Weight for It: Resistance Training Mitigates White Matter Hyperintensity-Related Disruption to Functional Networks in Older Females

Abstract

Background:

White matter hyperintensities (WMH) are associated with impaired cognition and increased falls risk. Resistance training (RT) is a promising intervention to reduce WMH progression, improve executive functions, and reduce falls. However, the underlying neurobiological process by which RT improves executive functions and falls risk remain unclear. We hypothesized that: 1) RT reduces the level of WMH-related disruption to functional networks; and 2) reduced disruption to the sensorimotor and attention networks will be associated with improved executive function and reduced falls risk.

Objective:

Investigate the impact of 52 weeks of RT on WMH-related disruption to functional networks.

Methods:

Thirty-two older females (65–75 years) were included in this exploratory analysis of a 52-week randomized controlled trial. Participants received either twice-weekly RT or balance and tone training (control). We used lesion network mapping to assess changes in WMH-related disruption to the sensorimotor, dorsal attention, and ventral attention networks. Executive function was measured using the Stroop Colour-Word Test. Falls risk was assessed using the Physiological Profile Assessment (PPA) and the foam sway test.

Results:

RT significantly reduced the level of WMH-related disruption to the sensorimotor network (p = 0.012). Reduced disruption to the dorsal attention network was associated with improvements in Stroop performance (r = 0.527, p = 0.030). Reduced disruption to the ventral attention network was associated with reduced PPA score (r = 0.485, p = 0.049)

Conclusion:

RT may be a promising intervention to mitigate WMH-related disruption to the sensorimotor network. Additionally, reducing disruption to the dorsal and ventral attention networks may contribute to improved executive function and reduced falls risk respectively.

Keywords

Cognition executive functions exercise falls risk functional connectivity resistance training

INTRODUCTION

A core component of cerebral small vessel disease, white matter hyperintensities (WMH), are present in over 90% of older adults [1]. WMHs are associated with impaired cognition [2] and increased falls risk [3, 4]. In the absence of any pharmaceutical treatment, the need to identify therapeutic interventions to mitigate WMH progression and its consequences is of vital importance.

WMHs are associated with impairments in global cognition, executive functions, processing speed, and memory [5 –8]. Importantly, a meta-analysis of 14 longitudinal studies showed that WMH progression was associated with cognitive decline over time, with a particular impact on executive functions [2]. In addition to cognition, both cross-sectional and longitudinal studies demonstrate the association between greater WMH volume and increased risk of falling [9, 10]. Of note, older adults with severe WMHs are twice as likely to fall than those with mild WMHs [4].

One potential mechanism by which WMHs elicit impaired cognition and increased falls risk can be explained by the disconnection hypothesis, which states that changes in brain structure have a subsequent effect on brain function [11]. As such, the presence of WMHs impairs communication between gray matter regions, resulting in disruptions to functional connectivity [12]. Reducing WMH progression, thereby mitigating aberrant functional connectivity, is an approach to promote cognitive and mobility outcomes in those with cerebral small vessel disease.

Resistance training (RT), a form of exercise aimed at improving muscular strength and endurance, is a promising strategy to slow WMH progression [13, 14]. The SMART trial showed six months of RT significantly regressed WMH volume and improved global cognition in older adults with mild cognitive impairment [14]. Similarly, our own randomized controlled trial [13] found significantly slowed WMH progression in older females who underwent 12 months of twice-weekly RT, which was associated with maintained gait speed. In the same trial, significantly greater improvements in executive function were also seen in those who underwent RT compared to an active control group [15].

It is hypothesized that RT may promote cognitive and mobility outcomes by mitigating WMH-related disruption to functional networks. Various patterns of aberrant connectivity have been identified in the large-scale functional networks of older adults with WMHs [16 –20]. For example, Ding and colleagues [16] identified aberrant between network functional connectivity involving the sensorimotor, dorsal attention, frontoparietal, and default mode networks in older adults with WMHs. Importantly, altered connectivity between the sensorimotor, frontoparietal, and dorsal attention networks with the auditory and visual networks was associated with cognitive impairment. Further, Benson and colleagues [17] found that higher functional connectivity of the frontoparietal network mitigated the impact of WMH volume on executive functions, but found no protective effects of default mode network connectivity on the relationship between WMHs and cognition. However, WMHs commonly manifest with coexisting pathologies such as global brain atrophy [21] and hypertension [22] that are known to significantly impact BOLD-derived functional connectivity [23]. Lesion network mapping is an image analyses technique that may reduce these confounders by mapping the location of lesions onto the functional connectivity of a human connectome of healthy adults [24]. This identifies the functional connectivity specifically associated with WMHs. It is then possible to calculate the level of WMH-related disruption to functional networks.

