As refractive surgery for myopia is performed more commonly, it is essential that the refractive surgeon have an accurate understanding of the natural course of adult myopia. This knowledge, as well as a knowledge of long-term surgical results, is central to surgery planning and patient counseling regarding the advisability of initial surgery and augmentation surgery. Because most long-term studies of refractive surgery do not include a control group, one must have an understanding of normal age-related shifts to interpret the data. Most commonly, analysts of long-term refractive results have assumed underlying stability of adult myopia and attributed any noted postoperative instability to a surgical effect.1*6
Recently, Sawelson and Marks analyzed their longterm radial keratotomy results and concluded that the hyperopic shift (decrease in amount of myopia) seen after radial keratotomy was similar to normal physiologic shifts in the general population.7
Current understanding of the long-term course of simple myopia is based on a limited number of studies of general populations.8"11 Reviewers of the subject have arrived at somewhat different conclusions regarding the age of stabilization of myopic refractive errors and whether or not a physiologic hyperopic shift occurs.12"13
We characterize the stability of adult simple myopia by retrospectively analyzing the records of myopic patients who have been followed for more than 20 years at the Spokane Eye Clinic and compare results to reanalyzed data from the Prospective Evaluation of Radial Keratotomy (PERK) Study 10year results.1"3
PATIEMTS AND METHODS
Records of patients who entered the practice of the Spokane Eye Clinic from 1967 through 1974 were examined. Approximately 74,000 records were surveyed to find adult patients with myopia who had been followed for at least 20 years and had at least two manifest refractions during each decade. In addition, a few patient charts were located by direct referral from the eight ophthalmologists and two optometrists currently staffing the Spokane Eye Clinic. Data were excluded on older patients after the development of clinically significant cataract was documented in the chart. Selected patients were required to have myopic refractive errors between - 1.00 and -6.00 D during some period of their adult life. Patients who wore contact lenses for more than 2 months were excluded. We identified 413 eyes of 208 patients for this study; 110 patients were female and 98 were male.
Racial demographics of the patients could not be determined from the charts, but the group probably was representative of the demographics of Spokane County. In 1974, Spokane County was 98% white and 1% black with 1% other minority groups.
Data were analyzed on a decade-by-decade basis, starting at age 20 and continuing through age 70. All refractions were manifest refractions and were converted to spherical equivalents. Eyes were included in a decade subgroup only if an initial refraction within 2 years of the start of the decade and follow-up refraction within 2 years of the end of the decade were available. For example, the subgroup for the decade of patient age 30s included all eyes for which an initial refraction between ages 28 and 32 and a follow-up refraction between ages 38 and 42 were available. This allowed for an average follow-up of approximately 10 years for each decade subgroup. The subgroup for patient age decade 60s was an exception in that the last available refraction was used for the follow-up refraction, even if it occurred before age 68. Because of this method of analysis, each decade subgroup represents an overlapping but different subgroup of eyes.
For each decade subgroup, the average age and average manifest refraction at the start of the decade, the average length of follow-up, and the average shift in manifest refraction during the decade were observed. The percentages of eyes that became more myopic by at least -1.00 D, more myopic by -0.62 to -0.88 D, and for whom the shift was -0.50 D or less were recorded. Similarly, percentages of eyes with shifts in the hyperopic direction were recorded.
The decade-by-decade manifest refraction shifts were compared with shifts observed in eyes operated on in the Prospective Evaluation of Radial Keratotomy (PERK) Study. Details of the background, design, patient selection, surgical technique, and clinical examination methods of the PERK Study have been published.14 Numerous articles have presented PERK results from 1 to 10 years after surgery.1,3,15-17 However, the age-related shift in manifest refraction in first operated eyes between 6 months and 10 years after surgery has not previously been presented.
The mean refractive shifts between patient age decade 20s and 40s and between the 50s and 60s were determined. All eyes that had a refraction in patient age decade 20s and another in the 40s were included in the calculation of the former mean total refractive shift, and all eyes with a refraction in patient age decade 50s and another in the 60s were used for the latter shift.
The mean shift in refraction from decade 30s to 50s was compared for eyes with low, moderate, and higher initial myopia. The shift was measured as the difference between the first refraction in the decade 30s and the last refraction in the decade 50s. Low initial myopia was defined as a manifest refraction in the decade 30s between 0 and -1.00 D, moderate initial myopia as a refraction between -1.00 and -3.88 D, and high initial myopia as a refraction greater than -4.00 D. The mean and standard deviation of the initial myopia, initial age in the decade 30s, length of follow-up until the last refraction in the decade 50s, and total shift in refraction were determined for each group.
