Early-Stage Alzheimer’s Disease Does Not Alter Pupil Responses to Colored Light Stimuli

Abstract

Background:

Pathologic changes in cerebral and retinal structures governing the pupillary light reflex occur in Alzheimer’s disease (AD). Analysis of pupillary responses originating from different retinal cells may allow for non-invasive detection of cerebral AD pathology.

Objective:

This study aimed to quantify the pupil light reflex using a portable chromatic pupillometer in patients with early stage AD and compare their responses to those of a healthy control group.

Methods:

Participants in this case-control pilot study were recruited from a well-characterized cohort of elderly people participating in a larger prospective study on early AD. Cognitive testing, volumetric brain imaging, and lumbar puncture were performed in all participants to define two groups: early AD, i.e., cognitively impaired subjects with biomarker-confirmed AD pathology, and control group of subjects with normal cognition and normal CSF biomarker profile. Pupil responses to red and blue light stimuli intended to activate cone photoreceptors and melanopsin ganglion cells were recorded under photopic conditions.

Results:

Sixteen patients with AD (mean age 77 years) and sixteen controls (mean age 71 years) were tested. Baseline pupil size was significantly smaller in AD patients. Pupillary contraction amplitude to all red and blue lights was also smaller in AD patients but did not reach statistical significance. The post-illumination pupillary response was the same between the two groups.

Conclusion:

Compared to healthy controls, we found only a smaller resting size of the pupil in patients with early AD. The pupillary dynamics to light stimulation remained relatively preserved.

Keywords

Alzheimer’s disease cerebrospinal fluid biomarkers dementia melanopsin ganglion cell pupil pupil light reflex

INTRODUCTION

Among patients with Alzheimer’s disease (AD), there is a high degree of variability in the presenting clinical manifestations, making diagnosis and prognosis challenging, in particular at early disease stages. This has spurred interest in identifying biologic indicators of the AD pathology, biomarkers which are both detectable at an early, predementia or preclinical stage of disease and correlated to severity of cerebral pathology. To date, several biomarkers in different categories have been identified, including neurochemical molecules in blood and in cerebrospinal fluid (CSF) and parameters derived from neuroimaging modalities [1, 2]. The biomarkers that are currently used in clinical practice are generally either invasive (CSF biomarkers) or costly with limited availability (amyloid imaging techniques).

The eye of patients with AD exhibits both disease-specific and non-specific pathologic changes [3 –8]. Amyloid plaques as well as amyloid-related neurodegeneration have been demonstrated in the retina of patients with AD [6]. In transgenic mice, retinal plaque formation precedes its appearance in the brain [6]. Other retinal abnormalities noted in patients with AD but not in age-matched controls include axonal thinning, retinal ganglion cell degeneration, reduced blood vessel density, and choroidal thinning [3 –8]. Additionally, the degree of retinal ganglion cell degeneration has been correlated with disease severity [9].

One subtype of retinal ganglion cell that is susceptible to AD-associated degeneration is represented by the melanopsin ganglion cells (mRGCs) [7]. mRGCs are retinal ganglion cells which express the photopigment melanopsin and hence are intrinsically photosensitive. Melanopsin relays light information for non-image forming physiologic functions such as mediation of pupil size and pupillary reactivity to light, secretion of melatonin hormone, stabilization of mood, and entrainment of circadian rhythm. In a recent histopathologic examination of mRGC density and morphology, the authors found that while mRGC number decreased as a function of increasing age between the 5^th and 7^th decade in control eyes, mRGC loss was more pronounced in AD eyes and did not correlate with age [7]. The mRGC loss occurred before degeneration of other types of RGCs and when axonal loss became evident, large fibers and superior retina were particularly affected. Additional alterations included reduced mRGC density, smaller soma, and thinner dendrites with patchy immunostaining for melanopsin.

In patients with AD, mRGC loss may explain disturbances in circadian rest-activity rhythms which contribute to increased night-time agitation and abnormal sleep patterns. Indeed, programmed dynamic lighting with higher blue light content during the day and lower illuminance without blue light in the evening has been shown to increase emotional expressions of pleasure and alertness and reduce non-physically aggressive behavior as well as improve nighttime sleep efficiency in patients with advanced dementia [10 –12].

Though postmortem studies demonstrated mRGCs loss as well as morphological changes in patients with AD, it is unknown when functional deficits of melanopsin and mRGCs might be detectable and whether those deficits are correlated to other biomarkers of AD. In this study, we aimed to assess the functional status of melanopsin from a quantified analysis of pupillary dynamics to a brief light flash in patients with early stage biomarker-confirmed AD and to compare their pupil responses against those from healthy controls. In addition, we examined associations between the pupil response mediated by melanopsin and other clinical parameters and biomarkers of AD.

