Effect of Short-Term, High-Dose Retinol on Dark Adaptation in Aging and Early Age-Related Maculopathy

Abstract

purpose. To examine the effect of a short course of high-dose retinol (preformed vitamin A) on dark adaptation in older adults with normal retinal health or early age-related maculopathy (ARM).

methods. The study design was a randomized, double-masked, placebo-controlled experiment. Adults ≥50 years of age whose fundus photographs for the eye to be tested psychophysically fell within steps 1 to 9 of the Age-Related Eye Disease Study (AREDS) Grading System were randomly assigned to a 30-day course of 50,000 IU oral retinol or a placebo. At baseline and 30-day follow-up, dark adaptation was tested and the Low Luminance Questionnaire (LLQ), an instrument for assessing difficulty with vision in reduced lighting, was administered. Primary outcomes of interest were rod- and cone-mediated parameters of dark adaptation, with scores on the LLQ’s six subscales as secondary outcomes.

results. The sample consisted of 104 participants with 52 each in the intervention and placebo groups. There were no group differences in baseline variables. At 30-days, the dark-adaptation parameters of cone time-constant, cone threshold, rod-cone break, and rod threshold did not differ. The retinol intervention group had significantly larger (i.e., steeper) rod slopes, indicating faster sensitivity recovery, than did the placebo group (P = 0.0419). There were no group differences in scores on the LLQ subscales driving, extreme lighting, emotional distress, general lighting, or peripheral vision. The retinol group had a higher score by five points on the mobility subscale compared with the placebo group (P = 0.0141). Those who had the most self-reported change on the mobility subscale at day 30 were more likely to have greater change in the speed of dark adaptation, as indicated by the rod slope parameter (r = 0.24, P = 0.0141).

conclusions. A short-term, high-dose course of retinol increased the rate of rod-mediated dark adaptation in older adults who were in the early phases of ARM or were exhibiting normal retinal aging. These results are consistent with the hypothesis that depositions and other structural changes in the retinal pigment epithelium and Bruch’s membrane in aging and early ARM cause a localized retinoid deficiency.

Photoreceptor function and survival are critically dependent on the retinal pigment epithelium (RPE) and Bruch’s membrane to regulate the transport of nutrients, fluid, ions, and metabolites to and from the subretinal space. ¹ Dysfunction and degeneration of the RPE are associated with photoreceptor death in both animal models and human donor eyes. ^{2 3} Aging leads to changes in the RPE-Bruch’s membrane complex that could compromise the metabolic exchange needed for normal photoreceptor function and survival. These changes include progressive thickening of Bruch’s membrane, ^{4 5 6} accumulation of extracellular material between the RPE and Bruch’s membrane, ^{7 8} reduced hydraulic conductivity of Bruch’s membrane, ⁹ and changes in the structure of RPE cells. ¹⁰ Similar but more severe changes have been observed in human donor retina with the clinical signs of early age-related maculopathy (ARM), including the accumulation of extracellular material in the RPE and Bruch’s membrane area, ^{11 12 13} such as cholesterol. The quantity of abnormal deposits directly relates to the degree of RPE and photoreceptor degeneration. ^{12 14} Studies have implicated both local intraocular cells and plasma lipoproteins as sources of these depositions, ^{8 15 16} and their build-up is presumably due to excessive production and/or impaired clearance to the choroid. ⁵  

These observations lead to the consideration that sub-RPE deposits in the aged retina and in early ARM may cause a diffusion barrier that disrupts metabolic exchange between the choroid and photoreceptors, leading to compromised photoreceptor function and even death. ¹⁷ Photoreceptor loss is characteristic of both aging and ARM. Rods are more vulnerable in the aging process than are cones ¹⁸ and also show earlier signs of ARM pathogenesis than do cones. ^{19 20} Functional studies are consistent with this histologic evidence. Scotopic (rod-mediated) function is impaired in older adults ^{21 22} and in those with early ARM, ^{23 24} including dark-adaptation delays that are particularly dramatic. ^{17 25 26}  

