Effects of Marathon Running on Cognition and Retinal Vascularization: A Longitudinal Observational Study : Medicine & Science in Sports & Exercise

Journal Logo


Effects of Marathon Running on Cognition and Retinal Vascularization: A Longitudinal Observational Study


Author Information
Medicine & Science in Sports & Exercise 53(10):p 2207-2214, October 2021. | DOI: 10.1249/MSS.0000000000002699


Recent estimates of the global burden of disease and its development from 2016 until 2040 indicate that three of the top 5 causes of death will be cardiovascular or neuropsychiatric diseases (1). Therefore, we need to further deepen our understanding of the association between the cardiovascular system and the central nervous system and their interaction under different conditions. Earlier studies identified a possible physiological relationship between neuropsychiatric and cardiovascular parameters at rest. For example, decreased cerebral blood flow is a possible modulator of cognitive functioning and can precede cognitive decline (2). Cerebrovascular dysfunction is a common mechanism of different types of cognitive decline (3).

Retinal fundus photography has been proposed as a tool for evaluating alterations in cerebrovascular functioning because retinal vascularization parameters, e.g., narrower arteriolar and wider venular diameters, serve as early biomarkers for the development of cognitive decline with and without dementia (4).

The beneficial effects of regular and moderate exercise on the cardiovascular system and the preventive effect on associated diseases are well established and appear to be mediated by vascular adaptations, among other factors (5). However, the amount and intensity of exercise required for lowering mortality rates is still under discussion (6).

Vascular adaptations through the regulation of retinal vascularization after different forms of exercise are also complex. In general, higher cardiorespiratory fitness is associated with narrower retinal veins, indicating beneficial effects of exercise on retinal circulation across the life span (7,8). Earlier studies by our group investigated the acute effects of marathon running on retinal circulation and showed an increase in the arteriolar-to-venular ratio (AVR) (9). Other studies found wider retinal arteries and veins after submaximal and maximal treadmill tests lasting 5 and 40 min (10), and in follow-up periods of up to 2.5 h (11), indicating rapid vascular adaptations to exercise.

As stated above, physical activity also has beneficial effects on cognition and improves the functioning of higher-order regions of the brain in particular. Moreover, it has beneficial effects on aging in general and on clinically relevant cognitive decline (12). The association between exercise intensity or duration and cognition seems to be an inverted U-shape: very strenuous exercise (e.g., ultramarathon running) can lead to decreased cognitive performance (13).

Thus, the existing literature emphasizes the mutual relationship of neuropsychiatric and cardiovascular parameters. Physical activity is a well-established moderator of both modalities on their own, but observations of their interaction during and after exercise remain controversial. Furthermore, the increasing knowledge on vascular alterations and cognitive decline (1) and their proposed relationship (3) warrants studies to identify possible moderators of both modalities. Therefore, the present study aimed to simultaneously assess cognitive and vascular parameters after different types of running. To the best of our knowledge, this is the first study to examine cognitive and vascular parameters during a training period, an acute bout of strenuous exercise (marathon race), and a short- and long-term follow-up period in one sample and to compare baseline values in marathon runners and sedentary controls. We hypothesized that over the study period, vascular adaptations to exercise would correlate with cognitive adaptations, indicating a possible underlying causal relationship.


The running effects on cognition and plasticity trial is a longitudinal observational study of marathon runners who registered for the Munich Marathon 2017. The detailed study protocol was previously published (14). Inclusion criteria for the marathon group were age between 18 and 60 yr and successful registration for the Munich Marathon 2017. All participants had to have completed at least one half-marathon, possess sufficient German language skills, and be able to provide written informed consent. Exclusion criteria were relevant neurological, cardiac, or psychiatric diseases; pregnancy; and cannabis abuse. The exclusion criteria also applied to the sedentary control group. A sedentary lifestyle was defined as performing fewer than 25 min of physical activity per day, and participants of the sedentary control group were screened for this criterion before inclusion in the study (15).

As detailed in the supplemental content, neurocognitive assessments included three different loads (1-back, 2-back, and 3-back) of the computer-based n-back task to assess continuous performance in a working memory task; Trail Making Tests A (TMT A) and B (TMT B) to evaluate executive functioning; visual scanning and graphomotor speed; and the d2 test to evaluate concentration and attention, assessed as concentration performance (CP) (see Document, Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334.)