The symmetry of refractive shifts in the two eyes of patients was investigated. This analysis was also done by decade. A patient was included in a decade subgroup if at least two refractions, at least 1 year apart, had been performed during that decade. Therefore, some of the patients in these decade subgroups were different than in the subgroups described above. For each decade subgroup, the percentage of patients for whom the asymmetry of refractive shift was within 0.50 D and within 1.00 D were recorded.
Decade-by-decade Shifts In Manifest Refraction
The decade-by-decade average shifts in manifest refraction are presented in Table 1. The eyes included in each decade subgroup were followed for approximately 10 years, except for the 88 eyes in the decade of the 60s, which were followed for an average of 7.14 years (SD = 1.55 yrs). Eyes experienced a myopic shift (increase in amount of myopia) by an average of -0.60 D (n = 26, SD = 0.81 D) during the decade of the 20s, -0.39 D (n = 78, SD = 0.60 D) during the 30s, and -0.29 D (n = 130, SD = 0.56 D) during the 40s. In the decades of the 50s and 60s, the average shifts were hyperopic shifts of +0.28 D (n = 110, SD = 0.54 D) and +0.41 D (n = 88, SD = 0.48 D), respectively.
Shift in Manifest Refraction for Each Decade Subgroup
Distribution of Shift (% of Eyes) in Manifest Refraction for Each Decade Subgroup
Table 2 presents the percentages of eyes with refractive shifts of 1 D or more, between 0.62 and 0.88 D, and within 0.50 D in either the myopic or hyperopic directions, for each decade subgroup. In patient decade 20s, the shifts were predominantly in the direction of increased myopia, with 30.8% of 26 eyes becoming more myopic by at least -1.00 D. The percentage of eyes having myopic shifts of at least -1.00 D decreased in subsequent decades (17.9% of 78 eyes in the 30s, 11.5% of 130 eyes in the 40s, 0.9% of 110 eyes in the 50s), while percentages of eyes with at least +1.00 D of hyperopic shift increased from the 40s onward (0.8% of 130 eyes in the 40s, 10.9% of 110 eyes in the 50s, 22.7% of 88 eyes in the 60s). Since the average follow-up was only 7.14 years for eyes included in the subgroup of the 60s decade, the percentages for the decade of the 60s are based on simple linear extrapolations of the shifts that were actually recorded. This linear extrapolation probably somewhat overestimates the myopic shifts and underestimates the hyperopic shifts during this decade.
Comparison with Shifts in PERK Study Eyes
In the PERK Study, the average shift in manifest refraction between 6 months and 10 years after radial keratotomy was a hyperopic shift for each decade subgroup studied. The average shift in manifest refraction for the 95 eyes in the decade of the 20s was a hyperopic shift of +0.20 D (standard error (SE) = 0.10 D), for the 169 eyes in their 30s, a hyperopic shift of +0.71 D (SE = 0.08 D), and for the 39 eyes in their 40s, a hyperopic shift of +1.28 D (SE = 0.14 D) (Figure). In comparison, the 10-year shifts observed in the sample for this paper (Table 1) were myopic shifts for each decade subgroup.
Shift in Refraction between Ages 20 and 40 and between Ages 50 and 60
Mean shift in refraction for 115 eyes examined in their middle 20s and again in their middle 40s was a myopic shift of -0.86 D (SD = 0.96 D) (Table 3). For 185 eyes followed, on average, from the decade of the early 50s to the middle 60s, the mean refractive shift was a hyperopic shift of +0.55 D (SD = 0.61 D).
Figure: Bar chart comparing the mean 10year refractive shifts in the PERK study eyes and our sample of unoperated myopic eyes. The results are stratified by age decade. Standard error (SE) bars represent the estimated variability of the means.
Shift in Manifest Refraction between Ages 20s to 40s and 50s to 60s
Comparison of Refractive Shifts between 30s to 50s According to Initial Myopia
Comparison of Refractive Shift for Low versus Higher Initial Myopia Groups
Table 4 shows the average refractive shift for eyes followed from their 30s to 50s according to initial myopia. The average refractive shift was a myopic shift of -1.33 D (SD = 0.53 D) for 15 eyes with low initial myopia, -0.34 D (SD = 0.79 D) for 100 eyes with moderate initial myopia, and -0.14 D (SD = 1.26 D) for 43 eyes with high initial myopia. The average age at the initial refraction in the 30s decade was similar for the three groups, and the eyes in each group were followed for an average of 21 years.