What mechanisms might underlie these dark-adaptation delays in aging and early ARM? One possibility is impaired transport of essential nutrients, such as vitamin A, across the abnormal extracellular matrix and RPE-Bruch’s membrane complex. Vitamin A deficiency causes preferential rod dysfunction and eventual photoreceptor death. ^{27 28 29 30 31} A scarcity of available vitamin A to combine with the protein opsin to form the visual pigment rhodopsin also leads to a specific change in the rate of rhodopsin regeneration and recovery of light sensitivity after light exposure. ³² Sorsby’s fundus dystrophy (SFD) ³³ is a rare human genetic disease that has structural and functional similarities to ARM, in that extracellular material accumulates between the receptors, and choroid and rod-mediated light sensitivity is impaired. After retinol (vitamin A) administration in relatively high doses for several days, four patients with SFD exhibited reversals in their rod-mediated light sensitivity deficit and/or dark-adaptation impairment. ^{34 35} A similar, although smaller, effect has been observed in autosomal dominant late-onset retinal degeneration (L-ORD), ³⁶ a condition also characterized by sub-RPE deposits. These findings raise the intriguing question of whether disrupted transport of vitamin A across Bruch’s membrane to the rod outer segments may also be operative in aging and early ARM. ¹⁷ Other potential mechanisms that may underlie retinal inefficiency in vitamin A use and deployment in persons with ARM include potential direct or indirect effects of retinol on the RPE itself, including modulation of RPE function ³⁷ and alterations in vitamin A metabolism, such as those caused by the RPE65 genetic abnormalities identified in other macular disorders. ³⁸  

We describe the results of a double-masked, placebo-controlled, randomized experiment on the effect of short-term, high-dose preformed retinol on dark adaptation in older adults in normal retinal health and those with early ARM. On the basis of the mechanisms just discussed, we hypothesized that the rate of rod-mediated dark adaptation would increase after a 30-day course of retinol compared with those receiving the placebo. Because cone photoreceptors are less vulnerable to ARM pathogenesis in its earlier stages than are rods, cone-mediated parameters of dark adaptation were expected to be less affected by retinol supplementation than rod-mediated parameters. 

Methods

This study was approved by the Institutional Review Board of the University of Alabama at Birmingham. The research followed the Tenets of the Declaration of Helsinki. Informed consent was obtained from all participants after the nature and possible consequences of the study were explained. 

Participants were recruited from the comprehensive ophthalmology and the retina services of the Department of Ophthalmology, University of Alabama at Birmingham, based on clinic patients seen over a 14-month period from August 2003 to September 2004. Eligibility criteria were as follows: at least 50 years of age; best corrected, distance visual acuity as listed in the medical record of 20/80 or better in at least one eye. Because the primary focus of the study was on normal retinal aging and early ARM, an acuity cutoff of 20/80 was used; the eye to be tested psychophysically had to have funduscopic grading steps between 1 and 9, as determined by the Age-Related Eye Disease Study (AREDS) Grading System. ³⁹ These steps are indicative of either normal retinal aging (step 1) or various levels of the early phases of ARM (steps 2–9). Stereoscopic color 30° fundus photographs were taken with a fundus camera (FF450 Plus; Carl Zeiss Meditec, Dublin, CA) at the baseline visit after dilation of the pupil to at least 6 mm. Photographs were evaluated according to the AREDS System by graders at the University of Wisconsin Reading Center. 

Participants were excluded if (1) the AREDS grading system indicated that they had advanced disease, either central geographic atrophy (step 10) or exudative disease (step 11), in the eye to be tested psychophysically. Advanced ARM was an exclusion, because severe photoreceptor damage would be highly probable, and thus the likelihood of visual improvement after a retinol intervention would be very low; (2) the medical record or a general health interview indicated that they had any of the following: glaucoma, optic neuropathy, or any ocular conditions other than ARM; refractive error (spherical equivalent) with an absolute value of more than 6 D; neurologic diseases such as Alzheimer’s disease, Parkinson’s disease, history of stroke; diabetes; serious frailty; or medical conditions expected to lead to death or disability within 12 months; (3) they had a vitamin A deficiency defined as ≤30 μg/dL in serum determined by a high-pressure liquid chromatography (HPLC) procedure ⁴⁰ ; (4) they had hypercalcemia or abnormal liver function, which can be exacerbated by high dose vitamin A; and/or (5) they could not perform the psychophysical task used to measure dark adaptation. 