METHODS

This pupillometric pilot study was a sub-study that was incorporated into an ongoing prospective study evaluating biomarkers in patients with early clinical stage AD. Outcome parameters from the main study which were used for correlational analysis in this sub-study were:

cognitive and functional assessment. This comprised the Mini-Mental State Examination (MMSE) to assess the global cognitive functioning and the Clinical Dementia Rating (CDR) scale which assesses six cognitive and functional domains: memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care [12]. The CDR sum of boxes (CDR-SB) was calculated. The CDR score was also determined. A CDR score 0 was considered as the absence of cognitive impairment; a CDR score of 0.5 is generally associated with a mild cognitive impairment (MCI), and a CDR score of one or more corresponds to dementia [13].

volumetric study of the brain with magnetic resonance imaging (MRI) using a 3D structural MPRAGE sequence [14]. Hippocampal volumes were estimated automatically using the MorphoBox prototype.

lumbar puncture for CSF biomarkers for AD. The amyloid-beta (Aβ_1-42), total tau, and tau phosphorylated at threonine 181 (ptau181) were measured in the CSF using ELISA (Fujirebio, Ghent, Belgium) [15]. A pathological CSF AD biomarker profile was defined as a ptau181/Aβ_1-42 ratio > 0.078 as previously described [15].

The pupillometric methodology was designed to comply with the pre-existing conditions of the main study protocol. For this reason and to avoid possible patient anxiety due to dark adaptation (20 min in a completely lightless room), we performed pupillometry under photopic conditions which allow assessment of pupil responses mediated primarily by cones and by melanopsin. Participants were recruited at the time of their usual study visit. Patients with diabetes, history of ocular trauma, age-related macular degeneration, glaucoma, or use of medicated eyedrops were excluded. Sixteen patients with AD confirmed by comprehensive neuropsychological testing and evaluation of activities of daily living [15] overall indicating MCI or mild dementia of AD type (CDR = 0.5 or CDR = 1), and an AD CSF biomarker profile comprised the patient group. Sixteen healthy older adults with normal neuropsychological testing (CDR = 0), and normal CSF biomarker profile comprised the control group.

The study was conducted according to the tenets of the Declaration of Helsinki and received authorization from the local ethical board committee for human research for the canton of Vaud of Switzerland (protocol 171/13). All study participants provided oral and written informed consent.

A best-corrected visual acuity was obtained in each eye. Pupils were tested using a handheld monocular pupillometer (Neurolight, IDMed, Marseilles) in a windowless room with a controlled illumination (110–150 lux). Each subject underwent two pupil tests, and each eye was tested independently with right eye always tested first. For each pupil test, the pupil was recorded continuously at 67 Hz starting 3 s before the first light stimulus and continuing until 20 s after the last light stimulus. The light stimuli were narrow bandwidth colored lights: blue (470 nm) and red (633 nm).

The first pupil test consisted of two blue light stimuli (1 s for each stimulus) having a different intensity (1.75 log cd/m² and 2.23 log cd/m²; 0 log = 1 cd/m²). There was a 15-s dark interval between the two blue light stimuli. Under photopic conditions, these two bright blue light stimuli predominantly activate S and M cone photoreceptors and the intrinsic melanopsin response in mRGCs. The second pupil test consisted of five red light stimuli (1 s for each stimulus) of increasing intensity. The first four stimulus intensities increased in half log-unit steps (from 0 to 1.5 log cd/m²; 0 log = 1 cd/m²) and the fifth red stimulus intensity was 2.6 log cd/m². The inter-stimulus dark interval was 3 s which was sufficient to allow the pupil to return to baseline size. The red light stimuli presented under photopic adaptation were considered to favor M and L cone photoreceptor activation, and not to favor melanopsin activation.

After each recording, pupil tracings were downloaded onto a PC. Each tracing was visually inspected by a trained assistant and artefacts induced by movement and/or blinking were removed by linear interpolation. To be considered a valid tracing for the first pupil test, the pupil response from both blue stimuli had to be well-recorded. To be considered a valid tracing for the second pupil test, pupil responses to at least 3 of the 5 red stimuli had to be well-recorded and then only the poorly recorded responses were removed from further analysis. In other words, if 3 or more pupil responses to red stimuli were considered poorly recorded, the entire tracing was considered invalid. Baseline pupil size was calculated as the averaged pupil size during 0.25 s before the first light stimulus. Tracings were normalized by converting absolute pupil size to pupil size relative to baseline pupil size: relative pupil size (RPS)=actual pupil size/baseline pupil size. The measure of the pupil light reflex was determined from the maximum contraction amplitude which was calculated as a % : [(baseline size – minimum RPS) / baseline size] x 100. The pupillary reaction after the light stimulus was terminated was also assessed. In previous publications, it has been determined that the post-illumination pupil response, or PIPR, is the dynamic phase that best reflects melanopsin contribution to the overall pupil light reflex [16 –20]. Thus for this study, we used the amount of persisting pupillary constriction at 6 s after light offset as the operational definition of PIPR [19, 21]. The PIPR was calculated as a % : [(baseline size – RPS at 6 s after light offset) / baseline size] x 100. PIPR was calculated for the two blue light stimuli (1.75 and 2.23 log cd/m²) and also for the brightest red light stimulus (2.6 log cd/m²). We used the PIPR from blue light stimulation as our pupil metric of melanopsin activity as melanopsin photopigment has a peak wavelength sensitivity around 480 nm which is in the blue color range. The PIPR from red light stimulation was considered a control and was expected to be essentially zero because melanopsin is relatively insensitive to red light.

