Patterns of Regional Cerebral Blood Flow as a Function of Age Throughout the Lifespan

Abstract

Background:

Understanding the influence of aging on the brain remains a challenge in determining its role as a risk factor for Alzheimer’s disease.

Objective:

To identify patterns of aging in a large neuroimaging cohort.

Methods:

A large psychiatric cohort of 31,227 individuals received brain SPECT at rest and during a concentration task for a total of 62,454 scans. ANOVA was done to identify the mean age trends over the course of the age range in this group, 0–105 years. A regression model in which brain SPECT regions of interest was used to predict chronological age (CA) was then utilized to derive brain estimated age (BEA). The difference between CA and BEA was calculated to determine increased brain aging in common disorders in our sample such as depression, dementia, substance use, and anxiety.

Results:

Throughout the lifespan, variations in perfusion were observed in childhood, adolescence, and late life. Increased brain aging was seen in alcohol use, cannabis use, anxiety, bipolar, schizophrenia, attention-deficit/hyperactivity disorder, and in men.

Conclusion:

Brain SPECT can predict chronological age and this feature varies as a function of common psychiatric disorders.

Keywords

Aging alcohol use attention-deficit/hyperactivity disorder bipolar brain SPECT cannabis use schizophrenia

INTRODUCTION

Alzheimer’s disease (AD) is the most common cause of dementia and age is recognized as the most powerful risk factor for the disorder [1]. Understanding the relationship between age and risk of AD has been attempted in multiple studies that have attempted to better characterize the phenomenon in humans with neuroimaging. Previous work has shown that age and AD have distinct main effects on the brain on magnetic resonance imaging (MRI) scans that, while distinct from AD, can overlap in the same areas of the brain such as the hippocampus [2]. These changes manifest on structural MRI scans as volume loss over time. Regional cerebral blood flow has also been used to track brain changes with aging. An early study of 27 subjects aged 19–76 showed age related decline in resting regional cerebral blood flow in frontal, temporo-sylvian, and parieto-occipital cortex [3]. Another positron emission tomography (PET) study with the O15 tracer study of 30 normal healthy volunteers age 30–85 showed age dependent decreases in regional cerebral blood flow in the cingulate, parahippocampal, superior temporal, medial frontal and posterior prefrontal, and posterior parietal cortices bilaterally, and in the left insula [4]. For brain single-photon emission computerized tomography (SPECT) using CERETEC, age predicted cerebral blood flow decline was detected on a voxel wise analysis in the insula, temporal lobes, frontal lobes, and parahippocampal gyrus in 52 healthy mean age 18–86 years [5]. A larger study in 309 subjects age 20–89 years old identified the frontal lobes as most closely linked with aging [6]. While many studies show decreases of cerebral blood flow in relation to aging, a systematic review identified that increased brain activity with aging can occur and this is thought to be a compensatory change [7] that can also be seen in the setting of AD [8].

While these studies have been important in establishing basic patterns of correlation seen with aging, they lack several key features. First, the sample sizes of such investigations are typically small decreasing their generalizability. Second, the age range of such studies while large typically do not encompass the full lifespan defined as birth to extreme old age of 90 or higher. Third, while studying the question of age-related brain imaging biomarkers in healthy individuals is important there is considerable value in studying this same question in common psychiatric conditions such as depression which itself is a separate risk factor for AD [9].

Better understanding of the relationship between aging and the brain is therefore needed to determine AD risk. Functional neuroimaging remains an important tool for identifying age related changes [10]. The purpose of this study is to therefore 1) characterize age related perfusion changes cross the lifespan; 2) identify the proportion of variance in chronological age accounted by regional cerebral blood flow; and 3) quantify the difference between chronological age (CA) from brain estimated age (BEA) from functional neuroimaging. We hypothesize that across the lifespan, regional cerebral blood flow will powerfully predict chronological age and will vary as a function of common psychiatric brain disorders.

MATERIALS AND METHODS

Subjects

Subjects were drawn from multiple branches of the Amen Clinics as described in prior work [11]. IRB approval for retrospective analysis of de-identified clinical and SPECT scan data was provided by accredited institutional review board, IntegReview (IRB# 004; http://www.integreview.com/). Inclusion criteria was widely broad to encompass the largest number of subjects for analysis for aging patterns across the lifespan in health and a variety of diseases. Subject demographics are detailed in Table 1.

