The Relationship Between Breastfeeding and Measurements of Refraction and Visual Acuity in Primary School Children

Introduction

Refractive errors are among the most common causes of visual impairment and the second leading cause of treatable blindness. ¹ While populations in different countries may have a comparatively different prevalence of refractive errors, the reasons for such differences have often been a subject of great interest. ² Since visual impairment can have a significant impact on a child's education and development, it is important that effective strategies are planned to explore the etiologic factors of refractive error and to reduce visual impairment.^3,4 Nutrition was taken into account by developing some refractive errors too. Studies from Africa revealed that children suffering from malnutrition have an increased prevalence of refractive error like high ametropia, anisometropia, and astigmatism. ² It has been also suggested that those initially breastfed have better vision and are less likely to be myopic in later life than those who are formula fed.^5,6 Docosahexaenoic acid (DHA) and arachidonic acid (AA), known to be long-chain polyunsaturated fatty acids, are found in breast milk, and their importance in infant nutrition can be recognized by the rapid accretion of these fatty acids in the brain during the first postnatal year. Some studies reported benefits of adding DHA or both DHA and AA to formulas, but other studies showed that visual and cognitive development were not changed after supplementation. ⁷ Since nutritional and environmental factors can have different impacts on eye development and disorders, i.e., visual acuity (VA) and refractive error, the present study was conducted to determine if there is any relationship between Breastfeeding and the results of refractive error and VA measurements in the child population of Sabzevar, Iran.

Methods

In this cross-sectional study, cluster sampling was used to select 400 children aged 1–5. In Sabzevar, there were 15 primary schools for girls and 12 for boys. One school for girls and another for boys were randomly selected through cluster sampling. In each cluster, subjects were randomly recruited proportionately considering the number and gender of students in each age group; the total number of samples came up to be 400 students with the age range of 6–10 years. By refraction, significant myopia was defined as −0.50 diopters (D) or less; hyperopia as +0.50D or more; and astigmatism was defined as at least a 0.50D difference between the two principal meridians. In addition, refractive error was converted to spherical equivalent (SE) (spherical dioptric power plus half of the cylindrical dioptric power). According to SE, myopia was defined as SE less than −0.5D and hyperopia more than +0.5D. These definitions were chosen to allow the comparison between findings in previous studies and present data. ⁸ Distance VA was measured using a chart with tumbling-E optotypes. Refraction was performed with a streak retinoscope (Welch Allyn, Inc.) at 50 cm, using a picture fixation target at 4 m for all children and just by one optometrist. One ophthalmologist was the director of the research team. Feeding category was ascertained by the question, “What was the type of feeding during infancy?” and categorized as follows: Breastfeeding, formula feeding, and combination of Breastfeeding and formula, as well as cattle, sheep, and goat milk. Duration of feedings was determined using a questionnaire with the cooperation of children's parents and the coauthored nutritionist. Breastfeeding was defined as feeding during the first 6 months of an infant's life. Informed consent was obtained from the school manager and the parents of students. The study procedures were also in accordance with the standards of the Ethics Committee of Sabzevar University of Medical Sciences, Iran. There were no legal and/or ethical effects since the study was descriptive and the school manager, participants, and their parents already signed an informed consent and expressed their inclination to participate. Statistical analyses were performed using SPSS Software 13 (Release 13.0; SPSS, Inc.). Descriptive statistics was calculated for sex, school, educational level, and feeding type. In addition, refractive error measurements were calculated, including myopia, hyperopia, and astigmatism, separately for right and left eyes. Correlations between refractive error, VA, and the feeding measures were calculated with Spearman correlation coefficient. All confidence intervals were 95%. In addition, one-way ANOVA and independent sample t-tests were used for comparing VA and feeding. Pearson chi-square test was also used for finding possible relationships between feeding type and mean SE in right and left eyes.

