Abstract
Background
Neurodegenerative diseases are public health challenges in aging populations. Early identification of people at risk for neurodegeneration might improve targeted treatment. Noninvasive, inexpensive screening tools are lacking but are of great potential. Optical coherence tomography (OCT) measures the thickness of nerve cell layers in the retina, which is an anatomical extension of the brain and might be indicative of common underlying neurodegeneration. We aimed to determine the association of macular ganglion cell-inner plexiform layer (mGCIPL) thickness with cognitive and sensorineural function in midlife.
Method
This cross-sectional study included 1,880 Beaver Dam Offspring Study participants (aged 27–93 years, mean 58) who participated in the 10-year follow-up examination. We assessed cognitive function and impairment, hearing sensitivity thresholds and impairment, central auditory processing, visual impairment, and olfactory impairment. We measured mGCIPL using the Cirrus 5000 HD-OCT Macular Cube Scan. Multivariable linear and logistic regression models adjusted for potential confounders were used to determine associations between mGCIPL thickness and cognitive and sensorineural functions, as well as for comparing participants with a thin mGCIPL (1 SD below average) to the remainder in those functions.
Results
Thinner mGCIPL was associated with worse cognitive function, worse central auditory function, and visual impairment. We found an association of mGCIPL thickness with hearing sensitivity in women only and no association with impairment in hearing, olfaction, and cognition. Results on the thin group comparisons were consistent.
Conclusions
mGCIPL thickness is associated with cognitive and sensorineural function and has the potential as a marker for neurodegeneration in middle-aged adults.
Keywords: Retinal thickness, Biomarker, Cognition, Vision, Senses
Neurodegenerative diseases are public health challenges given their expected increase in prevalence rates in aging populations (1–3). Since neurodegenerative diseases have a long preclinical phase with an onset of various neurodegenerative processes in midlife (4), early detection of people at risk has great potential for future targeted treatment and prevention strategies. Current biomarkers are invasive and expensive and we lack noninvasive, less expensive biomarkers that could reinforce the future development of screening tools.
The retina is an anatomical extension of the brain and physiologically similar (5). Optical coherence tomography (OCT) utilizes automated segmentation techniques to measure the thickness of the retinal nerve cell layers. The macular ganglion cell-inner plexiform layer (mGCIPL) represents a combination of cell bodies and dendrites in the macula (6) and has been investigated as a marker of neurodegenerative diseases. Layers are thinner in patients with Alzheimer’s Disease (7–9), Parkinson’s Disease (10) and Multiple Sclerosis (MS) (11) compared to healthy individuals and in older compared to younger adults (12,13). Thinner mGCIPL was also associated with smaller gray matter volume (14,15) and recent research found that mGCIPL thickness is associated cross-sectionally with cognitive function in healthy older adults (7,16). Despite the suggested early onset of neurodegenerative processes (4), research on midlife is scarce. MGCIPL could be a useful marker of early, preclinical decline within the central nervous system and its functions.
Neurodegeneration within the central nervous system is not limited to one system but affects various neuronal functions and can also be found in sensory brain regions (17). Declines in hearing, vision, and olfaction share many risk factors with cognitive decline (18), increase with increasing age and often precede cognitive symptoms in Alzheimer’s disease (17,19). They may, therefore, be early signs for neurodegenerative processes (17,19).
The aim of this study was to determine whether mGCIPL thickness is associated with cognitive and sensorineural function in midlife.
Material and Methods
Study Population
This cross-sectional study is based on participants of the 10-year follow-up of the Beaver Dam Offspring Study (BOSS), a prospective cohort study of aging. The adult offspring of the population-based Epidemiology of Hearing Loss Study (EHLS) participants were eligible for the baseline BOSS examination (conducted 2005–2008) (20) and have been followed every 5 years. OCT was obtained for the first time at the 10-year follow-up in 2015–2017 (N = 1,964). Participants were eligible for this study if they participated in an OCT scan in the BOSS 10-year follow-up examination. The study was approved by the University of Wisconsin Health Sciences Institutional Review Board with written informed consent from all participants before each examination.
Measurements
Optical coherence tomography
All assessments of the 10-year follow-up were conducted on the same day. We conducted an OCT Macular Cube 512 × 128 scan, centered on the fovea, using one Cirrus 5000 HD-OCT device (21) to scan both eyes. This scan generates data from a 6 mm square cube acquired from 128 horizontal line scans and 512 A-scans. Testing took place in a dimly lit room and participants were not dilated. The “FastTrac Retinal Tracking System” was enabled on our OCT. Retinal layer thicknesses, including the mGCIPL, were extracted using instrument-specific automatic segmentation algorithms with the Cirrus Software version 7.5.0.56. Each eye’s average mGCIPL thickness was calculated using six sectors in an elliptical annulus around the fovea. Six technicians were trained and certified in standard protocols to conduct the OCT scan with quarterly quality control review of a 10% random selection of scans and equipment quality control was conducted weekly by two quality control graders. Technicians reviewed the scans immediately for low signal strength (<6/10) and artifacts, such as saccades or banding. In cases where scans were not of sufficient quality based on these criteria, repeat scans were obtained. Final scans with signal strength of less than 6, poor alignment, or artifacts affecting measurement were excluded from the analyses. We used the average mGCIPL thickness (in the unit 10 μm) from the eye with the thinner mGCIPL (and in case of missing one eye, the available eye). Thin mGCIPL was defined as ≤70.3 μm, which was 1 SD below the mean for right eyes, which was the eye with more complete data in the cohort. We report our quantitative OCT data in line with the APOSTEL recommendations (22).
