The examinations included automated and clinical refraction in cycloplegia. A total of 192 students (100 females and 92 males) completed the study. Results: In ...
Clinical Science Refractive changes among Norwegian university students – A three-year longitudinal study Bettina Kinge and Anna Midelfart Department of Ophthalmology, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway
ABSTRACT. Purpose: The aim of this study was to investigate the changes in refractive error during a three-year period among university students exposed to high educational demands. Methods: A three-year longitudinal cohort study was performed among 224 Norwegian engineering students (mean age 20.6 years, 117 females and 107 males) measuring their refraction at the beginning and the end of the period. The examinations included automated and clinical refraction in cycloplegia. A total of 192 students (100 females and 92 males) completed the study. Results: In the student population under study the prevalence of myopia increased significantly from 48% to 65% (p∞0.001, right eye) and the mean refractive error increased significantly from ª0.64∫2.18 D (nΩ224) to ª1.21∫2.30 D (nΩ192) (p∞0.001, right eye) during the period. Of eyes emmetropic at the start of the study (nΩ49, right eye), 59% became myopic. Among the eyes initially myopic (nΩ92, right eye), 73% progressed further into myopia, with at least ª0.37 D during the three-year period. Of eyes initially hyperopic (nΩ51, right eye), 8% became myopic, while 14% became emmetropic. Conclusions: A shift in refraction towards myopia is frequent among university students during their study period. Key words: adult-onset myopia – myopic shift – students, university level. Acta Ophthalmol. Scand. 1999: 77: 302–305 Copyright c Acta Ophthalmol Scand 1999. ISSN 1395-3907
F
or over one hundred years there has been discussion on the etiology of myopia, a disorder that today affects a substantial part of the general population. Cross-sectional studies presenting the prevalence of refractive errors in different demographic strata throughout the world have revealed a considerable variation in the prevalence of myopia in different populations. In the Nordic countries, between 25% and 33% of the general adult population are reported to be myopic (Pa¨rssinen et al. 1985; Kinge et al. 1998) with at least 25–30% of these being adult-onset myopia, i.e. with an onset at the age of 18 years or older (Fledelius 1995).
302
Theories about a biological versus an environmental cause have been alternately dominant. Investigating possible environmental factors, several studies have indicated that there is a statistical correlation between myopia and the amount of near-work (Goldschmidt 1968, 1990). The attention was particularly directed towards young adults in demanding educational systems where a high prevalence of myopia was revealed (Shulkin & Bari 1986; Midelfart et al. 1992). Performing longitudinal studies, a refractive change in the direction towards myopia among adults with a high amount of close work was demonstrated in some studies, but these are relatively
scarce (Grosvenor & Scott 1993; McBrien & Adams 1997). Students at university level represent a group of special interest because of a possible risk of developing adult-onset myopia or adult progression of myopia during their education (Shulkin & Bari 1986; Zadnik & Mutti 1987; Lin et al. 1996). However, the results obtained until now show substantial variation in different populations. The aim of this study was to estimate the refractive change in a group of university students in our country exposed to high educational demands over a three-year period with special respect to adult-onset myopia and adult progression of myopia.
