Corneal Diameter, Axial Length, and Intraocular Pressure in Premature Infants

Susan M. Tucker, MD,I Robert W. Enzenauer, MD, MPH,z Alex V. Levin, MD,3 ]. Donald Morin, MD, 'I Jonathan Hellmann, MD5

Purpose: Seventy premature infants 25 to 37 weeks' postconceptional age were examined during their first week of life to determine the correlation of corneal diameter, axial length, and intraocular pressure with gestational age and birth weight.

Methods: Corneal diameter measurement was determined with corneal templates, total axial length with standardized A-scan ultrasound, and intraocular pressure with a Tonopen II tonometer.

Results: Corneal diameter and total axial length showed parallel linear increases from 6.2 mm to 9.0 mm and 12.6 mm to 16.2 mm, respectively; however, no significant correlation was found between intraocular pressure and gestational age or birth weight. The mean intraocular pressure was 10.3 mmHg (standard deviation, 3.5).

Conclusion: Normative values are established for corneal diameter and total axial length as they relate to birth weight and gestational age, and a mean and standard deviation for intraocular pressure in the premature newborn. These values will aid the ophthalmologist in assessing ocular dimensions in premature infants. Ophthalmology 1992;99: 1296-1300

The expanding role of the ophthalmologist in the neonatal nursery has made an understanding of the normal mat­uration of the visual system increasingly more important. There are, however, few reports in the literature about the changes in dimensions of ocular structures in the pre­mature neonate. 1

-7 In the current study, we examined 70

premature infants between 25 and 37 weeks' postconcep-

Originally received: October 24, 1991. Revision accepted: March 5, 1992.

I Department of Ophthalmology, Unive~sity of Toronto, Toronto.

2 Department of Ophthalmology, Fitzsimons Army Medical Center.

) Department of Pediatric Ophthalmology, Wills Eye Hospital, Philadelphia.

4 Department of Ophthalmology, The Hospital for Sick Children, Toronto.

5 Department of Neonatology, The Hospital for Sick Children, Toronto.

Presented in part as a poster at the American Academy of Ophthalmology Annual Meeting, Anaheim, October 1991.

Reprint requests to J. Donald Morin, MD, Department of Ophthal­mology, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario, Canada M5G IX8.

1296

tional age to determine the correlation of corneal diam­eter, axial length, and intraocular pressure (lOP) with postconceptional age and birth weight. We have developed normative values for corneal diameter, axial length, and a mean and standard deviation for lOP in the premature neonate.

Methods

We examined all premature infants who met our inclusion criteria of younger than 37 weeks' postconceptional age and who were admitted to the neonatal nursery at The Hospital for Sick Children in Toronto, Ontario, Canada between June 1990 and February 1991. Study protocol and consent forms were approved by the Human Subjects Review Committee at The Hospital for Sick Children in Toronto, and informed consent was obtained.

Exclusion criteria comprised the presence of major or­gan dysfunction (excluding lungs), syndromes, birth weight small for gestational age, and discrepancy of more

Tucker et al . Ocular Structures in Premature Infants

than 1 week in the clinical age and the age calculated by dates. A neonatologist determined the infants' gestational age based on maternal obstetric history, prenatal ultra­sound, and a modified Dubowitz score.7 We calculated postconceptional age by adding the actual age of the infant at the time of ocular measurement to the gestational age in weeks. For data analysis, postconceptional age was proximated to the nearest week.

All children were measured as soon as they were stable enough to undergo an ocular examination. For each in­fant, one examination was performed. The same examiner measured all infants in a supine position through the ports of the isolette. We eliminated parallax in the corneal di­ameter readings by bringing the baby right up to the port­hole or by momentarily lifting the isolette top while taking the measurement. A neonatal Barraquer wire speculum was used to open the lids after a drop of 0.5% proparacaine was applied. The cornea was kept moist with a balanced salt solution. Shortly after insertion of the speculum, the majority of infants became calm.

