Refractive changes in pregnancy
To determine the causes of any vision change reported during pregnancy.
An obstetrical practice in Southampton, New York.
Two hundred forty pregnant women were asked whether they had any alteration in vision. Those who agreed to take part in the study (83) and who complained of vision changes (12) were matched with the next patient seen in the practice who was asymptomatic.
All patients underwent a complete ophthalmic examination, including refraction. Those who had alterations in vision status were seen again after delivery.
Main outcome measures
Changes in visual acuity and refractive error during pregnancy.
All women who complained of visual changes were found to have experienced a myopic shift from pre-pregnancy levels. (0.87±0.3 diopters in the right eye (P<0.0001) and 0.98±0.3 diopters in the left eye (P<0.0001). Post partum, all subjects returned to near pre-pregnancy levels of myopia.
This report links worsening of myopia to pregnancy. The causes of this myopic shift are not readily evident and merit further investigation.
Anecdotal reports of changes in vision during pregnancy frequently are heard during residency training for obstetrics as well as ophthalmology. However, a review of the literature in the area failed to reveal any studies that have systematically looked at this issue. As part of a larger investigation into the role of pregnancy-related vision change, the author undertook a case–control study of pregnant women in a small community.
Materials and methods
The study group was selected from an obstetrical practice in Southampton, New York. This is a community of 5,000 on Long Island. The population is largely Caucasian and middle class. Every woman presenting to the obstetrical practice with a new pregnancy was asked to take part in the study. All principles of the Helsinki Declaration were followed. Those who agreed to participate answered a series of questions on a self-administered questionnaire. These dealt with demographic information, number of pregnancies, history of eye disease, presence of any vision or ocular changes associated with the pregnancy, and use of any dietary supplements. Any woman who reported a change in vision or had any other ocular complaint was invited to have an eye examination. For each woman who agreed to take part and who had a visual or ocular complaint, the next woman seen for obstetrical care in the practice who agreed to take part in the study and who did not have any visual or ocular symptoms was similarly invited to have the ocular examination.
The author performed all ophthalmic examinations. Visual acuity, both uncorrected and with current correction, was obtained with Snellen optotype. The room had subdued lighting that was kept constant for all examinations with a luminance of 17 cd/m2.Manifest refraction was performed using trial lenses and streak retinoscopy and the results recorded. The correction used prior to pregnancy was also documented and confirmed by contacting the previous ophthalmic or optometric practitioners. Ocular adnexa, pupillary reflex, ocular motility, slit-lamp examination of the cornea and anterior chamber, direct ophthalmoscopy, and indirect ophthalmoscopy to the mid-periphery were performed for each patient. The results of a routine 1-h glucose tolerance test performed on all patients at 24 weeks of gestation were recorded. The presence or absence of clinical edema and total weight gain during the pregnancy were abstracted by a review of each patient's chart by the author. Those reporting change in vision while pregnant (cases) were contacted after delivery and invited for a repeat eye examination including refraction.
Data were analyzed by repeated measure of analysis of variance (ANOVA) for continuous variables and chi-square testing for dichotomous variables. Contingency tables were created for dichotomous variables. Linear regression analysis was used to test for associations between changes noted in right and left eyes. The study had a power of greater than 90% between groups and greater than 80% within groups.
During the 4-month period between 1 February and 31 May 1999, 240 pregnant women receiving obstetrical care in a single group practice in Southampton, New York, were asked to participate in this study. Of these women, 83 (35%) consented to complete the questionnaire. Visual changes during pregnancy were noted by 21 women (25%). Each of these 21 women was contacted by telephone and offered a complete ophthalmologic examination. Thirteen women agreed to the examination and eight refused. The women who refused did not differ from those who agreed to participate with respect to age, parity, or duration of the current pregnancy. Five of these eight women reported reduced night vision in their questionnaire and all five with this symptom required corrective lenses.
Of the 13 women who reported for ophthalmologic evaluation, 12 noted reduced night vision and one did not. This woman complained of non-specific eye irritation without change in vision. Her ophthalmologic examination was normal.
Demographic and ophthalmologic characteristics (mean ± SD) of pregnant American women with PRNB (cases) and without PRNB (controls)
Number of subjects
Pregnancy duration (months)
Pre-pregnant spherical equivalent OD
−5.625 to +0.38
−13.0 to +1.75
Pre-pregnant spherical equivalent OS
−5.625 to 0
−5.875 to +1.00
Pregnant spherical equivalent OD
−6.38 to −0.50
−13.0 to +1.75
Pregnant spherical equivalent OS
−6.63 to −0.75
−6.38 to +1.00
Blood glucose (mg/dl)
Uncorrected and corrected visual acuity was similar in cases and controls. All patients obtained best-corrected visual acuity of 20/20 in each eye at every examination. Mean refractive error while pregnant did not differ between the groups (Table 1). Similarly, cases and controls did not differ with respect to weight gain during pregnancy or blood glucose measured at 6 months' gestation.
For each subject, chart review, analysis of corrective lenses, or both determined refractive error prior to pregnancy. Mean values for cases and controls did not differ. The prevalence of myopia (spherical equivalent >−0.25 diopters in either eye) in the groups before pregnancy was the same (50%) in cases and controls. Thus, there was a marked increase in the rate of myopia in the cases [from 50% to 100% (chi-square 6.316;P=0.037)] but not in the controls (from 50% to 58%). In fact, only one control subject had evidence of an increase of myopia during pregnancy, and the change was quite small (0.25 diopters) and limited to one eye.