Retinal vessel analyses were performed with a nonmydriatic, noninvasive fundus camera using the Static Retinal Vessel Analyzer (SVA-T; Imedos Systems UG, Jena, Germany). Per participant, two valid images per eye with a centered papilla were recorded in a darkened room. Intraocular pressure was not measured. The analysis of the images was conducted offline using analysis software Vesselmap 2 (Visualis, Imedos Systems UG). First, a papilla circle (inner circle) with a fixed diameter was marked. Following this, a thread circle with a center in the vessel origin was marked on the papilla, and the outer rings were outlined, representing the boundary of the analysis area. Veins (blue) and arteries (red) were marked between the outer and the middle rings in succession. All relevant vessels were included in the calculation. The vessel diameters were calculated automatically by the Vesselmap software. Vessels smaller than 45 μm were detected but not included in the calculation, and all marked vessels were visually verified for plausibility afterward. Figure 1 shows an exemplary image with a measurement area.

Exemplary analysis of the fundus images using analysis software Vesselmap 2 (Visualis, Imedos Systems UG). The inner circle represents a papilla circle with a fixed diameter. A thread circle with a center in the vessel origin is marked on the papilla, and the outer rings were outlined, representing the boundary of the analysis area. Veins are displayed in blue, and arteries are displayed in red.

Retinal vascular parameters included the mean values of left and right measurements of central retinal arteriolar equivalent (CRAE) and central retinal venular equivalents (CRVE) defined using the formula according to Parr-Hubbard and the AVR (16). The AVR was calculated from the quotients of CRAE and CRVE.

The study timeline, with six measurements over 6 months, is presented in the supplemental content (see Fig. 1, Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334) and described in detail elsewhere (14). In brief, participants were assessed 12 wk before (visit −1) and 1 wk before the marathon (visit 0), immediately after successful completion of the marathon (visit 1; all measurements at visit 1 were performed within 2 h of the individual completing the marathon), at 24 h (visit 2.1) and 72 h (visit 2.2) after completing the marathon, and 12 wk after the marathon (see Supplemental Digital Content, Appendix, for more details, https://links.lww.com/MSS/C334).

Statistical analysis

All analyses were performed in SPSS 25 (IBM, Armonk, NY), with a significance level of α = 0.05. Demographic and clinical differences between marathon runners and sedentary controls were assessed with independent t-tests and χ2 tests. Homogeneity of variances was tested with Levene’s test, and, if violated (P < 0.05), a degrees of freedom correction was applied. Changes of cognitive outcome for the within-subject factor time were evaluated by general linear mixed models (LMM) with a diagonal covariance matrix. In case of a significant main effect of time, within-subject differences were calculated by testing a given measurement against baseline and adjusting for multiple comparisons (Sidak correction). For analyses of correlations of cognitive and vascular parameters (with and without additional covariates), generalized LMM were used (general estimation equation [GEE]). In these models, in a second analysis, we included systolic blood pressure (RRsys) and age because previous studies established an association of AVR with these two parameters (17) and of cognitive parameters with age (18). Furthermore, we analyzed the acute (visit 0 vs visit 1) and chronic effects (all visits except visit 1) of exercise separately to differentiate between these two forms of exercise. For acute effects, we chose visit 0 because we wanted to minimize learning effects and effects of the intensive training period.


The baseline demographic characteristics of the marathon runners are presented in Table 1. A total of 100 participants were initially included in the study. Of these, 94 completed visit −1 with valid cognition data and 92 with valid vascular data. In total, 71 participants finished the marathon, and 58 completed the final follow-up visit with valid cognitive data and 57 with valid vascular data. Fifty-one participants completed all six visits, and 10 participants had one to two missing intermediate visits for time reasons (three of them missed the final follow-up). Reasons for participants to withdraw completely from the study after inclusion were as follows: time constraint (n = 18), internal diseases (n = 6), orthopedic disease (n = 7), marathon not finished (n = 4), termination of the study due to personal reasons (n = 1), and exclusion criterion for marathon at V-1 (n = 1).

