The relationship between cardiovascular risk and lifestyle activities on hippocampal volumes in normative aging

Abstract

Background

Despite the life-course perspective of popular aging models, few studies on healthy aging to date have examined both younger and older adulthood. The current study examined how cumulative vascular risk factors and self-reported levels of physical, social, and cognitive activity are associated with differences in hippocampal volumes in healthy younger and older adults.

Methods

34 neurologically healthy participants were separated into two age cohorts: a younger adult group (age 25–35, n = 17) and an older adult group (age 65–82, n = 17). Participants underwent a 3 T T1 MRI and completed a series of questionnaires. Voxel-based morphometry examined whole-brain grey matter density differences between groups. Hippocampal volumes were computed. Analyses examined the association between hippocampal volumes, cumulative vascular risk, and self-reported levels of physical, social, and cognitive activity, both within and across groups.

Results

Between-group comparisons revealed greater cortical atrophy in older relative to young adults in regions including the left and right hippocampus and temporal fusiform cortex. Across-group analyses revealed a significant negative association between cardiovascular risk scores and bilateral hippocampal volumes across age groups. A significant negative association was identified between frequency of social activities and bilateral hippocampal volumes in older adults only. No significant associations were found between left or right hippocampal volumes and total, cognitive, or physical activities in both within- and across-group analyses.

Conclusion

Greater cumulative vascular risk is associated with smaller hippocampal volumes across age cohorts. Findings suggest that social activities with low cognitive load may not be beneficial to structural brain outcomes in older age.

Introduction

In the last few decades, modern neuroimaging techniques have allowed for unprecedented advances in our understanding of structural changes in the brain that typify the aging process. Many of these research efforts have been devoted to identifying changes associated with age-related neurodegenerative diseases and their cognitive and physiological antecedents. Increasingly, there has been a focus on the study of successful aging, and the factors that contribute to, or detract from, cognitive health in later life. Contemporary models on aging, such as the revised Scaffolding Theory of Aging and Cognition (STAC-r; [32] have emphasized a life-course perspective, by which both endogenous and exogenous factors interact throughout the lifespan to impact structural and functional brain development. According to these models, external factors such as cardiovascular health, and psychosocial status, have the potential to significantly shape the brain through enrichment, or depletion, of neural resources [32].

In particular, vascular risk factors such as obesity, diabetes, smoking, hypertension, and hypercholesterolemia have been highlighted in several studies as key factors that contribute to faster rates of cognitive decline in later life and, ultimately, an increased risk of dementia [14], [37], [38]. Conversely, findings from several longitudinal studies have shown that engaging in regular physical activity, and particularly cardiovascular exercise, may be a protective factor associated with a decreased risk of cognitive decline with age [3], [5]. There is also evidence for a potential neuroprotective effect of social and cognitive activity (Marioni et al., 2015) [42], but findings have been mixed, with more research needed to identify the type and frequency required to observe beneficial effects [19], [17].

According to the STAC-r, the mechanisms underlying the relationship between life course factors and cognitive function are thought to be related to a bi-directional link between structure and function, reflecting both positive and negative effects of neuroplasticity throughout the lifespan [32]. Among the most extensively studied structures in aging is the hippocampus, which is widely recognized for its role in memory and its susceptibility to age-related neurodegeneration [7]; evidence in both human and animal models have highlighted the sensitivity of the hippocampus to several lifestyle factors, including stress [27], diet [15], and exercise [10], all of which contribute to an individual’s vascular health. Notably, the hippocampus has been shown to be sensitive to environmental factors in both younger and older age cohorts, which is in accordance with the life-course perspective of contemporary models of aging [23], [44].

Particularly well-documented is the decrease in hippocampal volumes with increased age, which is thought to relate to both atrophic processes and a decline in neurogenesis over the lifespan [2]. Considering the contribution of vascular risk to pathological aging, some have posited that the sensitivity of the hippocampus to age-related changes may be due to this structure’s greater susceptibility to vascular changes [9], [22]. Indeed, breakdown of the blood–brain barrier is thought to be one of the earliest signs of normative brain aging and has been shown to first arise in the region of the hippocampus [22]. Likewise, physical activity is also thought to promote neurogenesis and cell proliferation in the adult hippocampus, at least in part, due to its positive effects on brain vasculature [2], [11]. There is also evidence to suggest that social and cognitive stimulation may be related to hippocampal structural integrity in later life [45], [34], [35].