Since the validity of using corneal diameter templates instead of calipers has previously been established,2 we measured the corneal diameter with specially designed clear plastic templates, ranging in size from 6.0 mm to 9.0 mm in 0.25-mm increments (Fig 1). Total axial length was then determined with the Ophthascan (Biophysic Medical, Clermont-Ferrand, France) ultrasound, a stan­dardized A-scan ultrasound. The sound velocity used was 1532 m/sec from the cornea to the anterior lens and from the posterior lens to the retina, and 1640 m/sec across the lens. Probe alignment was adjusted to obtain maximal spikes from the acoustic interfaces to optimize the accu­racy of each measurement (Fig 2). The corneal surface was contacted as lightly as possible to minimize compres­sion of anterior segment structures. Readings off the op­tical axis (as determined by proper corneal, anterior len­ticular surface, posterior lenticular surface, and retinal peaks) were excluded. Intraocular pressure was obtained with a hand-held Tonopen II (Oculab, La Jolla, CA) ap­planation tonometer. 8 Only digital pressure display values that fell within a 5 percentile reliability range were in-

Figure 1. Photograph of the corneal templates in use on the neonate.

Figure 2. Photograph of total axial length determined from an A-scan ultrasound.

cluded. The reported corneal diameter, total axial length, and lOP represent the average value for each parameter measured in triplicate.

Results were analyzed statistically. The correlation be­tween left and right eye readings was measured with Pear­son's product-moment correlation. Stepwise linear regression was used to determine which of the factors­age, weight, and sex-were predictive of eye measurement. Regression analysis was performed to determine the pre­cise relationship of corneal diameter and axial length to postconceptional age and birth weight. Because our in­terest is in the prediction of the relationship for an indi­vidual infant, 95% prediction limits were calculated for the eye measurements. These limits are wider than the 95%-confidence limits for the mean eye measurement as given levels of predictor variables. Because prediction limits are used, ,2 (proportion of variability) is reported instead of, (correlation coefficient).

Results

Seventy of the 227 patients younger than 37 weeks' post­conceptional age examined met the inclusion criteria for this study. The others were excluded because of failure to obtain consent, major organ dysfunction, systemic syn­dromes, discrepancy between the infant's calculated age and clinical assessment of age, or because the infant's size for gestational age was less than that for those in the third percentile.

Figure 3 shows the age distribution of the 70 premature infants involved in the study. An effort was made to ex­amine the patients as soon as possible after birth. Age at examination ranged from the date of birth to 9 days (mean, 4.9 days). Forty-five of the 70 infants were male and 25 were female. The left eye was the first to be ex­amined in 38 infants, and the right eye was first in 32 infants. Corneal diameter and total axial length measure­ments were obtained for all infants in the study. Eleven infants continued to cry or were agitated throughout the examination; lOP measurements for these infants were

1297

Ophthalmology Volume 99, Number 8, August 1992

1/1 E .!!! .5 '0

~ E :J Z

12

10

8

6

4

2

o .111 1111 25 26 27 28 29 30 31 32 33 34 35 36 37

Postconceptlonal Age (wks)

Figure 3. Histogram of the number of infants examined at each postcon­ceptional week.

omitted because they were considered inaccurate because of presumed associated elevation ofIOP.

There was near perfect correlation for the corneal di­ameter of the left and right eyes (r = 0.99, P = 0.0001), as well as for the axial length of left and right eyes (r = 0.97, P = 0.0001). The maximum difference between axial measurements for each eye pair was 0.7 mm, which is within instrument error. Therefore, we used the com­bined values from both eyes to obtain the dependent vari­able.

Figures 4A and B show the relationship of corneal di­ameter to postconceptional age and birth weight, respec­tively. Corneal diameter increases by 0.5 mm every 15 days (corneal diameter [mm] = 0.23 X postconceptional age [weeks] + 0.43) and every 417 g (corneal diameter [mm] = 0.0012 X birth weight [g] + 5.97).

11 95th Percentile

10 Mean

E .§. 9 .. !

E .§.

5th Percentile ..

E III 8 '6

11 c 7 .. 0 0

6

5 I I

24 26 A

I I I I I 28 30 32 34 36

Postconceptlonal Age (weeks)

I 38

<II

~ ~

~ 8

Figures 5A and B show the relationship of total axial length to postconceptional age and birth weight. Total axial length increases by 1.0 mm every 23 days (total axial length [mm] = 0.30 X postconceptional age + 5.06) for every 714 g (total axial length [mm] = 0.00014 X birth weight [g] + 12.27).

The mean lOP, measured for the 59 quiet infants, was 10.4 mmHg in the right eye and 10.0 mmHg in the left. The upper 95% prediction limit was 18.0 mmHg for the right eye and 17.5 mmHg for the left. The correlation between right and left eyes for lOP was not as strong as that for corneal diameter and total axial length (r = 0.64, P = 0.0001). However, the difference was not statistically significant.