Table 1

Subject demographics (Total n = 31,227)

Variables	Statistics (χ±σ, Range, Percent (n))
Age	34.2±18.6, Age 0–105
Age Group (Pediatric/Adult/Geriatric)	23.3 (7277)/70.3 (21957)/6.4 (1993)
Gender (Male/Female)	60.6 (18925)/39.4 (12296)
All Cause Dementia (Including AD)	5.2 (1622)
Attention-Deficit/Hyperactivity Disorder	4.7 (1462)
Major Depression	15.2 (4753)
Bipolar Disorder	8.1 (2524)
Generalized Anxiety Disorder	20.7 (6457)
Traumatic Brain Injury	27.1 (8472)
Schizophrenia	2.1 (657)
Alcohol Use Disorder	4.7 (1457)
Cannabis Use Disorder	3.5 (1089)

Brain SPECT imaging

All subjects received intravenous administration of an age and weight-appropriate dose of technetium-99 m hexamethylpropylene amine oxime (99mTc-HMPAO) for brain SPECT imaging. Each subject received a resting, or baseline, scan and a task or concentration scan on different days. For baseline scans subjects were injected while sitting quietly in same setting with eyes open. Subjects were then scanned 30 minutes later using a high-resolution Picker Prism 3000 triple-headed gamma camera with fan beam collimators, acquiring data in 128×128 matrices, yielding 120 images per scan with each image separated by 3 degrees spanning 360 degrees. SPECT data was processed and attenuation correction performed using general linear (Chang) method for attenuation correction. Brain SPECT images were then reconstructed and resliced according to anterior-posterior commissure line so final images were similarly aligned for analysis. Cerebral perfusion was then estimated using on a region of interest basis using areas derived from the automated anatomical labeling atlas (AAL) [12]. As detailed in prior work [13], ROI Counts in each region of interest (ROI) were quantified using trimmed means. Calculation of these trimmed means was done using all scores in a 98% confidence interval (–2.58 <Z < –2.58). Perfusion for each region was then estimated with the trimmed mean using the following formula. T = 10*((subject ROI mean–trimmed regional avg)/trimmed regional stdev) + 50.

Statistical analysis

All statistical analyses were conducted using SPSS (Version 24, IBM, Armonk, NY). First, the relationship between age and regional cerebral blood flow was evaluated using on way ANOVA at both baseline and concentration tasks. Second, a linear regression model was constructed with chronological age as the response variable and the baseline brain SPECT region of interest blood flow metrics as the predictor variables. The beta coefficients from this model represent the predicted age for the brain SPECT regions, or the brain estimated age (BEA). The BEA was then subtracted from the patient’s chronological age (CA). A negative value was interpreted as increased aging of the brain compared to chronological age. A positive value was interpreted as decreased aging of the brain compared to chronological age. Two sample t-tests were then conducted to compare the difference between BEA and CA in 1) men compared to women, 2) anxiety disorder, 3) dementia, 4) attention-deficit/hyperactivity disorder (ADHD), 5) major depressive disorder, 6) bipolar disorder, 7) substance abuse disorder, 8) alcohol abuse, 9) cannabis use disorder, 10) obsessive compulsive disorder, and 11) traumatic brain injury. p-values <0.050 were considered statistically significant.

RESULTS

ANOVA results are detailed in the Supplementary Material. In summarizing this data, several trends emerge. First, multiple brain regions such as the anterior cingulate gyrus, inferior frontal operculum, orbital frontal cortex, frontal trigone, fusiform gyrus, the hippocampus, insula, olfactory gyrus, paracentral lobule, inferior parietal cortex, and pre and post-central gyri show increases in perfusion from birth to the adolescent years. This is then followed by a decrease in perfusion and a plateau in the remaining teenage years followed by an increase or plateau into the 60 s. This is followed by fluctuations in perfusion in late life characterized by both increases and decreased in regional cerebral blood flow.

Brain SPECT perfusion regions of interest at rest alone accounted for approximately 73% of the variance in reported chronological age (R² = 0.727). This relationship is shown in Fig. 1.

Fig.1

This figure shows the relationship between chronological age predicted by brain SPECT (brain estimated age) and reported chronological age.

Table 2 shows the comparisons between BEA and CA across gender and the major co-morbid conditions represented in this sample. The bivariate Pearson correlation between BEA and CEA was r = 0.85 (p < 0.001). Areas in bold are those in which BEA is higher than CA resulting in a negative value.