Figure 1summarizes the study design. The baseline visit consisted of serum collection, measurement of acuity and contrast sensitivity, dark-adaptation testing, questionnaires, and fundus photography. Randomization occurred approximately 3 days after the baseline visit, after the results of the serum tests were available, since they established medical eligibility. Participants were randomly assigned to a 30-day course of a daily tablet containing 50,000 IU vitamin A or a perceptually identical placebo (Tishcon, Westbury, NY) and were not informed of their group assignments. Vitamin A and the placebo were assigned code numbers by the manufacturer. Vitamin A levels in the tablet and placebo were confirmed by an independent laboratory. The safety of a short-term course of this level of vitamin A in healthy older adults has been established. ⁴¹ On the randomization day, the appropriate tablets were mailed by overnight delivery to the participant. Thus, day 1 of the 30-day course began the day after randomization. On or around day 30 ±2 days, the entire in-clinic protocol was repeated. The importance of compliance in taking the daily tablet was emphasized by the project coordinator on four occasions: at the baseline visit, in a follow-up letter delivered with the tablets, and in telephone calls at days 10 and 20. At the day-30 follow-up visit, the participant brought in any remaining tablets not taken during the 30-day course. A physician was on call 24 hours and available if the participants needed to report side effects encountered while taking the 30-day course of vitamin A or placebo. 

All study personnel including all investigators and staff involved in testing and interacting with participants were masked with respected to participant group assignment. Best corrected distance visual acuity was measured for each eye using the ETDRS chart ⁴² and expressed as logMAR (logarithm of the minimum angle of resolution). Contrast sensitivity was assessed for each eye with the Pelli-Robson chart and its standard administration protocol ⁴³ and scored by the letter-by-letter method. 

Dark adaptation was measured with techniques similar to those used in our earlier work. ^{22 25} Thresholds were measured (Humphrey Field Analyzer [HFA] model 640; Carl Zeiss Meditec), a computer-automated perimeter for measuring light sensitivity that was modified for scotopic testing and dark adaptometry, as described previously. ²² The HFA was adapted to include an infrared charge-coupled device (CCD) camera and light source to monitor fixation in the dark and an additional filter wheel to control target wavelength. To control for size, pupils were dilated with 1% tropicamide and 2.5% phenylephrine hydrochloride before testing. In all subjects a pupil diameter of ≥6 mm was achieved and verified under scotopic conditions before and after dark adaptometry. The subject’s head was positioned on a chin-forehead rest. The test eye was aligned to the fixation light using the HFA’s built-in camera. The fellow eye was patched. Subjects viewed the test target from a distance of 30 cm with their best optical correction for the test distance. The test eye underwent bleaching (11 ms) using an electronic flash of white light (7.65 log scotopic trolands-second) that produced an equivalent ∼98% bleaching of the area of the retina to be tested. ⁴⁴ The stimulus for measuring the cone-mediated threshold was a 650-nm circular test spot (Ealing 35-3961, full width at half maximum [FWHM] 11.4, peak 50%), and the stimulus for measuring rod-mediated threshold was a 500-nm circular test spot (Ealing 35-3508, FWHM 7.4, peak 50%). Stimuli were 1.7° in diameter and located 12° in the inferior visual field on the vertical meridian. During threshold measurement, the stimulus was presented every 2 to 3 seconds for a 200-ms duration. 

Cone-mediated threshold measurement began immediately after flash offset and terminated when sensitivity remained constant for 5 minutes. Rod-mediated threshold measurements began 5 minutes after flash offset. Threshold measurement continued until 60 minutes elapsed or the rod threshold was stable for 5 minutes after the rod-cone break. A three-down, one-up modified staircase procedure was used to estimate threshold. An external computer controlled the threshold measurement procedure and recorded responses. The subject was asked to press a response button when the target was visible and had 1500 ms to make a response after target onset. If the subject indicated the target was visible, the target intensity was decreased 0.3 log unit steps on successive trials until the subject stopped responding that the target was present. After the subject responded that the stimulus was invisible, target intensity was increased by 0.1 log unit until the subject responded that the target was once again visible. This target intensity was defined as the threshold. Successive threshold measurements started with the target intensity 0.3 log unit brighter than the previous threshold estimate. Threshold estimates were made twice every minute for the first 10 minutes and once every minute thereafter. 

Cone-mediated and rod-mediated dark-adaptation functions were expressed as log sensitivity as a function of time (minutes) after bleaching offset. The primary outcomes of interest on which the experimental and placebo groups were compared were the cone- and rod-mediated parameters of dark adaptation. Parameters were obtained by applying models of human dark adaptation to the thresholds as a function of time (described later), which describe the individual’s cone and rod dark-adaptation functions. ⁴⁵ This method avoids the potential for bias that occurs when using “hand-fitting” techniques. The analyst who fit these functions was masked to the participant’s group assignment and whether the data were from baseline or day 30. 

