Physical activity across the life course and neural biomarkers

Abstract

Background

Physical activity (PA) may decrease the risk of Alzheimer's disease and related disorders (ADRD), although most research has focused on exercise. In addition, ADRD prevention may be most effective when administered earlier in the life course, when changes in the brain may occur 20–25 years before the onset of symptoms. Finally, amyloid-β (A) and tau (T) accumulation are followed by neurodegeneration (N), key features of AD. However, there are additional biomarkers associated with ADRD. The association between different types of PA across the life course and biomarkers beyond the AT(N) framework is unknown.

Objective

Examine the association between different types of PA across the life course and biomarkers beyond the AT(N) framework.

Methods

PA (occupation/school, transportation, household) across the life course (school-age, adolescence, young adult, middle adult) was assessed. AD pathology, neurodegeneration, synaptic dysfunction, and gliosis were measured with the Roche NeuroToolKit, a panel of exploratory prototype assays for cerebrospinal fluid biomarkers. Significant bivariate correlations were followed by logistic regression.

Results

Young adult occupation/school PA was a significant predictor of neurofilament light protein (NfL, OR = 0.969, 95% CI = 0.940–0.999); the full model was significant, χ²(df = 2) = 15.57, p < 0.001, explaining 19.2–25.6% of the variance in NfL. Young adult occupation/school PA was a significant predictor of alpha-synuclein (OR = 0.968, 95% CI = 0.940–0.996); the full model was significant, χ²(df = 1) = 6.70, p = 0.01, explaining 8.8–11.7% of the variance in alpha-synuclein.

Conclusions

PA done as part of one's job or schooling during young adulthood, such as standing and walking, may be protective against dementia later in life.

Keywords

alpha-synuclein Alzheimer's disease dementia health promotion neurofilament proteins

Introduction

Dementia is currently the seventh leading cause of death and one of the major causes of disability and dependency among older people globally.¹ Alzheimer's disease (AD) is the most common cause of dementia, accounting for an estimated 60–80% of cases.² Amyloid-β (Aβ) and tau (T) accumulation are followed by dysfunction and damage to neurons, called neurodegeneration (N), which are key features of AD, as also outlined in the AT(N) framework for AD.³ In this National Institute on Aging and Alzheimer's Association Research Framework, AD and risk for AD are defined by underlying pathologic processes that can be documented in vivo by biomarkers,³ which is helpful in the preclinical phase where individuals do not exhibit signs and symptoms of AD.

Higher physical activity may decrease the risk of AD.⁴ In cognitively intact older adults, more physical activity was associated with lower Aβ₄₂/Aβ₄₀ ratio measured in plasma and with positron emission tomography (PET),⁵ while more vigorous physical activity was associated with less regional tau burden measured with PET.⁶ In older adults with and without cognitive impairment, physical activity was associated with less phosphorylated tau₂₁₇ (pTau₂₁₇) and neurofilament light protein (NfL), and higher cognition scores.⁷ Thus, physical activity is associated with attenuated AD pathology assessed via biomarkers in the AT(N) framework for AD. In all the aforementioned studies, four domains of physical activity were measured: (1) occupation physical activity done as part of one's job or work; (2) transportation physical activity done when traveling from place to place; (3) household physical activity done as part of housework, house maintenance, and caring for family; and (4) leisure-time physical activity done solely for recreation, sport, exercise or leisure.⁸ However, the previously cited studies combined all the physical activity domains into one. None measured each domain of physical activity separately, which is helpful for targeted interventions.

Most studies that examined a specific domain of physical activity and biomarkers measured leisure-time physical activity. Leisure-time physical activity was associated with attenuated regional Aβ but not tau burden measured with PET, and improved cognitive performance in cognitively unimpaired older adults.⁹ A meta-analysis found structured-based leisure-time physical activity significantly decreased blood NfL levels post-training in patients with multiple sclerosis, with no change in NfL levels in control arms.¹⁰ While impressive, 80% of physical activity occurs beyond the leisure-time physical activity domain, specifically during occupation and household activities.¹¹ Occupation and household physical activity provide additional ways for the population to increase their physical activity beyond leisure-time, which could prove essential if also found to decrease the risk for AD.

AD prevention may be most effective when administered earlier in the life course, when changes in the brain may occur as much as 20–25 years before the onset of symptoms.^12,13 More leisure-time physical activity over the span of 10 years was associated with attenuated AD pathology (Aβ₄₂, total Tau [tTau], pTau₁₈₁ in the cerebrospinal fluid [CSF]) of middle-aged and older cognitively intact adults.¹⁴ More leisure-time physical activity during the participants’ 60 s was associated with less tTau (measured in blood) and slower cognitive decline over the next 10 years.¹⁵ Regular leisure-time physical activity from adolescence to adulthood was associated with lower tTau and NfL levels (measured in blood) in young adults.¹⁶

Since most older individuals with dementia have brain changes associated with more than one cause,¹⁷ it may be advantageous to examine biomarkers beyond the AT(N) framework. More physical activity was not associated with plasma Aβ_42/40, ptau217, NfL or glial fibrillary acidic protein (GFAP) in cognitively unimpaired adults, although all physical activity domains were combined into one.⁷ Physical activity was not associated with Aβ₄₂, Aβ₄₀, Aβ_42/40, pTau, tTau or soluble triggering receptor expressed on myeloid cells 2 (sTREM2) in cognitively intact and impaired participants combined, with no physical activity domain specified.¹⁸ A 16-week randomized controlled trial in patients diagnosed with mild to moderate AD found moderate- to vigorous-intensity leisure time physical activity did not change levels of NfL, neurogranin or chitinase-3-like protein 1 (YKL-40) in the CSF,¹⁹ but did increase sTREM2 in the CSF and IL6 in the plasma.²⁰ More leisure-time physical activity was associated with less Aβ and α-synuclein, but not tTau, in the red blood cells of athletes compared to sedentary adults independent of age.²¹ Conversely, higher intensity leisure time physical activity has been shown to increase S100 calcium-binding protein B (S100B) in the blood immediately afterwards.^22,23 This suggests the importance of examining different domains of physical activity separately, along with neural biomarkers beyond the AT(N) framework. The purpose of this study was to examine the association between different domains of physical activity across the life course and biomarkers beyond the AT(N) framework in cognitively unimpaired middle-aged and older adults.

