Available online at www.sciencedirect.

com

Journal of Sport and Health Science 9 (2020) 74 81

www.jshs.org.cn

Review
Updated overview on interplay between physical exercise, neurotrophins,
and cognitive function in humans
Giuseppe Lippi a, Camilla Mattiuzzi b, Fabian Sanchis-Gomar c,*
a
Section of Clinical Biochemistry, University of Verona, Verona 37129, Italy
b
Service of Clinical Governance, Provincial Agency for Social and Sanitary Services, Trento 38123, Italy
c
Department of Physiology, Faculty of Medicine, University of Valencia and INCLIVA Biomedical Research Institute, Valencia 46010, Spain
Received 27 June 2019; revised 15 July 2019; accepted 16 July 2019
Available online 6 September 2019

Abstract
The many important benefits of physical exercise also encompass maintenance or improvement of cognitive functions. Among the various
mechanisms underlying the association between physical exercise and brain health, recent evidence attests that neurotrophin receptor signaling
may have an important role, because the activation of this pathway leads to growth and differentiation of new neurons and synapses, supports
axonal and dendritic growth, fosters synaptic plasticity, and preserves survival of existing neurons. In this review of published evidence, we high-
light that a positive relationship exists between physical exercise and circulating brain-derived neurotrophic factor levels and that the postexer-
cise variation of this molecule is associated with improvement of neurocognitive functioning. Less clear evidence has instead been published for
other neurotrophins, such as nerve growth factor, neurotrophin-3, and neurotrophin-4. Overall, promotion of adequate volumes and intensities of
physical exercise (i.e., approximately 3 months of moderate-intensity aerobic exercise, with 2 3 sessions/week lasting not less than 30 min)
may hence be regarded as an inexpensive and safe strategy for boosting brain-derived neurotrophic factor release, thus preserving or restoring
cognitive functions.
2095-2546/Ó 2020 Published by Elsevier B.V. on behalf of Shanghai University of Sport. This is an open access article under the CC BY-NC-ND
license. (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Cognitive function; Neurotrophins; Physical exercise; Sport

1. Introduction existing neurons.3 All these biological functions finally con-

verge to modulate personality traits, along with executive and
Neurotrophins are a heterogeneous class of evolutionarily
cognitive functions (e.g., learning, memory). Accordingly,
ancient nerve growth factors (NGFs) that exert a kaleidoscope
reliable evidence has been published that an impairment of
of important biological activities.1 The most important pro-
neurotrophins signaling, especially the BDNF pathway, may
teins belonging to the neurotrophins family include NGF,
be associated with some severe cognitive disorders (e.g.,
brain-derived neurotrophic factor (BDNF), neurotrophin-3
Alzheimer’s and Huntington’s diseases) or depression.4,5
(NT-3), and neurotrophin-4 (NT-4).2 Each of these molecules
Hence, it is obvious that strategies aimed at sustaining or even
can bind to 1 or more cellular receptors and mediate their bio-
amplifying production of neurotrophins may generate substan-
logical function in vivo (Table 1). Although a thorough discus-
tial cognitive benefits in humans.6
sion of the complexity of the neurotrophin receptor signaling
The many important benefits of physical exercise on human
system must be omitted in this article due to space constraints,
health have long been recognized and mostly include a
it is worthwhile mentioning here that the final effect is neuro-
decreased risk of cardiovascular disease, diabetes, cancer,
genesis, thus promoting growth and differentiation of new neu-
osteoporosis, and so forth.7,8 Only recently, however, it has
rons and synapses, supporting axonal and dendritic growth,
been acknowledged that physical exercise may also generate
fostering synaptic plasticity, and preserving survival of
substantial benefits on executive and cognitive functions
through a mechanism that seems to depend on exercise type,
Peer review under responsibility of Shanghai University of Sport.
* Corresponding author. intensity, and volume.9 In general, physical exercise of suffi-
E-mail address: fabian.sanchis@uv.es (F. Sanchis-Gomar). cient intensity and duration (i.e., approximately 3 months of
https://doi.org/10.1016/j.jshs.2019.07.012
Cite this article: Lippi G, Mattiuzzi C, Sanchis-Gomar F. Updated overview on interplay between physical exercise, neurotrophins, and cognitive function in
humans. J Sport Health Sci 2020;9:74 81.
Physical exercise, neurotrophins, and cognitive function 75