Cognitive measures
Participants completed a neurocognitive test battery at the 10-year follow-up that measured the cognitive domains of attention, speed and executive function (Trail-Making Tests A and B [TMTA, TMTB], Digit Symbol Substitution Test [DSST]), memory (modified Rey Auditory Verbal Learning Test [AVLT]), and language (Verbal Fluency Test [VFT]) (19). The Mini-Mental-State Examination was completed by participants aged >50 years (23). In TMTA, consecutive numbers are to be connected, and in TMTB, alternating consecutive numbers and letters are to be connected. The main outcome is completion time in seconds. Longer durations indicate poorer performance. Inability to complete the test in the allotted 5 minutes resulted in a score of 301 seconds (24). In DSST, participants had to convert numbers to symbols based on a key. The number of correct numbers converted in 90 seconds was used as outcome score. For the AVLT, subjects are asked to recall as many words as they can from a list of 15 verbally presented words. Three trials with the same 15-word list were administered followed by a new distractor list of 15 words. Immediately after recalling words from the distractor list, the participant was asked to recall as many words as they could from the first word list. The number of words correctly recalled from the first list in the final trial was used as a measure of memory function. In the VFT, the task is to provide as many words as possible that start with the letters F, A, and S, within 60 seconds for each letter. The total number of words provided for F, A, and S were summed as test outcome (25).
Memory concerns were ascertained by two questions. Cognitive impairment was defined as memory concerns and impairment in one or more cognitive domains (executive function, memory, verbal function), self-report or surrogate report of Alzheimer’s disease or dementia, or a Mini-Mental-State Examination score below 24 (19,26).
Sensorineural measures
Multiple sensorineural measures were obtained at the 10-year follow-up. Pure-tone air and bone conduction audiometry were conducted in either a sound-treated booth or in very few instances with insert earphones and followed American National Standards Institute standards for equipment (27,28) and American Speech-Language-Hearing Association guidelines (29). Pure-tone air conduction thresholds were obtained at 0.5, 1, 2, 3, 4, 6, and 8 kHz, and bone conduction thresholds at 0.5 and 2 kHz for both ears using clinical audiometers with TDH-50P earphones and ER-3A insert earphones (in cases of probable ear-canal collapse). When necessary, masking was done. The pure-tone average (PTA) at 0.5, 1, 2, and 4 kHz in decibel hearing level (dB HL) in the worse ear (and in case of missing one ear the available ear) was used as a measure of hearing function (20). Hearing impairment was defined as PTA > 25 dB HL.
To assess central auditory function, the free recall Dichotic Digits Test (DDT) and Word Recognition in Competing Message (WRCM) with the Northwestern University Auditory Test Number 6 were administered. The DDT was administered with 25 sets of triple-digit pairs. In this test, three digits (single-syllable numbers 1 through 10, excluding 7) are presented to each ear simultaneously, and the participant has to repeat as many of the 6 digits as possible. The presentation level was set at 70 dB HL. The sum of correctly repeated digits in all trials served as the main outcome and was converted to percent correct for analysis purposes (30). In the WRCM, 25 words were presented by a single female speaker to the better ear at 36 dB HL above the individual’s threshold at 2 kHz. If thresholds at 2 kHz were equal, the right ear was tested. The competing message (single male speaker) was added at a level 8 dB HL below the female speaker’s level in that same ear (31,32). The percentage of correctly repeated words was the main outcome measure.
The San Diego Odor Identification Test was used to evaluate olfaction. Eight common odorants were presented randomly with an interstimulus interval of 45 seconds. A picture board including the eight odorants plus 12 distractor items was used to aid identification. Participants were allowed to respond verbally or point to the item on the picture board. The number of correctly identified odors after two trials served as olfactory function outcome, with impairment defined as less than 6 out of 8 correctly identified odorants (33,34).
Vision was measured by Contrast Sensitivity (CS) using the Pelli-Robson letter chart for each eye separately (35). Participants were instructed to read as many letters as possible until two letters in a triplet were missed. We used the last triplet in which a participant correctly identified at least 2 of the 3 letters to assign a log CS score. Vision impairment was defined as a log contrast score less than 1.55 in the better eye (and in case of missing one eye, the available eye) (36). We asked participants for their history of age-related macular degeneration, glaucoma, diabetic retinopathy, and current cataract.