For the cone-mediated dark-adaptation function, a single exponential fit was used to estimate the cone time constant and cone sensitivity parameters. The cone time constant is the time constant of the exponential model and is an estimate of cone sensitivity recovery speed. The cone sensitivity is the plateau of the exponential function and is an estimate of the absolute threshold of the cone photoreceptors. 

For the rod-mediated dark-adaptation function, the parameters of interest were the rod-cone break, the rod slope, and the rod sensitivity. The rod-cone break is the time in minutes after bleaching offset at which the rods are more sensitive to the 500-nm stimulus than the cones. The rod slope parameter is the slope of sensitivity recovery during the second component of rod-mediated dark adaptation described by Lamb et al. ^{32 46 47} The rod sensitivity parameter is the average of the last three sensitivities of the rod slope. Because some AMD patients did not exhibit a third component before the end of the dark-adaptation test (60 minutes in duration), only thresholds through the second component were analyzed. To remove the third component, rod-mediated dark-adaptation functions were first fit with a three-linear-component model described previously. ^{25 45} The portion of the function identified as the third component ³² was discarded for analytic purposes. The remaining data were fit with the following bilinear equation, to estimate the rod parameters: log sensitivity = y intercept + a × minutes + max(minutes − rod-cone break,0) × b. The second slope equals a + b. 

Questionnaires were interviewer administered and included a review of demographic information (age, gender, race-ethnicity). General health was estimated through a questionnaire that asked about the presence of medical problems in 17 areas. The 32-item Low Luminance Questionnaire (LLQ) ⁴⁸ asked participants about the extent to which they experienced visual problems under low luminance and nighttime conditions. Subscales, scored from 0 to 100 (100, no difficulty; 0, so difficult that the person does not undertake the activity), consisted of driving, extreme lighting conditions, mobility, emotional distress, general dim lighting, and peripheral vision. 

Statistical Analysis

The t- and χ² tests were used to compare the study groups at baseline with respect to continuous and categorical variables respectively, including demographics, visual function, and health characteristics. Baseline dark-adaptation parameters and night-vision-low-luminance questionnaire scores were also compared by using t-tests. Paired t-tests were used to compare serum vitamin A levels within each study group at each of the study visits. For the 30-day measurements of these outcome variables, a linear regression model was used. The dependent variable for this analysis was the individual dark-adaptation parameter or low-luminance questionnaire score as measured at the follow-up visit. Each model contained two independent variables: the associated baseline dark-adaptation parameter or night-vision-low-luminance questionnaire score and a variable representing treatment group. This analytic approach is more attractive than simply computing the difference between baseline and follow-up, as it is better able to account for regression to the mean. ⁴⁹ Correlation coefficients (Pearson r) were calculated for the association between changes in variables between baseline and day 30. P ≤ 0.05 (two-tailed) was considered statistically significant. 

Results

There were 141 persons who had baseline visits. Results of the baseline visit indicated that 31 of these persons did not meet eligibility criteria and so were excluded (three had acuity worse than 20/80 in the better eye, eight had advanced ARM, four did not pass the liver and/or hypercalcemia screening tests, and 16 could not reliably make psychophysical judgments). No persons were ineligible based on serum vitamin A deficiency. Six persons withdrew from the study after baseline and randomization (five because of family illness or crisis and one because of migraine). Thus, the final sample consisted of 104 participants who were randomized to the vitamin A (n = 52) and placebo groups (n = 52) and who completed both the baseline and 30-day follow-up visits. During the course of the study, six persons in the treatment group reported transitory side effects (one had hot flashes, four nausea, one headache, two eye pain, one lethargy, and one blurry vision). Two persons in the placebo group reported side effects (one headache and one increased blood pressure). 

Table 1provides descriptive information at baseline for the two groups, with respect to demographic variables, visual acuity, contrast sensitivity, early ARM presence, serum vitamin A levels, general health, and compliance in taking tablets. There were no statistically significant differences in baseline variables between the two groups. Participants were, on average, in their early 70s, in large part white, split evenly between males and females, and found to have three to four comorbid conditions. The vitamin A group tended to have less severe ARM as measured by the AREDS fundus grade compared with the placebo group; however, the observed difference was not statistically significant. Acuity in the tested eye was modestly impaired, on average, in both groups. Compliance in taking the tablet each day was excellent in both groups, with approximately one pill missed by each subject during the 30-day period. Though the vitamin A group tended to be more compliant, the difference was not statistically significant. 