Using this method, our prior work showed greater WMH-related disruption to the sensorimotor and ventral attention networks was significantly correlated with poorer cognition in cerebral small vessel disease [25]. In the same population, greater disruption to the dorsal attention and ventral attention networks as a result of WMHs was also associated with greater falls risk [26]. Briefly, the sensorimotor network is primarily responsible for motor planning and execution [27], but recent evidence suggests a greater involvement in the performance of cognitive tasks [28]. The dorsal attention network is responsible for the top down orienting of attention, and the ventral attention network reorients attention to salient stimuli [29]. It is hypothesized that interventions that mitigate WMH-related disruption to these networks would subsequently improve executive functions and reduce falls risk.

Using lesion network mapping, we conducted an analysis of a 12-month randomized controlled trial to assess whether RT mitigates WMH-related disruption of functional networks that are relevant to executive functions and falls risk. We also explored the association between changes in WMH-related disruption of functional connectivity and changes in executive functions and falls risk. We hypothesize that compared to an active control group, participants who underwent 12 months, twice-weekly RT would show a greater reduction in WMH-related disruption to the sensorimotor, dorsal attention, and ventral attention networks. We further hypothesize that reduced disruption to these networks would be associated with improved executive functions and reduced falls risk.

METHODS

Participants

This study was an exploratory analysis of the Brain Power Study, a 52-week randomized controlled trial of RT [15]. To achieve our objective, we analyzed a sub-group of participants who were in the twice weekly RT and balance and tone (BAT) training groups, and who were identified as having WMHs at baseline (see Fig. 1).

Fig. 1

Consolidated Standards of Reporting Trial Flowchart of Participants. MRI, magnetic resonance imaging; WMH, white matter hyperintensities; RT, resistance training; BAT, balance and tone training.

The randomized controlled trial included females only, as evidence suggests there are sex differences in the cognitive response to exercise [30, 31]. Participants were: 1) aged 65–75 years old; 2) female; 3) living independently in their own homes; 4) scored ≥24 on the Mini-Mental State Examination (MMSE); 5) had a visual acuity of at least 20/40, with or without corrective lenses; and 6) provided informed consent. Participants were excluded if they: 1) had an absence of WMHs on magnetic resonance imaging (MRI); 2) had a medical condition for which RT is contraindicated; 3) had regularly participated in RT in the last six months; 4) had a neurodegenerative disease and/or stroke; 5) did not speak and understand English fluently; 6) had depression; 7) were taking medication that may negatively affect cognition, such as anticholinergics; or 8) were on estrogen replacement or testosterone therapy.

Ethical approval was obtained from the University of British Columbia Clinical Research Ethics Board (H06-03216) and the Vancouver Coastal Health Research Institute (V06-03216). All participants provided informed written consent.

Measurements

For descriptive purposes, age was recorded in years, height was measured in centimeters (cm) and weight in kilograms (kg). The functional comorbidity index was used to estimate the degree of comorbidity associated with physical functioning [32], and diastolic blood pressure was measured in millimeters of mercury (mmHg) and in accordance with the standard protocol of the American Heart Association [33].

Executive functions: selective attention and conflict resolution

The Stroop Colour-Word Test is used to assess the executive process of selective attention and conflict resolution and was the primary outcome of the Brain Power Study [15]. The test involves three conditions [34]. First, participants read aloud color-words (e.g., RED) printed in black ink (Stroop 1). Second, they are asked to name aloud the ink color of colored-X’s (Stroop 2). Finally, participants are asked to name the ink color of color-words printed in incongruent ink colors (e.g., the word RED printed in blue ink) (Stroop 3). To calculate Stroop interference, the time taken to complete Stroop 2 was subtracted from the time taken to complete Stroop 3 (Stroop 3-2). A smaller time difference between conditions is indicative of better selective attention and conflict resolution.

Falls risk

Falls risk was assessed using the Physiological Profile Assessment (PPA)© (Prince of Wales Medical Research Institute, Sydney, Australia) as well as postural stability. The PPA is a validated measure for quantifying falls risk in older adults [35] with a 75% predictive accuracy for falls [36]. A PPA score is calculated from a composite of five measures: 1) hand reaction time; 2) proprioception; 3) edge contrast sensitivity; 4) postural sway; and 5) knee extension strength. A greater PPA score is indicative of greater falls risk.

Postural instability is a major risk factors for falls [37, 38]. Thus, postural stability was independently assessed using the eyes open foam sway component of the PPA. Participants are asked to stand hip width apart on a 3-inch high-density foam cushion and stare straight ahead for 30 s. Postural sway (mm²) is recorded using the Lord sway-meter [39]. The task is stopped if the participant grabs for support. Greater sway is indicative of poorer postural stability.

Magnetic resonance imaging (MRI)

White matter hyperintensity quantification

The MRI scans were conducted at the UBC MRI Research Centre on a 3T Philips Achieva Scanner with an 8-channel SENSE neurovascular coil. T2-weighted and proton density (PD)-weighted structural MRI scans were acquired for each subject. The T2-weighted scans had a 5,431 ms repetition time (TR), a 90 ms echo time (TE), and a 256 x 198 acquisition matrix. The PD-weighted scan had a TR of 2,000 ms and TE of 8 ms. Both scans had a 256 x 204 acquisition matrix.