Methods

Participants

Participants were from visit one (baseline) of a study focused on white matter degeneration, which was a subset of the Wisconsin Registry for Alzheimer's Prevention study and the Wisconsin Alzheimer's Disease Research Center.²⁴ Of the 150 participants in the cohort, 99 participants consented to be interviewed with the Lifetime Total Physical Activity Questionnaire. Of the 99 participants, 24 did not have neural biomarkers. This was because the lumbar puncture (required to obtain the CSF from which the neural biomarkers were retrieved) were in some cases scheduled separately from other procedures. For the current analyses, lumbar puncture results collected at the same time as the comprehensive cognitive evaluation, cardiovascular risk factors and lifetime physical activity interview were used. Of the 75 remaining participants, two were missing multiple covariate data, resulting in a final sample size of 73. Participants were middle-aged (40‒64 years) and older adults (65+ years) with no cognitive symptomatology or physical limitations at baseline, with the sample enriched with participants who had a parental history of AD. Exclusion criteria were abnormal cognitive testing (more than 1.5 standard deviation below normal on the Rey Auditory Verbal Learning total learning or delayed free recall, or 2 standard deviations below IQ), known family autosomal dominant mutations, MRI scanner incompatibility, history of major psychiatric disease (i.e., schizophrenia, substance abuse) or major medical conditions (history of neurologic disorders including prior head trauma with loss of consciousness, cancer requiring chemotherapy or radiation, insulin-dependent diabetes, and untreated hypertension), and abnormal structural MRI indicating a disease or condition that would affect normalization procedures (tumor or hydrocephalus). Written informed consent was obtained under a protocol approved by the University of Wisconsin Health Sciences Institutional Review Board (2011–0907-CR002). This study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments.

Physical activity

Face-to-face interviews were conducted with the Lifetime Total Physical Activity Questionnaire (LTPAQ) using a recall calendar as previously described.²⁵ The LTPAQ has been used with health issues related to insufficient physical activity across the lifespan such as breast,^26–28 endometrial,^29–31 and prostate cancer³²; brain biomarkers,^33–35 cognitive function,³⁶ and general health outcomes.³⁷

The LTPAQ assesses four domains of physical activity; occupation/school, transportation to occupation/school, household, and leisure-time.²⁵ For occupation/school physical activity, the job title as well as up to three descriptors of the paid or volunteer activity (e.g., standing, sitting, walking) performed for at least eight hours per week for four months of the year were obtained (or 8 h × 16 weeks = 128 h total per year), starting with the first job in a participant's life.²⁵ Occupation/school physical activity also included school starting with kindergarten.³³ Transportation physical activity included walking, biking, running or rollerblading to and from one's occupation/volunteer/school if done at least two hours per week for four months of the year (or 2 h × 16 weeks = 32 h total per year).²⁵ Personal household physical activity such as gardening, yard work, do-it-yourself jobs around the home and childcare completed in a typical day or week for at least seven hours per week four months of the year (or 7 h × 16 weeks = 112 h total per year) were included.²⁵ Leisure-time physical activity such as sports, exercise, recess, and physical education/gym classes were recorded if they were done for at least two hours per week for four months of the year (or 2 h × 16 weeks = 32 h total per year) at least ten times in a lifetime.²⁵ The total number of hours per week for occupation/school, transportation, household, and leisure-time could not exceed 168 h (24 h × 7 days/week).³³

The LTPAQ was used to estimate the metabolic equivalent of task (MET) of activities abstracted from the Compendium of Physical Activities³⁸ within each of the domains, as previously described.³³ Sedentary, light, moderate and vigorous-intensity were characterized as 1.0–1.5 METs, 1.6–2.9 METs, 3.0–5.9 METs and ≥6 METs, respectively.³⁸ In order to assess physical activity across the life course, each physical activity domain was assessed across developmental stages based on Erikson's (1998) Theory of Psychosocial Development³⁹; school-age (6‒11 years old), adolescence (12‒18 years old), young (19‒39 years old), middle (40‒64 years old), and older adulthood (age 65+ years old). The average MET-hours/week/year in occupation/school, transportation, household and leisure-time physical activity during school-age, adolescence, young, middle and older adulthood were estimated as previously described.³³ Since only 44% of the sample were age 65+, the older adult developmental stage from the LTPAQ was also omitted from the current analysis. Investigators who performed the physical activity calculations did not have access to the neural biomarker results.

The face-to-face administration of the LTPAQ has evidence of construct validity⁴⁰ and test-retest reliability for moderate-intensity occupation/school and household physical activity during school-age, adolescent and young adulthood; and moderate-intensity transportation physical activity during young and middle adulthood.⁴¹ Although evidence of construct validity and test-retest reliability were found for vigorous-intensity occupation/school and household during school-age, and vigorous-intensity household during young and middle adult,⁴¹ zero amounts were reported and subsequently omitted from the current analysis.

CSF collection to obtain neural biomarkers

CSF collection was previously described⁴² and briefly described here. CSF samples were acquired with a uniform preanalytical protocol. Samples were collected in the morning after an 8- to 12-h fast using a Sprotte 24- or 25-gauge atraumatic spinal needle, along with 22 mL of fluid collected via gentle extraction into polypropylene syringes and combined into a single 30 mL polypropylene tube. After gently mixing, samples were centrifuged to remove red blood cells or other debris; 0.5 mL of CSF was aliquoted into 1.5-mL polypropylene tubes and stored at −80°C within 30 min of collection.