Other variables
We assessed participants’ age, sex, college education, waist circumference, smoking status, alcohol consumption, diabetes (history of diabetes diagnosis and/or glycated hemoglobin ≥ 6.5%), hypertension (systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg and/or use of antihypertensive drugs) and regular exercise (at least once a week, long enough to work up a sweat). Lead exposure (quintile 5 vs others) was assessed from whole blood samples (36) obtained at baseline and interleukin-6 and high-sensitivity C-reactive Protein from blood serum samples (36) obtained at 5-year follow-up.
Statistical Analyses
Data preparation and confounding
Statistical analyses were conducted using SAS software v.9.4 (SAS Institute, Inc, Cary, NC). The composite cognitive function score (in SD) was created using a principal component analysis with the factor procedure on neurocognitive test data (TMTA, TMTB, DSST, AVLT, VFT) (19).
A combined sensory-cognitive function (brain aging) score to represent the integrity of the entire sensorineural system was calculated by a composite of sensorineural (PTA, CS, San Diego Odor Identification Test) and neurocognitive (TMTA, TMTB, DSST, AVLT, VFT) function test data, which was created using principal component analysis with the factor procedure (18).
To evaluate confounding, those potential confounder variables that were significantly (p < .05) associated with thin and/or continuous mGCIPL and with cognitive function in age-sex-adjusted models were included in the models. Resulting covariates were waist circumference, alcohol consumption, diabetes, interleukin-6, high-sensitivity C-reactive Protein, regular exercise, and lead exposure. We additionally included hypertension, education, and smoking in the models given the strong literature suggesting their relevance. Missing values for confounder and outcome variables varied from 0% to 8% with highest missingness in laboratory markers. Participants with missing values were omitted from respective analyses.
Regression models
We used multivariable linear regression and logistic regression models to assess the strength of associations of the determinants mGCIPL average thickness and thin mGCIPL with the cognitive and sensorineural function outcomes. There was insufficient variation in olfactory test performance and CS to support continuous analyses in this middle-aged cohort. Models were first adjusted for age and sex only and then repeated adjusting for further confounders. Secondary analyses with the combined measure of brain aging score and with individual cognitive tests are shown in the Supplementary Tables. We tested for sex interactions in fully adjusted models and report sex-stratified results, where significant interactions were found. We repeated the analyses, additionally adjusting for OCT signal strength.
Sensitivity analyses
Age-sex-adjusted models were repeated in the smaller samples that had complete confounder variable data to investigate if effect sizes changed due to sample differences. We also repeated the models excluding participants with age-related macular degeneration, glaucoma, cataract, or diabetic retinopathy (N = 493) to confirm associations in healthy eyes.
Results
The analysis focused on 1,880 participants, which had a mean age of 58 (SD = 9, range 27–93) years and 54% were women (Table 1). We excluded participants with missing OCT data (N = 74) and insufficient scan quality (N = 10). There were not significant differences between the means in age, PCA cognition score, PTA, WRCM, DDT, probability of being female or having impaired cognition in the excluded (N = 84) compared to the analytic sample. There were small differences in the odds of having olfaction and CS impairment, with fewer cases in the analytic sample (data not shown).
Table 1.
Characteristics of the Analytic Sample of the Beaver Dam Offspring Study (N = 1,880)
| Age, y, M (SD) | 58.5 (9.3) |
| Sex, n (%) | |
| Women | 1,022 (54.4) |
| Men | 858 (45.6) |
| Education, n (%) | |
| <16 y | 1,203 (64.1) |
| 16 y and more | 673 (35.9) |
| Smoking, n (%) | |
| Never | 1,058 (56.7) |
| Former | 604 (32.4) |
| Current | 203 (10.9) |
| Waist circumference, cm, M (SD) | 103.2 (16.3) |
| Alcohol consumption, g/wk, n (%) | |
| None | 245 (13.3) |
| >0–14 | 699 (37.9) |
| 15–74 | 454 (24.6) |
| 75–140 | 248 (13.4) |
| >140 | 199 (10.8) |
| Hypertension, n (%) | 932 (49.6) |
| Diabetes, n (%) | 240 (12.8) |
| Exercise at least once a week, n (%) | 1,255 (67.3) |
| hsCRP, ln mg/L, M (SD)a | 0.4 (1.1) |
| IL-6, ln pg/mL, M (SD)a | 0.4 (0.8) |
| Lead, highest quintile, n (%)b | 322 (19.4) |
| mGCIPL, µm, M (SD) | 77.1 (9.0) |
| Thin mGCIPL, n (%) | 329 (17.5) |
| Cognitive impairment prevalence, n (%) | 78 (4.2) |
| Hearing function, PTA, dB HL, M (SD) | 20.9 (15.1) |
| DDT score, % correct, M (SD) | 78.7 (11.4) |
| WRCM score, % correct, M (SD) | 54.3 (17.1) |
| Hearing impairment prevalence, n (%) | 512 (27.3) |
| Visual impairment prevalence, n (%) | 122 (6.5) |
| Olfactory impairment prevalence, n (%) | 99 (5.3) |
Notes: db HL = decibel hearing level; DDT = Dichotic Digits Test; hsCRP = high-sensitivity C-reactive protein; IL-6 = interleukin-6; PTA = pure-tone average; mGCIPL = macular ganglion cell-inner plexiform layer; M = mean; SD = standard deviation; WRCM = word recognition in competing message test. If not stated otherwise, measure was collected at BOSS-3. Sample sizes differ slightly due to missing data. Thin mGCIPL was defined as an mGCIPL thickness 1 SD or more below the population mean (≤70.3 μm). Cognitive impairment was defined as memory concerns and impairment in one or more cognitive domains, self-report or surrogate report of Alzheimer’s disease or dementia, or a Mini-Mental-State Examination score below 24.
aAssessed at BOSS-2.
bAssessed at BOSS-1.