Material and Methods In 1992, a sample of 224 first-year students (107 males and 117 females, mean age 20.6∫1.1 years) at the Norwegian Institute of Technology (NTH), University of Trondheim, Norway, (renamed Norwegian University of Science and Technology in 1995), was examined at the Department of Ophthalmology, Faculty of Medicine at this university. They were selected at random and were a representative sample of the Norwegian engineering students at the NTH (Kinge & Midelfart 1994). Criteria for exclusion from the study were diabetes mellitus or previous or present eye disease or injury. However, no one was to be excluded according to these criteria. The first part of the examination consisted of the measurement of visual acuity (Baily-Lovie distance visual chart) and the power of corrective lenses, if any (Allergan Humphrey Lens Analyzer), as well
as slit-lamp examination of the anterior segment. Thereafter, one drop of 1% cyclopentolate (Cyclopentolate Minims, Chauvin) was instilled in each eye twice with a 5 minute interval and cyloplegic measurements were taken after 30 minutes. Cycloplegia was employed to rule out the possibility of latent hyperopia or pseudomyopia due to accommodative spasm. Subsequently, the refractive state was assessed with an automated refractor (Allergan Humphrey 500 autorefractor at the first examination and Nidek AR-820 autorefractor at the second examination) and thereafter by a clinical examination, using the subjective distance refractive method. The cycloplegic autorefractor results were considered a preliminary guidance only, as recommended in a study recently published (Kinge et al. 1996). All students participating in the study answered a questionnaire about their height, age, age at first lens prescription etc. Three years later, in 1995, the students were re-examined using the same procedure and by the same examiner as in 1992. In this second part of the study, 192 students (86%) were enrolled. Among the 32 (15 male and 17 female) students that dropped out of the study, 6 refused to participate, 14 had terminated their study, 8 had left the fourth-year, and 4 took their fourth-year abroad. Among the participating students, 92 (48%) were males and 100 (52%) females. The study was approved by the regional committee for medical research ethics and the students gave informed consent. Presentation of results For all data analysis, data for the right eye of each subject was used. The results obtained by the subjective distance refractive method were used in the analysis and the refractive values were converted to spherical equivalents (SER) for the analysis (i.e. sphere value π0.5 of the cylinder value) (Kinge et al. 1996). Myopia comprised a spherical equivalent (SER) equal to, or less than ª0.25 D, emmetropia an SER between ª0.25 D and π0.50 D and hyperopia equal to, or more than π0.50 D. The refractive error was regarded stable when the three year refractive change was within ∫0.37 D (McBrien & Adams 1997). Adult-onset myopia was defined as myopia with the onset at or after the age of 18, 19 or 20 years, respectively, according to the reply in the questionnaires about the first pre-
scription for glasses (Fledelius 1995). Three different onset-groups were established to make the results comparable to other studies. The data was stored on a personal computer and statistical analysis were performed using the Statistical Analyzing System (SAS) (SAS Institute Inc. 1987). Chi-square, Student’s, Wilcoxon Signed Rank and Wilcoxon Rank Sum tests were employed to determine the statistical differences. The variation was expressed as standard deviation (SD) and the level of statistical significance was 5%.
Results The prevalence of refractive errors in this population obtained at the start of the longitudinal study has been reported previously and will be discussed only briefly here (Kinge & Midelfart 1994). Analysing the data obtained initially among the 32 students that dropped out of the study, and comparing these to the data registered at the same time among the rest of the students, there was no significant difference regarding mean refractive error (pΩ0.09, right eye) or prevalence of my-
opia (pΩ0.45, right eye) between these two groups. In the enrolled population (nΩ192), there was no significant difference between the mean refractive error in right eye versus left eye, neither at the first nor at the second measurement session (p±0.8). Regarding the refractive changes observed among the participating students (nΩ192) during the longitudinal study, 33 of their 100 initially non-myopic right eyes had become myopic, with 85% of them now having a refractive error less than ª1 D. Thus, the prevalence of myopia had increased significantly from 48% to 65% (p∞0.001) during this period (Fig.1, Table 1). The prevalence of adult-onset myopia, based on the reported age at the first prescription for glasses and choosing the onset at or after the age of 18, 19 or 20 years, respectively, was found to be as high as 54%, 44% or 39% (right eye). There has been a distinct shift in the direction towards myopia in the total student population with 60% of the right eyes having a change of at least ª0.37 D and only 2% of π0.37 D or more. 38% were found to be stable with changes below these limits. A myopic change of at
Fig. 1. Prevalence of myopia, emmetropia and hyperopia in 1995 among the 192 students examined both in 1992 and 1995, and description of their previous refractive error (right eye).
Table 1. The changes of clinical refractive state during the three-year study among the total population for the initially myopic, emmetropic and hyperopic right eyes in percent.
Final clinical refractive error Myopic Emmetropic Hyperopic All
All (nΩ192)
Initially myopic (nΩ92)
65% 13% 22% 100%
100% 0% 0% 100%
Initially Initially emmetropic hyperopic (nΩ49) (nΩ51) 59% 37% 4% 100%
8% 14% 78% 100%
303
Table 2. The refractive changes (D) among the total population for the initially myopic, emmetropic and hyperopic right eyes.
Refractive change (D) ª1.00–ª2.38 ª0.50–ª0.99 ª0.25–ª0.49 0.00–ª0.24 0.00–0.24 0.25–0.75 All
All (nΩ192)
Initially myopic (nΩ92)
20% 30% 21% 22% 4% 3% 100%
32% 32% 16% 16% 3% 1% 100%
Initially Initially emmetropic hyperopic (nΩ49) (nΩ51) 12% 39% 18% 22% 4% 4% 99%
6% 18% 31% 33% 6% 6% 100%
Table 3. The mean refractive change (D) during the three-year period among the total population for the initially myopic, emmetropic and hyperopic right eyes.