CSF assays

Neural biomarkers were measured from CSF with the Roche NeuroToolKit, a panel of robust exploratory prototype assays designed to evaluate biomarkers associated with key pathologic events characteristic of AD and other neurological disorders, used for research purposes only, and not approved for clinical use.⁴² As previously described,⁴² CSF samples were re-assayed at the Clinical Neurochemistry Laboratory at Sahlgrenska University under strict quality control procedures. The following immunoassays were performed on a cobas^® e 601 analyzer: Elecsys ^® Aβ (1–42) [Aβ₄₂] CSF, Elecsys tau phosphorylated at threonine 181 [pTau₁₈₁] CSF, and Elecsys tTau CSF, S100B, and interleukin-6 (IL-6) (all Roche Diagnostics International Ltd, Rotkreuz, Switzerland). The remaining NeuroToolKit panel was assayed on a cobas e 411 analyzer: Elecsys Aβ (1-40) [Aβ₄₀] CSF, neurogranin, NfL, α-synuclein, GFAP, YKL-40, and sTREM2 (all Roche Diagnostics International Ltd, Rotkreuz, Switzerland). AD biomarkers include Aβ₄₂, Aβ₄₀, Aβ₄₂/Aβ₄₀, pTau₁₈₁ and pTau₁₈₁/Aβ₄₂, while a proinflammatory cytokine is IL-6.⁴³ Since Aβ_42, Aβ_40, and IL-6 did not differ between cognitively unimpaired, mild cognitive impairment, or dementia based on National Institute on Aging-Alzheimer's Association (NIA-AA) criteria in the current sample,⁴² they were omitted from the current analysis. Since Aβ₄₂/Aβ₄₀ and pTau₁₈₁/Aβ₄₂ exhibited 95% agreement in differentiating between cognitively unimpaired, mild cognitive impairment or dementia based on NIA-AA criteria in the current sample, and pTau₁₈₁/Aβ₄₂ simultaneously comprises both proteinopathies,⁴² Aβ₄₂/Aβ₄₀ was omitted from the current analysis. An AD-type neurodegeneration biomarker is tTau. Due to the high correlation between pTau₁₈₁ and tTau in the current sample (r = 0.974, p < 0.001), and pTau₁₈₁ proteinopathy is comprised in pTau₁₈₁/Aβ₄₂, pTau₁₈₁ was excluded from subsequent analyses. A general neurodegeneration biomarker is NfL; synaptic dysfunction biomarkers include neurogranin and α-synuclein; glial biomarkers (secreted from astrocytes and/or microglia) include YKL-40, GFAP, S100B, and sTREM2.

Covariates

Common risk factors for AD were chosen as covariates.² Demographic characteristics included age measured continuously in years, number of apolipoprotein ε4 (APOE ε4) alleles (zero, one, or two), and self-reported gender (male or female). The highest level of education achieved was measured as follows: (1) up to high school, (2) some college, (3) associate's degree, (4) bachelor's degree, (5) master's degree, (6) professional degree, i.e., PhD, MD, JD. For the purposes of statistical analysis, education was measured continuously in years (limiting degrees of freedom due to sample size). Cardiovascular risk factors were measured. Hypertension and dyslipidemia were assessed by whether participants were taking medication for these disorders. Diabetes was assessed by whether participants ever had a diagnosis of diabetes, as some participants were controlling their diabetes with lifestyle modifications (without medication). Body mass index was (BMI) was calculated by multiplying the measured weight in pounds by 703 and dividing by the height in inches squared.

Statistical analysis

Bivariate correlations examined the association between four demographic (age, number of APOE ɛ4 alleles, gender, education), four cardiovascular (hypertension, diabetes, dyslipidemia, body mass index), and eight physical activity measures (moderate-intensity occupation/school and household physical activity during school-age, adolescence, young adulthood; moderate-intensity transportation physical activity during young and middle adulthood) with the median of nine CSF assays (pTau₁₈₁/Aβ₄₂, tTau, NfL, neurogranin, α-synuclein, YKL-40, GFAP, S100B, sTREM2). The Benjamini-Hochberg Procedure was used to control for the False Discovery Rate < 0.05. Significant bivariate correlations were followed up with logistic regression. Univariate analyses were conducted in SPSS version 24 (Chicago, IL).

Results

Descriptive statistics

Sample characteristics are shown in Table 1. Participants were aged between 48 to 72 years, mostly white (98.6%), highly educated, and overweight but otherwise healthy with no APOE ε4 alleles (63.0%). Overall, participants reported relatively little physical activity during their school-age years. Occupation/school physical activity increased during adolescence, and again during young adulthood. There was relatively little change in household physical activity from school-age to adolescence, with an increase during young adulthood. Minimal amounts of transportation physical activity were reported.

Table 1.

Sample characteristics.

Characteristics	Total (n = 73)
	No.	%
Female	50	68.5
Number of APOE ε4 alleles
No APOE ε4 alleles	46	63.0
One APOE ε4 allele	25	34.2
Two APOE ε4 alleles	2	2.7
Highest level of education
High School	11	15.1
Some College	4	5.5
Associate's Degree	11	15.1
Bachelor's Degree	24	32.9
Master's Degree	18	24.7
Professional Degree, i.e., JD, MD, PhD	5	6.8
Hypertension medication	20	27.4

Diabetes diagnosis	3	4.1
Cholesterol medication	22	30.1
	Median	IQR
Age (years)	63.00	57.50–67.00
Body Mass Index (n = 72)	28.00	24.00–31.00
AD biomarker
pTau₁₈₁/Aβ₄₂	0.02	0.01–0.02
AD-type Neurodegeneration biomarker
tTau pg/mL	195.00	150.00–230.50
General Neurodegeneration biomarkers
NfL pg/mL	79.90	65.83–101.25
Synaptic dysfunction biomarkers
Neurogranin pg/mL	792.30	557.65–993.60
α-Synuclein pg/mL	134.00	111.25–181.20
Gliosis biomarkers
YKL-40 ng/mL	139.90	108.75–185.00
GFAP ng/mL	8.15	6.47–9.89
S100B ng/mL	1.11	0.88–1.23
sTREM2 ng/mL	7.69	6.18–9.21
School-age physical activity
Occupation/School moderate-intensity	0.00	0.00–1.76
Household moderate-intensity	0.00	0.00–0.87
Adolescent physical activity
Occupation/School moderate-intensity	6.67	5.14–9.49
Household moderate-intensity	0.00	0.00–1.40
Young adult physical activity
Occupation/School moderate-intensity	14.06	5.23–33.23
Transportation	0.34	0.00–1.02
Household moderate-intensity	2.66	0.00–22.33
Middle adult physical activity
Transportation	0.00	0.00–0.00

APOE ε4: apolipoprotein ε4 carrier; IQR: interquartile range; AD: Alzheimer's disease; Aβ: amyloid-β; pTau₁₈₁: tau phosphorylated at threonine 181; tTau: total tau; NfL: neurofilament light protein; YKL-40: chitinase-3-like protein 1; GFAP: glial fibrillary acidic protein; S100B: S100 calcium-binding protein B; sTREM2: soluble triggering receptor expressed on myeloid cells 2.

