaluated.

Findings

Socio-demographic characteristics

A total of 422 antenatal mothers were eligible and 380 (90%) responded. The mean age was 30.72 (+3.96) years and the majority (98.2%) were married. Most of the participants were Buddhist (95.5%) and Sinhala (98.2%). More than 70% of the participants resided in urban areas and 46.6% of the participants were employed in the government or private sector (table 1).

Table 1. Socio-demographic variables of the antenatal mothers (n=380).

Variables	Frequency (n)	Percentage (%)
Age (years)
Mean (SD)	30.72 (+3.96)
<21	5	1.3
22–25	36	9.5
26–29	96	25.3
>30	243	63.9
Marital status
Unmarried	7	1.8
Married	373	98.2
Religion
Buddhism	363	95.5
Christian	15	3.9
Other	2	0.5
Ethnic/linguistic groups
Sinhala	373	98.2
Tamil	7	1.8
Residential area
Urban	280	73.7
Semi-urban or rural	100	26.3
Occupation
Government/private	177	46.6
Housewife	167	43.9
Self-employed	36	9.5
Monthly income (LKR)
<45 000	135	35.5
45 000–60 000	81	21.3
>60 000	164	43.2

Dietary practices of antenatal mothers

The prevalence of dietary practices among antenatal mothers indicated that 64.7% had a high DDS with a mean score of 7.56 (SD±0.68), while 35.3% had a low DDS. Regarding the FVS, 52.1% of women had a high FVS, whereas 47.9% had a low FVS, with a mean score of 12.09 (SD±2.02). Similarly, 64.7% of participants had a high ASF, while 35.3% had a low ASF, with a mean score of 3.55 (SD±0.68). Overall, 64.7% of pregnant women demonstrated appropriate dietary practices, whereas 35.3% had inappropriate practices (table 2).

Table 2. Prevalence of dietary practices among antenatal mothers.

Variables	Frequency (n)	Percentage (%)
Dietary Diversity Score (DDS)
High	246	64.7
Low	134	35.3
Mean (SD)	7.56 (+0.68)
Food Variety Score (FVS)
High	198	52.1
Low	182	47.9
Mean (SD)	12.09 (+2.02)
Animal Source Food Score (ASF)
High	246	64.7
Low	134	35.3
Mean (SD)	3.55 (+0.68)
Dietary practices
Appropriate	246	64.7
Inappropriate	134	35.3

As shown in figure 1, food group consumption among antenatal mothers revealed that fruits were consumed by all participants (100%), followed by high consumption of vitamin A-rich vegetables (99.7%) and rice (98.4%). Other commonly consumed food items included fish (91.6%), eggs (91.1%) and lentils (90.5%). In contrast, curd (39.2%) and bread (48.9%) had the lowest consumption rates.

Figure 1. Different food items consumed among antenatal mothers within a 1 week. Percentage of different food items consumed by antenatal mothers within a 1-week period. Values represent the percentage of antenatal mothers who reported consuming each food item at least once during the preceding week, including rice, bread, vegetables (vitamin A-rich and other), dhal, green gram/cowpea, meat (chicken), fish, dried fish, eggs, fruits, milk, curd, cheese and yoghurt.

Factors associated with dietary practices

Based on the bi-variable analysis, age, MOH area, highest education level, marital status, religion, ethnic group, occupation, monthly income, parity and planned pregnancy were eligible for multi-variable analysis based on p<0.25.

Age was a significant predictor; the likelihood of an appropriate diet was higher for women aged 22 years and older than for women aged less than 21 years old (APR<1, p<0.001). Religion also influenced dietary practices, with Christians and others exhibiting higher adjusted prevalence ratios (APR 2.69 and 3.84, respectively, p<0.001) than Buddhists. Education level was an important factor, as degree holders and postgraduates were much more likely to practise appropriate dietary habits (APR 6.28 and 4.61, respectively, p<0.001) compared with other respondents with O/L education. Being a housewife or self-employed affected one’s dietary practice; they had lower APRs than women employed in either the private or government sector (p<0.001). Another important feature was location; women from Dehiwala had the lowest likelihood of having an appropriate diet (APR 0.48, p<0.001) as compared with women from Maharagama (table 3).

Table 3. Factors associated with dietary practices among pregnant women.