Table 2presents descriptive information at baseline for dark-adaptation parameters and subscale scores on the LLQ. The distributions of these variables did not differ between the two groups. LLQ subscale scores averaged for the most part in the 60s to the 80s, indicating moderate to serious problems in visual activities in low-lighting conditions. 

Dark-adaptation parameters at the 30-day visit are shown in Table 3 . Listed in the table for each group separately are the mean of each dark-adaptation parameter adjusted for that parameter’s baseline value. The groups did not significantly differ in cone time constant, cone threshold, rod-cone break or rod threshold. The vitamin A group had significantly larger rod slopes, indicating a faster sensitivity recovery, than did the placebo group (P < 0.0419). There were some differences in baseline demographic, visual function, and medical characteristics that may explain this difference, despite the lack of statistically significant group differences in the baseline variables. However, after adjustment for all the variables in Table 1 , the vitamin A group demonstrated larger rod slopes than the placebo group (0.16 vs. 0.14, respectively; P = 0.08). That the probability associated with this comparison increased was not unexpected, given the number of variables included in the model. The magnitude of the association remained unchanged (i.e., 0.02), suggesting a lack of confounding by the characteristics in Table 1 . To evaluate whether any observed association between dark-adaptation parameters and treatment group differed according to ARM status, an interaction term was introduced into the model. The interaction term was not statistically significant in any of the models (all P > 0.5), suggesting that any observed differences between treatment groups were similar in those with and without ARM. Figure 2shows composite dark adaptation at day 30 for the vitamin A group and the placebo group for cone-mediated responses (top) and rod-mediated responses (bottom). Visual acuity and contrast sensitivity in the tested eye were not different in the two groups at 30 days (P = 0.949 and P = 0.621, respectively). 

With respect to the LLQ (Table 3) , there were no group differences in scores on the subscales of driving, extreme lighting, emotional distress, general lighting, and peripheral vision. The vitamin A group had a higher score by 5 points on the mobility subscale compared with the placebo group (P < 0.0141). This 5-point difference between the vitamin A and placebo groups (88.4 vs. 83.6, respectively) and the statistical significance (P = 0.0224) remained after adjustment for the variables in Table 1 . Change from baseline to day 30 in the mobility subscale score on the LLQ was significantly associated with changes in the rod slope (Pearson r = 0.24, P = 0.0141). Those who had the most self-reported improvement on the mobility subscale at day 30 tended to have a greater increase in the rod slope parameter (Fig. 3) . Adjustment for the variables in Table 1had little influence on this association (Pearson r = 0.22, P = 0.0333). 

Serum vitamin A significantly increased over the 30-day period in the vitamin A group, from a mean at baseline of 62.47 ± 20.10 to 68.47 ± 19.18 μg/dL (SD) at 30 days (P = 0.0106). There was no change in serum vitamin A in the placebo group (baseline mean, 59.58 ± 14.17 μg/dL; 30-day mean, 59.59 ± 15.37 μg/dL). In the vitamin A group, the change in serum vitamin A over the 30-day period was not associated with the change in rod slope (P = 0.3764). there was a borderline association between change in serum vitamin A and change in scores on the mobility subscale of the LLQ (P = 0.0873). 

Discussion

In a sample of older adults with early ARM or in normal retinal health, the rate of rod-mediated dark adaptation became more rapid after a 30-day course of high-dose retinol, compared with those receiving a placebo. This was revealed by the treated group having a significantly steeper slope than the placebo group during the rod-mediated recovery of visual sensitivity after bleaching of the photopigment. From a human factor standpoint, the effect size in this study signifies that participants receiving the retinol, on average, detected a target after 30 minutes of dark adaptation that was twice as dim as a light detectable by individuals receiving the placebo. The retinol intervention, although having an effect on a rod-mediated parameter, did not have an effect on either of the cone-mediated parameters (cone time constant, cone sensitivity). This is consistent with the selective vulnerability of rod photoreceptors and comparative sparing of cone photoreceptors in aging and early ARM, based on findings in both histopathologic studies and functional studies. ^{17 18 19 20 21 22 23 24 25 26} The absence of an impact on retinol supplementation on cone photoreceptor function may also stem from the fact that cones, unlike rods, have an alternative source of retinol through the retinal vasculature and neurosensory retina ⁵⁰ and thus are supplied adequate retinoids in the early stages of ARM. 