Thicker mGCIPL was associated with better cognitive function (0.12 SD increase per 10 µm in thickness; 95% confidence interval (CI) 0.07,0.16; p < .0001) in age-sex-adjusted models. People with a thin mGCIPL had worse cognitive function scores (−0.25 lower in thin group; CI −0.35, −0.14; p < .0001) in age-sex-adjusted models. Effect sizes were comparable in fully adjusted models. We did not find an association of mGCIPL with cognitive impairment (Table 2). Analyses with single cognitive test scores revealed associations of mGCIPL thickness and thin mGCIPL with TMTA, TMTB, and DSST and not with VFT. There was a weak association between mGCIPL and AVLT that was attenuated in the fully adjusted model (Supplementary Table S1).
Table 2.
Associations Between Macular Ganglion Cell-Inner Plexiform Layer Thickness and Cognition
| Determinant | Cognitive Function Score, β in SD (95% CI) | Cognitive Impairment, OR (95% CI) | ||
|---|---|---|---|---|
| Age-Sex Adjusteda | Fully Adjusteda,c | Age-Sex Adjustedb | Fully Adjustedb,c | |
| Mean mGCIPL, 10 µm | 0.12 (0.07, 0.16), p < .0001 | 0.11 (0.06, 0.16), p < .0001 | 0.85 (0.68, 1.07), p = .17 | 0.99 (0.72, 1.35), p = .93 |
| Thin mGCIPL | −0.25 (−0.35, −0.14), p < .0001 | −0.23 (−0.34, −0.12), p < .0001 | 1.49 (0.86, 2.60), p = .15 | 1.38 (0.70, 2.71), p = .35 |
Notes: CI = confidence interval; mGCIPL = macular ganglion cell-inner plexiform layer; OR = odds ratio; SD = standard deviation.Thin mGCIPL was defined as an mGCIPL thickness 1 SD or more below the population mean (≤70.3 μm).
aMultivariable linear regression model adjusted for age, sex.
bMultivariable logistic regression model adjusted for age, sex.
cModel further adjusted for education, smoking, waist circumference, alcohol consumption, hypertension, diabetes, interleukin-6, high-sensitivity C-reactive protein, regular exercise, and lead exposure.
We found sex interactions with regards to hearing sensitivity (model with average mGCIPL p = .08, model with thin mGCIPL p = .004). Thicker mGCIPL and not being in the thin mGCIPL group were associated with better hearing sensitivity in women (−1.21dB decrease per 10 µm in thickness; CI −2.18, −0.25; p = .01; 4.05 dB higher in thin group; CI 1.90, 6.20; p < .001) but not men (−0.12 dB decrease per 10 µm in thickness; CI −1.18,0.94; p = .82; −0.87 dB lower in thin group; CI −3.44, 1.70; p = .51). Effect sizes were similar after further adjustment. There was no association between mGCIPL and hearing impairment. Thicker mGCIPL and not being in the thin mGCIPL group were associated with better central auditory performance: Thicker mGCIPL and not being in the thin mGCIPL group were associated with better DDT performance (0.67% increase per 10 µm thickness; CI 0.08, 1.25; p = .03; −2.03% lower in thin group; CI −3.40, −0.67; p = .003) in age-sex-adjusted models. Effect sizes were marginally different after further adjustment. There were significant sex interactions for WRCM performance (model with average mGCIPL p = .01; model with thin mGCIPL p < .05). Thicker mGCIPL and not being in the thin mGCIPL group were associated with better WRCM performance in men (1.50% increase per 10 µm in thickness; CI 0.23, 2.77; p = .02; −2.60% lower in thin group; CI −5.62, 0.43; p = .09), while effects were smaller in women (0.99% increase per 10 µm in thickness; CI −0.02, 2.00; p = .05; −1.97% lower in thin group; CI −4.22, 0.29; p = .09). Effect sizes minimally changed in fully adjusted models in men and slightly decreased after full adjustment in models for women (Table 3).
Table 3.