Variables	Dietary practice		CPR (95% CI)	APR (95% CI)	P value
	Appropriate (n=246)	Inappropriate (n=134)
Age
<21	05	0	1	1
22–25	20	16	0.56 (0.41 to 0.74)	0.50 (0.38 to 0.66)	<0.001
26–29	65	31	0.68 (0.59 to 0.78)	0.53 (0.4 to 0.69)
>30	156	87	0.64 (0.58 to 0.71)	0.50 (0.39 to 0.64)
Religion
Buddhist	231	132	1	1
Christians	13	2	1.36 (1.1 to 1.69)	2.69 (2.08 to 3.49)	<0.001
Others	2	0	1.57 (1.45 to 1.7)	3.84 (3.05 to 4.85)
Highest level of education
Up to O/L	14	36	1	1
Up to A/L	129	81	2.19 (1.39 to 3.47)	4.62 (2.82 to 7.56)	<0.001
Degree level	96	11	3.20 (2.05 to 5.02)	6.28 (3.73 to 10.57)
Postgraduate	7	6	1.92 (0.98 to 3.76)	4.61 (2.22 to 9.57)
Occupation
Private/government sector	138	39	1	1
Housewife	87	80	0.67 (0.57 to 0.79)	0.78 (0.65 to 0.94)	<0.001
Self-employed	21	15	0.75 (0.56 to 1.00)	0.81 (0.63 to 1.05)
MOH area
Maharagama	61	11	1	1
Nugegoda	48	14	0.91 (0.77 to 1.08)	0.91 (0.79 to 1.06)	<0.001
Homagama	85	39	0.81 (0.77 to 1.08)	0.74 (0.64 to 0.87)
Dehiwala	52	70	0.50 (0.4 to 0.63)	0.48 (0.38 to 0.59)

CPR was obtained from a bivariate Poisson regression analysis model with robust variance estimate, and APR, CI, and p value were obtained from a multivariable Poisson regression analysis model with robust variance estimate.

A/L, Advanced Level; APR, adjusted prevalence ratio; CPR, crude prevalence ratio; MOH, Medical Officer of Health; O/L, Ordinary Level.

Discussion

The study of dietary practices of antenatal mothers in Colombo District, Sri Lanka, provides significant socio-demographic characteristics and dietary behaviour that correspond with and diverge from existing literature. A total of 422 antenatal mothers were eligible, with a response rate of 90%, allowing for a sample size strong enough for analysis. The mean age of respondents was 30.72 (+3.96) years, with a predominantly demographic of married urban Sinhala Buddhists. This demographic profile is in conformity with previous studies indicating that pregnant women in Sri Lanka and other South Asian countries are from similar socio-cultural backgrounds.²³

The high percentage of married participants (98.2%) and the majority being Buddhist (95.5%) reflect the cultural norms of the region, where marriage is often a prerequisite for childbearing. This finding is corroborated by studies that emphasise the importance of marital status in maternal health outcomes, suggesting that married women may have better access to healthcare resources and support systems during pregnancy.²⁴

The study found that 64.7% of antenatal mothers achieved a high DDS, implying a fair level of consumption of food groups. The positive aspect of this finds its basis in that dietary diversity has been found associated with better nutritional status and health outcomes for mothers and their infants.²⁵ The mean DDS of 7.56 (+0.68) indicates that these women appear to have been consuming a variety of food groups, which are essential for meeting the increased nutritional needs throughout pregnancy. However, the finding that 35.3% of participants had a low DDS calls into question the nutritional adequacy of a sizeable share of the population, verifying other studies reporting inadequate dietary practices among pregnant women in similar socio-economic scenarios.²⁶

The DDS and FVS in this study are thus reported on the higher side when compared with reported studies done on similar populations in other countries. For instance, a study conducted in Bangladesh reported that only 50% of pregnant women had an adequate DDS, which was associated with lower socio-economic status and limited access to diverse food sources.²⁶ Factors such as socio-economic conditions and cultural dietary practices present in Sri Lanka may account for variance, as urbanisation and increased accessibility to a range of food groups might increase the dietary diversity experienced by pregnant women.²⁷

The FVS and ASF also illustrate the dietary patterns of the participants. According to the current findings, 52.1% of mothers had a high FVS and 64.7% had a high ASF; thus, mothers were giving more emphasis to high-protein foods, which are crucial for fetal development. Nonetheless, the lowest consumption rates were recorded for some food groups like curd and bread (39.2% and 48.9%, respectively), hence indicating specific areas that need to improve dietary practices. Culturally based preferences, along with the availability of certain foods, have been found in previous studies to influence dietary choices during pregnancy.²⁸