Table 2

Comparison of brain estimated age and chronological age

Group	BEA (χ±σ)	CA (χ±σ)	CA-BEA (χ±σ)
Men/Women	31.9±14.8/	31.5±17.9/	–0.43±9.27/
	37.79±16.3	38.2±18.9	0.49±10.3
Dementia	50.3±18.7	54.3±20.7	4.1±12.1
ADHD	22.6±13.3	21.2±15.5	–1.4±8.2
Depression	37.6±14.6	38.4±17	0.85±9.7
Bipolar	33.5±14.1	31.9±16.1	–1.65±9.4
Cannabis Use Disorder	29.4±9.8	26.6±10.4	–2.8±7.3
Alcohol Use Disorder	39±13.5	38.4±14.6	–0.6±9.4
Anxiety Disorder	33.1±14.9	32.6±17.3	–0.5±9.1
Traumatic Brain Injury	35.3±17.7	35.3±15.1	0.19±19.6
Schizophrenia	34.5±15.7	30.5±13.7	–4±9.3

DISCUSSION

This work has demonstrated the relationships between age and regional cerebral perfusion on brain SPECT across the lifespan from age 0–105 in the largest known imaging sample of 31,277 subjects for a total of 62,454 scans. Variations in blood flow and tracer uptake (as the retention of tracer is a dual process of blood flow and brain tissue extraction) across the lifespan demonstrated multiple areas with increase in earlier childhood years followed by a decline and plateau in the teenage years. Additional variations were observed in late life with both increases and decreases. Brain SPECT regions of interest alone accounted for approximately 73% of variance and chronological age, suggesting that perfusion imaging can predict chronological age. Additionally, this relationship varied as a function of common psychiatric conditions with several conditions showing higher BEA compared to CA and some showing lower BEA compared to CA. As such, several conditions show “older” appearing brains compared to persons while other afflictions did not appear to be characterized by accelerated aging.

The concept that perfusion and uptake can vary in early years of life and then decrease in the teenage years has been reflected in prior work. For example, one study using brain FDG PET showed that overall brain metabolism in children rises to twice that of adults by age 4-5 and remains high until about 9-10 years of age [14]. We observed this trends in multiple areas including bilateral angular, cingulate, hippocampi, and insula. The subsequent decrease and plateau in perfusion we observed in the teenage years is likely secondary to synaptic pruning [15] in which relative cerebral perfusion and uptake decreases were observed on ASL MRI in a smaller sample of 39 healthy participants age 7–17. Our larger pediatric sample of 7,277 participants further extends this work.

The plateau effect of perfusion in the 20 s and 30 s with some mild decrease further reflects prior work on a cohort of 232 subjects that received phase contrast flow velocity MRI [16]. The variation in perfusion observed in late life is independently confirmed by functional MRI literature in which both increases and decreases are notable in multiple networks such as the default mode network [17]. The additional variance observed may also vary as a function in the multiple psychiatric co-morbidities in our population.

The concept of accelerated aging refers to the morphological and functional appearance of the brain appearing smaller than expected for patient age and this has been typically described on longitudinal MR imaging studies such as with hypertension [18]. In our study, we inferred the presence of accelerated brain aging by the results of negative values between CA and BEA. Accelerated brain aging has been described in prior work by demonstrating magnitude of brain atrophy out of proportion to chronological age in different psychiatric and neurological disorders and has also been observed in persons who have undergone stressful life events [19, 20]. In applying this concept to different common psychiatric co-morbidities in our study, we confirmed as has been shown in these other studies [19, 21] that schizophrenia is associated with accelerated brain aging with SPECT perfusion predicting BEA as four years older compared to CA. Thus, with SPECT imaging, accelerated brain aging is denoted by magnitude of lower regional cerebral perfusion out of proportion of patient chronological age as a function of different psychiatric and neurological disorders. The main implication is that there can consequently be multi-modal imaging biomarkers of accelerated neural aging including regional cerebral blood flow imaging with brain SPECT. Accelerated aging has also observed in alcohol use disorder and this has also been quantified in prior voxel-based MR imaging study of 119 persons with alcohol use disorder compared to 97 controls [22]. We also observed accelerated brain aging in cannabis use disorder which further expands on our prior work of widespread perfusion decreases in this population [23]. The finding of accelerated aging in our population is also compatible with a theoretical framework of this disorder as a phenomenon of accelerated aging [24]. We also found ADHD has accelerated aging and to our knowledge this has not been reported in the literature before. The significance is therefore unclear but longitudinal studies may better elucidate this relationship. The increased brain aging in men compared to women reflects our earlier study showing the increased resting perfusion in women versus men [25]. The increased brain aging with anxiety is also supported by prior work mainly thought to act by mechanism of impaired hippocampal neurogenesis [26].