Correlations between demographic characteristics and neural biomarkers

There was a statistically significant correlation between age and NfL (r = 0.369, p = 0.001), YKL40 (r = 0.284, p = 0.015), and GFAP (r = 0.324, p = 0.005). There was a statistically significant correlation between the number of APOE ε4 alleles and pTau₁₈₁/Aβ₄₂ (r = 0.400, p < 0.001). There was a statistically significant correlation between gender and neurogranin (r = 0.334, p = 0.004). None of the correlations remained significant after the Benjamini-Hochberg Procedure. There were no statistically significant correlations between education and any of the neural biomarkers (Table 2).

Table 2.

Bivariate correlations with the median of neural biomarkers (n = 73 except where noted).

Characteristics	AD	Neurodegeneration		Synaptic Dysfunction		Gliosis
	AD	AD	General
	pTau₁₈₁/Aβ₄₂	tTau	NfL	Neurogranin	α-Synuclein	YKL-40	GFAP	S100B	sTREM2
Demographic Characteristics
Age (years)	0.13	0.20	0.37**	0.02	0.16	0.28*	0.32**	0.18	0.16
# of APOE ε4 Alleles	0.40**	0.02	0.02	0.12	0.02	−0.04	−0.09	0.17	−0.04
Gender	0.06	0.16	−0.02	0.33**	−0.02	0.04	−0.20	−0.20	−0.14
Education	−0.03	0.09	0.06	−0.01	0.19	0.01	−0.07	−0.16	−0.03
Cardiovascular Risk Factors
Hypertension	0.04	0.11	−0.01	−0.07	−0.01	0.05	0.24*	0.24*	−0.07
Diabetes	−0.08	0.07	0.07	−0.07	0.20	0.07	0.20	−0.07	0.07
Dyslipidemia	0.09	−0.01	0.11	−0.13	0.11	0.05	0.17	−0.01	0.05
Body Mass Index (n = 72)	−0.00	0.02	−0.08	0.02	0.02	−0.12	0.16	0.10	0.08
School-age physical activity
Occupation/School moderate-intensity	0.11	−0.10	−0.18	−0.06	−0.12	−0.07	−0.11	−0.15	−0.08
Household moderate-intensity	−0.05	0.03	0.06	0.06	0.01	0.17	0.10	0.05	0.23
Adolescent physical activity
Occupation/School moderate-intensity	−0.07	−0.19	−0.03	0.05	0.02	−0.03	−0.01	−0.12	0.04
Household moderate intensity	−0.12	0.09	0.06	0.16	−0.06	0.08	0.12	0.07	−0.02
Young adult physical activity
Occupation/School moderate-intensity	−0.07	−0.10	−0.27*	−0.04	−0.28*	−0.18	−0.05	−0.04	−0.04
Transportation moderate-intensity	0.08	−0.07	−0.11	−0.04	−0.06	−0.10	0.11	−0.10	−0.08
Household moderate-intensity	0.08	−0.00	0.09	0.10	0.01	0.01	0.04	−0.09	0.00
Middle adult physical activity
Transportation moderate-intensity	−0.03	0.18	0.09	0.18	0.14	0.15	0.10	0.12	0.14

AD: Alzheimer's disease; Aβ: amyloid-β; pTau₁₈₁: tau phosphorylated at threonine 181; tTau: total tau; NfL: neurofilament light protein; YKL-40: chitinase-3-like protein 1; GFAP: glial fibrillary acidic protein; S100B: S100 calcium-binding protein B; sTREM2: soluble triggering receptor expressed on myeloid cells 2; APOE ε4: apolipoprotein ε4.

*p < 0.05; **p < 0.01.

Correlations between cardiovascular risk and neural biomarkers

There were statistically significant correlations between hypertension with GFAP (r = 0.237, p = 0.043) and S100B (r = 0.237, p = 0.043), although none of the correlations remained significant after the Benjamini-Hochberg Procedure. There were no statistically significant correlations between diabetes, dyslipidemia, or body mass index with any of the neural biomarkers (Table 2).

Correlations between physical activity across the life course and neural biomarkers

There were no statistically significant correlations between the physical activity measures during school-age, adolescence, or middle adulthood and any of the neural biomarkers (Table 2). During young-adulthood, there was a statistically significant correlation between moderate-intensity occupation/school with NfL (r = −0.268, p = 0.022) and α-synuclein (r = −0.276, p = 0.018), although none of the correlations remained significant after the Benjamini-Hochberg Procedure.

Logistic regression

Young adult occupation/school physical activity was a significant predictor of NfL (OR = 0.969, 95% CI = 0.940–0.999) while controlling for age (see Table 3). (Age was controlled for since it was significantly correlated with NfL.) The full model was significant, χ² (df = 2, n = 73) = 15.57, p < 0.001, explaining between 19.2% (Cox & Snell R square) and 25.6% (Nagelkerke R Square) of the variance in NfL, correctly classifying 71.2% of cases.

Table 3.

Logistic regression predicting likelihood of above median amount of neural biomarkers.

Model	Predictor Variables	B	SE	Wald	df	p	Odds Ratio	95% CI for Odds Ratio
	NfL Regressed on Two Predictors (n = 73)
1	Young Adult Occupation/school moderate-intensity PA	−0.03	0.02	4.07	1	0.044	0.969	0.940–0.999
	Age	0.13	0.05	8.16	1	0.004	1.139	1.042–1.245
	Constant	−7.42	2.84	6.84	1	0.009	0.001
	α-Synuclein Regressed on One Predictor (n = 73)
2	Young Adult Occupation/school moderate-intensity PA	−0.03	0.02	5.12	1	0.024	0.968	0.940–0.996
	Constant	0.70	0.37	3.59	1	0.058	2.023

PA: physical activity; NfL: neurofilament light protein.