Symmetry of Refractive Shift
Table 5 shows the symmetry of refractive shifts in the two eyes of patients. There was a high degree of symmetry in all decade subgroups. The refractive shifts in the two eyes of patients were within 0.50 D for over 75% of patients, regardless of decade subgroup, and were within 1.00 D for over 97% of patients.
Previous studies of age-related shifts in refractive errors have usually analyzed mean refractive errors within a general population composed of hyperopic, emmetropic, and myopic individuals. In most of the analyzed general groups, one would expect myopic individuals to represent less than 25% of the group. A consensus seems to exist that within the general population, a myopic shift in refractive error can be identified from approximately age 7 until the early 308.8-10 From tne middle 30s to the middle 40s, a hyperopic shift can be identified.8'9,1820 Our myopic population seems to differ from the general population in that a mean myopic shift occurs until approximately age 50 and thereafter a hyperopic shift occurs. This suggests that emmetropic and/or hyperopic populations experience a mean hyperopic shift at an earlier age than the myopic population.
The fact that the myopic population seems to differ from the general population with regard to agerelated refractive shifts emphasizes the importance of using an appropriate control group when analyzing long-term refractive surgery data. Recently, Sawelson and Marks analyzed their 10-year follow-up data on a series of radial keratotomy patients.7 Comparing the hyperopic shift within their group to that within the general population, they concluded that much of the observed postoperative hyperopic shift was of a physiologic nature rather than surgically induced. Had they had an available myopic control group, they might have reached the opposite conclusion.
Previously, the authors of the PERK Study reported a mean long-term hyperopic shift of approximately +-0.87 D between 6 months and 2 years postoperatively using cycloplegic refractions. The hyperopic shift between 2 and 10 years postoperatively was +0.06 D per year.3 Further analysis of the long-term PERK 10-year data was performed to determine if any age-related response could be identified using manifest refractions. The results of this analysis are summarized in the Figure.
It is apparent that the observed hyperopic shift increases with increasing age, at least through the patient age decade 40s, which was the oldest decade in which there were adequate numbers of PERK patients for analysis. A comparison of the expected myopic shift from our study with the age-related shift from the PERK study clearly suggests that the observed PERK hyperopic shift is surgically induced. The difference between the mean myopic shifts noted in our myopic population and those hyperopic shifts noted in the PERK postoperative population might be termed the "surgical effect." This surgical effect is greater than the observed PERK hyperopic shift and seems to increase with increasing age. This seems to confirm the findings of Nizam and colleagues, who analyzed the unoperated eyes of PERK study patients and found a mean myopic shift in the unoperated eyes, while the operated eyes were experiencing a hyperopic shift.21 Whether or not the physiologic hyperopic shift that occurs within the patient age decade of the 50s and 60s will be additive to the surgically induced hyperopic shift is unknown.
Symmetry of Refractive Shifts in the Two Eyes of Patients
This information should be useful in counseling radial keratotomy patients with regard to desired results, the wisdom of augmentation in older patients, and the advisability of doing the procedure in older patients.
Certain individuals within the adult myopic population will experience fairly large physiologic hyperopic and myopic shifts. It would be useful if one could predict which individual will have what type of agerelated shift. An unexpected finding in this study was that individuals with smaller amounts of myopia were more likely to have a greater myopic shift. This group appears to be a small group of three to five percent of the myopic population who experience the onset of myopia in their late teens and 20s, and have relatively large myopic shifts from their late 20s through 40s. It is possible that individuals with this adult onset myopia may be part of the group that has shown a long-term myopic shift after refractive surgery.3
While this study provides clinically useful information, the population studied is not an exact control group for any published surgical series and suffers from many of the problems of a retrospective clinical study. Ideally, one would like to have data from a prospective study of large numbers of adult patients followed for over 50 years with cycloplegic refractions. In an age marked by high population mobility and increased contact lens wear, the completion of a meaningful prospective study would be a formidable task. A cooperative long-term retrospective analysis of all types of refractive errors could be accomplished and would be a fundamental cornerstone of information for refractive surgeons.
1. Waring GO, Lynn MJ, Nizam A1 Kutner MH, Cowden JW, Culbertson W, Laibson PR, McDonald MB, Nelson JD, Obstbaum SA, Rowsey JJ, SaIz JJ, Bourque LB, the PERK Study Group. Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study five years after surgery. Ophthalmology 1991;98:1164-1176.