Preprocessing of the MRI scans involved using: 1) the brain extraction tool (BET) [40], to remove all nonbrain matter; 2) a noise-removal filter (SUSAN) [41]; and 3) a multiscale version of nonparametric nonuniform intensity normalization (N3) [42]. An experienced radiologist identified the WMHs on the MRI of each participant and marked them with digital seeds for subsequent automated processing. Briefly, the seeding procedure involved: 1) marking all distinct WMHs regardless of size; 2) using additional points if more than one point would help define the extent of the lesion; 3) placing at least one point near the center of each lesion [43]. The WMHs were then automatically segmented by computing the extent of each lesion. This method is highly accurate and robust compared with radiologist segmentations [43]. Each lesion mask was then co-registered to MNI space with a voxel size of 2 mm³.

Lesion network mapping

Lesion network mapping was used to identify the functional connectivity associated with each lesion. Detailed procedures of this process have been provided elsewhere [25].

Briefly, the location of each WMH was used as a seed region of interest (ROI). The resting-state functional connectivity associated with each ROI was then extracted by cross-correlating the timeseries data from each ROI with every other voxel within the connectome. This was then converted to Fisher’s Z (Fz) scores and averaged across all ROIs to create an average Fz map per participant. These Fz maps represent the functional connectivity associated with the equivalent location of WMHs in the healthy adult brain. In other words, it identifies the functional connectivity most likely to be disrupted by the presence of WMHs (lesion network) in each participant.

To extract only the voxels functionally connected with WMHs, the average Fz maps were converted to binary maps whereby 1 = Fz value > 0.05 and 0 = Fz value ≤0.05. We then multiplied the sensorimotor, dorsal attention, and ventral attention network maps [44] by the binarized lesion network map. Thus, only the voxels that were positively functionally connected with both the lesion network and the networks of interest remained. We were then able to calculate the percentage overlap between the lesion network and each of the functional networks of interest:

$\frac{Total voxels overlapping with lesion network and network of interest}{Total voxels in network of interest} \times 100$

Intervention

The intervention protocol has been previously described in detail [15]. Briefly, participants were randomized to one of three groups: 1) once weekly RT; 2) twice weekly RT; or 3) twice weekly BAT (i.e., balance and tone training; active control). Each 60-min training session consisted of a 10-min warm up, 40 min exercise, and 10 min cool down. The class instructors were audited on a monthly basis for quality control. As changes in WMH volume were only identified between the twice weekly RT and BAT groups [13], this study investigated WMH-related changes in functional connectivity in these groups specifically.

Resistance training (RT)

The RT program consisted of exercises using a combination of free weights and Keiser^® Pressurized Air system machines. Machine-based exercises included leg press, biceps curls, triceps extension, seated row, latissimus dorsi pull down, hamstrings curls, and calf raises. Free weight exercises included mini-squats, mini-lunges, and lunge walks. Training intensity began at 2 sets of 6-8 repetitions for each exercise. The intensity was then increased using the 7-RM method, once 2 sets of 6-8 repetitions were completed with the correct form.

Balance and tone training (BAT)

The BAT program consisted of stretching, basic core/Kegel exercises, range of motion and balance exercises, deep breathing, and relaxation techniques. Key balance exercises included walking and standing in tandem, single leg stance (with the eyes open and closed), and Crane and Tree Tai Chi poses. Other than body weight, no additional loading was used. This training has not been shown to improve cognition [45], and thus served as a control for confounding variables, such as physical training gained from travel to the exercise sessions, social interaction, and changes to lifestyle secondary to the study participation

Data analysis

All analyses were based on complete cases, such that only participants with valid data at all time points were included. Analysis of covariance was used to assess differences in percentage network overlap between groups at trial completion controlling for baseline network overlap, diastolic blood pressure, WMH volume, and age. We controlled for diastolic blood pressure as it is strongly associated with WMHs [46]. Partial correlations explored associations between changes in network overlap, and changes in Stroop 3-2, PPA score, and foam sway. Due to sample size, participants were combined for the partial correlation analyses controlling for age, baseline outcome measure, and group. Change scores were calculated by subtracting baseline from trial completion. Bonferroni correction was used to control for multiple comparisons for all analyses.

RESULTS

Participants

Thirty-two older females from Metropolitan Vancouver were included in this study. Of this sample, eight withdrew before final assessment and four were removed due to having WMHs too close to the ventricles, causing errors during image processing and analysis. Thus, a final sample of 20 participants were analyzed in this study (see Fig. 1). The baseline mean (SD) age of this sample was 70.3 (2.7) years old, with a MoCA score of 26 (2.7), and WMH volume of 1507.4 (2118.3) mm³ (see Table 1). Participants in the RT group had significantly greater lesion network overlap with the dorsal (p = 0.047) and ventral (p = 0.029) attention networks compared with the BAT group at baseline (see Table 3). There were no other baseline differences between groups (see Table 2).