Associations Between Macular Ganglion Cell-Inner Plexiform Layer Thickness and Hearing, Olfaction and Vision
| Determinant | Hearing Sensitivity Score, β in dB (95% CI) | Word Recognition Score, β in % (95% CI) | ||
|---|---|---|---|---|
| Age Adjusteda | Fully Adjusteda,c | Age Adjusteda | Fully Adjusteda,c | |
| Mean mGCIPL, 10 µm | W: −1.21 (−2.18, −0.25), p = .01 M: −0.12 (−1.18, 0.94), p = .82d | W: −1.61 (−2.83, −0.390), p = .01 M: 0.05 (−1.18, 1.27), p = .94d | W: 0.99 (−0.02, 2.00), p = .05 M: 1.50 (0.23, 2.77), p = .02d | W: 0.53 (−0.73, 1.78), p = .41 M: 1.77 (0.30, 3.24), p = .02d |
| Thin mGCIPL | W: 4.05 (1.90, 6.20), p < .001 M: −0.87 (−3.44, 1.70), p = .51d | W: 4.56 (2.05, 7.08), p = .001 M: −1.27 (−4.20, 1.66), p = .40d | W: −1.97 (−4.22, 0.29), p = .09 M: −2.60 (−5.62, 0.43), p = .09d | W: −0.74 (−3.34, 1.85), p = .57 M: −2.82 (−6.31, 0.66), p = .11d |
| Hearing Impairment, OR (95% CI) | Dichotic Digits Test score, β in % (95% CI) | |||
| Age-Sex Adjustedb | Fully Adjustedb,c | Age-Sex Adjusteda | Fully Adjusteda,c | |
| Mean mGCIPL, 10 µm | 0.96 (0.85, 1.09), p = .50 | 0.98 (0.84, 1.14), p = .76 | 0.67 (0.08, 1.25), p = .03 | 0.54 (−0.13, 1.21), p = .11 |
| Thin mGCIPL | 1.08 (0.82, 1.44), p = .58 | 1.10 (0.79, 1.53), p = .57 | −2.03 (−3.40, −0.67), p = .003 | −1.52 (−3.03, −0.01), p = .048 |
| Determinant | Olfactory Impairment, OR (95% CI) | Visual Impairment, OR (95% CI) | ||
| Age-Sex Adjustedb | Fully Adjustedb,c | Age-Sex Adjustedb | Fully Adjustedb,c | |
| Mean mGCIPL, 10 µm | 1.02 (0.82, 1.27), p = .86 | 1.08 (0.83, 1.40), p = .55 | 0.62 (0.51, 0.73), p < .0001 | 0.55 (0.44, 0.70), p < .0001 |
| Thin mGCIPL | 0.97 (0.59, 1.61), p = .91 | 0.85 (0.47, 1.52), p = .58 | 2.66 (1.76, 4.03), p < .0001 | 2.90 (1.75, 4.79), p < .0001 |
Notes: CI = confidence interval; dB = decibel; OR = odds ratio; mGCIPL = macular ganglion cell-inner plexiform layer. Thin mGCIPL was defined as an mGCIPL thickness 1 SD or more below the population mean (≤70.3 μm).
aMultivariable linear regression model adjusted for age, sex.
bMultivariable logistic regression model adjusted for age, sex.
cModel further adjusted for education, smoking, waist circumference, alcohol consumption, hypertension, diabetes, interleukin-6, high-sensitivity C-reactive protein, regular exercise, and lead exposure.
dDue to sex interactions the results are presented stratified by sex with W = women and M = men.
No association of mGCIPL with olfactory impairment was found (Table 3). Thicker mGCIPL was strongly associated with lower odds of visual impairment (OR = 0.62 per 10 µm in thickness; CI 0.51, 0.73; p < .0001) and people with a thin mGCIPL had an increased odds of visual impairment (OR = 2.66 per 10 µm in thickness; CI 1.76, 4.03; p < .0001; Table 3). Effect sizes remained similar after adjustment.
Thicker mGCIPL was associated with higher brain aging scores. Individuals with a thin mGCIPL had lower brain aging scores (Supplementary Table S2).
All results were similar when additionally adjusting for OCT signal strengths.
Sensitivity Analyses
Effect sizes were similar in age-sex-adjusted models in the smaller sample with complete confounder variables compared to the initial models with the exception of the thin mGCIPL group comparisons in DDT and WRCM, where most effect sizes decreased (Supplementary Tables S3 and S4). Subsequent comparisons revealed that the samples were not different in age, sex, education, mGCIPL thickness, percentage of thin mGCIPL, DDT, or WRCM scores (data not shown). Results were similar when excluding participants with eye diseases (data not shown).
Discussion
We found that mGCIPL thickness was associated with cognitive and sensorineural function in healthy middle-aged individuals. This was also true when taking other traditional risk factors of metabolic function, vascular function, inflammation, and neurotoxin exposure into account. MGCIPL could, therefore, be a useful marker of early, preclinical decline within the central nervous system and its functionality.
Our results are consistent with previous research showing an association of thinner mGCIPL with neurodegenerative diseases (7,10,11), as well as with smaller gray matter volume (14,15) and with lower cognitive function (7) in healthy adults. We extend previous findings to a variety of neuronal functions and effects in midlife.