With increasing life expectancies and an associated rise in age-related neurodegenerative disorders, the need to better understand healthy aging and the factors that contribute to its divergence from pathological aging has become increasingly clear [13]. Although life course-focused models such as the STAC-r are now widely recognized, few studies have focused on examining the factors that contribute to successful aging across the lifespan, in both younger and older adulthood. Moreover, although the impact of lifestyle factors on hippocampal volumes has been widely examined in the context of advanced aging and disease, comparatively less is known about these relationships at different stages of adulthood.

In an effort to address this gap, the current investigation seeks to examine how lifestyle factors contribute to structural brain outcomes in a sample of healthy older and younger adults.

Specifically, the current study aims to (1) examine structural differences in a group of healthy older versus younger adults, with the aim of corroborating existing findings of structural differences in the hippocampus with age; (2) examine whether cumulative vascular risk factors and self-reported levels of physical, social, and/or cognitive activity are associated with differences in hippocampal volumes, both within and across groups. On the basis of existing research on brain aging, we hypothesized (1) greater atrophy of medial temporal lobe structures, including the hippocampus, in older versus younger adults and (2) that greater vascular risk, as well as less frequent participation in physical, social, and cognitive activities would be associated with smaller hippocampal volumes, both within and across age groups.

Methods

Recruitment

Participants were recruited through local advertisements posted in the greater Victoria community and through social media outlets. To facilitate contact with older adults in the community, recruitment efforts were also aided by e-mail advertisement through senior groups.

Participants

The study sample consisted of 34 neurologically healthy participants forming two separate age cohorts: a younger adult group (age 25–35, n = 17) and an older adult group (age 65+, n = 17). Self-reported sex and gender were congruent for all participants. For purposes of the current study, “neurologically healthy” referred to the absence of major neurological or psychiatric disorders known to significantly impact cognition (e.g., multiple sclerosis, psychotic spectrum disorders). However, to preserve ecological validity of the sample, participants were not excluded from the study on the basis of some existing non-neurological, medical, or psychological conditions (e.g., hypertension, self-reported history of anxiety or depression), nor if they were currently receiving pharmacological treatment for such conditions. Significant changes to an existing medication regime within the 3 months prior to recruitment did result in exclusion. Participants also needed to be capable of completing study tasks independently and to present with functional reading abilities necessary to complete questionnaires. For task administration reasons, all participants were required to be English-speaking adults. All participants were thoroughly screened to ensure lack of contraindication for MRI, including: any recent invasive procedures occurring ≤ 6 weeks prior to scanning; implanted pacemakers or other cardiac devices; and ferrous implants in close vicinity to the head and neck. Upon recruitment, participants provided written informed consent and were assigned an anonymized ID which was used to label all data collected during the remainder of the study. Ethics approval for this study was provided by the University of Victoria Research Ethics Board. Additional participant demographic information can be found in Table 2.

Table 2.

Participant demographic data, hippocampal volumes, and lifestyle variables of interest, stratified by age group. Column three shows group comparisons and associated p-values.

	YA	OA	YA vs. OA
Age	28.12 ± 2.8	70.88 ± 5.6	p = 0.000*
Females/Males	9/8	9/8
Education (years)	17.71 ± 3.2	17.18 ± 3.0	p = 0.621
R Hippocampal Volume (mm3)	4097.5 ± 485.7	3827.1 ± 578.8	p = 0.150
L Hippocampal Volume (mm3)	4004.4 ± 579.8	3733.5 ± 531.1	p = 0.165
CRAQ – Total	75.94 ± 59.59	249.94 ± 45.11	p = 0.000*
ALQ-R – Social	40.88 ± 8.08	44.29 ± 11.02	p = 0.312
ALQ-R – Physical	34.47 ± 8.88	30. 76 ± 11.82	p = 0.310
ALQ-R – Cognitive	106.29 ± 17.10	100.06 ± 19.85	p = 0.334
ALQ-R – Total	188.94 ± 28.79	183.00 ± 29.67	p = 0.558

YA = younger adults; OA = older adults.

* denotes statistical significance.

Study design

The data presented in the current study is part of larger data set collected during a 3-part multimodal investigation on normative aging outcomes. Data collection took place over two sessions scheduled no more than one week apart. During Session 1, participants in both groups participated in a total scan duration of 37 min; this included acquisition of a 3D-T1 anatomical image, T2* resting-state fMRI, and both 3D and 2D axial T2 FLAIR scans. Only the T1 anatomical data was examined for the current investigation. During session 2, participants completed a series of self-report questionnaires pertaining to demographic and lifestyle factors, as well as neuropsychological assessment measures.

