Abstract
Purpose.:
To investigate the pattern of diurnal variations in axial length (AL), choroidal thickness, intraocular pressure (IOP), and ocular biometrics over 2 consecutive days.
Methods.:
Measurements of ocular biometrics and IOP were collected for 30 young adult subjects (15 myopes, 15 emmetropes) at 10 different times over 2 consecutive days. Five sets of measurements were collected each day at approximately 3-hour intervals, with the first measurement taken at ∼9 AM and final measurement at ∼9 PM.
Results.:
AL underwent significant diurnal variation (P < 0.0001) that was consistently observed across the 2 measurement days. The longest AL was typically observed at the second measurement session (mean time, 12:26) and the shortest AL at the final session of each day (mean time, 21:06). The mean diurnal change in AL was 0.032 ± 0.018 mm. Choroidal thickness underwent significant diurnal variation (mean change, 0.029 ± 0.016 mm; P < 0.001) and varied approximately in antiphase to the AL changes. Significant diurnal variations were also found in vitreous chamber depth (VCD; mean change, 0.06 ± 0.029 mm; P < 0.0001) and IOP (mean change, 3.54 ± 0.84 mm Hg; P < 0.0001). A positive association was found between the variations of AL and IOP (r 2 = 0.17, P < 0.0001) and AL and VCD (r 2 = 0.31, P < 0.0001) and a negative association between AL and choroidal thickness (r 2 = 0.13, P < 0.0001). There were no significant differences in the magnitude and timing of diurnal variations associated with refractive error.
Conclusions.:
Significant diurnal variations in AL, choroidal thickness, and IOP were consistently observed over 2 consecutive days of testing.
It is well established that diurnal variations occur in a range of anatomic and physiological parameters of the eye, such as intraocular pressure (IOP),
1 –4 corneal thickness,
5 –7 corneal topography,
5,8 and anterior chamber biometrics.
4,9 The use of precise noncontact measurement techniques, such as partial coherence interferometry, has led to the finding that significant diurnal variation (range, 25–45 μm) also occurs in the axial length (AL) of the human eye, with the longest AL occurring during the day and the shortest during the night.
3,4,10 However, the relative consistency of diurnal AL rhythms across subjects and between days has been questioned.
3,10 In previous studies in which diurnal measures were performed on 2 separate days, the measurement days were separated by weeks and sometimes months,
3,10 which leaves open the possibility that longer term factors (e.g., seasonal variations) may have influenced some of the reported variability between days.
Although the presence of diurnal rhythms in the AL of human eyes is now well accepted, less is known about the mechanisms and ocular changes underlying these diurnal variations. One factor that is potentially involved in these changes is the eye's IOP. Surgically
11 –13 or mechanically
14 induced variations in IOP have been shown to be associated with changes in AL, which leaves open the possibility that the known diurnal variations in IOP
15 play a role in diurnal AL variations. The exact relationship between diurnal AL and IOP changes is unclear, with some recent studies suggesting that there is no relationship between the two variables
3 and others reporting a weak but significant positive association between the variations in AL and IOP.
4
As AL measurements have typically involved the determination of the distance from the cornea to the retinal pigment epithelium (RPE), diurnal changes in AL could be modulated by variations in the thickness of the choroid. There is evidence from research in chickens
16 –18 and marmosets
19 that choroidal thickness (CT) does undergo diurnal variations (increasing during the night and decreasing during the day). However, there have been only very limited investigations of diurnal variations of the choroid in human eyes. Brown et al.
20 in a retrospective analysis of partial coherence interferometry data, recently reported evidence of diurnal variations in CT in human eyes, and a trend for the choroid to fluctuate approximately in antiphase to AL in a small population of subjects.
As AL is the primary biometric determinant of refractive error, diurnal variations in this biometric parameter could be influenced by refractive error. Several animal studies have suggested a relationship between diurnal AL rhythms and refractive error.
16 –18,21 Studies of chickens have found that induced form deprivation
16 –18 or constant darkness
21 causes significant axial elongation and the development of myopia and also alters the normal diurnal rhythms of both AL and CT. Animal studies also suggest that phase differences between AL and choroidal rhythms may be an important factor in the regulation of ocular growth.