Young adult occupation/school physical activity was a significant predictor of α-synuclein (OR = 0.968, 95% CI = 0.940–0.996). (No other variables were controlled for, since no other variables were significantly correlated with α-synuclein.) The full model was significant, χ² (df = 1, n = 73) = 6.70, p = 0.01, explaining between 8.8% (Cox & Snell R square) and 11.7% (Nagelkerke R Square) of the variance in α-synuclein, correctly classifying 50.7% of cases. Both models remained significant with the Benjamini-Hochberg Procedure.

Results were repeated with age categorized as middle versus older adults, gender, APOE4, and education as covariates. Young adult occupation/school physical activity remained a significant predictor of NfL (OR = 0.966, 95% CI = 0.935–0.998). The full model remained significant, χ² (df = 6, n = 73) = 16.61, p = 0.011, explaining between 20.4% to 27.1% of the variance in NfL, correctly classifying 69.9% of cases. The only other variable that was a significant predictor of NfL was age (OR = 5.614, 95% CI = 1.798–17.528). Young adult occupation/school physical activity remained a significant predictor of α-synuclein (OR = 0.969, 95% CI = 0.940–0.998). However, the full model was not significant, χ² (df = 6, n = 73) = 9.728, p = 0.137, explaining between 12.5 to 16.6% of the variance in α-synuclein, correctly classifying 63.37% of cases. No other variables were significant predictors of α-synuclein.

Post hoc

The frequency of each moderate-intensity occupation/school physical activity during young adulthood was examined (Table 4). Over 80% of the activities involved standing or walking.

Table 4.

Frequency of each moderate-intensity occupation/school physical activity during young adulthood.

Percent	Activity
25.2	standing tasks with light effort, such as filing or bartending
18.4	walking moderately (3.0 miles per hour [mph]) carrying light objects weighing less than 25 pounds
10.3	standing tasks with light/moderate effort, such as packing boxes or nursing patient care
9	walking briskly (3.5 mph) carrying light objects weighing less than 25 pounds
5	walking slowly (2.5 mph) carrying light objects less than 25 pounds
4	standing with moderate effort lifting items 10–20 pounds continuously with limited walking or resting
4	walking moderately (3.0 mph) in office not carrying anything
2.7	Standing, moderate effort, intermittent lifting 50 pounds, hitch/twisting ropes
2.7	Coaching, football, soccer, basketball, baseball, swimming, etc.
2.7	Walking downstairs
2.2	Custodial work, moderate effort (e.g., electric buffer, feathering arena floors, mopping, taking out trash, vacuuming)
2.2	Walking or crawling on job, brisk, walking, not carrying anything, <3.5 mph
1.3	Bicycling (as part of one's job)
1.3	Construction, outside, remodeling, new structures (e.g., roof repair, miscellaneous)
1.3	Working out (as part of job, e.g., police)
0.9	Walking moderate pace
0.9	Painting
0.9	Walking/running, playing with child(ren), moderate effort, only active periods
0.5	Walking and carrying small child, child weighing 15 pounds or more
0.5	Child care standing (e.g., dressing, bathing, grooming, feeding, occasional lifting of child), moderate effort
0.5	Elder care, disabled adult, bathing, dressing, moving into and out of bed, only active periods
0.5	Shoveling, light (less than 10 pounds/minute)
0.5	Walking or walk downstairs or standing, carrying objects about 50–74 pounds
0.5	Farming, moderate effort (e.g., feeding animals, chasing cattle by walking and/or horseback, spreading manure, harvesting crops
0.5	Walking/running, playing with child(ren), vigorous effort, only active periods
0.5	Walking, 2.8 to 3.2 mph, level, moderate pace, firm surface
0.5	Milking a cow by hand
0.5	None reported

Discussion

The purpose of this study was to determine the association between different domains of physical activity across the life course and biomarkers beyond the AT(N) framework in cognitively unimpaired adults. The current results found that more occupation/school physical activity during young adulthood significantly predicted less NfL in middle and older adulthood, suggesting a potential protective benefit against neurodegeneration. This is consistent with previous research that found regular leisure-time physical activity from adolescence to adulthood was associated with lower NfL levels in young adults.¹⁶ This also aligns with a meta-analysis indicating structured-based leisure-time physical activity significantly decreased blood NfL levels post-training in patients with multiple sclerosis, with no change in NfL levels in control arms.¹⁰ Physical activity (irrespective of domain) in older adults was associated with less NfL and higher cognition scores.⁷ Furthermore, leisure-time physical activity has been associated with a diminished rate of cognitive decline among older adults with increased serum NfL concentrations.⁴⁴ In contrast, a 16-week randomized controlled trial did not find an effect of leisure-time physical activity on NfL in individuals with clinically diagnosed mild AD.¹⁹ Since the current sample comprised cognitively unimpaired middle-aged and older adults, this may explain the contradictory findings. The current study extends the literature by examining physical activity beyond leisure-time which comprises 7% of total physical activity, compared to occupation/school which comprises 40%.¹¹

The current results did not find an association between any domain of physical activity during any developmental stage and tTau. This coincides with studies examining cardiorespiratory fitness⁴⁵ and leisure-time physical activity in cognitively unimpaired older adults⁹ and patients with AD.⁴⁶ The current results contradict a study that found vigorous physical activity in cognitively unimpaired older adults was associated with less regional tau burden measured with PET imaging.⁶ Unfortunately, in the current sample, since evidence of construct validity and test-retest reliability were not found for vigorous-intensity physical activity, or zero amounts were reported,⁴¹ vigorous-intensity physical activity was omitted from the current analysis. The current results also contradict a study that found physical activity (measured with actigraphs) in older adults free of dementia reduced the impact of CSF-measured tTau on executive function.⁴⁷ Actigraph data do not differentiate between the different domains of physical activity and were not available for the current sample during school-age, adolescence, and young adulthood. The current study contradicts a longitudinal study that found more leisure-time physical activity during the participant's 60 s was associated with less tTau and slower cognitive decline over the next 10 years.¹⁵ Since only 44% of the current sample were age 65+, the older adult developmental stage from the LTPAQ was omitted from the current analysis. The current study also contradicts the literature that found regular leisure-time physical activity from adolescence to adulthood was associated with lower tTau levels in young adults.¹⁶ Unfortunately, the current study does not have tTau levels in the participants during young adulthood.