Table 1 authors included 29 studies in their analysis, representing a

Neurotrophins and their receptors. total of 1111 participants (no exclusion criteria were used for
p75NTR Trk-A Trk-B Trk-C the study populations), and concluded that a single session of
exercise was effective to immediately increase the serum or
Nerve growth factor Yes Yes — —
Brain-derived neurotrophic factor Yes — Yes — plasma concentration of BDNF (Hedges’ g = 0.46; 95% confi-
Neurotrophin-3 Yes Yes Yes Yes dence interval (CI): 0.29 0.62; p < 0.001). Postexercise
Neurotrophin-4 Yes — Yes — increase in BDNF values was then significantly amplified in
Abbreviations: p75NTR = 75; Trk = tyrosine kinase. subjects who had completed a program of regular exercise
training comprising 3 24 weeks (Hedges’ g = 0.58; 95%CI:
0.10 1.07; p = 0.02). The elimination of resistance training
moderate-intensity aerobic exercise, with sessions lasting not intervention (i.e., arm and elbow strength) did not significantly
less than 20 min) seems to be capable of triggering neurogene- attenuate the effect of acute exercise on BDNF (Hedges’
sis, which potentiates synaptic plasticity and creates new syn- g = 0.49; 95%CI: 0.34 0.64; p < 0.001), whereas strength/
apses and neural circuits that ultimately can contribute to the resistance training studies did generate a significant acute
optimization of brain plasticity and fitness.10 12 Notably, the impact on BDNF concentration (Hedges’ g = 0.57; 95%CI:
impact of acute exercise cannot be discounted because it is 0.24 to 1.37; p = 0.17), thus suggesting that aerobic exercise
essential for priming some specific cellular pathways that then may be more effective globally in boosting the circulating val-
translate into structural and functional adaptations typically ues of BDNF.
observed in subjects engaged in regular physical exercise.12 In a subsequent meta-analysis, Dinoff et al.17 explored the
Physical exercise can therefore be regarded as an enhancer of impact of exercise interventions lasting not less than 2 weeks
environmental factors that may actively promote neuroplastic- on the circulating values of BDNF. Overall, 29 studies met the
ity by repeated or chronic release of some important biochemi- criteria to be included in the analysis, representing a total 910
cal mediators. participants. In agreement with previous evidence, the resting
The combination of these 2 previous observations, that is, values of BDNF were found to be significantly increased
the essential contribution of neurotrophins to brain biology by exercise interventions (standardized mean difference
and the impact of physical exercise on cognitive function, (SMD) = 0.39; 95%CI: 0.17 0.60; p < 0.001), but a signifi-
leads to the hypothesis that these 2 aspects may somehow be cant impact was only observed after aerobic exercise
interrelated and that physical exercise may be regarded as a (SMD = 0.66; 95%CI: 0.33 0.99; p < 0.001) and not after
promising option for fostering neurotrophins production, ulti- resistance training (SMD = 0.07; 95%CI: 0.15 to 0.30;
mately improving human brain functioning.9,13 p = 0.52). No significant difference was noted among different
Although a large number of investigation studies have been age classes (p = 0.25), between women and men (p = 0.95),
carried out to explore the relationship between physical exer- between training programs of a duration of less than 12 weeks
cise and neurotrophins in cellular and animal models, the and of 12 weeks or greater (p = 0.39), or between short- and
results of these studies cannot be straightforwardly translated long-duration exercise (p = 0.97). The same group of authors
to humans, because substantial differences exist between performed a 2nd meta-analysis that investigated the impact of
human and animal brains in terms of anatomy, biochemistry, different types (e.g., running, cycling, and resistance training)
biology, and function.14,15 Moreover, the exercise protocols and intensities of exercise on circulating BDNF values.18 A
typically used in animal studies are very different from those total of 55 studies with 1180 total participants were included
used in human studies, so the exercise-induced biological in the final analysis. In accord with previous evidence, circu-
changes in animals must always be interpreted with caution. lating BDNF concentration was increased by approximately
The current article provides an updated overview of the inter- 60% immediately after acute exercise (SMD = 0.59; 95%CI:
play between physical exercise, neurotrophins, and cognitive 0.46 0.72; p < 0.001). Both aerobic exercise (SMD = 0.61;
function as reported in studies involving humans. 95%CI: 0.46 0.75; p < 0.001) and resistance training
(SMD = 0.48; 95%CI: 0.15 0.80; p = 0.004) were associated
2. Neurotrophins release triggered by physical exercise with immediate postexercise increase of circulating BDNF
values, albeit the variation was slightly higher after aerobic
The number of studies that have investigated the exercise-
training than after resistance training (i.e., 61% vs. 48%).
related biology of neurotrophins in healthy and diseased
Also, in this 2nd meta-analysis, no significant association was
humans has grown exponentially during the past decades. The
noted between BDNF variation and overall exercise duration
vast majority of studies have focused on BDNF, and data from
(b = 1.822; p = 0.075) and intensity (b = 1.766; p = 0.085).
those studies have been recently reviewed in several compre-
Nevertheless, when the exercise interventions were divided
hensive meta-analyses (Table 2).
between those lasting 30 min or less (SMD = 0.47; 95%CI:
0.31 0.63; p < 0.001) and those lasting more than 30 min
2.1. BDNF
(SMD = 0.81; 95%CI: 0.53 1.09; p < 0.001), a marginally
In 2015, Szuhany et al.16 carried out a meta-analysis of all significant difference was observed (p = 0.04), thus supporting
studies addressing the potential association between different the hypothesis that 30 min of exercise is the minimal threshold
exercise paradigms and BDNF concentration in humans. The for producing significant BDNF variations. Interestingly,
76 G. Lippi et al.