While we might assume that the histopathologic retinal changes of AD occur symmetrically in the two eyes, this has not been established. Thus, we chose to analyze pupil responses separately for right eyes and for left eyes. Interocular differences were determined.

Statistical analysis

Data was entered and analyzed using SPSS for Windows, version 23 (IBM Corp., Armonk, NY). For quantitative variables, we calculated means and standard deviations. For categorical variables, we determined absolute and relative frequencies. To compare proportions between groups (AD and controls), we used Fisher’s exact test. To compare means between groups, we used the Mann-Whitney U test. As our variables did not have a Gaussian distribution, we used non-parametric Spearman’s correlations to examine the links between the pupil parameters and the MMSE scores, the CDR-SB scores, the hippocampal volume, and the CSF biomarkers for AD. For the correlational analysis in the patient group, the pupil parameters (from each eye independently and also averaged together) were correlated to the AD clinical and biomarker measures MMSE score, CDR-SB score, MRI hippocampal volume, and single CSF biomarkers.

To account for the two-eye correlation from the same participant, we used mixed linear regression models with random participant intercepts and eye-by-PIPR interaction term; covariates in the models included AD clinical and biomarker measures mentioned above.

A p value ≤ 0.05 was considered significant.

RESULTS

There were 16 patients with AD (10 females, 6 males) with mean age 77 years, range 64–87 years. Sixteen controls (13 females, 3 males) were significantly younger with a mean age 71 years, range 58–82 years. All AD clinical and biomarker measures (cognitive scores, MRI, and CSF) were significantly different between the two groups. Table 1 summarizes the clinical and biomarker profiles of the two groups. One AD participant was taking donepezil at the time of the study. One control was taking a benzodiazepine which was at a stable dose for a long time. No AD patient or control subject was using any opioid-like substance.

Table 1

Clinical and biomarker profiles of patients with Alzheimer’s disease (AD) and controls

	AD patients	Controls	p
Age, y (m±SD)	77.2±6.0	70.7±7.0	0.014
Gender, n(%) women	10(62.5)	13 (81.2)	0.433
MMSE score, (m±SD)	25.6±3.6	28.5±1.2	0.024
CDR-SB (m±SD)	0.0±0.1	3.1±2.8	<0.001
Aβ_1-42, CSF, pg/mL (m±SD)	568±113	994±223	<0.001
Total tau, CSF, pg/mL (m±SD)	754±365	219±84	<0.001
Phosphorylated tau, CSF, pg/mL (m±SD)	103±63	48±16	<0.001
Hippocampal volume, absolute, mL (m±SD)	4.9±0.9	6.2±0.6	0.001
Hippocampal volume, relative (m±SD)	0.36±0.05	0.46±0.04	<0.001

M, mean; SD, standard deviation; CDR-SB, Clinical Dementia Rating scale sum of boxes; CSF, cerebrospinal fluid; MMSE, Mini-Mental State Examination.

Two AD patients had previously undergone bilateral cataract surgery and one had had refractive surgery for myopia. Two controls had previous bilateral cataract surgery and one control had amblyopia in the right eye. The mean visual acuity for the AD patients was 0.65 for right eyes and 0.78 for left eye and for controls was 0.77 for right eyes and 0.83 for left eyes.

The raw pupil tracings from the right eye of three controls and two patients showed excessive blink artefacts during the second (red light) pupil test and these recordings were removed from analysis. The raw pupil tracings from the left eye of two controls and three patients were not well-recorded for either pupil test (blue light and red light) and thus removed from analysis.

The baseline pupil size for the controls was 4.15±0.62 mm in the right eye (n = 13) and 4.17±0.71 mm in the left eye (n = 14) and for the AD patients was 3.44±0.55 mm in the right eye (n = 14); and 3.37±0.64 mm in the left eye (n = 13). The interocular difference in pupil size was not significant for either group, AD patients or controls. However, the group difference in baseline pupil size (patients versus controls) was significant for right eyes (p = 0.005) and for left eyes (0.006).

Figure 1A shows the raw tracing (mean) of the pupil response for AD patients and for control subjects during the first pupil test. The contraction amplitude was calculated for each of the two blue light stimuli. The mean contraction amplitude to the first blue stimulus for patients was 30.94±8.54% for right eyes and 31.05±8.82% for left eyes; for controls was 32.14±6.28% for right eyes and 32.08±6.29% for left eyes. The mean contraction amplitude to the second blue stimulus for patients was 34.65±7.33% for right eyes and 34.13±7.32% for left eyes; for controls was 36.57±5.25% for right eyes and 37.00±5.45% for left eyes. There was no significant difference between eyes or between groups.

