Introduction
Primitive reflexes (PR) are a group of stereotyped motor responses mediated by brainstem structures triggered by specific sensory inputs.Reference Delwaide and Dijeux1 PR are found in newborns, disappear after the first year of life, and may be released later in life. PR can be found in Alzheimer’s disease (AD) and vascular dementia,Reference Vreeling, Houx, Jolles and Verhey2 cerebrovascular disease,Reference Di Legge, Di Piero and Altieri3 Parkinsons’s disease,Reference Vreeling, Verhey, Houx and Jolles4 amyotrophic lateral sclerosis,Reference Tremolizzo, Susani, Lunetta, Corbo, Ferrarese and Appollonio5 other degenerative diseases,Reference Borroni, Broli and Costanzi6 and in healthy aging subjects.Reference Jacobs and Gossmann7 In clinical practice, PR are considered individually of poor diagnostic significance, whereas their various combinations are considered a warning sign of frontal lobe dysfunction.Reference Isakov, Sazbon, Costeff, Luz and Najenson8 Whether the reappearance of PR in healthy aging individuals is a consequence of age and/or of the accumulated load of degenerative and vascular lesions is still an open issue.
Tauopathy, β-amyloidosis (Aβ), and atrophy of frontal, parietal, and temporal cortices are the key elements of the underlying pathophysiology of AD and Mild Cognitive Impairment (MCI).Reference Albert, Dekosky and Dickson9 These elements also occur in cognitively normal (CN) aging individuals and are nearly inevitable by old age.Reference Jack, Wiste and Weigand10 Thinning of prefrontal, parietal, and temporal association cortices, cerebellum, and striatum occurs in normal aging.Reference Raz, Lindenberger and Rodrigue11 In CN individuals, elevated Aβ accumulation accelerates cortical atrophyReference Chételat, Villemagne and Villain12 and is associated with an increased burden of frontal, parietal, and occipital periventricular white matter hyperintensities (WMH-PV).Reference Marnane, Al-Jawadi and Mortazavi13 WMH and lacunes, i.e., two of the most common imaging hallmarks of cerebral small vessel disease (SVD),Reference Pantoni14 increase with age,Reference de Leeuw, de Groot and Achten15, Reference Chen, Wen, Anstey and Sachdev16 locally disrupt the structural integrity of white matter (WM), and induce thinning of the connected cortical regions through Wallerian degeneration.Reference Duering, Righart and Csanadi17, Reference Lambert, Narean, Benjamin, Zeestraten, Barrick and Markus18 Cortical atrophy induced by WMH drives, in turn, cognitive decline.Reference Rizvi, Narkhede and Last19 The pattern of cortical atrophy induced by WMH-PV overlaps substantially with the patterns of cortical atrophy age-related and AD-related.Reference Habes, Erus and Toledo20
Intact WM connections are necessary for the integration of information within the cerebral networks. Both vascular and atrophic changes occurring in the aging brain fragment over time cerebral networks into disconnected parts, which lead to the loss of the functional inter-relationships among cortical regions, and between them and subcortical structures. Therefore, it seems reasonable to suspect that the re-appearance of PR in healthy aging individuals may be the consequence of the loss of the brainstem inhibition due to the vascular and/or atrophic damage of the dedicated cerebral networks. Considering that extensive circuits interconnect cortex, basal ganglia (BG), thalamus, cerebellum, and brainstem,Reference Alexander, DeLong and Strick21, Reference Caligiore, Strick and Jörntell22 we hypothesized that global cerebral atrophy and atrophy of the caudate nuclei could also be related to the presentation of PR.
The aims of this cross-sectional study on a large sample of neurologically and cognitively healthy (NCH) adult-to-oldest old subjects are: 1) to evaluate whether PR are associated with WMH, lacunes, and with two linear measures of caudate atrophy and overall cerebral atrophy, i.e., bicaudate ratio (BCr) and lateral ventricles to brain ratio (LVBr), respectively; 2) to evaluate the relationship between PR and the topographical location of imaging-detected cerebrovascular lesions. In this paper, the terms “adult”, “elderly”, “old”, and “oldest old” are used to indicate people aged 45–64, 65–74, 75–84, and ≥85 years, respectively.