Table 1

Baseline Characteristics mean (SD) or n (%)

	RT (n = 17)	BAT (n = 15)	All (n = 32)
Age (y)	69.2 (3.1)	69.3 (2.8)	69.3 (3.0)
Height (cm)	161.3 (7.4)	162.9 (5.8)	162.1 (6.6)
Weight (kg)	68.1 (12.5)	69.5 (9.4)	68.8 (11.0)
Diastolic Blood Pressure (mmHg)	80.1 (10.9)	89.1 (24.1)	84.5 (18.7)
Education
High School Graduate	5 (29.4)	3 (20)	8 (25)
Trades or Professional Certificate	2 (11.8)	2 (13.3)	4 (12.5)
University Certificate	1 (5.9)	1 (6.7)	2 (6.3)
University Degree	8 (47.1)	7 (46.7)	15 (46.9)
Functional Comorbidity Index	2.4 (2.0)	1.9 (1.4)	2.2 (1.7)
MMSE (/30)	28.8 (1.1)	28.7 (1.3)	28.8 (1.2)
MoCA (/30)	25.9 (2.9)	26 (2.6)	25.1 (3.2)
WMH volume (mm³)	1470.9 (2225.2)	2306.2 (4508.4)	1858.1 (3453.1)

MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; WMH, white matter hyperintensity.

Table 2

Mean (SD) of behavioral outcome measures from participants included in the lesion network mapping analyses

		Baseline	Trial completion
RT Group (n = 10)
WMH volume (mm³)		2171.8 (2686.3)	2299.1 (2942.0)
Stroop (s)	Part 1	36.0 (8.2)	37.5 (7.3)
	Part 2	51.6 (11.6)	53.0 (9.8)
	Part 3	101.8 (21.7)	99.5 (18.4)
	Part 3-2	50.2 (19.1)	46.4 (15.9)
PPA		0.32 (1.3)	0.05 (1)
Foam Sway (mm)		135.4 (83.8)	131.8 (65.9)
BAT Group (n = 10)
WMH volume (mm³)		843.1 (1129.4)	990.7 (1368.0)
Stroop (s)	Part 1	37.5 (8.1)	39.9 (9.5)
	Part 2	54.8 (13.7)	53.7 (8.4)
	Part 3	98.1 (19.1)	101.9 (28.1)
	Part 3-2	43.3 (17.2)	48.1 (24.8)
PPA		0.12 (1.2)	–0.23 (0.8)
Foam Sway (mm)		143.5 (120.6)	106.1 (45.0)

RT, resistance training; WMH, white matter hyperintensity; PPA, Physiological Profile Assessment; BAT, balance and tone training.

Table 3

Mean (SD) Percentage Overlap Between Voxels Functionally Associated with the Lesion Network and the Networks of Interest

	Baseline (%)	Trial Completion (%)
RT Group
Sensorimotor	72.3 (29.0)	67.3 (35.4)
Dorsal Attention	46.7 (20.6)	55.6 (26.0)
Ventral Attention	83.6 (24.0)	81.1 (21.0)
BAT Group
Sensorimotor	57.4 (27.7)	67.4 (33.1)
Dorsal Attention	29.8 (14.2)	54.3 (27.3)
Ventral Attention	56.0 (27.7)	71.5 (31.0)

RT, resistance training; BAT, balance and tone training.

Change in network overlap between groups

Table 3 shows the average percentage of voxels that overlapped with both the lesion network and networks of interest for each group. We identified that the percentage overlap of each network across the sample was not normally distributed and thus log transformed the data prior to analyses. After correcting for multiple comparisons (p < 0.017), we found a significant difference in sensorimotor network disruption by intervention (F(1,14)=8.29, p = 0.012), such that RT mitigated WMH-related sensorimotor disruption while BAT increased network disruption. Although not significant, visual inspection of the data also suggests that while the BAT group showed an increase in disruption of all three networks, the RT group showed a smaller increase in disruption of the dorsal attention network and maintained the level disruption of the ventral attention networks (see Fig. 2).

Fig. 2

Change in percentage overlap between the lesion network and the sensorimotor network (SMN), dorsal attention network (DAN), and ventral attention network (VAN) by group. The error bars represent standard error. RT, resistance training; BAT, balance and tone training *p<0.05

Partial correlations between change in network overlap and change in executive function and falls risk

Change in the percentage overlap of the dorsal attention network was significantly correlated with change in Stroop Interference (r = 0.527, p = 0.030). Thus, reduced WMH-related disruption to the dorsal attention network was associated with improved selective attention and conflict resolution performance, as measured by the Stroop Colour-Word Test.

In addition, change in percentage overlap of the ventral attention network was also significantly correlated with change in PPA score (r = 0.485, p = 0.049). Thus, reduced WMH-related disruption to the ventral attention network was associated with reduced falls risk.

Although participants from both groups were combined in these analyses, there were no between group differences in the direction of these correlations (see Fig. 3). However, these findings were not significant after controlling for multiple comparisons (p < 0.008). Change in disruption with the sensorimotor network was not significantly associated with changes in executive functions or falls risk (see Table 4).