Self-report measures

Cardiovascular Risk Assessment Questionnaire (CRAQ; Metagenetics, Inc.). The CRAQ is a multiple-choice questionnaire developed for the purpose of quantifying an individual’s cardiovascular risk. The questionnaire allows for allocation of a “risk score” (positive, neutral, or negative) to each possible item endorsement, based on existing research on how each of the concerning factors contribute to, or reduce, risk for cardiovascular disease. The questionnaire consists of two parts: a patient self-report section and a physician questionnaire. Given that not all participants in the current study will have access to a physician or their medical records, the current study employed only the patient self-report portion. This section consists of questions spanning 10 key areas of potential risk based on existing research, namely: (1) age, (2) cardiovascular history, (3) family history, (4) lifestyle, (5) stress, (6) sleep, (7) bowel toxicity, (8) blood sugar, (9) inflammation and pain, and (10) diet. Participants were instructed to circle the response that best applies to them. The subsequent risk scores were then added to derive a total cardiovascular risk score, in which a higher score confers greater relate risk. Given that only Part 1 was completed, the scores were not assessed based on risk categorizations, but rather compared on a within and between-group basis for the purpose of the current study.

Revised activities and lifestyle questionnaire (ALQ-R;[16]).

The ALQ-R is a revised version of a self-report questionnaire initially developed for use in the Victoria Longitudinal Study. The initial goal of the ALQ was to examine the frequency of leisure activities in adults and the revised version improved upon the content validity of the scale by including additional items probing physical and social activities. The ALQ-R has been validated by confirmatory structure analyses in independent samples, the results of which support the use of its subscales as indicators of activities across the lifespan. The current version consists of three subscales: (1) The Physical Activity Subscale, (2) the Social Activity Subscale, and (3) the Cognitive Activity Subscale. The ALQ-R total score additionally includes scores from 3 items related to travel. Each subscale consists of a subset of items from the questionnaire that represent the subscale constructs. Each item on the ALQ-R is rated by participants on a 9-point scale (i.e., 0 = never, 1 = less than once a year, 2 = about once a year, 3 = 2 or 3 times a year, 4 = about once a month, 5 = 2 or 3 times a month, 6 = about once a week, 7 = 2 or 3 times a week, 8 = daily), with higher numbers indicating greater frequency of participation. For each of the subscales, a cumulative sum score is computed across the applicable items, resulting in 3 subscales and 1 total score that were subsequently used in the analysis.

Neuroimaging measures

Image Acquisition Parameters. Images were acquired on 3.0 T General Electric (GE) Signa Pioneer MRI scanner. Whole-brain anatomical MRI scans were acquired sagittally, with a T1-weighted 3D BRAVO sequence, with the following parameters: 1.0 mm slice thickness, 256 × 256 acquisition matrix, echo time (TE) of 3.3 ms, in-plane voxel dimension of 1 mm², a repetition time (TR) of 8.6 ms, and flip angle of 10°.

Whole-Brain Voxel-Based Morphometry. A confirmatory VBM analysis was first conducted to examine whole-brain grey matter density differences between the younger and older adult cohorts. Raw DICOM images were converted to NIFTI format using dcm2niix (via MRIcroGL; [20]. High-resolution anatomical images were brain extracted using FSL’s Brain Extraction Tool (BET), which was optimized for each individual and followed by manual verification. The VBM analysis was conducted according to the optimized FSL-VBM protocol using FSL tools [6], [36] as follows: brain-extracted structural images were grey matter-segmented and registered to standard space via non-linear registration [1] to create a study-specific grey matter template. Participant grey matter images were then non-linearly registered to the study-specific template and smoothed with a Gaussian kernel (3 mm).

Hippocampal Volumetry. In the current study, hippocampal volumes were computed for each participant using FSL’s sub-cortical segmentation pipeline, FSL-FAST [28]. Briefly, FSL-FAST utilizes Bayesian modelling to analyze the shape and intensity of sub-cortical structures as identified in T1 anatomical images. For the current study, hippocampal volumes were computed bilaterally for each participant, resulting in a left and right hippocampal volume for each subject.