Methods
Participants
Data were used from the Cognitive Impairment through Aging (CogItA) study, a hospital-based prospective study focused on normal and pathological aging beginning on January 2, 2000 and closing on December 30, 2015. Participants in the study were outpatients self-referred or referred by general practitioners for health screening to the clinics, including the memory clinic, of the Department of Neurology and Cognitive Disorders of the teaching Hospital (AOUP “P. Giaccone”) of the School of Medicine of the University of Palermo, Italy. Details of the inclusion and exclusion criteria of the CogItA study can be found elsewhere.Reference Camarda, Torelli, Camarda, Gagliardo and Monastero23 The local Ethics Committee approved the study, and all participants provided written informed consent.
Data of CogItA NCH participants aged 45–91 years were used. These subjects did not complain about cognitive concerns as corroborated by the informants, showed normal global cognition at the Mini-Mental State Examination,Reference Folstein, Folstein and McHugh24 no impairment in the domains of memory, attention, executive functions, language, and constructional and visuospatial skills, and normal scores on the Activities of Daily Living (ADL)Reference Katz, Ford, Moskowitz, Jackson and Jaffe25 and Instrumental Activities of Daily Living (IADL)Reference Lawton and Brody26 scales. Moreover, they scored a 0 on the Clinical Dementia Rating (CDR) scaleReference Morris27 and did not convert to MCI or dementia during follow-up (mean follow-up =79.15 ± 33.54 months).
Baseline Clinical Assessment
Clinical assessment included medical history, neurological, neuropsychological, and behavioral examinations, as well as blood tests, carotid ultrasonography, and brain imaging. The vascular risk factors (VRF) evaluated were cigarette smoking, arterial hypertension, diabetes mellitus, hypercholesterolemia, hypertriglyceridemia, obesity (Body Mass Index [BMI] ≥ 30 Kg/m2), and chronic obstructive pulmonary disease (COPD). The vascular diseases (VD) evaluated were ischaemic heart diseases, atrial fibrillation, cardiac valvulopathies, chronic heart failure, history of transient ischemic attacks, and lower limb arteriopathy. VRF and VD were assessed as reported elsewhere.Reference Camarda, Torelli, Camarda, Gagliardo and Monastero23 APOE genotypes were assessed using standard methodsReference Hixson and Vernier28 on 94.2% (n = 1174) of the sample. Subjects with at least 1 APOE ε4 allele were considered as APOE ε4 carriers.
Neurological Examination
NCH participants underwent a standardized neurological examination including a careful evaluation of PR. Although many PR are recognized, we focused on three nociceptive PR (nPR), i.e., glabellar tap, snout reflex, and palmomental reflex frequently observed in normal adults, and aging subjects.Reference Jacobs and Gossmann7, Reference Isakov, Sazbon, Costeff, Luz and Najenson8 Examination of nPR was performed independently by two neurologists blinded to the patients history and neuroimaging. Each reflex was considered present according to a combined decision, and only when a reproducible reaction was obtained.
Functional and Behavioral Assessments
Functional status was assessed using the ADL,Reference Katz, Ford, Moskowitz, Jackson and Jaffe25 and the IADLReference Lawton and Brody26 scales administered to the informants, and somatic multimorbidity was quantified through the Cumulative Illness Rating Scale (CIRS).Reference Parmelee, Thuras, Katz and Lawton29 Depressive symptoms were evaluated using the Cornell Scale for DepressionReference Alexopoulos, Abrams, Young and Shamoian30 and the depression subscale of the Hospital Anxiety and Depression Scale.Reference Zigmond and Snaith31 Anxiety symptoms were evaluated through the anxiety subscale of the Hospital Anxiety and Depression ScaleReference Zigmond and Snaith31 and the Hamilton Anxiety Rating Scale.Reference Hamilton32 Depression and anxiety were deemed to be present if at least one of the depression and anxiety scores was found to be above the cut-off level.