TABLE 1 - Baseline characteristics, cognitive parameters, and vascular parameters of marathon group and sedentary controls.
Variable MA SC χ 2 df P n (MA/SC)
Sex, n, male:female 80:20 35:11 0.288 1 0.591 146 (100/46)
Smoking, yes:no 1:84 10:31 18.706 1 <0.001 126 (85/41)
Mean SD Mean SD t df P
Age, yr 43.6 10.0 40.8 11.0 1.529 140 0.128 142 (96/46)
Education, yr 15.0 3.9 14.3 3.3 1.040 128 0.300 130 (88/42)
BMI, kg·m−2 23.5 2.7 24.8 2.7 −2.544 133 0.012 135 (93/42)
RRsys 122.4 12.1 126.3 13.8 −1.673 133 0.097 135 (93/42)
RRdia 79.8 7.6 81.0 7.7 −0.786 133 0.434 135 (93/42)
IPAQ score visit −1 7410.2 7450.4 3378.1 6526.7 2.963 130 0.004 132 (92/40)
IPAQ score visit 3 4995.5 5550.4 55 (55/0)
Finishing time, min 236.8 37.6 71 (71/0)
V˙O2max visit −1, mL·kg−1⋅min−1 46.5 6.6 92 (92/0)
V˙O2max visit 0, mL·kg−1⋅min−1 48.3 6.3 66 (66/0)
d prime 3-back 1.55 0.58 1.47 0.70 0.728 133 0.468 135 (94/41)
d prime 2-back 2.82 0.88 2.87 1.07 −0.282 133 0.778 135 (94/41)
d prime 1-back 4.11 0.61 3.85 0.81 1.792 60.442 0.078 135 (94/41)
d2 CP 182.71 36.20 192.78 54.44 −1.083 56.361 0.283 135 (92/41)
TMT A 26.57 9.21 27.24 11.84 −0.355 131 0.723 135 (92/41)
TMT B 59.30 22.51 64.65 31.33 −0.993 61.277 0.324 135 (92/41)
Mean AVR 0.89 0.07 0.85 0.06 2.745 85.155 0.007 129 (92/37)
Mean CRAE 194.22 17.96 199.40 20.59 −1.419 127 0.158 129 (92/37)
Mean CRVE 219.82 18.82 234.04 21.32 −3.735 127 <0.001 129 (92/37)
MA, marathon group; SC, sedentary controls; BMI, body mass index; RRsys, blood pressure systolic; RRdia, bood pressure diastolic; IPAQ, International Physical Activity Questionnaire; Finishing time, finishing time in Munich Marathon 2017; V˙O2max, maximal oxygen uptake in spiroergometry; d prime, discriminability index; d2 CP, d2 test concentration performance.

In the retinal vessel analysis, four participants of the marathon group and one participant of the sedentary control group only provided data for one eye (one participant because of unilateral eye treatment, and four participants because of impaired picture quality). In these cases, AVR, CRAE, and CRVE mean values were calculated from the data from one eye.

Descriptive results of cognitive and vascular parameters over the complete study period are presented in Table 2. Figures 2 and 3 display d prime 3-back and AVR over the complete study period.

TABLE 2 - Descriptive results in cognitive and vascular parameters over time in marathon runners.
V-1 n V0 n V1 n V2.1 n V2.2 n V3 n
d prime 3-back 1.55 ± 0.58 94 1.67 ± 0.68 78 1.67 ± 0.74 65 2.16 ± 0.93 56 1.98 ± 0.89 56 1.94 ± 0.86 58
d prime 2-back 2.82 ± 0.88 94 3.18 ± 0.89 78 3.10 ± 1.04 65 3.47 ± 0.88 56 3.46 ± 0.94 56 3.50 ± 0.84 58
d prime 1-back 4.11 ± 0.61 94 4.24 ± 0.50 78 3.98 ± 0.66 65 4.36 ± 0.44 56 4.20 ± 0.67 56 4.29 ± 0.49 58
TMT A (s) 26.57 ± 9.21 91 22.02 ± 9.30 81 19.84 ± 6.87 66 20.80 ± 11.82 63 19.11 ± 6.80 56 20.28 ± 7.42 57
TMT B (s) 59.30 ± 22.5 91 52.20 ± 20.60 81 47.75 ± 14.06 66 43.84 ± 17.78 63 40.07 ± 18.21 56 44.27 ± 19.20 57
d2 CP 182.71 ± 36.20 92 197.66 ± 46.37 79 211.97 ± 44.05 63 219.73 ± 46.81 59 230.42 ± 43.44 57 220.93 ± 43.69 58
Mean AVR 0.89 ± 0.07 92 0.88 ± 0.07 77 0.89 ± 0.06 65 0.90 ± 0.07 60 0.89 ± 0.07 58 0.88 ± 0.06 57
Mean CRVE 219.82 ± 18.82 92 221.81 ± 19.17 77 231.89 ± 20.12 65 215.98 ± 16.47 60 216.80 ± 19.03 58 220.58 ± 19.99 57
Mean CRAE 194.22 ± 17.96 92 195.28 ± 18.80 77 205.04 ± 16.40 65 193.29 ± 16.02 60 192.66 ± 16.94 58 193.64 ± 17.10 57
RRsys (mm Hg) 122.37 ± 12.15 93 121.97 ± 16.02 76 109.21 ± 16.26 67 124.76 ± 12.91 62 126.02 ± 15.33 59 120.92 ± 12.47 60
All values are presented as mean ± SD; d prime, discriminability index; RRsys, systolic blood pressure; d2 CP, d2 test concentration performance.
V-1, baseline (10 to 12 wk before marathon); V0, 2 wk before marathon; V1, immediately after the marathon; V2.1, 24 h after the marathon; V2.2, 72 h after the marathon; V3, 10 to 12 wk after the marathon.