The current results found that more occupation/school physical activity during young adulthood significantly predicted less α-synuclein in middle and older adulthood, suggesting a potential protective benefit against synaptic dysfunction. This is consistent with previous research that found higher leisure-time physical activity associated with less α-synuclein in older adults and adults independent of age,²¹ although α-synuclein in platelets did not correlate with physical activity.⁴⁸

The current results did not find an association between any domain of physical activity during any developmental stage and neurogranin, pTau₁₈₁/Aβ₄₂, YKL-40, GFAP, S100B, or STREM2. The current results are consistent with a randomized controlled trial that found 16 weeks of leisure-time physical activity was not associated with changes in neurogranin or YKL-40 levels in patients with AD.¹⁹ Physical activity (all four domains combined) has been associated with GFAP in crude models, but not when adjusting for confounders.⁷ The current results are also consistent with studies that found no association between cardiorespiratory fitness and pTau₁₈₁/Aβ₄₂⁴⁵ or physical activity (no domain mentioned) and STREM2.¹⁸ The current results contradict a 16-week randomized controlled trial in patients diagnosed with mild to moderate AD that found moderate to high-intensity exercise increased STREM2.²⁰ In addition, higher intensity exercise has been associated with elevated S100B, especially with sports which may involve head trauma,²² although a cross-sectional study in athletes found running but not cycling increased S100B, presumably resulting from skeletal muscle injury.²³ Unfortunately, in the current sample, since evidence of construct validity and test-retest reliability were not found for leisure-time physical activity,⁴¹ leisure-time physical activity was omitted from the current analysis.

Limitations and strengths

Limitations to this current study include a small, predominantly White, highly educated convenience sample without disability. Attempts should be made to replicate the current results in a larger, more diverse representative sample with disabilities, as results may differ in a less healthy sample. Individual income or household wealth were not measured, which may impact occupation/school physical activity across the life course. Physical activity was self-reported, which may be biased. However, there are no objective measures of physical activity across the life course in this sample of middle-aged and older adults. As the current study was cross-sectional, it is possible that participants with fewer “at risk” biomarkers were more likely to remember physical activity. However, the participants in this study had no cognitive symptomatology and had normal cognitive testing per the Rey Auditory Verbal Learning. Relatively small amounts of physical activity were reported during school-age, adolescence, and middle adulthood, which may explain the lack of significant findings. Future studies should be conducted with participants who report larger amounts of physical activity across the life course. Significant bivariate correlations were followed up with logistic regression, which may bias the results. Finally, CSF was sampled at an age where AD pathological biomarker signatures may not have been fully developed yet; longitudinal follow-up of these study participants is planned, with biomarkers re-assessed at more advanced ages.

The strengths of this study include examining multiple domains of physical activity across developmental stages across the life course using a physical activity measure with evidence of validity⁴⁰ and reliability.⁴¹ The current study also utilized a panel of robust exploratory prototype assays marking AD, neurodegeneration, synaptic dysfunction, and glial activation.

Conclusion

To date, most research has focused on reducing the risk of ADRD by increasing leisure-time physical activity, which only comprises 7% of total physical activity.¹¹ In contrast, the current study extends the literature by examining occupation/school which comprises 40% of total physical activity.¹¹ More occupation/school physical activity during young adulthood significantly predicted less NfL and α-synuclein in middle and older adulthood, suggesting a potential benefit against neurodegeneration and synaptic dysfunction. Most of the activities involved standing or walking. Physical activity-led workplace health promotion interventions which have been shown to increase physical activity⁴⁹ may be a method to decrease the risk of ADRD. Examples include structured sit-stand workstations, work break exercises, and redesigning office environments by strategically placing waste bins and printers to increase routine physical activity.⁴⁹ Irrespective of the method, these findings suggest that being physically active during young adulthood while at work or school, such as standing and walking, may be protective against dementia later in life.

Footnotes

Acknowledgements

COBAS and ELECSYS are trademarks of Roche. All other product names and trademarks are the property of their respective owners.

ORCID iDs

Elisa R. Torres

Kaj Blennow

Ethical considerations

Prior to recruitment, approval was obtained from the Institutional Review Board at the University of Wisconsin-Madison (2011-0907-CR002).

Consent to participate

Written informed consent was obtained.

Consent for publication

Not applicable

Author contribution(s)

Elisa R. Torres: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Visualization; Writing – original draft; Writing – review & editing.

Eloise D. Lopez: Methodology; Writing – review & editing.

Lydia McBurrows: Methodology; Writing – review & editing.

Barbara B. Bendlin: Data curation; Funding acquisition; Methodology; Project administration; Resources; Software; Supervision; Writing – review & editing.

Gwendlyn Kollmorgen: Data curation; Methodology; Resources; Writing – review & editing.

Margherita Carboni: Data curation; Methodology; Resources; Writing – review & editing.

Sterling C. Johnson: Data curation; Funding acquisition; Methodology; Resources; Software; Writing – review & editing.

Henrik Zetterberg: Data curation; Funding acquisition; Methodology; Resources; Software; Writing – review & editing.