Statistical analyses

For the VBM analysis, a voxelwise general linear model (GLM) was applied using randomise, FSL’s permutation-based non-parametric testing, to compare whole-brain grey matter densities between the two groups. Differences were examined at the p < 0.01 level with threshold-free cluster enhancement and correction for multiple comparisons. Results were covaried for sex. Univariate general linear models were then applied in IBM SPSS Statistics to examine the association between left and right hippocampal volumes and self-reported lifestyle variables, both within and across groups. All analyses were covaried for sex and across-group analyses were also covaried for age cohort (group).

Results

Participant demographics and lifestyle measures

The younger and older adult groups did not differ statistically in years of education or activity levels, as defined by the ALQ-R. On average, participants in the older adult group had smaller left and right hippocampal volumes than participants in the younger adult group, though this difference did not reach statistical significance (Fig. 2). As expected, older adults present with significantly higher cardiovascular risk than younger adults, as defined by CRAQ total scores (p < 0.05). Additional information can be found in Table 2.

Fig. 2.

Scatterplot showing the distribution of left and right hippocampal volumes in the younger adult (YA) and older adult (OA) groups. Blue dots represent left hippocampal volumes and red dots represent right hippocampal volumes. Sex is coded by shape. Y-axis values are listed in cubic millimetres. Dotted lines represent the mean left (blue) and right (red) hippocampal volume within each group. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Regional differences in grey matter density in younger versus older adults

Between-group comparisons were performed to examine differences in whole-brain grey matter density in healthy younger versus older adults. Consistent with ample empirical data on normative aging, results revealed significant differences in grey matter density (p < 0.01, corrected for multiple comparisons) in younger versus older adults in several localized regions, including the medial temporal lobes, bilaterally. Specifically, contrasts showed greater cortical atrophy in older relative to young adults in the left and right hippocampus and temporal fusiform cortex, as well as in the right precentral gyrus (Fig. 1). Additional grey matter regions of significance and their associated coordinates are listed in Table 1.

Fig. 1.

Results of between-group comparison of whole-brain grey matter density, showing regions of cortical atrophy in older relative to young adults (p < 0.01, corrected for multiple comparisons). Images on overlaid on T1-weighted MNI152_T1_1mm standard template provided FSL. The color bar shows the range of intensities across the regions of significance: colors in the lighter range indicate a greater statistical robustness (smaller p-value threshold) of the observed results, whereas colors in the darker range indicate a lower statistical robustness (larger p-value threshold).

Table 1.

Regions showing statistically significant atrophy in grey matter density in older relative to younger adults. Coordinates are listed in Montreal Neurological Institute (MNI152) standard stereotaxic space.

Brain Region	Laterality	MNI152 Coordinates
		x	y	z
Hippocampus	L	−30	−24	−11
Hippocampus	R	29	−24	−12
Temporal Fusiform Cortex	L	−36	−21	−26
Temporal Fusiform Cortex	R	35	–22	−26
Middle Temporal Gyrus	L	−66	−19	−21
Middle Temporal Gyrus	R	66	−19	−16
Parahippocampal Gyrus	R	29	−19	−27
Precentral Gyrus	R	3	−19	55
Postcentral Gyrus	L	−57	−9	36
Cerebellum	R	38	−68	−36
Cerebellum	L	−34	−54	−27

Association between hippocampal volumes and cerebrovascular risk scores

A univariate general linear model was applied to examine the individual associations between left and right hippocampal volumes and cardiovascular risk assessment scores as derived by the CRAQ, both within and across age cohorts (Fig. 3). Results revealed a statistically significant negative association between CRAQ scores, and both left (p = 0.006, F = 8.987; B = -5.167, η² = 0.237; Table 3) and right (p = 0.012; F = 7.190; B = -4.487; η² = 0.199; Table 4) hippocampal volumes across groups, when covarying for the effects of both sex and age cohort. Follow-up within-group analyses revealed a significant negative linear relationship between CRAQ scores and left hippocampal volumes in the older adult group only (p = 0.033; F = 5.591; B = -6.290; η² = 0.285), with trend-level results for the right hippocampus (p < 0.1). Neither left nor right hippocampal volumes surpassed trend level results in the younger adult group alone (p < 0.1)

Fig. 3.