d prime 3-back over the study period. Box plots of LMM analysis of the marathon cohort including all six visits (descriptive statistics in Table 2). V-1, baseline (10 to 12 wk before marathon); V0, within 2 wk before marathon; V1, within 2 h after the marathon; V2.1, 24 h after the marathon; V2.2, 72 h after the marathon; V3, 10 to 12 wk after the marathon.
AVR over the study period. Box plots of LMM analysis of the marathon cohort including all six visits (descriptive statistics in Table 2). V-1, baseline (10 to 12 wk before marathon); V0, within 2 wk before marathon; V1, within 2 h after the marathon; V2.1, 24 h after the marathon; V2.2, 72 h after the marathon; V3, 10 to 12 wk after the marathon.

Correlation of chronic alterations of vascular and cognitive parameters

The most consistent chronic effects (including all visits except for visit 1 immediately after the marathon) were an improvement of cognitive parameters and correlations with vascular parameters.

After correction for the described covariates (RRsys and age; see Table 2, Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334), d prime 3-back correlated positively with AVR (P = 0.024, regression coefficient B = 1.861, SE = 0.824) and negatively with CRVE (P = 0.05, B = −0.006, SE = 0.003). Similarly, TMT B correlated negatively with CRAE (P = 0.025, B = −0.155, SE = 0.069). Age had the greatest effect on d2 CP (P < 0.001, B = −1.555, SE = 0.424 with AVR) but also affected other high-demand cognitive tasks (d prime 3-back, P = 0.011, B = −0.021, SE = 0.008 with CRVE; d prime 2-back, P = 0.033, B = −0.018, SE = 0.009 with CRVE; TMT B, P = 0.022, B = 0.407, SE = 0.178 with AVR). In the analyses that excluded visit 1, RRsys correlated with none of the parameters.

Results and details of the GEE analyses with (Table 2, Supplemental Digital Content, https://links.lww.com/MSS/C334) and without (Table 1, Supplemental Digital Content, https://links.lww.com/MSS/C334) covariates can be found in the supplemental content (Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334).

Separate analysis of chronic cognitive and vascular processes

Regarding the longitudinal changes of cognitive performance, significant effects of time were observed in all LMM analyses. Supplemental Table 3 (Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334) displays descriptive statistics and the results of the post hoc tests in case of a significant effect of time (respective visit vs baseline [V-1]). Again, analyses showed the greatest improvement in d prime 3-back at the visit 24 h after the marathon (V2.1); d prime 2-back and TMT A/B also showed the greatest improvement at V2.1 and remained stable afterward.

Correlation of acute alterations of vascular and cognitive parameters

The acute effects (V0 vs V1 immediately after the marathon) showed a different pattern. In the GEE analysis after correction for RRsys and age (see Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334), we observed a positive correlation of CRAE with d2 CP (P = 0.020, B = 0.492, SE = 0.211) and a negative correlation of CRVE with d prime 1-back (P = 0.016, B = −0.007, SE = 0.003) and TMT A (P = 0.028, B = −0.075, SE = 0.034). In these analyses, age had the highest effect on d2 CP (negative association, P < 0.001, B = −1.855, SE = 0.474 with AVR and P < 0.001, B = −1.861, SE = 0.494 with CRVE), whereas RRsys correlated strongest with TMT A (P = 0.003, B = 1.124, SE = 0.043 with AVR, positive association).

Results and details of the analyses with (Table 5, Supplemental Digital Content, https://links.lww.com/MSS/C334) and without (Table 4, Supplemental Digital Content, https://links.lww.com/MSS/C334) covariates can be found in the supplemental content (Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334).

Separate analysis of acute cognitive and vascular processes

Significant effects of time between visits V0 and V1 were seen in d prime 1-back (impairment, P = 0.011, F1,118.331 = 6.747), CRAE (wider diameter, P = 0.00, F1,139.836 = 10.9231), CRVE (wider diameter, P = 0.003, F1,133.609 = 9.239), and RRsys (reduced values, P < 0.001, F1,138.227 = 22.249). Supplemental Table 6 shows the descriptive and LMM results with post hoc tests (Sidak) in case of a significant effect of time (see Supplemental Digital Content, Appendix, https://links.lww.com/MSS/C334).