We found an association of mGCIPL with the composite cognitive function score in middle-aged adults, similar to results in older adults from the Rotterdam Study (7). Previous research suggests that particularly processing speed starts to decline early (37) and speed tests might be more prone to detect early changes. This might explain why we found the strongest effects on cognitive tests for TMTA, TMTB, and DSST, which are measures of speed, attention, executive functioning. There were weaker effects for the memory test (AVLT) and no effects on language function (VFT). We did not find an association with cognitive impairment, which might be due to limited power since we had only few cognitive impairment cases.
Degeneration within the central nervous system affects various neuronal functions. Researchers assume that sensory brain regions are affected by neurodegeneration such as Alzheimer pathology and that deficits in hearing, vision, and olfaction precede cognitive symptoms in Alzheimer’s disease (17). We found associations with hearing sensitivity and with central auditory functions, which are measured by more complex hearing tests (38). Some central auditory effect sizes declined in the fully adjusted models. We revealed that those effects were already attenuated in the smaller sample with complete confounders, and therefore, the decline in the fully adjusted models may not be due to confounder adjustment but to sample issues. Finding no substantial differences between the study sample and the smaller sample with complete confounder data, suggests that we might be lacking the power to consistently detect an association between mGCIPL and central auditory functions after confounder adjustment. There were sex differences among hearing sensitivity and WRCM results. Sex differences in hearing impairment prevalence and incidence have been previously shown (20,39). Associations might be different in men and women because of different sex-specific pathophysiological mechanisms or because effects might be obscured differently due to differences in exposures to recreational and occupational noise between men and women (40).
Few studies investigated retinal nerve fiber layers and olfaction. In MS patients, neuroaxonal degeneration—as measured by peripapillary retinal nerve fiber layer thickness was associated with odor identification (41). We did not find an association between mGCIPL and olfaction impairment, which might be due to limited power because of low prevalence of olfaction impairment in our middle-aged sample. An olfactory threshold test might be more sensitive to detect early olfactory changes in middle-aged adults.
We identified strongest associations between mGCIPL and visual impairment, which is consistent with an MS patient study showing a relationship between thinner mGCIPL and impaired CS (42). Since these nerve layers constitute the primary neuronal substrate in the visual pathway, the direct structural-functional relation between the mGCIPL and vision could explain that these were the strongest effects in our study.
We also found an association of mGCIPL thickness with the combined measure of brain aging. This corresponds to the idea that decline within the central nervous system is not limited to one system but simultaneously affects various neuronal functions.
Limitations
Due to limited data availability, we used lead and inflammation markers from earlier BOSS waves. Importantly, mGCIPL was a marker for sensory and cognitive function with and without adjusting for those factors. Our middle-aged sample showed few cases of cognitive and olfactory impairment, which may have limited our power to detect effects. Moreover, there were slight differences in vision and olfaction impairment prevalence in the excluded compared to the included participants. Excluding more severe cases may have led to a slight underestimation of the effects. Although we used a combined measure of various cognitive tests, some of those tests relied on visual stimuli, which may have led to an overestimation of effects on cognition. However, results were similar in sensitivity analyses, excluding participants with eye diseases. Due to the cross-sectional design, we were not able to investigate the prognostic value of change in mGCIPL. The BOSS cohort is primarily non-Hispanic White, which may limit the generalizability of these results to other populations.
Conclusion
Thickness of mGCIPL is associated with cognitive and sensorineural function in midlife and could, therefore, become an inexpensive, noninvasive marker of early neurodegeneration in aging individuals. Longitudinal studies are needed to determine whether the change in mGCIPL might be an even more sensitive indicator of neurodegenerative processes.
Funding
This work was supported by the National Institute on Aging of the National Institutes of Health (grant number R01 AG021917 to K.J.C.); and an Unrestricted Grant from Research to Prevent Blindness, Inc. to the UW-Madison Department of Ophthalmology and Visual Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funding agencies had no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Supplementary Material
Acknowledgments
We would like to acknowledge the contributions of Dr. Ronald Klein to this study. He passed away before this paper was written but was a valued member of our investigative group. He contributed his expertise in ocular epidemiology to the design and conduct of this cohort study.
Conflict of Interest
None reported.
References
- 1. Prince M, Wimo A, Guerchet M, Ali G-C, Wu Y-T, Prina M. World Alzheimer Report 2015 The Global Impact of Dementia an analysis of prevalence, incidence, cost and trends.https://www.alz.co.uk/research/WorldAlzheimerReport2015.pdf. Published 2015. Accessed February 10, 2020.