This dark-adaptation rate improvement is rather modest in size, but similar in magnitude to effects reported in a study of a retinal disorders characterized by sub-RPE deposits, which can provide some context for the present findings. In a study on L-ORD, high-dose, ³⁶ short-term retinol supplementation accelerated rod-mediated dark-adaptation kinetics to a similar degree, as reported in the present study. In addition, in a study on scotopic vision in SFD, ³⁴ which tested dark adaptation in two patients, one patient’s dark-adaptation data were unaffected by retinol supplementation, whereas the other patient showed a dramatic improvement in rod kinetics. The earlier studies on L-ORD and SFD were not randomized, placebo-controlled experiments, and thus effect sizes are not referenced against the results of a control group, as they are in the present study. 

At a biological level, the responsiveness of rod-mediated dark adaptation to a short course of high-dose retinol is consistent with the hypothesis that depositions and other structural changes in the RPE/Bruch’s membrane complex in aging ^{4 5 6 8 51 52} and early ARM ^{11 12 15} cause a diffusion barrier that disrupts normal metabolic exchange, leading to a local shortage of vitamin A. In this framework, a scarcity of retinol in turn causes rod photoreceptor dysfunction and eventually death, ^{27 28 30} which should be noted are also features of early ARM pathogenesis. ^{19 20} Although this experiment does not provide direct evidence of in vivo retinoid deficiency, our results would be predicted by this hypothesis. These data cannot inform us of the site of the dysfunction that limits the availability of retinoids necessary for visual sensitivity recovery. Increased systemic vitamin A concentrations may force additional vitamin A across Bruch’s membrane into the RPE cells, via mass action. Alternatively, increased levels of vitamin A may overcome possible impaired transport between the RPE cells and rod outer segments. 

Excessive accumulation of lipofuscin fluorophores in the RPE, especially A2E, which relies on retinol for its biosynthesis, is characteristic of degenerative diseases of the macula including Stargardt disease and ARM. ^{53 54} Photoreceptor dysfunction in Stargardt disease results from photoreceptor degeneration caused by A2E-mediated toxicity to the RPE. ⁵⁵ The gene affected in Stargardt disease is ABCR, ⁵⁶ and it has been shown that in the ABCR^−/− mouse, isotretinoin treatment reduces A2E and lipofuscin accumulation. ⁵⁷ Thus, based on these findings, one might predict that a retinol intervention in ARM may actually exacerbate the condition, including the dark-adaptation deficits. However, the preponderance of evidence is that allelic variation in ABCR, even though associated with Stargardt disease, does not have a role in ARM. ⁵⁸ Furthermore, although the accumulation of lipofuscin plays a role in ARM-related cell injury and death, ⁵⁹ rod photoreceptor dysfunction and loss in ARM is not greatest in the retinal areas where lipofuscin is most concentrated, ^{19 23 60} implying that any role for lipofuscin in rod dysfunction is indirect, at best. 

The data presented in this article highlight the potential of this pathway in inciting early ARM pathogenesis, and as such, suggest a possible avenue for prevention and/or intervention. At present there are no proven ways to prevent early ARM or to slow its progression once it has emerged. Although long-term vitamin A at the high dose administered in the present study is not advisable given its toxicity, it may be worthwhile to investigate the impact of a chronic, lower dose of vitamin A, perhaps on an intermittent schedule, on changes in rod-mediated dark adaptation in older adults with normal retina or early ARM, predicated on the prior establishment of the medical safety of the intervention. Regardless of whether one views retinol as a candidate interventional strategy for early ARM or not, our results are significant because they suggest that consideration be given to a potential role for retinoid deficiency in early ARM pathogenesis. 

Previous epidemiologic work on preformed vitamin A (retinol) did not reveal a protective association with ARM. ⁶¹ However, in the earlier study advanced ARM was the outcome of interest in contrast to the focus in the current study on early disease and normal aging, with photoreceptor function as the outcome of interest. Fundus appearance may be too crude a measure to reveal retinol effects in aging and early ARM. Previous work has demonstrated that early ARM is present (e.g., significant accumulation of basal laminar deposits, altered photoreceptor morphology) before lesions associated with the disease are clinically visible in the fundus. ^{13 60} In fact, in our own data there were no differences in the fundus changes over 30 days as indexed by the AREDS grading system in the treated group compared with the placebo group (P = 0.3613), despite the fact that the treatment group exhibited faster rod-mediated dark adaptation and reported less difficulty with mobility at low luminance at 30 days. This pattern of findings illustrates the inadequacy of case definitions of early ARM that exclusively rely on fundus appearance as revealed in photographs and suggests that physiological or functional characteristics may be useful in defining the earliest stages of the disease. 