Interestingly, none of the more demanding cognitive parameters (d prime 2-back and 3-back, TMT B, d2 CP) showed a change immediately after the marathon.

Comparison of cognitive and vascular parameters between marathon runners and sedentary controls

In a subsequent analysis, we compared the results of our marathon runners with those of an age- and sex-matched sedentary control group (Table 1). The analyses showed no difference in cognitive functioning, but they did find a difference in AVR (P = 0.007, t85.155 = 2.745) and CRVE (P < 0.001, t127 = −3.735): The sedentary controls had wider retinal veins than the marathon runners, resulting in reduced AVR values.


In this study, we investigated for the first time the interaction of vascular and cognitive parameters over a 6-month period in people performing acute (marathon race) and chronic exercise (training and recovery periods for the marathon) and were able to demonstrate a close relationship, especially during the chronic exercise period.

Interaction during chronic exercise

In the analysis of the adaptations after chronic exercise (training period, short- and long-term recovery but without acute changes immediately after the marathon), we observed a positive correlation between high-demand working memory tasks (d prime 3-back) and both AVR and, to a smaller (and inverse) extent, CRVE. Improved retinal perfusion (higher AVR, caused mainly by narrower retinal veins [CRVE]) was associated with improved performance in high-demand working memory tasks. Moreover, wider retinal arteries (CRAE) were associated with improvements in other domains of higher demand executive functioning (TMT B task). These findings remained unchanged even after correcting for age and systolic blood pressure as two major possible moderators (18,19).

Few studies have examined the effects of chronic exercise interventions on cognition and retinal vasculature, and the results of these studies are partly inconclusive: A cluster-randomized interventional study in adolescents (N = 36) found an association between wider retinal arteries and improved inhibitory control (Stroop test) but no changes in CRVE (20). Although our study replicated this finding to some extent, at the same time our aforementioned finding of narrower veins (CRVE) after regular exercise is in contrast to the finding of this study (7). The comparison of our study population and the adolescent study has to be discussed cautiously, as our participants were older with less female participants (43.6 ± 10.0 yr and 20% female vs 12.5 ± 0.2 yr and 36% female) and the intervention was different (8 wk with 20 min of aerobic and coordinative exercise session on each school day vs preparation for a marathon). Both age (increasing arteriolar diameters until adulthood and slow decrease after the age of 45) and sex (with mainly higher values of retinal arteriolar caliber and AVR in female participants) can affect the retinal circulation (21). Still, both studies investigated the chronic effects of exercise on cognitive and retinal parameters, and both studies point to a positive interaction in different population. A recent systematic review summarized the effects of exercise on retinal vasculature across the life span and found positive effects in all age-groups (8). For better comparability, future studies should investigate different age cohorts with similar interventions.

Interaction after acute exercise

When examining acute effects immediately after the marathon, our analyses revealed a positive correlation of CRAE with d2 CP (concentration and attention) and a negative correlation of CRVE with d prime 1-back (low-demand working memory) and TMT A (low-demand executive functioning). These findings point in different directions because higher values of CRAE indicate improved vascular status (and correlated with improved performance in the cognitive task d2 CP), whereas reduced values of CRVE (which also indicate improved vascular status) correlated with worse performance in the d prime 1-back and also improved performance in the TMT A.

Previous research investigating the relationship of cognitive performance and retinal vascular status in response to acute exercise obtained different results in that they did not show associations of restored central perfusion and cognitive task performance during exercise (ergometer training) (22). By contrast, our results of acute interactions showed both impairments and improvements in cognitive functioning that correlated with improved retinal vascular status. A possible explanation for these ambiguous observations could be that immediately after a marathon race, multiple other influencing factors (e.g., inflammation, cardiac markers, and fatigue) contribute to both vascular adaptations and cognitive performance and therefore complicate the validity of the cognitive–vascular interaction.

Separate analysis of cognitive and retinal parameters after acute exercise and in the chronic exercise period

We measured retinal vascular status immediately after the marathon and found vasodilation in both CRAE and CRVE. These results are in line with previous studies, which found that different vasoactive mechanisms resulted in transient retinal vasodilatation; however, these studies did not measure cognitive parameters (10,23).

Retinal vascular adaptations in response to acute exercise can be described with different mechanisms: on the one hand vasodilatory effects (e.g., nitric oxide–mediated shear stress) and on the other hand vasoconstrictive effects (e.g., the Bayliss effect, hypocapnia). During high-intensity acute exercise, vasoconstrictive mechanisms seem to override vasodilatory effects to ensure constant blood flow (with increasing blood pressure during exercise). After exercise and a normalization of hemodynamic and metabolic stimuli, vasodilatory mechanisms overbalance (10).