Discussion

Awareness of the dimensions and lOP of the maturing eye in the neonate is important when premature infants are assessed for developmental anomalies and syndromes. We have established normative values with 95% prediction limits for corneal diameter and total axial length as related to postconceptional age and birth weight (Figs 4 and 5). Corneal diameter and total axial length show a parallel linear relationship with postconceptional age and birth weight. From 25 to 37 weeks, corneal diameter and total axial length increased from 6.2 mm to 9.0 mm and from 12.6 mm to 16.2 mm, respectively. Postconceptional age and birth weight had no predictive value for lOP.

Only a few studies have attempted to determine mor­phometric values for premature infants. Musarella and Morin3 correlated corneal diameter and lOP with birth weight and postgestational age in premature babies. When they examined 37 infants with a mean postconceptional age of34 weeks (standard deviation, 2 weeks), they found

11 95th Percentile

10 Mean

9 5th Percentile

8

7

6

5 I I I I I I I I I I

500 1000 1500 2000 2500 3000 3500

B Weight (gms)

Figure 4. A, mean corneal diameter plotted against postconceptional age with 95% prediction limits (~ = 0.82; P = 0.0001; CD = 0.23 X PCA + 0.23). B, mean corneal diameter plotted against birth weight with 95% prediction limits (~ = 0.79; P = 0.0001; CD = 0.0012 X birth weight + 5.97). CD = corneal diameter; PCA = postconceptional age.

1298

Tucker et al . Ocular Structures in Premature Infants

18

17

E 16 .s .c 15 & c GI ...I i;j 14 >< ct

13

12

11 d4 2~ ~8 3b a2 i4 a6 a8

A Postconceptlonal Age (weeks)

95th Percentile

Mean

5th Percentile E .s .c C, c

oS i;j

~

18

17

16

15

14

13

12

11

95th Percentile

Mean

5th Percentile

I I I I I I I I I I I I I 500 1000 1500 2000 2500 3000 3500

B Weight (gms)

Figure 5. A, mean total axial length plotted against postconceptional age with 95% prediction limits (~ = 0.82; P = 0.0001; TAL = 0.30 X PCA + 5.06). B, mean axial length plotted against birth weight with 95% prediction limits (~ = 0.75; P = 0.0001; TAL = 0.00014 X birth weight + 12.27). TAL = total axial length; PCA = postconceptional age.

that corneal diameter correlated better with birth weight (Pearson correlation coefficient, 0.74) than with postcon­ceptional age (Pearson correlation coefficient, 0.57). However, we found postconceptional age to be a more significant predictor than birth weighqr = 0.91, r2 = 0.82). Our results yield corneal diameter values 0.5 mm lower (correlation coefficient r = 0.89 and proportion of variability r2 = 0.79). The greater statistical significance of our results may reflect the increased number of infants in our study, as well as our accurate determination of postconceptional age.

In a study of fetal cadaver specimens, Harayama et al1.2 reported corneal diameter, as well as sagittal, trans­verse, and vertical diameters of the eye taken from fetuses 12 to 40 weeks old. The mean corneal diameter values we obtained correlate well with the average (horizontal plus vertical) corneal diameters reported by Harayama et al2 for infants 25 to 28 weeks' gestational age. Similarly, our mean values for axial length in infants 25 to 37 weeks' gestational age generally fell within 0.5 mm of the mean values reported by Harayama et all for corresponding weeks.

Additionally, our total axial lengths correlate well with the ultrasonic values that Blomdahl9 obtained for newborn infants weighing between 2000 and 4500 g. When data from our graphs are extrapolated to values for 40 weeks' gestation, they correspond directly to Larsen'slO ultrasonic measurements of the axial length of the eye expressed as a function of weight in newborn infants. As with corneal diameter, we found postconceptional age to be a more significant predictor than birth weight.