Each of the cases returned after delivery for a follow-up examination. The average period of time between delivery and the follow-up evaluation was 15±5 weeks (range 5–24 weeks). All reported resolution of night-blindness. Ten subjects had a decrease in myopia in both eyes. One subject improved in one eye (+0.75 diopters) and had a slight deterioration in the other (−0.25 diopters). One subject, seen at 9 weeks post partum, remained the same in both eyes.
However, significant post-partum refractive error remained when compared with the pre-pregnant measurement. In the right eye, the post-partum measurement was −1.65±1.6 versus −1.18±1.7 diopters pre-partum (P <0.0001). In the left eye, the post-partum measurement was −1.62±1.6 versus −1.32±1.7 diopters pre-partum (P<0.0001). There was no relationship between the degree of recovery and the number of weeks between delivery and the post-partum examination.
Perhaps the most striking findings of this study were the number of pregnant women in a well-nourished population taking vitamin A supplementation that reported symptoms of PRNB and the transient worsening of myopia in affected subjects. The most conservative estimate of the frequency of PRNB in Southampton, New York would be based upon the number of consecutive patients seen in the obstetrical practice (n=240) and the number examined with PRNB (n=12). Using these figures, 5% of the patients in this practice admitted to this symptom upon specific questioning. In fact, by far the commonest complaint of the women was difficulty driving at night. Both the limitations that PRNB places on activities of pregnant women and the increased risk of traffic accidents while driving at night make the findings of this study significant from the public health perspective of developed countries.
A review of the literature regarding visual changes associated with pregnancy revealed little mention of changes in refractive error. Duke-Elder in his "System of Ophthalmology"  discusses many ocular manifestations of pregnancy. However, there is no mention of change in refractive error. Hilton  carefully followed a group of women during pregnancy and found no change in refraction. Sunness  in a review of the pregnant woman's eye that cited 200 references, did not mention a systematic change in refractive error. Wang , in "Clinical Ophthalmology", reported no changes in refraction during pregnancy. Manges and colleagues  studied the effects of pregnancy on nearly 100 women. They measured a number of parameters including refraction, corneal thickness and curvature, and accommodation. They found no difference between pregnant and non-pregnant women in any category. The only alteration they described was a change in accommodation during the third trimester that they ascribed to swelling of the lens. Several studies of the effect of birth-control pills on corneal thickness and curvature showed no change during therapy [7, 8]. Thus this problem is not widely recognized in the ophthalmology literature.
In the obstetrical literature, there is a similar dearth of information regarding visual impairment during pregnancy. There is only a single case report of severe transient myopia in a pregnant woman that resolved spontaneously within 4 weeks of onset while the patient was still pregnant . A major review by Weinreb et al. in 1987 did not discuss changes similar to those observed in this study [3, 12]. A review by Offret and co-workers  in the French literature, mentioned a myopic shift in pregnancy, but the reference cited makes no mention of such a phenomenon. Christian et al. reported a small shift in myopia among a group of women in Nepal . To the best of my knowledge, this is the first systematic study that describes myopic changes that occur during pregnancy.
In this study, the amount of the shift in refractive status to some extent appeared related to prior refractive status. However, myopia did not appear to be a risk factor for PRNB since both cases and controls had similar rates of myopia prior to pregnancy (Fig. 2). Approximately half the subjects in each group were myopic and the degree of myopia was similar.
Why was there a shift in myopia in the pregnant women studied? While not a formal part of the investigation, it would be remiss not to consider some potential explanations. Vaughan and Asbury's "General Ophthalmology"  lists three principal causes of acquired myopia: cataract, diabetes mellitus, and accommodative spasm. No woman in the study had any sign of cataract on examination. Blood sugar measurements were available for each woman during the pregnancy. Only one patient developed gestational diabetes. Pupillary responses were uniformly normal and did not show any sign of accommodation spasm. In fact every investigation performed failed to reveal the etiology for the refractive changes seen in the study. This is an area for further investigation.
This study has limitations which are worth mentioning. The relatively small sample size, while adequate for a case–control study, does not permit wide generalization of the results. One examiner performed all refractions and he was not masked as to subject status. Also, the pre-pregnant refractive error was determined from the patient's medical record or existing glasses. Since none of the women reported a change in vision prior to the onset of pregnancy using their existing glasses and since the change in refractive status was so dramatic, it would seem that despite these limitations, the observations remain valid.
The author would like to thank the staff of Hamptons Obstetrics and Gynecology for their invaluable assistance in this study. They include Drs. Hunt, Diaz, Murphy and Rolston as well as Ruth Lazars. He would also like to thank Jane Bender, Sandy Hiltner, Tracy Kohnken, Polly Stark, Kyra Viet and Janet Wright of Peconic Ophthalmology. In addition, Dr. Elizabeth Shane assisted in manuscript preparation.
- 2.Duke-Elder S (1976) System of ophthalmology. Mosby, St LouisGoogle Scholar
- 4.Hilton GF (1958) Some effects of pregnancy on the eye. Am J Optom Arch Am Acad Optom 35: 117–124Google Scholar
- 7.Soni PS (1980) Effects of oral contraceptive steroids on the thickness of the cornea. Am J Optom Physiol Optics 57: 825–834Google Scholar
- 8.Soni PS (1982) Effects of oral contraceptive steroids on corneal curvature. Am J Optom Physiol Optics 59: 199–201Google Scholar
- 10.Vaughan D, Asbury T (1974) General ophthalmology. Lange, Los Altos, CaliforniaGoogle Scholar
- 11.Wang FM (1984) Perinatal ophthalmology. In: Duane TD, Jaeger EA (eds) Clinical ophthalmology. Harper and Row, PhiladelphiaGoogle Scholar
- 13.Weinstock FJ (1971) Transient severe myopia. JAMA 217(9):1245–1246Google Scholar