In our study, cognitive performance remained mainly unchanged, except for slight impairment in low-demand working memory (1-back task) after the marathon. Previous studies showed improvements in executive functioning, cognitive speed, and working memory, but they also found reduced reaction times and accuracy after acute exercise (24–26). However, no earlier study has evaluated these factors after moderate-intensity training and prolonged and strenuous exercise, such as a marathon race.

Vascular adaptations in response to chronic exercise include a narrowing of the retinal veins (7). We could replicate this finding when we compared the marathon group with the sedentary control group (i.e., we found wider retinal veins in the control group). The duration of exercise required to cause vascular adaptations seems to be relatively short, and different intensities seem to create similar effects (27). The adaptations throughout our study period were mainly observed around the marathon event. However, a recent investigation of other vascular parameters in response to the training period and a marathon found that central blood pressure and aortic stiffness in 138 first-time marathon completers were significantly reduced, with greater benefit in older, male participants (28). However, our results are not fully comparable with this study because we examined different parameters (retinal vasculature and brachial blood pressure), and our study was not performed exclusively in first-time marathon runners.

With regard to cognition, studies consistently showed beneficial effects of regular aerobic exercise (29,30) and confirmed age as a possible covariate (19). In our study, the observed cognitive adaptations differed when high-demand cognitive tasks (with d prime 3-back, d2 CP, and TMT B) were compared with lower demand cognitive tasks (d prime 1-back and TMT A). Learning effects are well established in cognitive tasks, and low-demand tasks are known to be more susceptible to such learning effects (31,32). The pattern of improvements in the cognitively more demanding tasks showed a peak 24 and 72 h after the marathon, and in these tasks we did not see a continuous increase in performance over time like we did in the less demanding tasks. Thus, our results indicate that the course of the changes in cognitive performance was likely to be independent from potential learning effects.

Physiology of the interaction between acute and chronic physical activity, cognition, and retinal vasculature

Acute and chronic physical activity influences cognitive abilities by modulating multiple pathways, e.g., the adrenaline threshold after incremental exercise (33) and vessel compliance (34). Research provides increasing evidence that neither single bouts of acute exercise nor regular exercise alone fosters cognitive improvements but that a combination of both seems to be important. Earlier findings point to a possible priming effect of chronic exercise and resulting cognitive improvements after acute bouts of exercise, but results are inconclusive (35,36). Some researchers assumed that regular exercise could provide the necessary stimuli and therefore lay the foundation for an increased responsiveness to acute exercise (37,38). Chronic exercise improves retinal vascularization, and both chronic exercise and improved retinal vascularization can be linked to improved cognitive performance (3,4,9,39). Our comparison of the marathon runners with the sedentary control group supports this hypothesis because we did not find any differences in cognitive performance between the two groups, but we did find a significant difference in retinal vascular status (with wider retinal venules in the sedentary group). The better retinal vascular status could potentially represent priming in the marathon group, with a facilitating effect of acute exercise after the marathon (and resulting cognitive improvement). Figure 4 displays this causal relationship graphically.

(Central figure) Effects of chronic and acute exercise on retinal vascularization and cognition. Chronic exercise leads to improved retinal (central) vascularization, as displayed by AVR differences between the marathon cohort and the sedentary control group at baseline. Chronic exercise combined with acute exercise (e.g., marathon running) promotes improved cognitive performance (e.g., working memory).


As mentioned above, learning effects are established in cognitive assessments and should be taken into account; however, the course of changes throughout the study period, especially in the high-demand parameters, point to the intervention as the driving factor. Future studies should include a control group for all visits to further differentiate between exercise-related and learning effects.

Our cohort of marathon runners contained more male than female participants, and the subsample of female runners was too small for subanalyses. Nevertheless, our cohort represents common distributions of marathon participants.

Last, our cohort was not homogenous regarding their previous experience with running marathons, i.e., it included first-time and very experienced marathon runners. The composition of our cohort improves the generalizability of our results, but the inclusion of experienced runners might result in training effects being underestimated.


In this study, we simultaneously assessed for the first time different exercise modalities and their effects on neurocognitive and vascular parameters in a longitudinal design and were able to demonstrate a relationship between cognitive improvements in complex cognitive tasks and positive adaptations in vessel structures and function in responses to exercise. We hypothesize that chronic exercise could prime the central nervous system for possible facilitating effects of acute bouts of exercise. The central vascular status—represented by the retinal measurements of CRAE, CRVE, and their quotient AVR—could play a pivotal role in such a priming process of cognitive improvements. Future studies should further focus on this interaction in larger samples and consider additional possible influencing factors (e.g., sex, catecholamines) or the effect of inflammation to differentiate between the significance of their contributions.