- 2. Wallin MT, Culpepper WJ, Nichols E, et al. Global, regional, and national burden of multiple sclerosis 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2019;18(3):269–285. doi: 10.1016/S1474-4422(18)30443-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Dorsey ER, Elbaz A, Nichols E, et al. Global, regional, and national burden of Parkinson’s disease, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2018;17(11):939–953. doi: 10.1016/S1474-4422(18)30295-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Jack CR Jr, Knopman DS, Jagust WJ, et al. Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical model of dynamic biomarkers. Lancet Neurol. 2013;12(2):207–216. doi: 10.1016/S1474-4422(12)70291-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. London A, Benhar I, Schwartz M. The retina as a window to the brain-from eye research to CNS disorders. Nat Rev Neurol. 2013;9(1):44–53. doi: 10.1038/nrneurol.2012.227 [DOI] [PubMed] [Google Scholar]
- 6. Mwanza J-C, Oakley JD, Budenz DL, Chang RT, Knight OJ, Feuer WJ. Macular ganglion cell–inner plexiform layer: automated detection and thickness reproducibility with spectral domain–optical coherence tomography in glaucoma. Investig Opthalmology Vis Sci. 2011;52(11):8323. doi: 10.1167/iovs.11-7962. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Mutlu U, Colijn JM, Ikram MA, et al. Association of retinal neurodegeneration on optical coherence tomography with dementia: a population-based study. JAMA Neurol. 2018;75(10):1256–1263. doi: 10.1001/jamaneurol.2018.1563 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Chan VTT, Sun Z, Tang S, et al. Spectral-domain OCT measurements in Alzheimer’s disease: a systematic review and meta-analysis. Ophthalmology. 2019;126(4):497–510. doi: 10.1016/j.ophtha.2018.08.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Choi SH, Park SJ, Kim NR. Macular ganglion cell-inner plexiform layer thickness is associated with clinical progression in mild cognitive impairment and Alzheimer’s disease. PLoS One. 2016;11(9):e0162202. doi: 10.1371/journal.pone.0162202 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sari ES, Koc R, Yazici A, Sahin G, Ermis SS. Ganglion cell–inner plexiform layer thickness in patients with Parkinson disease and association with disease severity and duration. J Neuro-Ophthalmology. 2014;35(2):1. doi: 10.1097/WNO.0000000000000203. [DOI] [PubMed] [Google Scholar]
- 11. Petzold A, Balcer LJ, Calabresi PA, et al. ; ERN-EYE IMSVISUAL Retinal layer segmentation in multiple sclerosis: a systematic review and meta-analysis. Lancet Neurol. 2017;16(10):797–812. doi: 10.1016/S1474-4422(17)30278-8 [DOI] [PubMed] [Google Scholar]
- 12. Mauschitz MM, Holz FG, Finger RP, Breteler MMB. Determinants of macular layers and optic disc characteristics on SD-OCT: the Rhineland Study. Transl Vis Sci Technol. 2019;8(3):34. doi: 10.1167/tvst.8.3.34 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Khawaja AP, Chua S, Hysi PG, et al. Comparison of associations with different macular inner retinal thickness parameters in a large cohort. Ophthalmology. 2020;127(1):62–71. doi: 10.1016/j.ophtha.2019.08.015. [DOI] [PubMed] [Google Scholar]
- 14. Mutlu U, Bonnemaijer PWM, Ikram MA, et al. Retinal neurodegeneration and brain MRI markers: the Rotterdam Study. Neurobiol Aging. 2017;60:183–191. doi: 10.1016/j.neurobiolaging.2017.09.003 [DOI] [PubMed] [Google Scholar]
- 15. Ong YT, Hilal S, Cheung CY, et al. Retinal neurodegeneration on optical coherence tomography and cerebral atrophy. Neurosci Lett. 2015;584:12–16. doi: 10.1016/j.neulet.2014.10.010 [DOI] [PubMed] [Google Scholar]
- 16. Liu Y-L, Hsieh Y-T, Chen T-F, et al. Retinal ganglion cell–inner plexiform layer thickness is nonlinearly associated with cognitive impairment in the community-dwelling elderly. Alzheimer’s Dement Diagnosis, Assess Dis Monit. 2019;11:19–27. doi: 10.1016/j.dadm.2018.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Albers MW, Gilmore GC, Kaye J, et al. At the interface of sensory and motor dysfunctions and Alzheimer’s disease. Alzheimers Dement. 2015;11(1):70–98. doi: 10.1016/j.jalz.2014.04.514 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Schubert CR, Fischer ME, Pinto AA, et al. Brain aging in midlife: the beaver dam offspring study. J Am Geriatr Soc. 2019;67(8):1610–1616. doi: 10.1111/jgs.15886 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Schubert CR, Cruickshanks KJ, Fischer ME, et al. Sensorineural impairments, cardiovascular risk factors, and 10-year incidence of cognitive impairment and decline in midlife: the Beaver dam offspring study. J Gerontol A Biol Sci Med Sci. 2019;74(11):1786–1792. doi: 10.1093/gerona/glz011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Nash SD, Cruickshanks KJ, Klein R, et al. The prevalence of hearing impairment and associated risk factors. Arch Otolaryngol Neck Surg. 2011;137(5):432–439. doi: 10.1001/archoto.2011.15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Carl Zeiss Meditec Inc. Cirrus HD-OCT User Manual – Models 500, 5000. Carl Zeiss Meditec, Inc. Dublin, CA, USA. 2014. [Google Scholar]