It is well known that normal lOP in children is lower than the normal levels in adults. I 1-13 The normal lOP for premature infants, however, has been much less exten­sively studied. Brockhurst,14 using a McLean tonometer, obtained values varying from 6.5 to 33.0 mmHg (average,

24.5 mmHg) in 59 premature infants 27 to 34 weeks' postconceptional age. Dolcet, 15 using a Schi0tz tonometer, obtained an average lOP of 35.0 mmHg in 5 premature infants 900 to 2000 g compared with 41.0 to 56.0 mmHg in full-term infants. However, scleral rigidity in infants is tonometrically higher than in adults, 16 making indentation tonometer readings less reliable in this population. This factor has less influence in applanation tonometry. I? As well, problems inherent in the measurement of lOP in premature infants also could account for differences in results: tensions fluctuate and elevate with a rise in venous pressure caused by crying and straining or with increased extraocular muscle tonus caused by squeezing eyelids. To ensure accuracy, lOP should be measured when a preterm infant is not crying. Brockhurstl4 reported that most of the infants in his study were quiet and relaxed; Dolcetl 5

did not comment on their behavior. All our lOP mea­surements were taken in quiet, awake infants. Musarella and Morin3 measured lOP in 37 quiet infants 29 to 38 weeks' gestation using a hand-held Perkins applanation tonometer. They obtained average values of 18.0 (standard deviation, 2.3) and 18.6 mmHg (standard deviation, 2.3) in the right and left eyes, respectively. They found no significant correlation between lOP and postconceptional age or birth weight (Pearson correlation coefficient, 0.35; P = 0.05). Although our observations confirm no corre­lation between lOP and postconceptional age or birth weight, we found lower lOP measurements. This may re­flect the increased number of infants in oW" study. The mean lOP in our study population was 10.3 mmHg (stan­dard deviation, 3.5 mmHg). Hence, we conclude that 97.5% of premature infants should have an lOP of 18.0 mmHg or less.

We have established normative values for corneal di­ameter, total axial length, and lOP in premature new­borns. We have not, however, created growth curves for

1299

Ophthalmology Volume 99, Number 8, August 1992

these parameters since we did not follow our patients lon­gitudinally. These normative values should greatly aid the ophthalmologist in evaluating the ocular dimensions of premature babies.

References

I. Harayama K, Amemiya T, Nishimura H. Development of the eyeball during fetal life. J Pediatr Ophthalmol Strabismus 1981 ;18(4):37-40.

2. Harayama K, Amemiya T, Nishimura H. Development of the cornea during fetal life: comparison of corneal and bulbar diameter. Anat Rec 1980; 198:531-5.

3. Musarella MA, Morin JD. Anterior segment and intraocular pressure measurements of the un anaesthetized premature infant. Metab Pediatr Syst Ophthalmol 1985;8(2-3):53-60.

4. Isenberg SJ, McCarty JW, Rich R. Growth of the conjunc-tival fornix and orbital margin in term and premature in-fants. Ophthalmology 1987;94: 1276-80.

5. Sivan Y, Merlob P, Reisner SH. Eye measurements in pre-term and term newborn infants. J Craniofac Genet Dev Bioi 1982;2:239-42.

6. Jeanty P, Dramaix-Wilmet M, Van Gansbeke D, e t al. Fetal ocular biometry by ultrasound. Radiology 1982; 143: 513-6.

1300

7. Dubowitz LMS, Dubowitz V, Goldberg C. Clinical assess­ment of gestational age in the newborn infant. J Pediatr 1970;77: 1-10.

8. Kao SF, Lichter PR, Bergstrom TJ, et al. Clinical compar­ison ofthe Oculab Tono-Pen to the Goldmarin applanation tonometer. Ophthalmology 1987;94:1541-4.

9. Blomdahl S. Ultrasonic measurements of the eye in the newborn infant. Acta Ophthalmol 1979;57: 1048-56.

10. Larsen JS. The sagittal growth of the eye. IV. Ultrasonic measurement of the axial length of the eye from birth to puberty. Acta Ophthalmol 1971 ;49:873-86.

II.

12.

13.

14.

15 .

16.

17.

Youn DH, Yu YS, Park IW. Intraocular pressure and axial length in children. Korean J Ophthalmol 1990;4:26-9. Dominguez A, Banos SM, Alvarez MG, et al. Intraocular pressure measurement in infants under general anesthesia. Am J OphthalmoI1974;78:11O-6. Goethals M, Missotten L. Intraocular pressure in children up to five years of age. J Pediatr Ophthalmol Strabismus 1983;20:49-51. Brockhurst RJ. The intraocular pressure of premature in­fants. Am J Ophthalmol 1955;39:808-11. Dolcet L. Tension ocular del recien nacido. Arch Soc Of tal Hispano-Am 1952; 12: 1057-63. Ytteborg J. Investigations of the rigidity coefficient in chil­dren's eyes. Acta Ophthalmol 1960;38:658-74. Manzitti E, Darnel A. Valores de la tonometria aplamitica en ellactante normal. Arch Oftal B Air 1964;39:360-2.