Fig.1

A) Raw tracing (mean) of the pupil response of AD patients and controls to two blue lights. The shaded blue bars indicate the timing and intensity of the two blue light stimuli. Solid lines indicate control subjects and dotted lines indicate patients with AD. Right eyes (RE) and left eyes (LE) are shown separately. The baseline pupil size of the AD patients is significantly smaller compared to controls. The pupil constricts immediately following each blue light stimulus. The contraction amplitude is smaller in AD patients compared to controls but does not reach statistical significance when compared as % contraction from baseline. The arrows indicate the PIPR (post-illumination pupil response) which is the remaining contraction amplitude measured 6 s after the light stimulus is terminated. B) Raw tracing (mean) of the pupil response of AD patients and controls to 5 red lights. The shaded red bars indicate the timing and intensity of the 5 red light stimuli. As in the A, the baseline pupil size of the AD patients is significantly smaller compared to controls and the contraction amplitude is smaller in AD patients compared to controls but does not reach statistical significance. The arrow indicates the PIPR measured following the last and brightest red light stimulus.

Figure 1B shows the raw tracing (mean) of the pupil response for AD patients and for controls during the second pupil test. The contraction amplitude was calculated for each of five red light stimuli. Though the contraction amplitudes appear smaller in the AD patients compared to the controls, the differences were not significant. Table 2 shows the values of the contraction amplitude to all red and blue stimuli for both groups.

Table 2

Contraction amplitude (%) to all red and blue stimuli for two groups

Stimulus		AD patients		Controls
		RE	LE	RE	LE
Blue	1.75 log cd/m²	30.94±8.54	31.05±8.82	32.14±6.28	32.06±6.29
	2.23 log cd/m²	34.65±7.33	34.13±7.32	36.57±5.25	37.00±5.45
Red	0 log cd/m²	6.77±5.32	6.85±5.81	9.36±6.34	9.56±5.61
	0.5 log cd/m²	12.12±5.38	9.96±5.92	15.10±5.97	14.17±6.98
	1 log cd/m²	20.76±6.75	20.48±7.06	23.27±5.04	23.56±6.42
	1.5 log cd/m²	27.57±6.10	26.40±5.43	27.98±3.81	29.96±4.31
	2.6 log cd/m²	32.87±5.03	33.46±6.58	34.65±3.82	36.28±4.96

AD, Alzheimer’s disease; RE, right eye; LE, left eye.

The mean PIPR to the red light, as expected, was zero for both groups. Except for one outlier, all individual values of PIPR to the red light were 6% or less. This confirms that the red light stimulus used in this study is not effective for activating melanopsin. The PIPR to the blue light stimuli 1.75 log cd/m² was slightly larger than the red light PIPR, indicating some degree of melanopsin activation, but the mean values for both groups was still small. For AD patients, it was 4.71±5.63% for right eyes and 5.11±5.88% for left eyes; for controls, it was 5.15±4.80% for right eyes and 4.98±4.83% for left eyes. The mean PIPR to the brighter blue light 2.23 log cd/m² showed the largest PIPR values; for AD patients it was 9.19±5.94% for right eyes and 9.11±6.37% for left eyes; for controls, 9.61±5.70% for right eyes and 9.77±7.29% for left eyes. Table 3 shows all PIPR values. There was a high variability of PIPR to both blue light stimuli. The mean PIPR for right eyes and for left eyes was nearly identical between the two groups for both blue lights as seen on the Figure 2 scatterplot.

Table 3

PIPR (%) to bright red and blue stimuli for two groups

Stimulus		AD patients		Controls
		RE	LE	RE	LE
Blue
	1.75 log cd/m²	4.71±5.63	5.11±5.88	5.15±4.80	4.98±4.83
	2.23 log cd/m²	9.19±5.94	9.11±6.37	9.61±5.70	9.77±7.29
Red
	1.5 log cd/m²	0.02±3.55	–0.77±5.32	–1.53±4.28	–1.89±4.13

PIPR, post-illumination pupil response; AD, Alzheimer’s disease; RE, right eye; LE, left eye.

Fig.2

Scatterplot of the PIPR values for AD patients and controls. Right eye (RE) and left eye (LE) values are shown separately. A negative value of PIPR indicates that the pupil had recovered beyond baseline size at 6 s after termination of the light stimulus and has the same implication as PIPR = zero. The mean PIPR to the red light stimulus (2.6 log cd/m²) is essentially zero for both eyes for both groups. This is an expected value as red light does not effectively stimulation melanopsin and the zero value serves as control value. The PIPR to the blue lights is greater than zero as blue light is a more effective stimulus of melanopsin.

Within the AD group, there was no association between any of the pupil parameters with any of the AD CSF biomarkers (Table 4). There was, however, a trend between the PIPR from the brighter (2.23 log cd/m²) blue light and the absolute hippocampal volume (p = 0.08). Within the AD group, mixed linear models did not find any significant associations between PIPR and any of the AD CSF biomarkers, cognitive test scores (MMSE and CDR-SB), or the hippocampal volume.