18,21 However, no previous study has prospectively investigated the potential differences in the magnitude and timing of diurnal AL rhythms between myopic and emmetropic human subjects.
In this study we sought to further investigate the underlying origins, relative consistency, and influence of refractive error on the diurnal variation of AL in human eyes, through measurements of ocular biometry (using an instrument capable of determining a comprehensive range of ocular biometric measures including CT) and IOP performed over 2 consecutive days, in populations of young adult emmetropic and myopic subjects.
Thirty young adult subjects aged between 18 and 30 years (mean ± SD, 25.16 ± 3.32) were recruited for the study. Seventeen of the subjects were male. None of the participants had any history of significant ocular or systemic disease, ocular injury, or surgery. Before the study, each subject underwent an initial ophthalmic screening to ensure good ocular health and to determine their refractive status. The subjects were classified according to their spherical equivalent refraction (SER) as either emmetropes (SER, +0.75 to −0.75 DS, n = 15; mean, −0.15 ± 0.31 DS) or myopes (SER, ≥ −1.00 DS, n = 15; mean, −3.95 ± 1.41 DS). No subject exhibited anisometropia greater than 1.00 DS or cylindrical refraction greater than 1.25 DC. All subjects had normal logMAR visual acuity of 0.00 or better.
Among the myopes, three subjects wore soft contact lenses. These subjects discontinued wearing the lenses for 1 week before the study and abstained from wearing them for the duration of the study. No wearers of rigid gas-permeable (RGP) contact lenses were included. As the intake of alcohol
22 and caffeine
23 have been found to influence IOP, all participants were asked to abstain from consuming alcoholic and caffeinated beverages from the evening before and for the 2-day duration of the study. Subjects were instructed to maintain consistent sleep/wake cycles for 7 days before commencement of the study, which was confirmed through completion of a modified version of the Pittsburgh Sleep Quality Index (PSQI) questionnaire by each participant.
24 All subjects had an average sleep duration of >5 hours and average sleep efficiency >65%. Approval from the university human research ethics committee was obtained, and written informed consent was obtained from all subjects. All subjects were treated in accordance with the Declaration of Helsinki.
To investigate ocular diurnal variations in each subject, we took a series of measurements of ocular biometrics and IOP over 2 consecutive days. On each day, five measurement sessions were conducted at 2.5- to 3-hour intervals, with the first measurement taken at approximately 9 AM (∼1–2 hours after the subjects had awakened) and the final measurement at approximately 9 PM. One emmetropic subject was unable to attend session 3 only on the first day of measurements. Each session took approximately 10 to 15 minutes to complete all the measurements and undertook their regular daily activities between measurement sessions.
Collecting the first measurement 1 to 2 hours after waking avoided the potential confounding of the results by the large changes in anterior eye parameters that are typically observed immediately after waking (as observed in central corneal thickness, [CCT] and anterior chamber depth [ACD]).
4 Noncontact techniques were used for all measurements, to avoid any corneal epithelial disruption as a result of instruments that contact the eye or the use of any anesthetic eye drops.
25 The order of clinical measurements was randomized at each measurement session for each subject, to minimize the risk of systematic bias.
A noncontact optical biometer (Lenstar LS 900; Haag-Streit AG, Köniz, Switzerland) was used to obtain the measurements of AL and other ocular biometric parameters. This instrument works on the principle of optical low-coherence reflectometry (OLCR) and recent studies have found that it provides highly repeatable results (reported intra- and inter-session repeatability for AL is 0.016 and 0.006 mm, respectively) that are comparable with other validated instruments.
26,27 The following ocular biometric measures were collected: CCT (the distance from the anterior to posterior corneal surface), ACD (the distance from the posterior corneal surface to the anterior crystalline lens surface), lens thickness ([LT], the distance from the anterior to posterior lens surface), vitreous chamber depth ([VCD], the distance from the posterior lens surface to the inner limiting membrane), and AL (the distance from the anterior corneal surface to the RPE). Seven measurements were obtained for each biometric parameter at each measurement session and were later averaged.