A regression plot illustrating the relationship between left (top) and right (bottom) hippocampal volumes (in cubic millimetres) and cardiovascular risk scores, as derived from each participant’s raw CRAQ total score. Groups are differentiated by colour: older adults (OA) are shown in black, whereas younger adults (YA) are shown in dark blue. Sex is coded by shape. Lines represent the linear fit for each group. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 3.

Table showing the results of the general linear model examining the association between left hippocampal volumes and cardiovascular risk assessment questionnaire (CRAQ) scores across groups.

	Mean Square	F	Sig.	Partial Eta Square
CRAQ Score	2286206.41	8.987	0.006	0.237
Sex	151554.95	0.596	0.446	0.020
Age Cohort (Group)	81645.39	0.321	0.575	0.011

Table 4.

Table showing the results of the general linear model examining the association between right hippocampal volumes and cardiovascular risk assessment questionnaire (CRAQ) scores across groups.

	Mean Square	F	Sig.	Partial Eta Square
CRAQ Score	1718463.99	7.190	0.012	0.199
Sex	417257.02	1.746	0.197	0.057
Age Cohort (Group)	26420.42	0.111	0.742	0.004

Association between hippocampal volumes and frequency of activity

A univariate general linear model was applied to examine the individual associations between left and right hippocampal volumes and frequency of social, cognitive, physical, and total activity levels according to the ALQ-R. Additional information about ALQ-R score distributions can be found in Table 2.

Within-group results revealed a significant negative association between frequency of social activities as reported by the ALQ-R and both left (p = 0.013, F = 8.186) and right (p = 0.002, F = 14.532) hippocampal volumes in older adults (Fig. 4). This association was not significant in the younger adult group nor when examined across both groups. No significant associations were found between left or right hippocampal volumes and total, cognitive, or physical activities in both within- and across-group analyses. However, a trend-level positive association was identified between frequency of physical activity as reported by the ALQ-R and left hippocampal volumes in older adults (p = 0.066; F = 3.982).

Fig. 4.

A regression plot illustrating the relationship between left (top) and right (bottom) hippocampal volumes (in cubic millimetres) and frequency of social activity in older adults, as derived from participant ALQ-S scores. Groups are differentiated by colour: older adults (OA) are shown in black, whereas younger adults (YA) are shown in dark blue. Sex is coded by shape. Lines represent the linear fit for each group. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Discussion

Despite a renewed focus on life-course models of aging, few studies to date have examined the lifestyle and neuroimaging correlates of successful aging in a sample that includes both younger and older adults. This is despite ample evidence suggesting that the integrity of structures in the brain involved in pathological aging, such as the hippocampus, may be susceptible to environmental factors throughout the lifespan. The goal of the current study was therefore to address a gap in the literature by (1) examining structural differences in a group of healthy older versus younger adults, with a focus on validating previous findings of hippocampal structural changes with age, and (2) examining whether cumulative vascular risk factors and self-reported levels of physical, social, and cognitive activity are associated with differences in hippocampal volumes both within and across different age cohorts.

Hippocampal structural differences in older versus younger adults

In line with our hypothesis, whole-brain grey matter density comparisons revealed statistically significant atrophy in older versus younger adults in several regions of the medial temporal lobes, including the bilateral hippocampus (Fig. 1). These findings are consistent with several longitudinal studies on normative aging that have revealed mild to moderate rates of hippocampal atrophy with increasing age. Longitudinal studies have found that both healthy older and younger adults exhibit some degree of hippocampal atrophy over a 5-year period, with the rate of decline increasing significantly in those above the age of 50 years [31].

In addition to the VBM analysis, between-group hippocampal comparisons were also examined in the current study using structural segmentation methods, which yielded a total right and left hippocampal volume for each subject (Fig. 2). In contrast to the grey matter density comparisons, while, on average, older adults had smaller hippocampal volumes relative to younger adults, no statistically significant differences were found in global left or right hippocampal volumes between age cohorts. This discrepancy may, in part, be explained by findings suggesting that age-related hippocampal changes may differentially affect certain sub-regions [7], [12], [30], [43]. For instance, Gordon et al. [12] observed greater age-related atrophy in the anterior hippocampus relative to the posterior tail in cognitively normal adults; this type of structural difference may not be captured by cross-sectional measures of global hippocampal volume. Likewise, other studies on normative aging have also failed to find a cross-sectional aging effect on hippocampal atrophy when examining segmentation-based hippocampal volumes, despite observing a significant longitudinal atrophy rate [8]. Taken together, these findings suggest that a multi-analysis approach may be best suited at capturing the more nuanced age-related hippocampal changes seen in non-clinical aging populations.