Fig. 3

A) Partial correlation between change in overlap of the dorsal attention network and change in Stroop performance controlling for age and baseline Stroop. B) Partial correlation between change in overlap of the ventral attention network and change in Physiological Profile Assessment (PPA) score controlling for age and baseline PPA. Blue: balance and tone group; Red: Twice-weekly resistance training group.

Table 4

Partial correlations between change in network overlap and change in cognition and falls risk, controlling for age

		Sensorimotor	Dorsal Attention	Ventral Attention
Stroop Part 3-2	r	–0.187	0.527*	0.360
	p	0.472	0.030	0.156
PPA	r	–0.140	0.004	0.485*
	p	0.592	0.988	0.049
Foam Sway	r	0.303	–0.067	0.279
	p	0.238	0.798	0.279

PPA, Physiological Profile Assessment; *p<0.05.

DISCUSSION

This study is the first to use lesion network mapping to investigate the impact of RT versus a BAT control group on WMH-related disruption to functional networks. We showed that 52 weeks of twice weekly RT significantly reduced WMH-related disruption to the sensorimotor network compared to twice weekly BAT. In addition, while not surviving Bonferroni correction, there was a trend towards reduced disruption to the dorsal attention network being associated with improved response inhibition and conflict resolution performance; and reduced disruption of the ventral attention network being associated with reduced falls risk.

In addition to the statistically significant difference in disruption to the sensorimotor network between groups, the data showed trends that suggest 12 months twice-weekly RT may mitigate the level of disruption to the ventral attention and dorsal attention networks as well. These findings support the notion that RT may target WMH-related disruption to functional networks. It is important to recognize that the RT group had significantly greater overlap of both the ventral attention, and dorsal attention networks with the lesion network at baseline. Thus, although we did control for baseline in our analyses, it is possible that this may have influenced the extent to which the effect of RT on these networks would be evident compared with the less affected control group.

While WMHs are considered progressive, twelve months, twice weekly RT was previously shown to slow the progression of WMH volume and improve executive functions compared with the BAT group [13]. There was also a trend towards an association between WMH volume and executive function performance, as measured by the Stroop Colour-Word Test [13]. The results from the current study support these findings by suggesting that slowed WMH progression may improve executive functions by mitigating WMH-related disruption to the dorsal attention network. The Stroop Colour-Word Test requires response inhibition, conflict resolution, and sustained selective attention for task performance [47]. As the dorsal attention network is responsible for the top-down orienting of attention [29], it is reasonable to expect that reduced disruption to this network could improve performance on the Stroop Colour-Word Test.

Reduced disruption to the ventral attention network was associated with reduced falls risk. As we age, risk for falling increases due to declines in physical function, including proprioception, muscle strength, and balance. This results in a greater need for attentional resources to maintain mobility. The ventral attention network is responsible for the bottom-up control of attention [29], meaning that it is stimulus driven. It is activated when the focus of attention needs to be reoriented to salient stimuli. The association between disruption of the ventral attention network and falls risk may be due to conflicting patterns of network activation. While under activation of the network may lead to slowed or missed responses to relevant salient stimuli, over activation may cause unnecessary distraction from the task and thus, impaired performance. Both the RT and BAT groups have the potential to reduce falls risk. Therefore, it may be that both exercise programs targeted similar neurobiological mechanisms (i.e., the ventral attention network), which contributed to reduced falls risk. This may also explain why we did not observe significant between-group differences in the change of disruption to the ventral attention network as well.

Our previous work highlighted the significance of disruption to both the dorsal and ventral attention networks to poorer cognition and increased falls risk [25, 26]. These two attention networks work in concert with each other to effectively control attention but are functionally distinct [29]; our results support this prior finding. Specifically, our data suggest reduced WMH-disruption in the dorsal attention network versus the ventral attention network may be more important for improving executive functions. Conversely, for reducing falls risk, it may be more important to reduce disruption in the ventral attention network versus the dorsal attention network. As RT may mitigate WMH-related disruption to both of these networks, it should be considered an important intervention to improve executive functions and falls risk.

Contrary to our hypothesis, the change in WMH-related disruption to the sensorimotor network after 12 months of RT was not associated with changes in executive functions or falls risk. The lack of an association may be, in part, due to sex differences in the contribution of the sensorimotor network to cognitive function [48]. Specifically, Stumme and colleagues [48] showed that the sensorimotor network was significantly more integrated in males, while the ventral attention network had higher within-network connectivity in females. Importantly, they identified that between-network connectivity of the sensorimotor network mediated the sex-related differences in cognitive function. Thus, they concluded that the sensorimotor network may have greater contribution to cognition in males than females. As the present study exclusively involved females, sex differences may have contributed to the lack of associations between change in the sensorimotor network and behavioral measure. Alternatively, between network connectivity of the sensorimotor network with other functional networks has been associated with cognitive status and falls risk [49 –51]; this study only examined within functional network connectivity.