In addition to these automatically derived ocular measurements, manual analysis of the biometer data was performed to determine retinal thickness ([RT], distance from inner limiting membrane to RPE) and CT (distance from RPE to choroid–sclera interface). Previous studies using instruments based on similar principles have shown that the A-scan data originating from the posterior eye typically contain a series of peaks corresponding to retinal and choroidal structures,
20,28 with the anterior peak (P1) thought to originate from the inner limiting membrane of the retina, the central peak (P3) from the RPE, and the posterior peak (P4) from the choroid–sclera interface. Manually determining interpeak distances from the posterior portion of the A-scan by adjusting the retinal cursors with the biometer's software allows the determination of RT (distance from P1 to P3) and CT (distance from P3 to P4) from each subject's biometry data (method detailed elsewhere).
29 –31 RT and CT derived from the biometer (Lenstar; Haag-Streit) have been shown to correlate closely with RT and CT measured with spectral domain OCT.
31 An independent, masked observer performed the manual analysis to determine RT and CT for each subject.
A noncontact tonometer (Ocular Response Analyzer [ORA]; Reichert, Depew, NY) was used to measure IOP at each session. This tonometer
32 has been found to provide IOP measures that agree closely with Goldmann tonometry.
33,34 The instrument provides two estimates of IOP: IOPg, which is a Goldmann-correlated IOP measurement, and IOPcc, which takes corneal biomechanical properties into account and has been reported to be less affected by corneal properties than other tonometric techniques.
32,33 Given that corneal parameters such as CCT are known to vary diurnally,
4 we assumed that using IOPcc should provide a more reliable assessment of diurnal IOP changes. The mean IOPcc was calculated for each subject at each measurement session from a total of four readings at each session.
Figure 3 illustrates the diurnal variation in posterior eye biometrics (VCD, RT, and CT). Significant diurnal variations (
P < 0.0001) were found in VCD. Analogous to the AL results, the longest VCD was typically observed at the second session (mean time, 12:26) and the shortest during the night at the final session (mean time 21:06) on both the days of testing in the 27 subjects with available LT data. The repeated-measures ANOVA (between-subjects effect of refractive error) revealed that the group mean VCD of 17.19 ± 1.19 mm was significantly deeper in the myopic subjects (18.10 ± 0.97 mm) than in the emmetropic subjects (16.35 ± 0.62 mm;
P < 0.0001). The mean amplitude of change in VCD was 0.06 ± 0.029 mm (range, 0.15–0.02 mm), which was not significantly different between the refractive error groups (0.066 ± 0.033 and 0.073 ± 0.024 mm for the emmetropes and the myopes, respectively;
P = 0.786). The change in VCD also exhibited a significant effect of day of measurement (
P = 0.042; repeated-measures ANOVA).
Consistent P4 choroidal peaks could not be detected by the independent observer in all measurements for six subjects (three from each refractive error group), therefore the CT analysis represents data from 24 subjects. The group mean CT of 0.256 ± 0.049 mm was not significantly different between myopes (mean, 0.242 ± 0.020 mm) and emmetropes (mean, 0.269 ± 0.065 mm; P = 0.201). Significant diurnal variations (P = 0.011) were observed in CT. The choroid was typically found to be thicker at night and thinnest in the morning on both measurement days. The mean amplitude of change in CT was 0.029 ± 0.016 mm (range, 0.079–0.011 mm), which was not significantly different between the myopic (mean, 0.029 ± 0.013 mm) and the emmetropic (mean, 0.030 ± 0.019 mm) subjects (P = 0.231). Repeated-measures ANOVA revealed no significant diurnal variation in RT for 30 subjects (P = 0.162) over the 2 days of the experiment.
In summary, AL, CT, anterior chamber biometrics, and IOP all underwent significant diurnal variations that were consistently observed over 2 consecutive days of testing. AL was typically longest during the day and shortest during the night, and the changes appear to be largely due to variations in the posterior segment of the eye. CT changed by similar measured amplitude and was approximately in antiphase to the AL variations.
Presented at the 13th Scientific Meeting in Optometry and 7th Optometric Educators Meeting, Sydney, Australia, September 2010.
Disclosure:
R. Chakraborty, None;
S.A. Read, None;
M.J. Collins, None
The authors thank Emily Woodman and Fan Yi for assistance with data analysis procedures.