Kaj Blennow: Data curation; Funding acquisition; Methodology; Resources; Software; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Elisa Torres was supported by grants from the Alzheimer's Association (grant #AARGD-NTF-21-848187); Michigan Alzheimer's Disease Research Center Early Career Investigator Mentorship Program (grant #P30AG072931); Clinical and Translational Science Award program through the NIH National Center for Advancing Translational Sciences (grant #UL1TR000427 & #KL2TR000428); and the Mississippi Center for Clinical and Translational Research (grant #5U54GM115428). Barbara Bendlin was supported by a grant from the National Institute on Aging (grant #R01AG037639). Sterling Johnson was supported by a grant from the National Institute on Aging (grant #P50AG033514). Henrik Zetterberg is a Wallenberg Scholar supported by grants from the Swedish Research Council (grant #2022-01018 and #2019-02397); European Union's Horizon Europe research and innovation program (grant #101053962); Swedish State Support for Clinical Research (grant #ALFGBG-71320); Alzheimer Drug Discovery Foundation (ADDF), USA (grant #201809-2016862); AD Strategic Fund and Alzheimer's Association (grant #ADSF-21-831376-C, #ADSF-21-831381-C, and #ADSF-21-831377-C); Bluefield Project; Olav Thon Foundation; Erling-Persson Family Foundation; Stiftelsen för Gamla Tjänarinnor; Swedish Brain Foundation (Hjärnfonden), (grant #FO2022-0270); European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement (grant #860197) (MIRIADE); European Union Joint Programme – Neurodegenerative Disease Research (grant #JPND2021-00694); the National Institute for Health and Care Research University College London Hospitals Biomedical Research Centre; and the UK Dementia Research Institute at UCL (grant #UKDRI-1003). Kaj Blennow is supported by the Swedish Research Council (grant #2017-00915 and #2022-00732); Swedish Alzheimer Foundation (grant #AF-930351, #AF-939721 and #AF-968270); Swedish Brain Foundation (Hjärnfonden), Sweden (grant #FO2017-0243 and #ALZ2022-0006); Swedish state under the agreement between the Swedish government and the County Councils, the ALF-agreement (grant #ALFGBG-715986 and #ALFGBG-965240); European Union Joint Program for Neurodegenerative Disorders (grant #JPND2019-466-236); Alzheimer's Association 2021 Zenith Award (grant #ZEN-21-848495); and the Alzheimer's Association 2022-2025 Grant (grant #SG-23-1038904 QC). The funding sources had no role in the study design, collection, analysis, interpretation of data, writing of the report, or the decision to submit for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Drs. Bendlin and Zetterberg are Editorial Board Members of this journal but were not involved in the peer-review process of this article nor had access to any information regarding its peer-review.

Dr Zetterberg has served at scientific advisory boards and/or as a consultant for Abbvie, Acumen, Alector, Alzinova, ALZPath, Annexon, Apellis, Artery Therapeutics, AZTherapies, Cognito Therapeutics, CogRx, Denali, Eisai, Nervgen, Novo Nordisk, Optoceutics, Passage Bio, Pinteon Therapeutics, Prothena, Red Abbey Labs, reMYND, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave, has given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure, Biogen, and Roche, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program (outside submitted work).

Dr Blennow has served as a consultant and on advisory boards for Acumen, ALZPath, BioArctic, Biogen, Eisai, Lilly, Moleac Pte. Ltd, Novartis, Ono Pharma, Prothena, Roche Diagnostics, and Siemens Healthineers; has served at data monitoring committees for Julius Clinical and Novartis; has given lectures, produced educational materials and participated in educational programs for AC Immune, Biogen, Celdara Medical, Eisai and Roche Diagnostics; and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program, outside the work presented in this paper.

Data availability statement

The data supporting the findings of this study are available on request from .

References

World Health Organization. Dementia, https://www.who.int/news-room/fact-sheets/detail/dementia (2025, accessed 18 August 2025).

Alzheimer’s Association. 2025 Alzheimer’s disease facts and figures. Alzheimers Dement 2025; 21: 7.

Jack

Lowe

Weigand

, et al. Serial PiB and MRI in normal, mild cognitive impairment and Alzheimer’s disease: implications for sequence of pathological events in Alzheimer’s disease. Brain 2009; 132: 1355–1365.

Livingston

Huntley

Sommerlad

, et al. Dementia prevention, intervention, and care: 2020 report of the lancet commission. Lancet 2020; 396: 413–446.

Brown

Peiffer

Taddei

, et al. Physical activity and amyloid plasma and brain levels: results from the Australian imaging, biomarkers and lifestyle study of ageing. Mol Psychiatry 2013; 18: 875–881.

Brown

Rainey-Smith

Dore

, et al. Self-reported physical activity is associated with tau burden measured by positron emission tomography. J Alzheimers Dis 2018; 63: 1299–1305.

Kim

Shin

Ham

, et al. Physical activity, Alzheimer plasma biomarkers, and cognition. JAMA Netw Open 2025; 8: e250096.

Craig

Marshall

Sjöström

, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003; 35: 1381–1395.

Aslanyan

Ortega

Fenton

, et al. Protective effects of sleep duration and physical activity on cognitive performance are influenced by β-amyloid and brain volume but not tau burden among cognitively unimpaired older adults. Neuroimage Clin 2023; 39: 103460.

10.

Salamatullah

Alkhiri

Ezzi

, et al. The interaction between exercise and neurofilament light chain in multiple sclerosis: a systematic review and meta-analysis. Mult Scler Relat Disord 2024; 90: 105809.

11.

Welk

Kim

. The context of physical activity in a representative sample of adults. Med Sci Sports Exerc 2015; 47: 2102–2110.

12.

Bateman

Xiong

Benzinger

, et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med 2012; 367: 795–804.

13.

Villemagne

Burnham

Bourgeat

, et al. Amyloid β deposition, neurodegeneration, and cognitive decline in sporadic Alzheimer’s disease: a prospective cohort study. Lancet Neurol 2013; 12: 357–367.

14.

Liang

Mintun

Fagan

, et al. Exercise and Alzheimer's disease biomarkers in cognitively normal older adults. Ann Neurol 2010; 68: 311–318.

15.

Desai

Evans

Dhana

, et al. Longitudinal association of total tau concentrations and physical activity with cognitive decline in a population sample. JAMA Netw Open 2021; 4: e2120398.

16.

Ezzatvar

López-Gil

Yáñez-Sepúlveda

, et al. Physical activity is related to lower levels of neurofilament light and plasma tau protein: a 22-year follow-up study from adolescence to adulthood. J Adolesc Health 2025; 76: 1055–1061.

17.

Kapasi

DeCarli

Schneider

. Impact of multiple pathologies on the threshold for clinically overt dementia. Acta Neuropathol 2017; 134: 171–186.

18.

Wang

Sheng

, et al. Associations among healthy lifestyle characteristics, neuroinflammation, and cerebrospinal fluid core biomarkers of Alzheimer's disease in cognitively intact adults: the CABLE study. J Alzheimers Dis 2024; 102: 855–865.