Neuropsychological Assessment
The neuropsychological assessment included the Mini-Mental State ExaminationReference Camarda, Torelli, Camarda, Gagliardo and Monastero23 as a test of global cognition and a neuropsychological battery of 12 tests covering 6 cognitive domains, i.e., memory, attention, executive function, language, constructional ability, and visuospatial skill, as previously described.Reference Camarda, Pipia and Azzarello33
Carotid Ultrasonography and Imaging Assessments
Intimal-medial thickness (IMT) and stenosis of internal carotid arteries (SICA) were assessed as reported elsewhere.Reference Camarda, Torelli, Camarda, Gagliardo and Monastero23 Participants were scanned on a 1.5T system (Signa HDxt; GE Medical System, Milwaukee, Wisconsin, USA). Details of the protocol employed can be found elsewhere.Reference Camarda, Torelli, Camarda, Gagliardo and Monastero23 MR images included T1-w and T2-w, and fluid-attenuated inversion recovery (FLAIR) sequences and were recorded and analyzed over time by the Radiology Department of the teaching Hospital affiliated to the School of Medicine of the University. However, for the purpose of the present study, brain scans were re-evaluated independently by two neurologists (CC and IB) trained in neuroimaging and blinded to subjects data. Caudate atrophy, global cerebral atrophy, WMH, and lacunes were evaluated. The BCr and the LVBr were calculated as reported elsewhere.Reference Camarda, Torelli, Camarda, Gagliardo and Monastero23 WMH and lacunes were evaluated according to published criteria.Reference Wardlaw, Smith and Biessels34 WMH were assessed according to the Wahlund scale.Reference Wahlund, Barkhof and Fazekas35 To measure WMH severity, the scores of frontal, parieto-occipital, temporal, infratentorial, and basal ganglia regions of both hemispheres were summed in order to obtain the partial score of deep/subcortical WMH (WMH-SC) (range 0–18), of the infratentorial WMH (WMH-INF) (range 0–6), of the basal ganglia WMH (WMH-BG) (range 0–6), and the WMH total score (WMH-T) (range 0–30). Lobes were distinguished according to anatomical landmarks. Since the Wahlund scale does not evaluate WMH-PV, we also used the Fazekas scale.Reference Fazekas, Chawluk, Alavi, Hurtig and Zimmerman36 WMH-PV severity was graded as mild, moderate, and severe and scored 1, 2, and 3, respectively. WMH-PV scores were dichotomized as 0–1 vs 2–3. Lacunes were identified on T-2w images and scored topographically according to the Whalund regions used to score WMH. Lacunes were categorized as lacunes-SC, lacunes-INF, lacunes-BG, and lacunes-T and were recorded into a binary variable with two levels: the absence of lacunes = 0 and the presence of at least 1 lacuna = 1. Interrater reliability for the assessment of BCr, LVBr, WMH, and lacunes in random samples was excellent.
Statistical Analysis
Descriptive statistics were used. Continuous variables were compared between nPR using one-way analysis of variance. Differences of categorical variables between nPR were analyzed using χ2 test. Results were summarized as absolute numbers with percentages, mean and SD, median, and IQR. The association between putative risk factors/diseases, imaging-detected lesions, and nPR was assessed using logistic regression models with adjustments for age, sex, and education (years). The full model (Model 1) included all the variables resulted significant in the univariate analysis. Parsimonious models were obtained using a stepwise approach, and the Akaike information criterionReference Akaike37 was used to select the reduced model (Model 2). Logistic regression models were also used to evaluate the association of glabellar tap, snout reflex, and palmomental reflex with BCr, LVBr, and the topographical locations of lacunes and WMH evaluated according to the Wahlund scale. By using tetrachoric correlation,Reference Kirk38 we found that depression and anxiety were strongly correlated (ρ =1; p < 0.001). By using the Pearson’s correlation test, we also found that BCr and LVBr were highly correlated (ρ = 0.93, p < 0.001). Therefore, due to multicollinearity problems, association of depression, anxiety, BCr, and LVBr with nPR, and association of BCr and LVBr with individual nPR were estimated using logistic regression analysis separately with adjustments for age, sex, and education (years). Results are presented as odd ratio with 95% confidence interval (95% CI). Level of significance was set at p ≤ 0.05. All analyses were performed using R (3.4.1) statistical software.