Results

From a total of 400 selected students for study, there were 33 children with missing data and also, there was one missing value for refractive error. Therefore, total participation rate was 367 children (92%): 156 boys (42.5%) and 211 girls (57.5%). Findings revealed that 311 of the participants (84.7%) were breastfed and 56 subjects (15.3%) had experienced other kinds of feeding. A high correlation was found between right and left eye refractions, including myopia (r = −0.997), hyperopia (r = 0.885), and astigmatism (r = 0.792). In addition, correlation between right and left eye VA was high (r = 0.90). In addition, mean VA was different in both eyes across feeding types, but independent sample t-test did not show any statistically significant difference. Mean cylinder power in right and left eyes across feeding types is shown in Table 1. In addition, mean cylinder power was different across various feeding types, but independent sample t-test did not show any significant difference. Overall, 3.6% of the children were myopic, 28.1% hyperopic, and 11.2% were astigmatic in the right eye. As for the left eye, 3.6% of the children were myopic, 27.6% hyperopic, and 10.9% were astigmatic. The mean SE refraction in right eye across various feeding types was +0.16D ± 0.71 SD (range −3.5 to +5.00) and 0.08D ± 0.58 SD (range −2.88 to +1.5.00) (Table 2). However, mean SE (MSE) refraction measurements in left eye were +0.13D ± 0.61 SD (MSE; range −3.5 to +4.63.00) and +0.12D ± 0.54 SD (MSE; range −2.88 to +1.5.00).

Frequency of refractive error measurements (i.e., emmetropia, myopia, and hyperopia) according to MSE for both eyes in relation with feeding types is shown in Table 3.

Although MSE findings according to the kind of feeding were different, Pearson chi-square test did not show any statistically significant difference between two groups. One-way ANOVA for comparison of VA did not reveal statistically significant differences in right and left eyes across the five school grades. Furthermore, ANOVA and independent sample t-test for comparing VA in both eyes with the kind of feeding during infancy revealed no statistically significant differences between them.

Discussion

Health services in Sabzevar are provided with both the public sector and private practitioners. Children at primary school level offer a better cooperation with researchers upon eye examination and that was the reason why we focused on this population and not younger than this age. There might be some doubt on the cooperation of children at age 6 (grade 1) and a study by Naidoo et al. ⁹ in South Africa revealed that testing VA in 5-year-olds, and to some extent in 6-year-olds, was difficult because of poor attention, lack of understanding, and restlessness. However, in Iran, the public school and private system admit children at 6 years of age and children in urban areas often have practice sessions with tumbling-E optotypes chart for screening VA in public and private nursery, from age 3 to 6. Thus, ability of children to cooperate in the screening process is not significantly problematic and results of VA tests in many 6-year-olds can be considered as reliable.

In the present study, refractive error test was done with retinoscopy without cycloplegic agents. Frequency of myopia (SE of at least −0.50D) in both eyes measured with retinoscopy was 5.2%, but based on the feeding type, myopia was less observed in children with Breastfeeding than those with alternative kinds. In a study, myopia (SE of at least −0.50D in either eye), measured by retinoscopy, in children aged between 5 to 15 years in Guangzhou, China affected 73.1% of children aged 15 years, and the prevalence of myopia was 3.3% in 5-year-olds with retinoscopy. ⁴ This finding revealed that the myopia prevalence in young children is low. Therefore, myopia was significantly associated with age, and other studies have shown the same results. ¹⁰

Although significant myopia in these two studies is slightly different according to age groups, the effect of feeding type in the Chinese study is not known and different type of feeding in these two studies may have a different effect on study results. In addition, age plays an important role in the progression of myopic refractive error since the Guangzhou study states the role of aging upon myopic refractive error. Astigmatism (cylinder of 0.75D or higher) was present in 33.6% of Guangzhou children, with retinoscopy; whereas in the present study, astigmatism (cylinder of −0.5D or higher) was 11.2% in the right eye and 10.9% in the left eye. It seems, in comparison with the Guangzhou study, refractive error in Sabzevar was not an important public health problem and other factors like age and ethnicity might play an important role in this regard.