This study is limited by its small sample size and the inability to perform between-network analyses based on the network maps used. Lesion network mapping is an effective technique for extracting the functional connectivity specifically associated with WMHs. However, it is important to acknowledge that the BOLD signal derived from white matter is much weaker than that of the gray matter. In addition, when converting WMH masks to standard space in lesion network mapping, WMHs on the very edge of the ventricles can bleed into the ventricle space and cause the lesion network mapping to fail. This may have introduced a sampling bias. Thus, it is possible that we did not capture the full impact of WMHs on functional networks. There were also notable differences in baseline WMH volume between the groups, which, despite being controlled for in our ANCOVA analyses, may have impacted the results. Therefore, it would be important for future studies with larger sample sizes to verify our current results. Further, using a human connectome of healthy adults does not account for the individual differences in functional connectivity. Thus, our results derived using lesion network mapping should be interpreted with these limitations in mind. The BAT group was initially selected as the active control for the parent study due to it having no impact on cognitive function [45], the primary outcome in that study [15]. However, it may impact falls risk measures assessed in the current study. While we did not assess between group differences in falls risk, future research aimed at comparing the effects of RT on functional connectivity and falls risk should aim to use a more appropriate control group. Finally, our sample only included those who were living independently in their own homes and able to participate in exercise classes. Thus, our findings may not be generalizable to older adults with advanced cognitive impairment and/or who are more physically frail. Due to sex differences in response to exercise, the results of this study may also not be generalizable to males. Larger randomized controlled trials of RT stratified by sex are required to confirm our current findings.

In conclusion, RT can significantly mitigate WMH-related disruption to the sensorimotor network and may also mitigate disruption to the dorsal and ventral attention networks in older females with cerebral small vessel disease. Reduced WMH-related disruption to the dorsal attention and ventral attention networks may promote executive functions and reduce falls risk, respectively. These preliminary findings suggest RT may be an efficacious intervention to target WMH-related disruption to functional networks that are associated with cognitive and mobility outcomes in cerebral small vessel disease.

Footnotes

ACKNOWLEDGMENTS

We extend our sincere gratitude to Dr. Michael Fox from the Laboratory for Brain Network and Imaging Modulation at Harvard Medical School for providing his expertise and support with lesion network mapping. RAC and TL-A were involved in study conception, design, analyses, and interpretation of the results. RAC also wrote the first draft of the manuscript. CLH, ED, JJE, and TH were involved in the analyses and interpretation of the findings. RT was involved in data collection and interpretation. All authors critically evaluated and approved the manuscript.

Funding was provided by the Vancouver Foundation (BCM06–0035), the Michael Smith Foundation for Health Research (MSFHR; CI-SCH-063 [05 –]), and the Jack Brown and Family Alzheimer Research Foundation Society to TLA. RAC is a recipient of the UBC Rehabilitation Sciences Doctoral Award, CLH is a Canadian Institutes of Health Research Postdoctoral Fellow, and ED is a Michael Smith Foundation of Health Research Post-doctoral Fellow. JJE is a Canada Research Chair (Tier I) in Neurological Rehabilitation and TL-A is a Canada Research Chair (Tier II) in Physical Activity, Mobility, and Cognitive Neuroscience.

Authors’ disclosures available online ().

References

de Leeuw

, de Groot

, Achten

, Oudkerk

, Ramos

, Heijboer

, Hofman

, Jolles

, van Gijn

, Breteler

(2001) Prevalence of cerebral white matter lesions in elderly people: A population based magnetic resonance imaging study. The Rotterdam Scan Study. J Neurol Neurosurg Psychiatry 70, 9–14.

Kloppenborg

, Nederkoorn

, Geerlings

, van den Berg

(2014) Presence and progression of white matter hyperintensities and cognition: A meta-analysis. Neurology 82, 2127–2138.

Zheng

, Delbaere

, Close

, Sachdev

, Lord

(2011) Impact of white matter lesions on physical functioning and fall risk in older people: A systematic review. Stroke 42, 2086–2090.

Srikanth

, Beare

, Blizzard

, Phan

, Stapleton

, Chen

, Callisaya

, Martin

, Reutens

(2009) Cerebral white matter lesions, gait, and the risk of incident falls: A prospective population-based study. Stroke 40, 175–180.

Pantoni

, Poggesi

, Inzitari

(2009) Cognitive decline and dementia related to cerebrovascular diseases: Some evidence and concepts. Cerebrovasc Dis 27 Suppl 1, 191–196.

Smith

, Salat

, Jeng

, McCreary

, Fischl

, Schmahmann

, Dickerson

, Viswanathan

, Albert

, Blacker

, Greenberg

(2011) Correlations between MRI white matter lesion location and executive function and episodic memory. Neurology 76, 1492–1499.

van Dijk

, Prins

, Vrooman

, Hofman

, Koudstaal

, Breteler

(2008) Progression of cerebral small vessel disease in relation to risk factors and cognitive consequences: Rotterdam Scan study. Stroke 39, 2712–2719.

Pantoni

, Leys

, Fazekas

, Longstreth

Jr. , Inzitari

, Wallin

, Filippi

, Scheltens

, Erkinjuntti

, Hachinski

(1999) Role of white matter lesions in cognitive impairment of vascular origin. Alzheimer Dis Assoc Disord 13 Suppl 3, S49–54.