Table 4

Non-parametric correlations between blue light PIPR and biomarkers of Alzheimer’s disease

		CDR-SB	MMSE	CSF Amyloid-β	CSF hTau	CSF ptau181	Absolute hippocampal volume	Relative hippocampal volume
PIPR Blue 1.75 log cd/m²	Rho	–0.169	–0.172	–0.056	–0.056	0.038	0.441	0.462
	p	0.532	0.524	0.837	0.837	0.888	0.152	0.131
PIPR Blue 2.23 log cd/m²	Rho	–0.059	–0.019	0.088	0.150	0.065	0.524	0.385
	p	0.827	0.944	0.745	0.579	0.812	0.080	0.217

CDR-SB, Clinical Dementia Rating scale sum of boxes; CSF, cerebrospinal fluid; MMSE, Mini-Mental State Examination.

DISCUSSION

This pilot study examined the state of the pupil at rest and its response to colored light stimuli in patients with early AD. While the baseline pupil size in roomlight (photopic adaptation) was significantly smaller in the patients with AD, we found no significant differences in the pupil response (contraction amplitude and post-illumination response) to various intensities of colored light stimulation between patients with early AD and normal controls. Furthermore, the pupil response was not significantly correlated with MMSE score, MRI hippocampal volume, or any CSF biomarker measures of amyloid, neuronal injury, and tau pathology.

We were particularly interested to examine the pupil response mediated by melanopsin as recent histopathological studies have demonstrated early changes and loss of retinal ganglion cells which express melanopsin (mRGCs) in patients with AD [7]. We wondered if cerebral pathology that results in cognitive dysfunction is paralleled by clinical evidence of retinal dysfunction, specifically reduced melanopsin activity. While most clinical tests of melanopsin activity are cumbersome, pupillometry is a rapid, non-invasive, and objective test which, for patients with AD, makes the pupil an attractive potential biomarker of the disease. Besides AD, primary ocular pathology such as glaucoma has been shown to cause loss of mRGCs and in patients with glaucoma, reduction of mRGC function has been demonstrated by hormonal, circadian, and pupillary markers [21 –25]. Although the rapid and transient pupillary constriction that immediately follows a light stimulus is a well-established marker of cone photoreceptor activity, the melanopsin activation of mRGCs is better assessed in the pupillary dynamics after termination of the light stimulus, i.e., the post-illumination pupil response or PIPR. Following a high intensity short pulse of blue light, the pupil does not return promptly to baseline size when melanopsin activated as it sustains pupillary constriction, even after light termination, and acts against the usual rapid re-dilation movement of the pupil in darkness [16 –20]. Several studies have shown that the spectral sensitivity of the PIPR matches that of melanopsin and persistently small pupils following blue light is a generally considered a sign of melanopsin influence on the pupillary post-light dynamics [16 –20].

How does one measure pupillary post-light dynamics? As stated above, the pupil tends to stay constricted when melanopsin is activated and thus pupillary re-dilation after light termination is slowed. Various parameters of the post-light response have been measured and compared, and one reliable commonly used metric of PIPR in clinical studies is the relative pupil size, i.e., the contraction amplitude, at 6 s after light termination [19]. Using this metric, we noted that the PIPR to blue light showed a wide range of values in both AD patients and control subjects (Fig. 2) but the mean PIPR was quite similar between the two groups. Our findings echo those from a recently published study by Oh et al. which used chromatic pupillometry with similar light stimulus conditions (blue light 450 nm, intensity 2.3 log cd/m², 1 s duration) and measured the same parameter of PIPR (pupil size between 6 and 8 s post-light) in 10 patients with pre-clinical AD and 10 age-matched controls [26]. The authors did not find any difference of the blue light PIPR between patients and controls and, like our study, their results suggest that intrinsic mRGC activation is not reduced in the early stages of AD. In addition, the smaller pupil size in patients with AD in our study suggests that steady-state function of mRGCs, which largely dictates pupil size, is also not compromised.

Then how might we explain the smaller pupils in patients with early AD as, theoretically, disease-specific cholinergic neuronal dysfunction at the level of the Edinger Westphal nucleus would, like loss of mRGCs, lead to a larger baseline pupil size. One possible explanation for the smaller pupils in AD patients in this study is older age. It is known that baseline pupil size progressively decreases with increasing age after 20 years [27]. This age effect is less pronounced, however, on pupils measured under light-adapted conditions, as in our study. From the study by Winn et al. the expected difference in photopic pupil size at age 77 years (mean age of AD group) versus at age 71 years (mean age of control group) is roughly 0.25 mm [28]. The difference in pupil size between our AD patients and control subjects exceeded 0.25 mm so we must consider other factors. Another potential explanation for the smaller pupils in AD patients is a medication effect. However, only one AD patient and one control subject were taking a medication that could potentially influence baseline pupil size so this explanation does not explain the smaller pupils of AD patients in this study. Finally, there is an association between greater cognitive ability and larger pupillary dilation [29, 30] and so we may conjecture that loss of cognitive faculties, increased mental fatigability, and other changes in supranuclear and autonomic pathways act to decrease baseline pupil size in AD patients, as previously proposed by Prettyman et al. [31].