Interestingly, we did not observe accelerated aging in dementia. However, given the regional specificity of various dementias such as temporal parietal loci of pathology in AD, this disorder is actually not thought to be a one of accelerated aging [27]. These findings may also be a function of disease duration which, while not in the scope of this current study, would be a valuable variable to investigate in future work. Also, while traumatic brain injury is also characterized by atrophy, this is also not thought to necessarily be a disorder of increased brain aging and our results also support this assertion [28]. While depression has been associated with accelerated aging [29], we did not observe this in our study. This could be due to the confounding effects of disease duration and medication effects which were not separately modeled in this work.

Strengths of this study were its large sample size, data on multiple co-morbidities, and availability of quantitative functional neuroimaging data. However, caveats include potential confounding from disease duration and medications regimens. An additional weakness is lack of longitudinal data.

However, the support of our findings with multiple studies in the prior literature even across modalities suggests that our findings can be applied in a variety of settings. Implications of these findings are applications to prevention programs for diseases that can impair cognition such as drugs and alcohol. Additionally, a unique feature of our study is the estimation of brain age in such a large sample. This has implications for using a simple brain SPECT scan to predict brain age and, by comparing it to chronological age, determine if a patient’s brain is undergoing accelerated aging. Such information is actionable for patients to further enroll them earlier in programs for prevention and management of cognitive decline [9, 30, 31 , 9, 30, 31].

DISCLOSURE STATEMENT

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/18-0598r1).

Footnotes

The supplementary material is available in the electronic version of this article: .

References

Ganguli

, Rodriguez

(2011) Age, Alzheimer’s disease, and the big picture. Int Psychogeriatr 23, 1531–1534.

Raji

, Lopez

, Kuller

, Carmichael

, Becker

(2009) Age, Alzheimer disease, and brain structure. Neurology 73, 1899–1905.

Pantano

, Baron

, Lebrun-Grandié

, Duquesnoy

, Bousser

, Comar

(1984) Regional cerebral blood flow and oxygen consumption in human aging. Stroke 15, 635–641.

Martin

, Friston

, Colebatch

, Frackowiak

RSJ

(1991) Decreases in regional cerebral blood flow with normal aging. J Cereb Blood Flow Metab 11, 684–689.

Matsuda

, Ohnishi

, Asada

, Li

, Kanetaka

, Imabayashi

, Tanaka

, Nakano

(2003) Correction for partial-volume effects on brain perfusion SPECT in healthy men. J Nucl Med 44, 1243–152.

De Vis

, Peng

S-L

, Chen

, Li

, Liu

, Sur

, Rodrigue

, Park

, Lu

(2018) Arterial-spin-labeling (ASL) perfusion MRI predicts cognitive function in elderly individuals: A 4-year longitudinal study: CBF predicts cognition in healthy aging., J Magn Reson Imaging, doi: 10.1002/jmri.25938

Eyler

, Sherzai

, Kaup

, Jeste

(2011) A review of functional brain imaging correlates of successful cognitive aging. Biol Psychiatry 70, 115–122.

Alsop

, Casement

, de Bazelaire

, Fong

, Press

(2008) Hippocampal hyperperfusion in Alzheimer’s disease. Neuroimage 42, 1267–1274.

Barnes

, Yaffe

(2011) The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol 10, 819–828.

10.

Hays

, Zlatar

, Wierenga

(2016) The utility of cerebral blood flow as a biomarker of preclinical Alzheimer’s disease. Cell Mol Neurobiol 36, 167–179.

11.

Amen

, Raji

, Willeumier

, Taylor

, Tarzwell

, Newberg

, Henderson

(2015) Functional neuroimaging distinguishes posttraumatic stress disorder from traumatic brain injury in focused and large community datasets. PLoS One 10, e0129659.

12.

Tzourio-Mazoyer

, Papathanassiou

, Crivello

, Etard

, Delcroix

, Mazoyer

, Joliet

(2002) Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. NeuroImage 15, 273–289.

13.

Amen

, Taylor

, Meysami

, Raji

(2018) Deficits in regional cerebral blood flow on brain SPECT predict treatment resistant depression. J Alzheimers Dis 63, 529–538.

14.