Callisaya

, Beare

, Phan

, Blizzard

, Thrift

, Chen

, Srikanth

(2015) Progression of white matter hyperintensities of presumed vascular origin increases the risk of falls in older people. J Gerontol A Biol Sci Med Sci 70, 360–366.

10.

Zheng

, Lord

, Close

, Sachdev

, Wen

, Brodaty

, Delbaere

(2012) Brain white matter hyperintensities, executive dysfunction, instability, and falls in older people: A prospective cohort study. J Gerontol A Biol Sci Med Sci 67, 1085–1091.

11.

Friston

(1998) The disconnection hypothesis. Schizophr Res 30, 115–125.

12.

Ter Telgte

, van Leijsen

EMC

, Wiegertjes

, Klijn

CJM

, Tuladhar

, de Leeuw

(2018) Cerebral small vessel disease: From a focal to a global perspective. Nat Rev Neurol 14, 387–398.

13.

Bolandzadeh

, Tam

, Handy

, Nagamatsu

, Hsu

, Davis

, Dao

, Beattie

, Liu-Ambrose

(2015) Resistance training and whitematter lesion progression in older women: Exploratory analysis of a12-month randomized controlled trial.. J Am Geriatr Soc 63, 2052–2060.

14.

Suo

, Singh

, Gates

, Wen

, Sachdev

, Brodaty

, Saigal

, Wilson

, Meiklejohn

, Singh

, Baune

, Baker

, Foroughi

, Wang

, Mavros

, Lampit

, Leung

, Valenzuela

(2016) Therapeutically relevant structural and functional mechanisms triggered by physical and cognitive exercise. Mol Psychiatry 21, 1645.

15.

Liu-Ambrose

, Nagamatsu

, Graf

, Beattie

, Ashe

, Handy

(2010) Resistance training and executive functions: A 12-month randomized controlled trial. Arch Intern Med 170, 170–178.

16.

Ding

, Ding

, Hua

, Xiong

, Wen

, Ding

, Wang

, Thompson

(2018) Altered connectivity patterns among resting state networks in patients with ischemic white matter lesions. Brain Imaging Behav 12, 1239–1250.

17.

Benson

, Hildebrandt

, Lange

, Schwarz

, Kobe

, Sommer

, Floel

, Wirth

(2018) Functional connectivity in cognitive control networks mitigates the impact of white matter lesions in the elderly. Alzheimers Res Ther 10, 109.

18.

Sun

, Qin

, Zhou

, Xu

, Qian

, Tao

, Xu

(2011) Abnormal functional connectivity in patients with vascular cognitive impairment, no dementia: A resting-state functional magnetic resonance imaging study. Behav Brain Res 223, 388–394.

19.

Vipin

, Loke

, Liu

, Hilal

, Shim

, Xu

, Tan

, Venketasubramanian

, Chen

, Zhou

(2018) Cerebrovascular disease influences functional and structural network connectivity in patients with amnestic mild cognitive impairment and Alzheimer’s disease. Alzheimers Res Ther 10, 82.

20.

Atwi

, Metcalfe

AWS

, Robertson

, Rezmovitz

, Anderson

, MacIntosh

(2018) Attention-related brain activation is altered in older adults with white matter hyperintensities using multi-echo fMRI. Front Neurosci 12, 748.

21.

Appelman

, Exalto

, van der Graaf

, Biessels

, Mali

, Geerlings

(2009) White matter lesions and brain atrophy: More than shared risk factors? A systematic review. Cerebrovasc Dis 28, 227–242.

22.

van Dijk

, Breteler

, Schmidt

, Berger

, Nilsson

, Oudkerk

, Pajak

, Sans

, de Ridder

, Dufouil

, Fuhrer

, Giampaoli

, Launer

, Hofman

, Consortium

(2004) The association between blood pressure, hypertension, and cerebral white matter lesions: Cardiovascular determinants of dementia study. Hypertension 44, 625–630.

23.

Schulz

, Malherbe

, Cheng

, Thomalla

, Schlemm

(2021) Functional connectivity changes in cerebral small vessel disease - a systematic review of the resting-state MRI literature. BMC Med 19, 103.

24.

Fox

(2018) Mapping symptoms to brain networks with the human connectome. N Engl J Med 379, 2237–2245.

25.

Crockett

, Hsu

, Dao

, Tam

, Eng

, Handy

, Liu-Ambrose

(2021) Painting by lesions: White matter hyperintensities disrupt functional networks and global cognition. Neuroimage 236, 118089.

26.

Crockett

, Hsu

, Dao

, Tam

, Alkeridy

, Eng

, Handy

, Liu-Ambrose

(2021) Mind the gaps: Functional networks disrupted by white matter hyperintensities are associated with greater falls risk. Neurobiol Aging 109, 166–175.

27.

Solodkin

, Hlustik

, Chen

, Small

(2004) Fine modulation in network activation during motor execution and motor imagery. Cereb Cortex 14, 1246–1255.

28.