The authors thank Jacquie Klesing, Board-certified Editor in the Life Sciences (ELS), for editing assistance with the manuscript.

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.

The authors declare no conflicts of interest related to the study.

Results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. Results of the present study do not constitute endorsement by the American College of Sports Medicine.

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. The study protocol had been approved by the ethics committees of both the Ludwig-Maximilians University Munich (approval number 17-148) and the Technical University Munich (approval number 218/17S). The study was registered at https://www.drks.de/ (DRKS-ID: DRKS00012496). All participants provided written informed consent before inclusion in the study.


1. Foreman KJ, Marquez N, Dolgert A, et al. Forecasting life expectancy, years of life lost, and all-cause and cause-specific mortality for 250 causes of death: reference and alternative scenarios for 2016-40 for 195 countries and territories. Lancet. 2018;392(10159):2052–90.
2. Ogoh S. Relationship between cognitive function and regulation of cerebral blood flow. J Physiol Sci. 2017;67(3):345–51.
3. Raz L, Knoefel J, Bhaskar K. The neuropathology and cerebrovascular mechanisms of dementia. J Cereb Blood Flow Metab. 2016;36(1):172–86.
4. Dumitrascu OM, Qureshi TA. Retinal vascular imaging in vascular cognitive impairment: current and future perspectives. J Exp Neurosci. 2018;12:1179069518801291.
5. Pinckard K, Baskin KK, Stanford KI. Effects of exercise to improve cardiovascular health. Front Cardiovasc Med. 2019;6:69.
6. Blond K, Brinkløv CF, Ried-Larsen M, Crippa A, Grøntved A. Association of high amounts of physical activity with mortality risk: a systematic review and meta-analysis. Br J Sports Med. 2019;54(20):1195–201.
7. Braun G, Hafner B, Königstein K, et al. Association of cardiorespiratory fitness with retinal vessel diameters as a biomarker of cardiovascular risk. Microvasc Res. 2018;120:36–40.
8. Streese L, Guerini C, Bühlmayer L, et al. Physical activity and exercise improve retinal microvascular health as a biomarker of cardiovascular risk: a systematic review. Atherosclerosis. 2020;315:33–42.
9. Pressler A, Suchy C, Friedrichs T, et al. Running multiple marathons is not a risk factor for premature subclinical vascular impairment. Eur J Prev Cardiol. 2017;24(12):1328–35.
10. Nussbaumer M, Donath L, Fischer M, et al. Effects of acute bouts of endurance exercise on retinal vessel diameters are age and intensity dependent. Age (Dordr). 2014;36(3):9650.
11. Rueddel T, Kneser M, Tost F. Impact of exercise on retinal microvascular regulation measured by dynamic vessel analysis in healthy individuals. Clin Physiol Funct Imaging. 2012;32(2):158–61.
12. Daskalopoulou C, Stubbs B, Kralj C, Koukounari A, Prince M, Prina AM. Physical activity and healthy ageing: a systematic review and meta-analysis of longitudinal cohort studies. Ageing Res Rev. 2017;38:6–17.
13. Hurdiel R, Riedy SM, Millet GP, et al. Cognitive performance and self-reported sleepiness are modulated by time-of-day during a mountain ultramarathon. Res Sports Med. 2018;26(4):482–9.
14. Roeh A, Bunse T, Lembeck M, et al. Running effects on cognition and plasticity (ReCaP): study protocol of a longitudinal examination of multimodal adaptations of marathon running. Res Sports Med. 2020;28(2):241–55.
15. Cabrera de León A, Rodríguez-Pérez MdC, Rodríguez-Benjumeda LM, et al. Sedentarismo: tiempo de ocio activo frente a porcentaje del gasto energético [Sedentary lifestyle: physical activity duration versus percentage of energy expenditure]. Rev Esp Cardiol. 2007;60(3):244–50.
16. Hubbard LD, Brothers RJ, King WN, et al. Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the Atherosclerosis Risk in Communities Study. Ophthalmology. 1999;106(12):2269–80.
17. Liew G, Sharrett AR, Kronmal R, et al. Measurement of retinal vascular caliber: issues and alternatives to using the arteriole to venule ratio. Invest Ophthalmol Vis Sci. 2007;48(1):52–7.