Results
The frequency of nPR was computed on 1246 NCH subjects aged 45–91 years (Table 1). Overall, 33.1% (n = 413) of all subjects showed at least one nPR. Glabellar tap, snout reflex, and palmomental reflex were present in all age classes and progressively increased with age showing the highest frequency in the oldest old individuals. The snout reflex was the nPR with the highest frequency (23.5%), followed by glabellar tap (15.7%) and palmomental reflex (5.9%). With the progressive increase of age, nPR were seen in various combinations, particularly in the old and oldest old subjects. Given that the latter were few, subsequent analysis was conducted pooling together the old and oldest old classes in the class of old–oldest old. According to the presence of at least one nPR, subjects were subdivided into those having at least one nPR (nPR+, n = 413), and those without nPR (nPR−, n = 833). Baseline characteristics of NCH subjects with and without nPR are shown in Table 2. Females significantly outnumbered males in both groups, particularly in the nPR- group. Compared to subjects nPR-, subjects nPR+ were significantly older and more educated and showed worse functional scores, and higher CIRS comorbidity and severity indexes. Former and current smokers were significantly more common within subjects nPR+. Compared to subjects nPR-, subjects nPR+ showed significant higher frequency of almost all VRF and VD evaluated, suggesting a higher vascular burden in the latter group than in the former. Subjects nPR+ were significantly more APOE ε4 carriers than subjects nPR-. All subjects performed well above age- and education-corrected cut-offs for cognitive normality (Table 3). However, subjects nPR+ performed less in tests evaluating global cognition, executive functions, attention, and language. Depressive and anxiety symptoms were significantly more frequent in subjects nPR+ than in subjects nPR-. IMT and SICA were more frequent in subjects nPR+ than in subjects nPR-. WMH and lacunes were highly frequent in subjects nPR+ compared to subjects nPR-. BCr and LVBr were significantly higher in subjects nPR+ than in subjects nPR- (Table 4). Because nPR increase with age as many of the factors associated with nPR do, we stratified these factors according to age classes. Table S1 in the supplementary material shows all the variables resulted significant in the univariate analysis of NCH subjects with and without nPR according to age classes. The factors that increased with increasing age were WMH-PV, WMH-BG, depression, anxiety, BCr, and LVBr. Therefore, in subsequent analyses, age was entered into all statistical models to account for this confound.
Table 1: Number and types of nociceptive primitive reflexes (nPR) according to age classes
Data presented are number (%).
Table 2: Baseline characteristics of adult-to-oldest old NCH subjects (n = 1246) with and without nociceptive primitive reflexes (nPR)
a Missing data: nPR+ = 5 (1.2%); nPR- = 67 (8.7%).
f.l., functions lost.
Data presented are number (%) for categorical, and mean (SD) for continuous data.
Bold values indicate significance at p ≤ 0.05.
Table 3: Neuropsychological and behavioral performances of adult-to-oldest old NCH subjects (n = 1246) with and without nociceptive primitive reflexes (nPR)
MMSE, Mini-Mental State Examination.
Data presented are number (%) for categorical, and median (interquartile range, IQR) for continuous data.
Bold values indicate significance at p ≤ 0.05.
Table 4: Carotid ultrasonography, and imaging findings of adult-to-oldest old NCH subjects (n = 1246) with and without nociceptive primitive reflexes (nPR)
a Missing data: IMT, nPR+ = 17; nPR − = 106; SICA, nPR+ = 17; nPR − = 106.
IMT, Intimal-medial thickness; SICA, Stenosis of the internal carotid arteries; WMH, White matter hyperintensities.
Data presented are number (%) for categorical and mean (SD) for continuous data.
Bold values indicate significance at p ≤ 0.05.
The association between nPR and VRF, VD, carotid ultrasonography findings, WMH, lacunes, APOE ε4 carriers, depression, anxiety, BCr, and LVBr, was evaluated through logistic regression models (Table 5). At Model 2 (reduced), APOE ε4 allele, WMH-SC, WMH-PV, lacunes-SC, depression, anxiety, BCr, and LVBr were significantly associated with nPR. When analyzing the association of each nPR with BCr, LVBr, and the topographical location of lacunes and WMH scored according to the Wahlund scale (Table 6), at Model 2, the risk of glabellar tap resulted related to BCr, LVBr, and frontal, parieto-occipital, and temporal WMH. The risk of snout reflex resulted related to frontal lacunes, temporal WMH, BCr, and LVBr. The risk of palmomental reflex resulted related to parieto-occipital WMH, basal ganglia lacunes, BCr, and LVBr.