Relationship between hippocampal volume and cardiovascular risk

In addition to group structural comparisons, a primary aim of the current study was to examine how cumulative life-course factors, and particularly cardiovascular risk factors, impact hippocampal structural outcomes in both healthy younger and older adults. Based on existing research showing a greater susceptibility of the hippocampus to vascular compromise, we hypothesized that a greater cumulative vascular risk score would be associated with smaller hippocampal volumes. In support of this hypothesis, the current findings revealed that greater cardiovascular risk (i.e., poorer vascular health) was uniformly associated with smaller hippocampal volumes when examined across age cohorts. However, subsequent within-group analyses found that this relationship only reached statistical significance in the older adult group, and for the left hippocampus only. While the negative linear relationship between greater vascular risk and hippocampal volumes remained tenable in the right hippocampus and in the younger adult group, these results did not surpass trend-level (Fig. 3).

Overall, the findings of an association between poorer vascular health and reduced hippocampal structural integrity is commensurate with existing studies in the pathological aging literature which strongly support a link between vascular risk factors and age-related neurodegenerative diseases [39]. In addition, more recent longitudinal studies on normative aging populations have also provided evidence of a link between greater cumulative cardiovascular risk and lower hippocampal structural integrity [38].

While the exact mechanisms underlying the relationship between vascular risk factors and pathological aging remain unclear, it is thought to involve a combination of cerebrovascular, inflammatory, and neurotoxic processes [17], [33]. Notably, adequate vascular perfusion to the brain has been shown to be associated with both cognitive and structural differences in older adults [4], [25]. The hippocampus in particular has been shown to be sensitive to vascular compromise and hypoxia, in part due to the size of its vessels and their vulnerability to cerebrovascular small vessel disease [29]; (Wu et al., 2008). Older adults with a more robust vascular supply to the hippocampus have been shown to perform better on medial temporal lobe tasks and have a higher anterior hippocampal volume, relative to older adults with more limited hippocampal vascularization [29]. Thus, the current findings of a negative association between cumulative vascular risk and hippocampal volumes across younger and older adults are supported by existing cognitive and physiological data showing a specific susceptibility of the medial temporal lobe to vascular compromise.

While it is notable that the relationship between cardiovascular risk and hippocampal volumes failed to reach significance within the younger adult group alone, several factors may account for this finding. First, as reflected in our own demographic data (Table 2), incidence of vascular illness and risk of cardiovascular disease is significantly lower in younger populations, relative to middle age and older [40]. Thus, when examined in a smaller sample, it is possible that the comparatively weaker association lacked the power to surpass trend level findings, as it did when examined across-groups. Second, it is important to note that the difference in findings between age cohorts in the current study is consistent with several population-based studies showing that life-course factors in adolescence and young adulthood are typically predictive of vascular outcomes arising in mid-life [40]. Thus, while structural changes may not yet be evident in younger adults below age 35, such as the younger group in the current study, the finding of a significant effect in cognitive healthy older adults nonetheless has important implications for preventative public health measures aimed at younger populations.

Relationship between hippocampal volume and lifestyle activities

Another aim of the current study was to examine whether self-reported levels of physical, social, and/or cognitive activity are associated with differences in hippocampal volumes, both within and across groups. Contrary to our hypothesis, no significant association was identified between left or right hippocampal volumes and total, cognitive, or physical activity levels in both within- and across-group analyses, with only a trend-level positive association between frequency of physical activity and left hippocampal volumes in older adults.

Interestingly, however, within the older adult group only, a significant negative association was found between frequency of social activity and both left and right hippocampal volumes. This suggests that older adults who endorsed more frequent participation in social activities, such as interacting with friends and relatives or attending social gatherings, tended to have smaller hippocampal volumes, bilaterally. This finding is contrary to existing studies showing that socially engaging activities may be associated with improved structural outcomes in brain aging, including the hippocampus [45]. Moreover, social isolation has also been suggested as a potential risk factor for cognitive impairment and neurodegenerative disease [21]. However, as with the current findings, other studies have also found opposite effects on brain health outcomes. For instance, a large multi-site study of over 2,000 community-dwelling older adults in Asia found that participants who engaged in more social activity over a period of 10 years were at greater risk of having impaired cognition [19]. The authors suggest that engaging in more social activities may result in a reduced frequency of participation in more challenging physical and intellectual activities, the latter of which have been more reliably associated with improved cognitive reserve [19]. For instance, passive social engagement, such as spending time with friends and family, has a relatively low cognitive demand and therefore may not be as protective of cognition, especially in average community-based older adults [19]. Further studies are needed to better understand the effects of different types of social engagement on cognitive and structural outcomes in normative aging groups.