Retinol supplementation improved the recovery speed of the rods as measured by the rod slope, but no improvement (i.e., no decrease) was found in the rod-cone break parameter. Because the rod-cone break parameter depends on both cone and rod photoreceptors, the parameter must be interpreted carefully. The cone plateau of the rod-mediated dark-adaptation function improved, on average (i.e., cones became more sensitive), although the change did not reach statistical significance. Increased cone sensitivity causes a downward shift of the cone plateau of the rod-mediated dark-adaptation function. This downward shift causes a delay in the appearance of the rod-mediated portion of the dark-adaptation function, even if the rod recovery rate is moderately increased. Thus, improved cone function can mask an improvement in rod function as assessed by the rod-cone transition break parameter. 

The self-report data also revealed a positive effect of the vitamin A intervention on rod-mediated vision, with persons in the intervention group reporting decreased difficulty in mobility tasks under low lighting (e.g., social events in the evening, mobility within a darkened theater, seeing furniture when lighting is poor, concern about falling at night). What adds strength to these self-report data is that those persons who at day 30 had the greatest improvements in the speed with which they dark adapted also reported the most improvement in their ability to carry out mobility behavior under low luminance. This association extends earlier reports that scotopic function in the elderly is associated with the extent of self-reported night-vision problems, including fall risk. ^{62 63}  

This study has several strengths. Its randomized, placebo-controlled design allowed for rigorous protection from bias and the ability to make inferences about causation. Normal retinal aging and ARM severity were defined by the implementation of a standard and accepted fundus grading system based on stereofundus images. Persons with serum vitamin A deficiency, who might be expected to show dramatic dark-adaptation rate improvements after a vitamin A regimen, were excluded from the study. The main finding of the study—the positive impact of a high-dose, 30-day course of retinol on rod-mediated vision—was confirmed both psychophysically and through self-report. Potential limitations must also be addressed. Younger adults were not enrolled, however, many studies have demonstrated that dark adaptation cannot be improved by retinol supplementation in healthy young and middle-aged adults when serum vitamin A levels are within the normal ranges. ⁶⁴ One may wonder if the dark-adaptation effect size was modest in this study because of a ceiling effect in the dark-adaptation functions. A certain degree of impairment in rod-mediated dark adaptation was not an eligibility criterion of the study. However, when the baseline dark-adaptation functions of the participants in this study were compared with published data for young and middle-aged adults, dark-adaptation parameters of study participants were dramatically impaired (all P < 0.0001), ²² suggesting ample room for improvement in study participants. Another limitation is that the sample size precluded subgroup analyses on different AREDS grades; however, future research can explore this question. 

In summary, a short-term, high-dose course of retinol increased the rate of rod-mediated dark adaptation in older adults who were in the early phases of ARM or were exhibiting normal retinal aging. These results are consistent with the hypothesis that depositions and other structural changes in the retinal pigment epithelium and Bruch’s membrane in aging and early ARM cause a localized retinoid deficiency. Previous research indicates that a scarcity of retinol causes rod photoreceptor dysfunction and death and thus the present data highlight the possibility that retinoid deficiency may contribute to the pathogenesis of early ARM—a possibility worthy of further study. 

 Supported by National Institute on Aging Grant R01-AG04212, National Eye Institute Grant R21-EY14071, Research to Prevent Blindness, Inc., and the EyeSight Foundation of Alabama. CO is a Research to Prevent Blindness Senior Scientific Investigator.

 Submitted for publication September 28, 2005; revised October 27 and November 7, 2005; accepted January 3, 2006.

 Disclosure: C. Owsley, (P); G. McGwin, None; G.R. Jackson, (P); D.C. Heimburger, None; C.J. Piyathilake, None; R. Klein, None; M.F. White, None; K. Kallies, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Study design and steps in the protocol.

Table 1.

 

Baseline Demographic, Visual Function, and Medical Characteristics Stratified for the Vitamin A and Placebo Groups