- 22. Cruz-Herranz A, Balk LJ, Oberwahrenbrock T, et al. ; IMSVISUAL consortium The APOSTEL recommendations for reporting quantitative optical coherence tomography studies. Neurology. 2016;86(24):2303–2309. doi: 10.1212/WNL.0000000000002774 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189–198. doi: 10.1016/0022-3956(75)90026-6 [DOI] [PubMed] [Google Scholar]
- 24. Reitan RM. Trail Making Test Manual for Administration and Scoring. Tucson, AZ: Reitan Neuropsychology Laboratory; 1992. [Google Scholar]
- 25. Strauss E, Sherman EMS, Spreen O, eds. A Compendium of Neuropsychological Tests. 3rd ed.New York, NY: Oxford University Press; 2006. [Google Scholar]
- 26. Wichmann MA, Cruickshanks KJ, Carlsson CM, et al. Long-term systemic inflammation and cognitive impairment in a population-based cohort. J Am Geriatr Soc. 2014;62(9):1683–1691. doi: 10.1111/jgs.12994 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. American National Standards Institute. Maximum Permissible Ambient Noise Levels for Audiometric Test Rooms. New York, NY: ANSI, S3.1; 1999. [Google Scholar]
- 28. American National Standards Institute. Specification for Audiometers. New York, NY: ANSI, S3.6; 2010. [Google Scholar]
- 29. American Speech-Language-Hearing Association (ASHA). Guidelines for manual pure-tone threshold audiometry. ASHA. 1978;20:297–301. [PubMed] [Google Scholar]
- 30. Fischer ME, Cruickshanks KJ, Dillard LK, et al. An epidemiologic study of the association between free recall Dichotic Digits Test performance and vascular health. J Am Acad Audiol. 2019;30(4):282–292. doi: 10.3766/jaaa.17079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Wilson RH, Zizz CA, Shanks JE, Causey GD. Normative data in quiet, broadband noise, and competing message for Northwestern University Auditory Test No. 6 by a female speaker. J Speech Hear Disord. 1990;55(4):771–778. doi: 10.1044/jshd.5504.771 [DOI] [PubMed] [Google Scholar]
- 32. Wiley TL, Cruickshanks KJ, Nondahl DM, Tweed TS, Klein R, Klein BE. Aging and word recognition in competing message. J Am Acad Audiol. 1998;9(3):191–198. [PubMed] [Google Scholar]
- 33. Zhong W, Cruickshanks KJ, Schubert CR, et al. Carotid atherosclerosis and 10-year changes in cognitive function. Atherosclerosis. 2012;224(2):506–510. doi: 10.1016/j.atherosclerosis.2012.07.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Murphy C. Prevalence of olfactory impairment in older adults. JAMA. 2002;288(18):2307. doi: 10.1001/jama.288.18.2307. [DOI] [PubMed] [Google Scholar]
- 35. Pelli DG, Robson JG, Wilkins AJ. The design of a new letter chart for measuring contrast sensitivity. Clin Vis Sci. 1988;2(3):187–199. [Google Scholar]
- 36. Paulsen AJ, Schubert CR, Johnson LJ, et al. Association of cadmium and lead exposure with the incidence of contrast sensitivity impairment among middle-aged adults. JAMA Ophthalmol. 2018;136(12):1342–1350. doi: 10.1001/jamaophthalmol.2018.3931 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Hedden T, Gabrieli JD. Insights into the ageing mind: a view from cognitive neuroscience. Nat Rev Neurosci. 2004;5(2):87–96. doi: 10.1038/nrn1323 [DOI] [PubMed] [Google Scholar]
- 38. Gates GA, Beiser A, Rees TS, D’Agostino RB, Wolf PA. Central auditory dysfunction may precede the onset of clinical dementia in people with probable Alzheimer’s disease. J Am Geriatr Soc. 2002;50(3):482–488. doi: 10.1046/j.1532-5415.2002.50114.x [DOI] [PubMed] [Google Scholar]
- 39. Dalton DS, Schubert CR, Pinto A, et al. Cadmium, obesity, and education, and the 10-year incidence of hearing impairment: the beaver dam offspring study. Laryngoscope. 2020;130(6):1396–1401 . doi: 10.1002/lary.28244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Gates GA, Cobb JL, D’Agostino RB, Wolf PA. The relation of hearing in the elderly to the presence of cardiovascular disease and cardiovascular risk factors. Arch Otolaryngol Head Neck Surg. 1993;119(2):156–161. doi: 10.1001/archotol.1993.01880140038006 [DOI] [PubMed] [Google Scholar]
- 41. Bsteh G, Berek K, Hegen H, et al. Smelling multiple sclerosis: different qualities of olfactory function reflect either inflammatory activity or neurodegeneration. Mult Scler J. 2020;26(1):57–68. doi: 10.1177/1352458518814113. [DOI] [PubMed] [Google Scholar]
- 42. Narayanan D, Cheng H, Tang RA, Frishman LJ. Multifocal visual evoked potentials and contrast sensitivity correlate with ganglion cell-inner plexiform layer thickness in multiple sclerosis. Clin Neurophysiol. 2019;130(1):180–188. doi: 10.1016/j.clinph.2018.10.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.