Mean refr. change (D)
All (nΩ192)
Initially myopic (nΩ92)
Initially emmetropic (nΩ49)
Initially hyperopic (nΩ51)
ª0.51∫0.49
ª0.66∫0.51
ª0.48∫0.47
ª0.27∫0.35
least ª0.37 D over the three years was found in 73% of the initially myopic right eyes, in 61% of the initially emmetropic right eyes, and in 37% of the initially hyperopic right eyes. Among the progressing myopic eyes, 30% progressed more than ª1 D with a maximum progression of ª2.38 D and 70% progressed by ª1 D or less. Table 2 illustrates the distribution of refractive change (right eye) in the total population, as well as among the initially myopic, emmetropic and hyperopic eyes. When first evaluated, the mean refractive error in the total group was ª0.64∫2.18 D (nΩ224, right eye) with a refractive range from ª8.88 D to π7.00 D (nΩ224, right eye) compared to ª1.20∫2.30 D and a refractive range from ª7.50 D to π6.50 D (nΩ192, right eye) three years later. The maximum myopia was reduced from ª8.88 D to ª7.50 D due to one drop-out. The median refractive value was ª1.75 D in the myopic group at the final examination. The mean change of ª0.51∫0.49 D (nΩ192) was statistically significant (pΩ 0.0001) (Table 3). The differences in mean refractive change between the three initial groups were statistical significant (pΩ 0.0001) (Table 3). The three-year difference of ª0.59∫0.52 D in the right eye of females was greater than ª0.43∫0.43 D for the male eyes, but the difference in the amount of change between male and female eyes was not statistically signifi-
304
cant (pΩ0.09). Among the 38 right eyes that had a mean refractive change of at least ª1 D during the three-years study, 66% were in females. This gender difference came close to statistical significance (pΩ0.059). Regarding the use of corrective glasses, 47% of the students with a myopic right eye at the end of the three-year study had corrective glasses with appropriate power, this means within ∫0.29 D of the refractive error measured by the subjective distance refractive method. Among the rest, 28% of the myopic right eyes were undercorrected by at least 0.30 D, 11% were overcorrected by at least 0.30 D and 14% had no correction at all. The mean refractive error in the latter group was ª0.60∫0.29 D (nΩ17).
Discussion In cohort studies a high follow-up rate is desirable to minimize the chance of biased estimates. In the present study, the follow-up rate of 86% is regarded as high. Furthermore, analysing the data obtained among the 32 students that dropped out of the study, their refraction was not significantly different from the participants. Thus, the risk of introducing selection bias is considered to be low. The definition of both myopia and hyperopia will naturally influence the prevalence of the refractive errors. Obviously,
a cut-off point at ª0.50 D or ª0.75 D would give a lower prevalence of myopia. In the present study made on young cycloplegic eyes, the cut-off point for myopia was chosen at ª0.25 D, which is in accordance with other studies, both cross-sectional and longitudinal (Teasdale et al. 1988; Midelfart et al. 1992; Lin et al. 1996). The results obtained in the present study are in agreement with other studies demonstrating that the prevalence of myopia increases in young adults exposed to high educational demands over several years (Dunphey et al. 1968; Zadnik & Mutti 1987; Lin et al. 1996). In the present study, an overall increase in the prevalence of myopia of 17% during a three-year university study was revealed. This is considerably more than the increase of myopia by 3% among medical students at Taiwan who were followed for 5 years (Lin et al. 1996). This is, however, explained by the very high initial prevalence of myopia of 93% among these medical students, and thereby a much smaller potential for myopic shift in the group. Law students in California followed for 6 months had a 38% myopic shift (ت0.5 D) (Zadnik & Mutti 1987), which is a relatively high value compared to the 49% (ت0.50 D) registered in the present study over a threeyear period. With respect to adult-onset myopia, and choosing 19 years as the limit age of onset, the prevalence of 44% in the present study is in close accordance with the prevalence of 43% obtained previously among medical students in Norway (Midelfart et al. 1992). Choosing 18 years as the age of onset, the prevalence of 54% in the present university based study is substantially higher than the 35% previously reported in a group of educationally less selected adults in Denmark (Fledelius 1995). This observation supports the theory of an increased risk of developing myopia during a university course (Shulkin & Bari 1986). Naturally, it would be of interest to know the kind and degree of refractive error change over time in a group of nonacademic controls, but so far no such detailed follow-up study based on a similar ethnic group has come to our knowledge. There are some other studies of refractive changes to demonstrate the same tendency of myopic shift as in the present study among other professions with sightdemanding conditions for the eyes. A longitudinal study among clinical microscopists with a mean age of 30 years revealed that 45% developed myopia during
a two-year period (McBrien & Adams 1997). Furthermore, a case-control study among a small group of textile workers in Norway exposed a high prevalence of myopia and adult-onset myopia among those having a short working distance of 30 cm (Simensen & Thorud 1994). These results support the theory that near-work such as reading or close working distance is a risk factor for the development of myopia. It is well known that the cost of correcting myopia with glasses or contact lenses is quite high. During the last years, the public interest in refractive surgery is increasing in many countries and many people have been treated with this procedure. However, the risk of development of retinal complications and vision loss in the enlarged myopic eye will probably not change after such surgery. Until further clarified, it seems to be of great value to investigate all possible factors contributing to the development of myopia, and particularly to follow groups longitudinally which might be at a relatively high risk of becoming myopic or progressing their myopia.