In a correlational analysis of all pupil parameters to other biomarkers of AD, a trend between absolute hippocampal volume and the PIPR from the brighter blue (2.23 log cd/m²) light was noted. This suggested relationship between the PIPR and hippocampal volume is attractive as it suggests that, while the mRGC degeneration in the retina is not detectable in the absolute measure of PIPR in early stages of disease, perhaps the PIPR is an indirect marker of hippocampal atrophy.

We also examined the rapid pupillary constriction that immediately follows light stimulation. This rapid and transient pupillary constriction, the well-known pupil light reflex, is a motor function mediated by the cholinergic parasympathetic pathway to the eye which has origin at the Edinger-Westphal nucleus of the rostral midbrain. In AD, structural and functional failure of the central cholinergic system is an important pathophysiologic contributor to the clinical manifestations of the disease, in particular the memory and cognitive deficits [32]. Histopathologic studies of AD patients have demonstrated altered neuronal morphology, decreased neuronal count, and deposition of amyloid plaques and neurofibrillary tangles in the Edinger-Westphal nucleus, even in patients with preclinical disease [33 –35]. However, whether or not AD is associated with loss of the rapid pupillary constriction to a light stimulus remains debated. Some clinical studies have reported alterations in the pupil light reflex with AD [36 –38], whereas others studies have not [39 –41]. The variability in the results may be, in part, due to a lack of standardized test protocols and thus the different light stimulus conditions and recording devices render difficult an exacting interpretation of the data and a comparative analysis between studies. In our study, the pupillary constriction (determined from the maximal contraction amplitude from baseline) to a range of red and blue light stimuli was smaller in AD patients and this was readily observable in the raw pupil tracings (Fig. 1). However, the difference was small and not statistically significant. Thus a cholinergic deficit measured from pupillary constriction is not reduced in the early stages of AD

Besides the inclusion of subjects with clinically-diagnosed and biomarkers-confirmed AD, a major distinction of our study to other clinical pupillometric studies in AD is the use of a strictly defined normal control group. All controls were community-dwelling volunteers and underwent the same neurocognitive, neuroradiologic, and CSF testing as patients and only those with normal results in all clinical aspects and the absence of cerebral AD pathology as indicated by CSF biomarkers were retained in the study as controls. This strictly defined control group defers any criticism of subclinical AD patients being inadvertently included in the control group and thus would strengthen our results that neither pupillary constriction to light or the post-illumination pupillary response can be used to differentiate between patients with early AD from controls. In addition, we selected only AD patients with early cognitive impairment defined by CDR = 0.5 or CDR = 1, i.e., clinical AD. We intentionally chose this subgroup because development of amyloid pathology and CSF abnormalities related to AD starts many years before the symptoms onset. Changes in mRGCs are, however, likely related to neurodegeneration which is more closely related to the beginning cognitive decline. Accordingly, changes in mRGCs may not be present in all subjects with preclinical AD, but only in those subjects that are relatively close to symptom onset [42].

However, we recognize certain limitations of our study. This was a pilot study with limited number of subjects (n = 32) and may have been underpowered to detect differences between the two groups. Also, our patients had early clinical stage disease and without a structural marker of mRGCs or Edinger-Westphal neurons in vivo, it cannot be known that our patients actually had AD pathology in these pupil-related structures at this time. Finally, it is possible that the pupil is not the most sensitive marker of early alterations in mRGC activity due to AD pathology. Other markers of melanopsin function and mRGC activity, for example the suppression of nocturnal melatonin with light, should be measured simultaneously with the pupil response in order to compare the sensitivity of various biomarkers of mRGC activity in patients at various stages of AD.

In conclusion, this study has shown that patients with early stage AD have a significantly smaller baseline pupil size compared to controls but the pupillary response to light is not significantly altered. Our findings do not support the hypothesis that either the cholinergic deficit or the mRGC loss that occurs in AD is detectable using pupillometry in patient with early stage disease. Longitudinal studies of the pupil light reflex in AD patients and/or pathologic examination of pupil-mediating brain and retinal structures are needed to more definitively define if relationship exists.

Footnotes

ACKNOWLEDGMENTS

The authors would like to thank Myriam Ladaique, Aikaterini Oikonomidi, Domile Tautvydaite, and Barbara Moulet for technical assistance with pupillometric testing and Bénédicte Maréchal and Jonas Richiardi for volumetric analysis of MRI images. This study was supported by the Swiss National Science Foundation (grant: 320030L_141179) and Synapsis Foundation – Alzheimer Research Switzerland (2017-PI01) to JP.

Authors’ disclosures available online ().

References

Blennow

, Zetterberg

(2018) The past and the future of Alzheimer’s disease fluid biomarkers. J Alzheimers Dis 62, 1125–1140.