Tau

, Peterson

(2010) Normal development of brain circuits. Neuropsychopharmacology 35, 147–168.

15.

Kilroy

, Liu

, Yan

, Kim

, Dapretto

, Mendez

, Wang

DJJ

(2011) Relationships between cerebral blood flow and IQ in typically developing children and adolescents. J Cogn Sci 12, 151–170.

16.

, Xu

, Rodrigue

, Kennedy

, Cheng

, Flicker

, Hebrank

, Uh

, Park

(2011) Alterations in cerebral metabolic rate and blood supply across the adult lifespan. Cereb Cortex 21, 1426–1434.

17.

Grady

(2012) The cognitive neuroscience of ageing. Nat Rev Neurosci 13, 491–505.

18.

Tzourio

, Laurent

, Debette

(2014) Is hypertension associated with an accelerated aging of the brain? , Hypertension 63, 894–903.

19.

Nenadić

, Dietzek

, Langbein

, Sauer

, Gaser

(2017) BrainAGE score indicates accelerated brain aging in schizophrenia, but not bipolar disorder. Psychiatry Res Neuroimaging 266, 86–89.

20.

Hatton

, Franz

, Elman

, Panizzon

, Hagler

, Fennema-Notestine

, Eyler

, McEvoy

, Lyons

, Dale

, Kremen

(2018) Negative fateful life events in midlife and advanced predicted brain aging. Neurobiol Aging 67, 1–9.

21.

Schnack

, van Haren

NEM

, Nieuwenhuis

, Hulshoff Pol

, Cahn

, Kahn

(2016) Accelerated brain aging in schizophrenia: A longitudinal pattern recognition study. Am J Psychiatry 173, 607–616.

22.

Guggenmos

, Schmack

, Sekutowicz

, Garbusow

, Sebold

, Sommer

, Smolka

, Wittchen

H-U

, Zimmermann

, Heinz

, Sterzer

(2017) Quantitative neurobiological evidence for accelerated brain aging in alcohol dependence. Transl Psychiatry 7, 1279.

23.

Amen

, Darmal

, Raji

, Bao

, Jorandby

, Meysami

, Raghavendra

(2017) Discriminative properties of hippocampal hypoperfusion in marijuana users compared to healthy controls: Implications for marijuana administration in Alzheimer’s dementia. J Alzheimers Dis 56, 261–273.

24.

Rizzo

, Costa

, Mansur

, Swardfager

, Belangero

, Grassi-Oliveira

, McIntyre

, Bauer

, Brietzke

(2014) The theory of bipolar disorder as an illness of accelerated aging: Implications for clinical care and research. Neurosci Biobehav Rev 42, 157–169.

25.

Amen

, Trujillo

, Keator

, Taylor

, Willeumier

, Meysami

, Raji

(2017) Gender-based cerebral perfusion differences in 46,034 functional neuroimaging scans. J Alzheimers Dis 60, 605–614.

26.

Perna

, Iannone

, Alciati

, Caldirola

(2016) Are anxiety disorders associated with accelerated aging? A focus on neuroprogression. Neural Plast 2016, 1–19.

27.

Ghosh

, Haggerty

, Agarwal

(2011) Alzheimer’s disease - Not an exaggeration of healthy aging. Indian J Psychol Med 33, 106.

28.

Ledig

, Kamnitsas

, Koikkalainen

, Posti

, Takala

RSK

, Katila

, Frantzén

, Ala-Seppälä

, Kyllönen

, Maanpää

H-R

, Tallus

, Lötjönen

, Glocker

, Tenovuo

, Rueckert

(2017) Regional brain morphometry in patients with traumatic brain injury based on acute- and chronic-phase magnetic resonance imaging.PLoS One 12, e0188152.

29.

Wolkowitz

, Reus

, Mellon

(2011) Of sound mind and body: Depression, disease, and accelerated aging. Dialogues Clin Neurosci 13, 25–39.

30.

Bredesen

(2014) Reversal of cognitive decline: A novel therapeutic program. Aging 6, 707–717.

31.

Ngandu

, Lehtisalo

, Solomon

, Levälahti

, Ahtiluoto

, Antikainen

, Bäckman

, Hänninen

, Jula

, Laatikainen

, Lindström

, Mangialasche

, Paajanen

, Pajala

, Peltonen

, Rauramaa

, Stigsdotter-Neely

, Strandberg

, Tuomilehto

, Soininen

, Kivipelto

(2015) A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): A randomised controlled trial. Lancet 385, 2255–2263.