19.

Jensen

Portelius

Høgh

, et al. Effect of physical exercise on markers of neuronal dysfunction in cerebrospinal fluid in patients with Alzheimer's disease. Alzheimers Dement (NY) 2017; 3: 284–290.

20.

Jensen

Bahl

Østergaard

, et al. Exercise as a potential modulator of inflammation in patients with Alzheimer's disease measured in cerebrospinal fluid and plasma. Exp Gerontol 2019; 121: 91–98.

21.

Daniele

Pietrobono

Fusi

, et al. α-Synuclein aggregates with β-amyloid or tau in human red blood cells: correlation with antioxidant capability and physical exercise in human healthy subjects. Mol Neurobiol 2018; 55: 2653–2675.

22.

Koh

Lee

. S100b as a marker for brain damage and blood-brain barrier disruption following exercise. Sports Med 2014; 44: 369–385.

23.

Stocchero

Oses

Cunha

, et al. Serum S100B level increases after running but not cycling exercise. Appl Physiol Nutr Metab 2014; 39: 340–344.

24.

Johnson

Koscik

Jonaitis

, et al. The Wisconsin registry for Alzheimer's prevention: a review of findings and current directions. Alzheimers Dement (Amst) 2017; 10: 130–142.

25.

Friedenreich

Courneya

Bryant

. The lifetime total physical activity questionnaire: development and reliability. Med Sci Sports Exerc 1998; 30: 266–274.

26.

Friedenreich

Courneya

Bryant

. Relation between intensity of physical activity and breast cancer risk reduction. Med Sci Sports Exerc 2001; 33: 1538–1545.

27.

Friedenreich

Gregory

Kopciuk

, et al. Prospective cohort study of lifetime physical activity and breast cancer survival. Int J Cancer 2009; 124: 1954–1962.

28.

Tam

Friedenreich

, et al. Lifetime physical activity in postmenopausal Caucasian and Chinese-Canadian women. Eur J Cancer Prev 2014; 23: 90–95.

29.

Friedenreich

Biel

Lau

DCW

, et al. Case-control study of the metabolic syndrome and metabolic risk factors for endometrial cancer. Cancer Epidemiol Biomarkers Prev 2011; 20: 2384–2395.

30.

Friedenreich

Cook

Magliocco

, et al. Case-control study of lifetime total physical activity and endometrial cancer risk. Cancer Causes Control 2010; 21: 1105–1116.

31.

Friedenreich

Cook

Wang

, et al. Prospective cohort study of pre- and postdiagnosis physical activity and endometrial cancer survival. J Clin Oncol 2020; 38: 4107–4117.

32.

Friedenreich

McGregor

Courneya

, et al. Case-control study of lifetime total physical activity and prostate cancer risk. Am J Epidemiol 2004; 159: 740–749.

33.

Torres

Bendlin

Kassahun-Yimer

, et al. Transportation physical activity earlier in life and areas of the brain related to dementia later in life. J Transp Health 2021; 20: 100992.

34.

Torres

Hoscheidt

Bendlin

, et al. Lifetime physical activity and white matter hyperintensities in cognitively intact adults. Nurs Res 2019; 68: 210–217.

35.

Sohn

Byun

, et al. Late-life physical activities moderate the relationship of amyloid-β pathology with neurodegeneration in individuals without dementia. J Alzheimers Dis 2022; 86: 441–450.

36.

Gill

Friedenreich

Sajobi

, et al. Association between lifetime physical activity and cognitive functioning in middle-aged and older community dwelling adults: results from the brain in motion study. J Int Neuropsychol Soc 2015; 21: 816–830.

37.

Boisvert-Vigneault

Payette

Audet

, et al. Relationships between physical activity across lifetime and health outcomes in older adults: results from the NuAge cohort. Prev Med 2016; 91: 37–42.

38.

Ainsworth

Haskell

Herrmann

, et al. 2011 Compendium of physical activities: a second update of codes and MET values. Med Sci Sports Exerc 2011; 43: 1575–1581.

39.

Erikson

. The life cycle completed. New York: WW Norton, 1998.

40.

Torres

. Validation of the lifetime total physical activity questionnaire (LTPAQ) in midlife and older adults with a history of late-onset depression. Arch Psychiatr Nurs 2018; 32: 580–584.

41.

Torres

Lopez-Lambert

Bendlin

. Psychometrics of the lifetime total physical activity questionnaire in adults with a history of depression. J Alzheimers Dis 2025; 105: 666–676.

42.

Van Hulle

Jonaitis

Betthauser

, et al. An examination of a novel multipanel of CSF biomarkers in the Alzheimer's disease clinical and pathological continuum. Alzheimers Dement 2021; 17: 431–445.

43.

Zetterberg

. Biofluid-based biomarkers for Alzheimer's disease-related pathologies: an update and synthesis of the literature. Alzheimers Dement 2022; 18: 1687–1693.

44.

Desai

Dhana

DeCarli

, et al. Examination of neurofilament light chain serum concentrations, physical activity, and cognitive decline in older adults. JAMA Netw Open 2022; 5: e223596.

45.

Schultz

Boots

Darst

, et al. Cardiorespiratory fitness alters the influence of a polygenic risk score on biomarkers of AD. Neurology 2017; 88: 1650–1658.

46.

Steen Jensen

Portelius

Siersma

, et al. Cerebrospinal fluid amyloid beta and tau concentrations are not modulated by 16 weeks of moderate- to high-intensity physical exercise in patients with Alzheimer disease. Dement Geriatr Cogn Disord 2016; 42: 146–158.

47.

Dougherty

Soldan

Pettigrew

, et al. Physical activity modifies associations between cerebrospinal fluid tau measures and executive function. Alzheimers Dement (NY) 2025; 11: e70085.

48.

Daniele

Pietrobono

Fusi

, et al. α-Synuclein aggregated with tau and β-amyloid in human platelets from healthy subjects: correlation with physical exercise. Front Aging Neurosci 2018; 10: 17.

49.

Zhang

Nie

Peng

, et al. Effectiveness of physical activity-led workplace health promotion interventions: a systematic review. Healthcare (Basel) 2025; 13: 1292.