In the present study, hyperopia (SE of +0.5D or higher) was the highest refractive error with 11.2% in the right eye and 10.4% in the left eye, followed by astigmatism (4.4%) and myopia (3.3%). One study in Singapore ⁵ to investigate the relationship between breastfeeding and the likelihood of myopia in children (grades 1–3 of a single junior school) revealed that from 797 subjects, myopia was observed in 521 children (65.4%) and 8.5% of them were breastfed. However, in the present study, the study population included 367 children (in grades 1–5) and Breastfeeding was defined as 6 months or more feeding, and 311 subjects (85%) were breastfed. Frequency of myopia (SE of at least −0.50D) in both eyes measured with retinoscopy was 5.2%. One important cause of difference between these two studies can be the nature of study, different range of age group and ethnicity. In the Singapore study, the participants were reexamined for myopia when they reached ages 10–12 years; whereas in the present research, our participants were randomly recruited from 6- to 10-year-old children and, therefore, not reexamined in a longitudinal study. In their study, children who were breastfed experienced lower myopia prevalence (259 [62.0%] of 418) than children who were not breastfed (262 [69.1%] of 379) (p = 0.04). However, in the present study, Pearson chi-square test did not show any statistically significant difference between Breastfeeding and non-Breastfeeding groups.

Sham et al. ¹⁰ reported that the breastfeeding status and the duration of breastfeeding were independently associated with the SE refraction of the right eye. The mean refractive error for children with a history of being breastfed was significantly more hyperopic than for those who were non-breastfed after adjustment with some sociodemographic factors.

The Singapore findings revealed that the type of Breastfeeding (exclusive, mostly, or partly) was not significantly associated with myopia. Other studies have also been done on breastfed children for visual function. For instance, study by Rudnicka et al., ¹¹ showed no associations between infant feeding and vision after adjustment for age.

In human milk, the fatty acids usually contain 8–30% linoleic acid and 0.5–2.0% a-linolenic acid, with about 0.5–0.8% arashidonic acid and 0.1–0.4% docosahexaneoic acid (DHA). The variability in fatty acids is dependent on variations in the fat composition of the mother's diet. Breastfeeding has been reported to benefit visual development in children. A higher concentration of DHA in breast milk than in formula has been proposed as one explanation for this association and as a rationale for adding DHA to infant formula, but few long-term data support this possibility. ¹²

DHA (22:6 ω-3) is now recognized as a physiologically essential nutrient in the brain and retina of the eye where it is required in high concentrations for providing optimal mental performance (neuronal functioning) and visual activity, respectively. The highest concentration of DHA per unit tissue weight was found in the membrane phospholipid components of the photoreceptor outer segments of the retina. The fluidity of retinal membrane due to DHA helps in faster response to stimulation. The optimal functioning of rhodopsin, the photopigment necessary for initiating visual sensation, is considered to be supported by the presence of DHA in the retinal membranes. ¹³

The retina, functionally an extension of the brain, contains rods and cones with the most fluid membranes of all the body's cell types; they are also highly enriched in DHA. The laboratory animals with experimentally induced omega-3 deficiencies showed deficits in retinal structure, visual activity development, and cognitive performances. ¹⁴

In addition, the results of Bhargava's study demonstrated the beneficial effect of orally administered O3FAs in alleviating dry eye symptoms, decreasing tear evaporation rate, and improving Nelson grade in patients suffering from computer vision syndrome-related dry eye. ¹⁵

Therefore, dependence of milk fatty acids on maternal diet makes it very difficult for us to determine the fatty acid requirements of young infants, ¹⁶ as well as the exact effect of feeding type on VA and/or refractive error. Although in the present study mean VA in both eyes was different across feeding types, independent sample t-test did not show any statistically significant difference. Thus, dependence of milk fatty acids on maternal diet may be one cause with no significant association between refractive error and VA with Breastfeeding or alternative types of feeding. In conclusions based on the results, there was no significant relationship between kind of feeding during the first 6 months of infancy and VA and refractive errors; however, for exact judgment about these findings, more studies are suggested with a larger sample size.