Table 5: Association between VRF, VD, APOE ε4 Carriers, IMT, SICA, WMH, lacunes, depression, anxiety, BCr, LVBr, and nociceptive primitive reflexes in adult-to-oldest old NCH subjects (n = 1246)
VRF, vascular risk factors; VD, vascular diseases; APOE, apolipoprotein E; IMT, intimal-medial thickness; SICA, stenosis of internal carotid artery; WMH, white matter hyperintensities; BCr, bicaudate ratio; LVBr, lateral ventricles to brain ratio; NCH, neurological and cognitively healthy; CI, confidence interval; COPD, chronic obstructive pulmonary disease; TIA, transient ischemic attacks. Model 1 (full) is composed by all the variables found significant by the univariate analysis, and adjusted for age, sex, and education (years). Model 2 (reduced) is obtained by a stepwise model selection procedure with additional adjustments for age, sex, and education (years). Association of depression, anxiety, bicaudate ratio, and lateral ventricles to brain ratio with nPR was estimated using logistic regression analysis separately with adjustments for age, sex, and education (years). Bold values indicate significant associations.
Table 6: Association between glabellar tap, snout reflex, palmomental reflex, and topographical location of WMH scored according to Wahlund scale, lacunes bicaudate ratio, and lateral ventricles to brain ratio in adult-to-oldest old NCH subjects (n = 1246)
CI, confidence interval; NCH, neurological and cognitively healthy; WMH, white matter hyperintensities.
Model 1 (full) is composed by all the variables found to be significant univariate analysis, and adjusted for age, sex, and education (years). Model 2 (reduced) is obtained by a stepwise model selection procedure with additional adjustments for age, sex, and education (years). Association of BCr and LVBr with glabellar tap, snout reflex and palmomental reflex was estimated using logistic regression analysis separately with adjustments for age, sex, and education (years).
a Lesions too infrequent for model to converge.
Bold values indicate significant associations.
Discussion
Glabellar tap, snout reflex, and palmomental reflex are frequently observed in NCH adult, elderly, and old-oldest old subjects and increase with age either in isolation or in combination. VRF, VD, and CIRS comorbidity and severity indexes, APOE ε4 carriers, imaging-detected cerebrovascular lesions, IMT, and SICA were significant more frequent among subjects nPR+ than subjects nPR-. The cognitive normality of our participants supports previous conclusion that the presence of nPR is unrelated to cognition.Reference van Boxtel, Bosma, Jolles and Vreeling39 However, the fact that NCH subjects nPR+ performed less in tests evaluating global cognition, executive function, attention, and language than those nPR- suggests that the presentation later in life of nPR may precede any evidence of cognitive impairment.
The association of nPR with WMH and lacunes is not surprising, given that these two hallmarks of cerebral SVD are frequently observed in healthy aging individualsReference Pantoni14, Reference de Leeuw, de Groot and Achten15 and are associated with nPR and other neurological abnormalities.Reference Camarda, Torelli, Camarda, Gagliardo and Monastero23, Reference Poggesi, Gouw and van der Flier40 The association of nPR with BCr and LVBr in adult-to-oldest old subjects deserves, instead, some considerations. BCr is considered a reliable marker of caudate atrophy in normal aging.Reference Doraiswamy, Patterson and Na41 It may be that in our cohort, the increased intercaudate distance was the consequence of the age-related shrinkage of the striatum,Reference Raz, Lindenberger and Rodrigue11 or of the basal ganglia lacunes, or of the degeneration of cortico-striatal connections secondary to the cortical thinning induced by frontal lacunes and frontal, parietal, and temporal WMH-SC, or by WMH-PV, or by the cortical Aβ deposition. Considering that age and APOE ε4 allele increase the risk of Aβ deposition in CN individuals,Reference Jack, Wiste and Weigand10 that the majority of subjects nPR+ (62%; n = 257) were elderly-to-oldest old, and that the presence of APOE ε4 allele increased of about 73% the risk of having nPR, it is reasonable to suspect that the subjects nPR+ were in high percentage amyloid positive. The strong association of WMH-PV with nPR is likely to reflect the disruption of long associating tracts induced by WMH-PV and the consequent thinning of frontal, parietal, and temporal association cortices.Reference Lambert, Narean, Benjamin, Zeestraten, Barrick and Markus18–Reference Habes, Erus and Toledo20 One of these tracts could be the superior longitudinal fasciculus (SLF). SLF is a long associating tract, which runs near the lateral ventricles, interconnects