Laforce

Jr , Soucy

, Sellami

, Dallaire-Théroux

, Brunet

, Bergeron

, Miller

, Ossenkoppele

(2018) Molecular imaging in dementia: Past, present, and future. Alzheimers Dement 11, 1522–1552.

Hinton

, Sadun

, Blanks

, Miller

(1986) Optic nerve degeneration in Alzheimer’s disease. N Engl J Med 315, 485–487.

Blanks

, Torigoe

, Hinton

, Blanks

(1996) Retinal pathology in Alzheimer’s disease. I. Ganglion cell loss in foveal/parafoveal retina. Neurobiol Aging 17, 377–384.

Blanks

, Schmidt

, Torigoe

, Porrello

, Hinton

, Blanks

(1996) Retinal pathology in Alzheimer’s disease.II. Regional neuron loss and glial changes in GCL. Neurobiol Aging 17, 385–395.

Koronyo-Hamaoui

, Koronyo

, Ljubimov

, Miller

, Ko

, Black

, Schwartz

, Farkas

(2011) Identification of amyloid plaques in retinas from Alzheimer’s patients and noninvasive in vivo optical imaging of retinal plaques in a mouse model. Neuroimage 54(Suppl 1), S204–S217.

La Morgia

, Ross-Cisneros

, Koronyo

, Hannibal

, Gallassi

, Cantalupo

, Sambati

, Pan

, Tozer

, Barboni P Provini

, Avanzini

, Carbonelli

, Pelosi

, Chui

, Liguori

, Baruzzi

, Koronyo-Hamaoui

, Sadun

, Carelli

(2016) Melanopsin retinal ganglion cell loss in Alzheimer’s disease. Ann Neurol 79, 90–109.

Lim

, Li

, He

, Vingrys

, Wong

, Currier

, Mullen

, Bui

, Nguyen

(2016) The eye as biomarker for Alzheimer’s disease. Front Neurosci 10, 536.

Blanks

, Hinton

, Sadun

, Miller

(1989) Retinal ganglion cell degeneration in Alzheimer’s disease. Brain Res 501, 364–372.

10.

Wahnschaffe

, Nowozin

, Haedel

, Rath

, Appelhof

, Münch

, Kunz

(2017) Implementation of dynamic lighting in a nursing home: Impact on agitation but not on rest-activity patterns. Curr Alzheimers Res 14, 1076–1083.

11.

Münch

, Schmieder

, Bieler

, Goldbach

, Fuhrmann

, Zumstein

, Vonmoos

, Scartezzini

, Wirz-Justice

, Cajochen

(2017) Bright Light Delights: Effects of daily light exposure on emotions, rest-activity cycles, sleep and melatonin secretion in severely demented patients. Curr Alzheimer Res 14, 1063–1075.

12.

Figueiro

(2017) Light, sleep and circadian rhythms in older adults with Alzheimer’s disease and related dementias. Neurodegener Dis Manag 7, 119–145.

13.

Morris

(1993) The clinical dementia rating (CDR): Current version and scoring rules. Neurology 43, 2412–2414.

14.

Brant-Zawadzki

, Gillan

, Nitz

(1992) MP RAGE: A three-dimensional, T1-weighted, gradient-echo sequence–initial experience in the brain. Radiology 182, 769–775.

15.

Oikonomidi

, Tautvydaitė

, Gholamrezaee

, Henry

, Bacher

, Popp

(2017) Macrophage migration inhibitory factor is associated with biomarkers of Alzheimer’s disease pathology and predicts cognitive decline in mild cognitive impairment and mild dementia. J Alzheimers Dis 60, 273–281.

16.

Gamlin

, McDougal

, Pokorny

, Smith

, Yau

, Dacey

(2007) Human and macaque pupil responses driven by melanopsin-containing retinal ganglion cells. Vision Res 47, 946–954.

17.

McDougal

, Gamlin

(2010) The influence of intrinsically-photosensitive retinal ganglion cells on the spectral sensitivity and response dynamics of the human pupillary light reflex. Vision Res 50, 72–87.

18.

Kankipati

, Girkin

, Gamlin

(2010) Post-illumination pupil response in subjects without ocular disease. Invest Ophthalmol Vis Sci 51, 2764–2769.

19.

Park

, Moura

, Raza

, Rhee

, Kardon

, Hood

(2011) Toward a clinical protocol for assessing rod, cone, and melanopsin contributions to the human pupil response. Invest Ophthalmol Vis Sci 52, 6624–6635.

20.

Adhikari

, Zele

, Feigl

(2015) The post-illumination pupil response (PIPR). Invest Ophthalmol Vis Sci 56, 3838–3849.

21.

Münch

, Leon

, Collomb

, Kawasaki

(2015) Comparison of acute non-visual bright light responses in patients with optic nerve disease, glaucoma and healthy controls. Sci Rep 5, 15185.

22.

Drouyer

, Dkhissi-Benyahya

, Chiquet

, WoldeMussie

, Ruiz

, Wheeler

, Denis

, Cooper

(2008) Glaucoma alters the circadian timing system. PLoS One 3, e3931.

23.