	Vitamin A Group (n = 52)	Placebo Group (n = 52)	P
Age, mean y (SD)	71.8 (9.3)	71.7 (8.1)	0.9821
Gender, % (n)
Female	53.9 (28)	61.5 (32)	0.4272
Male	46.1 (24)	38.5 (20)
Race, % (n)
White	94.2 (49)	98.1 (51)	0.6176
African American	5.8 (3)	1.9 (1)
Visual acuity, mean logMAR (sd)
Tested eye	.14 (.15)	0.15 (.13)	0.5472
Fellow eye	.38 (.35)	0.38 (.34)	0.9954
Contrast sensitivity, mean (sd)
Tested eye	1.37 (.17)	1.41 (.14)	0.1388
Fellow eye	1.24 (.35)	1.25 (.40)	0.8665
AREDS fundus grade, % (n)
1 (normal retinal aging)	42.3 (22)	36.5 (19)	0.1147
2–6 (early ARM)	48.1 (25)	38.5 (20)
7–9 (intermediate ARM)	9.6 (5)	25.0 (13)
General Health, mean (sd)
Co-morbid medical conditions (n)	4.2 (2.1)	3.5 (2.1)	0.1222
Smoking, % (n)
Never	44.2 (23)	48.1 (25)	0.7336
Previous	46.2 (24)	42.3 (22)
Current	7.7 (4)	9.6 (5)
Unknown	1.9 (1)	0.0 (0)
Serum vitamin A (μg/dL), mean (sd)	62.47 (20.10)	59.58 (14.17)	0.3997
Tablets not taken, mean (sd)	1.2 (1.1)	1.5 (1.1)	0.1123

Table 2.

 

Baseline Dark Adaptation and LLQ Results for the Vitamin A and Placebo Groups

	Vitamin A Group (n = 52)	Placebo Group (n = 52)	P
Dark adaptation, mean (sd)
Cone time constant	5.73 (11.45)	3.18 (4.01)	0.1322
Cone sensitivity	2.40 (.48)	2.40 (.33)	0.4272
Rod-cone break	19.68 (10.50)	19.92 (9.21)	0.9034
Rod slope	0.15 (.07)	0.14 (.06)	0.4338
Rod sensitivity	2.89 (.68)	2.82 (.72)	0.6473
LLQ, mean (sd)
Driving subscale	61.35 (31.43)	66.96 (30.55)	0.3582
Extreme lighting subscale	71.22 (16.54)	73.52 (18.78)	0.5297
Mobility subscale	86.39 (16.74)	88.65 (11.73)	0.4271
Emotional distress subscale	88.82 (15.62)	91.47 (13.27)	0.3575
General dim lighting subscale	82.58 (16.27)	86.31 (14.48)	0.2193
Peripheral vision subscale	83.17 (18.78)	86.86 (16.61)	0.2916
Composite score subscale	78.94 (14.93)	82.29 (14.33)	0.2543

Table 3.

 

30-Day-Visit Dark Adaptation and LLQ Results for Vitamin A and Placebo Groups, Adjusted for Each Variable’s Baseline Value

	Vitamin A Group (n = 52)	Placebo Group (n = 52)	P
Dark adaptation, adjusted mean (sd)
Cone time constant	2.89 (5.51)	4.08 (6.51)	0.2869
Cone sensitivity	2.46 (.42)	2.53 (.39)	0.3225
Rod-cone break	20.87 (12.78)	20.72 (11.97)	0.9381
Rod slope	0.17 (.07)	0.15 (.07)	0.0419
Rod sensitivity	3.03 (.78)	2.94 (.75)	0.4241
LLQ, adjusted mean (sd)
Driving subscale	68.38 (31.79)	67.78 (29.79)	0.8320
Extreme lighting subscale	75.21 (18.67)	73.05 (19.58)	0.3615
Mobility subscale	91.38 (13.54)	86.46 (16.19)	0.0141
Emotional distress subscale	90.49 (15.34)	89.56 (13.64)	0.6473
General dim lighting subscale	84.80 (16.86)	82.13 (16.38)	0.2178
Peripheral vision subscale	86.61 (19.39)	84.54 (18.46)	0.5178
Composite score	83.11 (16.13)	80.29 (16.59)	0.1045

Figure 2.

 

Cone-mediated (top) and rod-mediated (bottom) dark-adaptation functions and associated 95% confidence bands for the vitamin A group (blue) and placebo groups (red). Marginal functions for the vitamin A and placebo groups were obtained by applying models (e.g., cones: exponential; rods: bilinear) to grouped subject-specific data in a mixed statistical model. The cone-mediated function presents the cone time constant and the cone sensitivity, both of which were not significantly different. The rod-mediated function presents the rod slope and rod sensitivity. The rod slope in the vitamin A group was significantly steeper than in the placebo group (P = 0.0419; see Table 3 ).

Figure 3.

 

The association between the change in the rod slope parameter (day-30 visit minus baseline) and the change in the score on the mobility subscale of the LLQ (day-30 visit minus baseline); Pearson r = 0.24, P = 0.0141.