Bagarinao

, Watanabe

, Maesawa

, Mori

, Hara

, Kawabata

, Yoneyama

, Ohdake

, Imai

, Masuda

, Yokoi

, Ogura

, Taoka

, Koyama

, Tanabe

, Katsuno

, Wakabayashi

, Kuzuya

, Ozaki

, Hoshiyama

, Isoda

, Naganawa

, Sobue

(2019) Reorganization of brain networks and its association with general cognitive performance over the adult lifespan. Sci Rep 9, 11352.

29.

Vossel

, Geng

, Fink

(2014) Dorsal and ventral attention systems: Distinct neural circuits but collaborative roles. Neuroscientist 20, 150–159.

30.

Colcombe

, Kramer

(2003) Fitness effects on the cognitive function of older adults: A meta-analytic study. Psychol Sci 14, 125–130.

31.

Barha

, Hsu

, Ten Brinke

, Liu-Ambrose

(2019) Biological sex: A potential moderator of physical activity efficacy on brain health. Front Aging Neurosci 11, 329.

32.

Groll

, To

, Bombardier

, Wright

(2005) The development of a comorbidity index with physical function as the outcome. J Clin Epidemiol 58, 595–602.

33.

Pickering

, Hall

, Appel

, Falkner

, Graves

, Hill

, Jones

, Kurtz

, Sheps

, Roccella

, Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research (2005) Recommendations for blood pressure measurement in humans and experimental animals: Part 1: Blood pressure measurement in humans: A statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research. Hypertension 45, 142–161.

34.

Graf

, Uttl

, Tuokko

(1995) Color- and picture-word Stroop tests: Performance changes in old age. J Clin Exp Neuropsychol 17, 390–415.

35.

Lord

, Menz

, Tiedemann

(2003) A physiological profile approach to falls risk assessment and prevention. Phys Ther 83, 237–252.

36.

Lord

, Ward

, Williams

, Anstey

(1994) Physiological factors associated with falls in older community-dwelling women. J Am Geriatr Soc 42, 1110–1117.

37.

Ambrose

, Paul

, Hausdorff

(2013) Risk factors for falls among older adults: A review of the literature. Maturitas 75, 51–61.

38.

Tinetti

, Kumar

(2010) The patient who falls: “It’s always atrade-off”. JAMA 303, 258–266.

39.

Lord

, Clark

, Webster

(1991) Postural stability andassociated physiological factors in a population of aged persons. J Gerontol 46, M69–76.

40.

Smith

(2002) Fast robust automated brain extraction. Hum Brain Mapp 17, 143–155.

41.

Smith

, Brady

(1997) SUSAN - A new approach to low level image processing. Int J Comput Vis 23, 45–78.

42.

Sled

, Zijdenbos

, Evans

(1998) A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE Trans Med Imaging 17, 87–97.

43.

McAusland

, Tam

, Wong

, Riddehough

, Li

(2010) Optimizing the use of radiologist seed points for improved multiple sclerosis lesion segmentation. IEEE Trans Biomed Eng 57. doi: 10.1109/TBME.2010.2055865.

44.

Yeo

, Krienen

, Sepulcre

, Sabuncu

, Lashkari

, Hollinshead

, Roffman

, Smoller

, Zollei

, Polimeni

, Fischl

, Liu

, Buckner

(2011) The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J Neurophysiol 106, 1125–1165.

45.

Colcombe

, Kramer

, Erickson

, Scalf

, McAuley

, Cohen

, Webb

, Jerome

, Marquez

, Elavsky

(2004) Cardiovascular fitness, cortical plasticity, and aging. Proc Natl Acad Sci U S A 101, 3316–3321.

46.

Caunca

, Simonetto

, Cheung

, Alperin

, Lee

, Elkind

MSV

, Sacco

, Rundek

, Wright

(2020) Diastolic blood pressure is associated with regional white matter lesion load: The Northern Manhattan Study. Stroke 51, 372–378.

47.

Cohen

, Dunbar

, McClelland

(1990) On the control of automatic processes: A parallel distributed processing account of the Stroop effect. Psychol Rev 97, 332–361.

48.

Stumme

, Jockwitz

, Hoffstaedter

, Amunts

, Caspers

(2020) Functional network reorganization in older adults: Graph-theoretical analyses of age, cognition and sex. Neuroimage 214, 116756.

49.

Crockett

, Hsu

, Best

, Liu-Ambrose

(2017) Resting state default mode network connectivity, dual task performance, gait speed, and postural sway in older adults with mild cognitive impairment. Front Aging Neurosci 9, 423.

50.

Crockett

, Hsu

, Best

, Beauchet

, Liu-Ambrose

(2019) Head over heels but I forget why: Disruptive functional connectivity in older adult fallers with mild cognitive impairment. Behav Brain Res 376, 112104.

51.

Wang

, Zhou

, Yao

, Zhan

, Zhang

, Cui

, Xu

, Ma

, Wang

, An

, Zhang

, Liu

, Jiang

(2015) Aberrant intra- and inter-network connectivity architectures in Alzheimer’s disease and mild cognitive impairment. Sci Rep 5, 14824.