bidirectionally prefrontal and premotor frontal cortices and parietal, temporal, and occipital association cortices, and subserves memory, attention, executive, audiospatial, and visuospatial processing.Reference Makris, Kennedy and McInerney42 The possibility that WMH-PV may have damaged SLF disconnecting the above association cortices could explain why subjects nPR+ performed less than subjects nPR- in tests evaluating global cognition, executive functions, attention, and language. LVBr was also associated with nPR. LVBr is a measure of ventricular enlargement that is commonly observed in healthy older adults.Reference Longstreth43 Ventricular expansion reflects gray matter (GM) and WM volumetric reduction throughout multiple brain regions, is associated with thinner GM in frontal, parietal, and temporal regions, i.e., AD signature cortical regions,Reference Madsen, Gutman and Joshi44 and is considered a feasible surrogate, but non-specific, marker of neurodegeneration in MCI and AD.Reference Nestor, Rupsingh and Borrie45 Another finding of our study is that anxiety and depressive symptoms were strongly associated with nPR in adult-to-old-oldest old subjects. The association of depression with nPR is in agreement with the “vascular depression” hypothesis,Reference Alexopoulos, Meyers, Young, Campbell, Silbersweig and Charlson46 whose central mechanism is the disconnection process of cortical and subcortical regions driven mainly by the disruption of the prefrontal connectivity due to SVD. In CN older people, depressive symptoms have been associated with WMH.Reference O’Brien, Firbank and Krishnan47
The reappearance later in life of nPR has been related to the loss of the physiological inhibitor control of lower brainstem centres.Reference Delwaide and Dijeux1 Our findings support the hypothesis that their reappearance could be the consequence of the vascular and degenerative damage of the cortical-basal ganglia-thalamo-cortical circuits, causing the disconnection of the cerebral networks subserving this control. In this hypothetical context, caudate atrophy and the likely damage of the internal circuits of BG could have led to an excessive inhibition to its output nuclei with consequent disinhibition of lower brainstem structures.
Limitations of the study
Some limitations of our study need to be acknowledged. First, among CN subjects aged 50–89 years, different subgroups have been identified on the basis of a various combination of imaging biomarkers of amyloidosis and neurodegeneration of AD type.Reference Jack, Wiste and Weigand10 We were unable to perform in our cohort Aβ and tau imaging and to evaluate quantitatively brain regional structural and metabolic abnormalities, so we do not know to which of these subgroups our subjects nPR+ belong to. Second, caudate atrophy and global cerebral atrophy were evaluated using simple linear measures. We are aware of the crudity of the method used, but BCr and LVBr are reliable markers of caudate atrophy and global cerebral atrophy, respectivelyReference Doraiswamy, Patterson and Na41, Reference Nestor, Rupsingh and Borrie45 and have the advantage of being inexpensive, not labor-intensive, and easy to use in clinical practice without any technical assistance. Third, we did not evaluate cerebral microbleeds, which are present in 18.6% of CN subjects and are associated with age, APOE ε4, Aβ deposition, lacunes, and severe WMHReference Yates, Desmond and Phal48 underestimating, probably, the magnitude of SVD in our sample. Lastly, due to the cross-sectional design of our study, we cannot assess causality but only correlations.
Conclusion
In the modern neurological era, the clinical usefulness of eliciting nPR could seem outdated. However, we have shown that the presence of nPR in adult-to-oldest old NCH individuals is a sign of cortical and subcortical vascular and atrophic changes, and, probably, a warning sign of incipient cognitive decline. Therefore, NCH subjects presenting nPR should manage their VRF/VD rigorously in order to prevent or delay progression of SVD, and future neurological and cognitive disabilities.
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/cjn.2018.388.
Acknowledgements
We gratefully thank all participants and relatives, as well as all neurologists and neuropsychologists who over time collected patients data.
Statement of Authorship
CC and RC were responsible for the study’s concept and design, data management, and record linkage.
GS and GC did the statistical analysis. CC, PT, CP, DA, IB, and RC contributed to the analysis and interpretation of the data.
CC wrote the paper. All co-authors edited the paper and approved its final version.
Disclosures
All the authors hereby declare that they have nothing to disclose.