Acknowledgements We would like to thank all the participating students. Furthermore, we thank Jarand Rystad, M.Sc., for help with data analysis. The abstract was presented in part at the Meeting, Norw. Ophthalmol Society, Nov. 97.
References Dunphey EB, Stoll MR & King SH (1968): Myopia among American male graduate students. Am J Ophthalmol 65: 518–521. Fledelius HC (1995): Myopia of adult onset. Can analysis be based on patients memory? Acta Ophthalmol Scand 73: 394–396. Goldschmidt E (1968): In: On the Etiology of Myopia. An epidemiological study. Munksgaard, Copenhagen. Goldschmidt E (1990): Myopia in humans: can progression be arrested? Wiley, Chichester (Ciba Foundation Symposium 155), p. 222– 223. Grosvenor T & Scott R (1993): Three-year changes in refraction and its components in youth-onset and early adult-onset myopia. Optom Vis Sci 70: 677–683. Kinge B & Midelfart A (1994): Refractive errors among Norwegian engineering students. Ophthalmic Epidemiology 1: 5–13. Kinge B, Midelfart A & Jacobsen G (1996): Clinical evaluation of the Allergan Humphrey 500 autorefractor and the Nidek AR-1000 autorefractor. Br J Ophthalmology 80: 35–39. Kinge B, Midelfart A & Jacobsen G (1998): Refractive errors among students and controls. Acta Ophthalmol Scand 76: 692– 695. Lin LL-K, Shih Y-F, Lee Y-C, Hung P-T & Hou P-K (1996): Changes in ocular refraction and its components among medical students – a 5-year longitudinal study. Optom Vis Sci 73: 495–498. McBrien NA & Adams DW (1997): A longitudinal investigation of adult-onset and adult-progression of myopia in an occu-
pational group. Invest Ophthalmol Vis Sci 38: 321–333. Midelfart A, Aamo B, Sjøhaug KA & Dysthe BE (1992): Myopia among medical students in Norway. Acta Ophthalmol (Copenh) 70: 317–322. Pa¨rssinen O, Leskinen A-L, Era P & Heikkinen E (1985): Myopia, use of eyes, and living habits among men aged 33–37 years. Acta Ophthalmol (Copenh) 63: 395–400. SAS Institute Inc. (1987): SAS/STAT guide for personal computers, Version 6 Edition. Cary, NC: SAS Institute Inc. Shulkin DJ & Bari MM (1986): Deteriorating Vision: An Occupational Risk for the Medical Student. Arch Ophthalmol 104: 1274. Simensen B & Thorud LO (1994): Adult-onset myopia and occupation. Acta Ophthalmol (Copenh) 72: 469–471. Teasdale TW, Fuchs J & Goldschmidt E (1988): Degree of myopia in relation to intelligence and educational level. Lancet 10: 1351–1354. Zadnik K & Mutti DO (1987) Refractive Error Changes in Law Students. Am J Optom Phys Opt 64: 558–561. Received on August 26th, 1998. Accepted on January 23rd, 1999.
Corresponding author: Bettina Kinge MD Ivar Aasens vei 29 N-0373 Oslo Norway. Fax: 47 22 13 67 79 E-mail: bkinge/online.no
305