The findings from this study about the FVS indicate better performance than those from Ethiopia, where it was found that only 40% of pregnant women had a high FVS, mainly owing to economic constraints and the absence of an education programme for dietary practices.²⁹ The high FVS in the Colombo District indicates that better outreach in education and availability of a mixed food basket is better in urban settings, which dovetails with findings of Amugusi and Dimbuene,³⁰ who emphasised the role of sociodemographic factors in dietary practices.³⁰

Although it shall be noted that overall dietary practices appear adequate, the study further showed that a substantial number of participants (35.3%) had a low DDS and were involved in inappropriate dietary practices. This corresponds with another study done in India, which revealed that due to socio-economic barriers and a lack of nutritional knowledge, a considerable percentage of pregnant women have inadequate dietary diversity.³¹ These challenges for women point to the need for targeted intervention programmes directed at improving dietary practices among this vulnerable population.

According to the current findings, factors influencing dietary practices were age, level of education, religion and occupation. Findings indicated that young mothers (less than 21 years) were less likely to have dietary practices considered appropriate, as the literature suggests that younger mothers lack the knowledge or resources to maintain a healthy diet during pregnancy.³² There was a strong association between mothers with higher education levels having favourable dietary practices. Education reinforces the need to promote maternal nutrition because such mothers are more likely to understand the nutritional needs of pregnancy and make informed dietary choices.³³

Religion also played a significant role; Christians and other religious groups exhibited higher APRs for appropriate dietary practices than Buddhists. Such a finding may reflect the presence of differences in cultural dietary restrictions and/or food availability among different religious groups, a phenomenon documented in a huge number of studies of dietary habits across different cultures.²⁹ Furthermore, maternal employment influenced dietary practices, with mothers in government or private sector employment having better dietary practices compared with housewives. This is likely due to possible access to more financial resources and information that employed mothers possess, and these help them make healthy food choices.³⁴

Furthermore, geographical area significantly influenced dietary practices. Apparently, women from Dehiwala were significantly less likely to observe proper dietary norms. This suggests there might be differences in access to healthy foods or awareness of nutritional education between locations, a theme echoed by other studies examining urban-rural divides in maternal health.^{35 36}

Limitations

The study’s descriptive cross-sectional design provides only a snapshot of the dietary practices of pregnant women at a single point in time, limiting the ability to establish causal relationships. Additionally, the research was conducted in only four purposively selected MOH areas in the Colombo district, which may not fully represent the socio-economic and cultural diversity of the entire district, potentially introducing sampling bias. Despite the use of statistical adjustments, there may still be unmeasured confounders that could influence the relationship between socioeconomic factors and dietary practices, further complicating the interpretation of the findings.

Conclusions

The current study provides a clear insight into the dietary practices of pregnant women in the Colombo District, Sri Lanka. The results indicated that although a large number of antenatal mothers consumed sufficient dietary diversity and variety of foods, a significant percentage of them had poor dietary habits, therefore calling for some nutritional interventions. The high DDS noted among the participants indicates that a high proportion of antenatal mothers observe the consumption of a range of important food groups, positively affecting maternal and fetal health. Conversely, the occurrence of low DDS indicates that there are existing nutrition gaps requiring intervention.

Socio-demographic factors and pregnancy-related factors were also identified in the study to influence dietary practices. Educational attainment, regular employment and religious beliefs all proved to be strong predictors of better food habits with greater knowledge and exposure to nutritional facts presumably playing a part. Younger age, lower education and being a housewife are closely linked with poor dietary practices. The results stressed the relevance of education and economic empowerment in improving dietary behaviours among antenatal mothers. Moreover, the geographic variations in dietary behaviour suggest that access to a variety of foods and nutritional knowledge also varies geographically.

Overall, while the dietary diversity and FVSs in the current study are relatively higher than those of comparable populations in other countries, the prevalence of nutritional inadequacies suggests that continued efforts in maternal nutrition education are necessary. Policy developers and healthcare professionals must prioritise the creation of nutritional literacy, food availability and the implementation of targeted provision programmes to allow all antenatal mothers to access the necessary dietary inputs for a healthy pregnancy.

Data availability statement

Data are available upon reasonable request.