Feigl

, Mattes

, Thomas

, Zele

(2011) Intrinsically photosensitive (melanopsin) retinal ganglion cell function in glaucoma. Invest Ophthalmol Vis Sci 52, 4362–4367.

24.

Gracitelli

, Duque-Chica

, Roizenblatt

, Moura

, Nagy

, Ragot de Melo

, Borba

, Teixeira

, Tufik

, Ventura

, Paranhos

Jr (2015) Intrinsically photosensitive retinal ganglion cell activity is associated with decreased sleep quality in patients with glaucoma. Ophthalmology 122, 1139–1148.

25.

Obara

, Hannibal

, Heegaard

, Fahrenkrug

(2016) Loss of melanopsinexpressing retinal ganglion cells in severely staged glaucoma patients. Invest Ophthalmol Vis Sci 57, 4661–4667.

26.

, Amore

, Sultan

, Asanad

, Park

, Romagnoli

, La Morgia

, Karanjia

, Harrington

, Sadun

(2019) Pupillometry evaluation of melanopsin retinal ganglion cell function and sleep-wake activity in pre-symptomatic Alzheimer’s disease. PLoS One 14, e0226197.

27.

Lowenfeld

(1979) Pupillary changes related to age. In Topics in Neuro-Ophthalmology, Thompson

, ed. Williams and Wilkins, Baltimore, pp. 124–150.

28.

Winn

, Whitaker

, Elliott

, Phillips

(1994) Factors affecting light-adapted pupil size in normal human subjects. Invest Ophthalmol Vis Sci 35, 1132–1137.

29.

Van Der Meer

, Beyer

, Horn

, Foth

, Bornemann

, Ries

, Wartenburger

(2010) Resource allocation and fluid intelligence: Insights from pupillometry. Psychophysiology 47, 158–169.

30.

Murphy

, Robertson

, Balsters

, O’Connell

(2011) Pupillometry and P3 index the locus coeruleus-noradrenergic arousal function in humans. Psychophysiology 48, 1532–1543.

31.

Prettyman

, Bitsios

, Szabadi

(1997) Altered pupillary size and darkness and light reflexes in Alzheimer’s disease. J Neurol Neurosurg Psychiatry 62, 665–668.

32.

Hampel

, Mesulam

, Cuello

, Farlow

, Giacobini

, Grossberg

, Khachaturian

, Vergallo

, Cavedo

, Snyder

, Khachaturian

(2018) The cholinergic system in the pathophysiology and treatment of Alzheimer’s disease. Brain 141, 1917–1933.

33.

Scinto

, Frosch

, Wu

, Daffner

, Gedi

, Geula

(2001) Selective cell loss in Edinger-Westphal in asymptomatic elders and Alzheimer’s patients. Neurobiol Aging 225, 729–736.

34.

Mavroudis

, Manani

, Petrides

, Petsoglou

, Njau

, Costa

, Baloyannis

(2014) Dendritic and spinal alterations of neurons from Edinger-Westphal nucleus in Alzheimer’s disease. Folia Neuropathol 52, 197–204.

35.

Prettyman

, Bitsios

, Szabadi

(1997) Altered pupillary size and darkness and light reflexes in Alzheimer’s disease. J Neurol Neurosurg Psych 62, 665–668.

36.

Fotiou

, Brozou

, Haidich

A-B

, Tsiptsios

, Nakou

, Kabitsi

, Giantselidis

, Fotiou

(2007) Pupil reaction to light in Alzheimer’s disease: Evaluation of pupil size changes and mobility. Aging Clin Exp Res 19, 364–371.

37.

Fotiou

, Stergiou

, Tsiptsios

, Lithari

, Nakou

, Karlovasitou

(2009) Cholinergic deficiency in Alzheimer’s and Parkinson’s disease: Evaluation with pupillometry. Int J Psychophysiol 73, 143–149.

38.

Frost

, Robinson

, Rowe

, Ames

, Masters

, Taddei

, Rainey-Smith

, Martins

, Kanagasingam

(2017) Evaluation of cholinergic deficiency in preclinical Alzheimer’s disease using pupillometry. J Ophthalmol 2017, 7935406.

39.

Bittner

, Wieseler

, Wilhelm

, Riepe

, Müller

(2014) Repetitive pupil light reflex: Potential marker in Alzheimer’s disease? J Alzheimers Dis 42, 1469–1477.

40.

Ferrario

, Molaschi

, Villa

, Varetto

, Bogetto

, Nuzzi

(1998) Is videopupillography useful in the diagnosis of Alzheimer’s disease? Neurology 50, 642–644.

41.

Van Stavern

, Bei

, Shui

Y-B

, Huecker

, Gordon

(2019) Pupillary light reaction in preclinical Alzheimer’s disease subjects compared with normal ageing controls. Br J Ophthalmol 103, 971–975.

42.

Mormino

, Papp

(2018) Amyloid accumulation and cognitive decline in clinically normal older individuals: Implications for aging and early Alzheimer’s disease. J Alzheimers Dis 64, S633–S646.