18. Cabeza R, Albert M, Belleville S, et al. Maintenance, reserve and compensation: the cognitive neuroscience of healthy ageing. Nat Rev Neurosci. 2018;19(11):701–10.
19. Stern Y, MacKay-Brandt A, Lee S, et al. Effect of aerobic exercise on cognition in younger adults: a randomized clinical trial. Neurology. 2019;92(9):e905–16.
20. Ludyga S, Köchli S, Pühse U, Gerber M, Hanssen H. Effects of a school-based physical activity program on retinal microcirculation and cognitive function in adolescents. J Sci Med Sport. 2019;22(6):672–6.
21. Dumitrescu AG, Voinea L, Badarau IA, Paun VA, Schowe M, Ciuluvica R. Update on retinal vascular caliber. Rom J Ophthalmol. 2017;61(3):171–80.
22. Komiyama T, Tanoue Y, Sudo M, et al. Cognitive impairment during high-intensity exercise: influence of cerebral blood flow. Med Sci Sports Exerc. 2019;52(3):561–8.
23. Pournaras CJ, Rungger-Brändle E, Riva CE, Hardarson SH, Stefansson E. Regulation of retinal blood flow in health and disease. Prog Retin Eye Res. 2008;27(3):284–330.
24. Lefferts WK, DeBlois JP, White CN, Heffernan KS. Effects of acute aerobic exercise on cognition and constructs of decision-making in adults with and without hypertension. Front Aging Neurosci. 2019;11:–41.
25. Moreau D, Chou E. The acute effect of high-intensity exercise on executive function: a meta-analysis. Perspect Psychol Sci. 2019;14(5):734–64.
26. McMorris T, Sproule J, Turner A, Hale BJ. Acute, intermediate intensity exercise, and speed and accuracy in working memory tasks: a meta-analytical comparison of effects. Physiol Behav. 2011;102(3–4):421–8.
27. Hanssen H, Minghetti A, Faude O, et al. Effects of different endurance exercise modalities on retinal vessel diameters in unipolar depression. Microvasc Res. 2018;120:111–6.
28. Bhuva AN, D’Silva A, Torlasco C, et al. Training for a first-time marathon reverses age-related aortic stiffening. J Am Coll Cardiol. 2020;75(1):60–71.
29. Stern Y, Lee S, Predovan D, Sloan RP. Sex moderates the effect of aerobic exercise on some aspects of cognition in cognitively intact younger and middle-age adults. J Clin Med. 2019;8(6):886.
30. Northey JM, Cherbuin N, Pumpa KL, Smee DJ, Rattray B. Exercise interventions for cognitive function in adults older than 50: a systematic review with meta-analysis. Br J Sports Med. 2018;52(3):154–60.
31. Brickenkamp R. Handbuch Psychologischer und Pädagogischer Tests [Handbook of Psychological and Educational Tests]. 2nd ed. Göttingen: Hogrefe; 1997.
32. Soveri A, Lehtonen M, Karlsson LC, Lukasik K, Antfolk J, Laine M. Test-retest reliability of five frequently used executive tasks in healthy adults. Appl Neuropsychol Adult. 2018;25(2):155–65.
33. Brisswalter J, Collardeau M, René A. Effects of acute physical exercise characteristics on cognitive performance. Sports Med. 2002;32(9):555–66.
34. Voss MW, Vivar C, Kramer AF, van Praag H. Bridging animal and human models of exercise-induced brain plasticity. Trends Cogn Sci. 2013;17(10):525–44.
35. Ludyga S, Gerber M, Brand S, Holsboer-Trachsler E, Pühse U. Acute effects of moderate aerobic exercise on specific aspects of executive function in different age and fitness groups: a meta-analysis. Psychophysiology. 2016;53(11):1611–26.
36. Li L, Zhang S, Cui J, et al. Fitness-dependent effect of acute aerobic exercise on executive function. Front Physiol. 2019;10:902.
37. Roig M, Nordbrandt S, Geertsen SS, Nielsen JB. The effects of cardiovascular exercise on human memory: a review with meta-analysis. Neurosci Biobehav Rev. 2013;37(8):1645–66.
38. Berchtold NC, Chinn G, Chou M, Kesslak JP, Cotman CW. Exercise primes a molecular memory for brain-derived neurotrophic factor protein induction in the rat hippocampus. Neuroscience. 2005;133(3):853–61.
39. Shalev I, Moffitt TE, Wong TY, et al. Retinal vessel caliber and lifelong neuropsychological functioning: retinal imaging as an investigative tool for cognitive epidemiology. Psychol Sci. 2013;24(7):1198–207.


Supplemental Digital Content

Copyright © 2021 by the American College of Sports Medicine