2. Waring GO, Lynn MJ, Strahlman ER, Kutner MH, Culbertson W, Laibson PR, Lindstrom RL, McDonald MB, Myers WD, Obstbaum SA, Rowsey JJ, Smith RE, PERK Study Group. Stability of refraction during four years after radial keratotomy in the prospective evaluation of radial keratotomy study [see comments]. Am J Ophthalmol 1991;111:133-144.
3. Waring GO, Lynn MJ1 McDonnell PJ, PERK Study Group. Results of the prospective evaluation of radial keratotomy (PERK) Study ten years after surgery. Arch Ophthalmol 1994;112:1298-1308.
4. Deitz MR, Sanders DR, Raanan MG, DeLuca M. Long-term (5 to 12 year) follow-up of metal-blade radial keratotomy procedures. Arch Ophthalmol 1994;112:614-620.
5. Sawelson H, Marks RG. Five-year results of radial keratotomy. Refract Corneal Surg 1989;5:8-20.
6. Arrowsmith PN, Marks RG. Visual, refractive, and keratometric results of radial keratotomy. Five-year follow-up [see comments]. Arch Ophthalmol 1989;107:506-511.
7. Sawelson H1 Marks RG. Ten year refractive and visual results of radial keratotomy. Ophthalmology 1995;102:18921901.
8. Tassman I. Frequency of the various kinds of refractive errors. Am J Ophthalmol 1932;15:1044-1053.
9. Slataper FJ. Age norms of refraction and vision. Arch Ophthalmol 1950;15:1044-1053.
10. Brown EVL. Net average yearly shifts in refraction in atropinized eyes from birth to beyond middle life. Arch Ophthalmol 1938;19:719-734.
11. Hirsch MJ. The longitudinal study of refraction. Am J Optom 1964;41:137.
12. Bores LD. Refractive Eye Surgery. Boston, MA: Blackwell Scientific Publications; 1993:32-34.
13. Waring GO. Refractive Keratotomy for Myopia and Astigmatism. St. Louis, MO: Mosby Year Book; 1992:11-13.
14. Waring GO1 Moffitt S, Gelender H, Gelender H, Laibson PR, Lindstrom RL, Myers WD, Obstbaum SA, Rowsey JJ, Safir A, Schanzlin DJ, Bourque LB, PERK Study Group. Rationale for and design of the National Eye Institute Prospective Evaluation of Radial Keratotomy (PERK) Study. Ophthalmology 1983;90:40-58.
15. Waring GO, Lynn M, Gelender H, Laibson PR, Lindstrom RL, McDonald MB, Schanzlin DJ, Sperduto RD1 Bourque LB, PERK Study Group. Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study one year after surgery. Ophthalmology 1985;92:177-198.
16. Waring GO, Lynn M, Culbertson W, Laibson PR, Lindstrom RL, McDonald MB, Myers WD, Obstbaum SA, Rowsey JJ, Schanzlin DJ, PERK Study Group. Three-year results of the Prospective Evaluation of Radial Keratotomy (PERK) Study. Ophthalmology 1987;94:1339-1354.
17. Waring GO, Lynn M, Fielding B, Asbell PA, Balyeat HD1 Cohen EA, Culbertson W1 Doughman DJ, Fecko P, McDonald MB, Smith RE, Wilson LB, PERK Study Group. Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study 4 years after surgery for myopia. JAMA 1990;263:1083-1091.
18. Exford J. A longitudinal study of refractive trends after age forty. Am J Optom 1965;42:685-692.
19. Hirsch MJ. Changes in refractive state after the age of fortyfive. Am J Optom 1958;35:229-237.
20. Bucklers M. Changes in refraction during life. Brit J Ophthalmol 1953;37:587-592.
21. Nizam A, Waring GO III, Lynn MS, Ward M, Asbell PA, Balyeat HD, Cohen E, Culbertson W, Doughman D, Fecko P, McDonald MB, Smith R, PERK Study Group. Stability of refraction and visual acuity during five years in eyes with simple myopia. Refract Corneal Surg 1992;8:439-444.
Shift in Manifest Refraction for Each Decade Subgroup
Distribution of Shift (% of Eyes) in Manifest Refraction for Each Decade Subgroup
Symmetry of Refractive Shifts in the Two Eyes of Patients