Despite existing evidence linking physical and cognitive activity to improved structural and cognitive outcomes [10], [42], the current study did not find a significant association between either type of activity and hippocampal volumes in older or younger adults. However, as aforementioned, a trend-level positive association was identified in the older adult group only, in which the expected positive association between physical activity and hippocampal volume was observed in the left hippocampus. One particularly important variable that may account for the discrepancy in findings relates to the type of physical activity examined. The current study’s ALQ-R physical subscale is an aggregate measure that reflects total frequency of strength-based, aerobic, and other sports-related activities; thus, if hippocampal neurogenesis is primarily driven by aerobic forms of exercise [10], this association may be dampened by accounting for other forms of non-aerobic physical activity. Another intriguing possibility relates to the idea that different types of activity may differentially affect the hippocampus [2]: while physical activity is thought to increase cell proliferation through improvements in brain vasculature, cognitive stimulation is thought to affect cell survival in the dentate gyrus [26]. Thus, like age-related structural changes, measures of total hippocampal volume may not adequately capture sub-regional differences attributed to life-course factors in healthy aging groups.

Limitations and future directions

By examining the impact of life-course factors on hippocampal structural changes in healthy older and younger adults, the current study aims to address important gaps in the current aging literature. However, there are several limitations to consider. First, the current study examines hippocampal volume as a measure of structural integrity; this is based on ample research showing that age-related reductions in hippocampal volume are typically associated with cognitive decline and age-related neuropathology [24], [41]. However, other studies have cautioned that the association between structure and function in the hippocampus may also be linked to differential recruitment of brain regions with age [18]. Thus, it is not possible to ascertain that a smaller hippocampal volume necessarily implies worse cognitive outcomes. Second, as mentioned previously, many of the variables examined in the current study are summary variables that may not adequately capture some of the nuances in how life course variables are associated with hippocampal structure. For instance, while widely used, measures of hippocampal volume fail to account for sub-regional changes within the hippocampus. As such, further studies by the authors are currently underway to examine these associations using more sensitive methods that can delineate hippocampal sub-regions. In addition, life-course measures such as the ALQ-R and CRAQ are aggregate measures of cardiovascular risk and activity that do not allow for more specific conclusions about which specific vascular risk or activity may more strongly account for these associations. Thus, further studies in normative aging are needed to examine hippocampal changes in relation to specific life-course factors, which in turn would help better inform preventative lifestyle interventions.

Finally, another limitation of the current study pertains to characteristics of the sample and study design. While our study examines younger adults, who are often excluded from studies on brain aging, middle-age adults were not represented in the current study. Moreover, the sample size is small and limited to a single time point, which makes it difficult to account for confounding variables that could impact an individual’s hippocampal volume over time, including heterogeneity in hippocampal volumes in early life [2]. As a result of these factors combined, causal inference cannot be established. Thus, future studies on normative aging should employ a longitudinal model, with an emphasis on better understanding the structural trajectory of the hippocampus across younger, middle, and older adulthood and its association with individual variability in life-course factors.

Conclusion

A large portion of neuroimaging research in aging to date has focused on age-related pathologies. Increasingly, there has been a renewed focus on the study of healthy aging and better understanding how normative aging trajectories deviate from disease outcomes. However, despite the popularity of life-course aging models such as the STAC-R, few studies to date include early adulthood. To address these important gaps, the current study examined how cumulative vascular risk factors and self-reported levels of physical, social, and cognitive activity are associated with differences in hippocampal volumes, in a sample of healthy younger and older adults. In accordance with increasing emphasis on the importance of vascular health, results of our study suggest that greater cumulative vascular risk is associated with smaller hippocampal volumes across age cohorts. Notably, the current study also adds to emerging evidence suggesting that social activities with a low cognitive load may not be beneficial to older adults with respect to preserving hippocampal structural integrity. Additional studies on normative aging are needed to inform preventative lifestyle interventions, with an emphasis on longitudinal models that examine structural changes in response to life-course factors from early to late adulthood.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.