Table 2
Summary of meta-analyses that have explored the effects of physical exercise on BDNF.

Authors Study population n (studies/total Conclusions

participants included)
Szuhany et al. (2015)16 No exclusion criteria 29/1111 Acute increase in BDNF after exercise
Chronic increase in BDNF after exercise programs
Regular exercise training amplifies the BDNF response to
acute exercise
Aerobic exercise more effective than resistance training
No dependency on BDNF variation based on age
Dinoff et al. (2016)17 Healthy subjects 29/910 Chronic increase in BDNF after exercise programs
Aerobic exercise more effective than resistance training
No association between BDNF variation and exercise dura-
tion and intensity
No dependency on BDNF variation based on age, sex, and
body mass index
Dinoff et al. (2017)18 Healthy subjects 55/1180 Acute increase in BDNF after exercise
Slightly greater effect on acute BDNF variation from aero-
bic exercise than from resistance training
Association between BDNF variation and aerobic capacity
No association between BDNF variation and exercise dura-
tion and intensity (but significant variation only observed
>30 min)
BDNF variation higher in men than in women
No dependency on BDNF variation based on age or body
mass index
Feter et al. (2019)19 No exclusion criteria 25/2152 Chronic increase in BDNF after exercise programs
Association between BDNF variation, aerobic capacity, and
exercise volume
No dependency on BDNF variation based on age, sex,
health status, or exercise intensity
Dinoff et al. (2018)20 Patients with major depressive 6/1176 No chronic increase in BDNF after exercise programs
disorders
Kurebayashi and Otaki (2018)21 Patients with major depressive 5/199 No chronic increase in BDNF after exercise programs
disorders
Mackay et al. (2017)22 Patients with neurologic disorders 11/303 Chronic increase in BDNF after exercise programs
Hirsch et al. (2018)23 Patients with Parkinson’s disease 4/100 Chronic increase in BDNF after exercise programs
Abbreviation: BDNF = brain-derived neurotrophic factor.

postexercise BDNF variation seemed to be higher in men than and Statistical Manual of Mental Disorders. Overall, 6 studies
in women (SMD = 0.75 vs. 0.13; p < 0.001), but the variation totaling 1176 participants were included in the final analysis,
was not influenced by age (b = 0.877; p = 0.38) or body mass which revealed that exercise interventions were not effective
index (b = 1.000; p = 0.33). Finally, greater variations in post- in modifying the baseline concentration of BDNF in these
exercise circulating BDNF values were observed in patients patients (SMD = 0.43; 95%CI: 0.06 to 0.92; p = 0.09). Inter-
with higher levels of cardiorespiratory fitness (b = 3.548; estingly, this evidence was then confirmed in another recent
p = 0.002). meta-analysis published by Kurebayashi and Otaki.21 The
More recently, Feter et al.19 published another meta-analy- authors identified 5 studies with 199 total participants who suf-
sis of 25 studies with 2152 total participants aimed at explor- fered from major depressive disorders. The authors investi-
ing the chronic effects of exercise on circulating BDNF. gated the effects of physical exercise on circulating BDNF
No exclusion criteria were applied for selecting the study pop- concentration and concluded that the no significant postexer-
ulations. In the final meta-analysis, exercise was associated cise variation could be noted in these patients (SMD = 0.05;
with a significant increase in BDNF concentration (mean 95%CI: 0.23 to 0.32; p = 0.75).
increase = 596 pg/mL; 95%CI: 472 721 pg/mL; p < 0.001), Unlike the evidence garnered in studies on patients with
and this effect was found to be independent of age, sex, and major depressive disorders, different results were recently pub-
health status. The major contributors to postexercise BDNF lished by Mackay et al.22 in patients with neurologic disorders.
variation in the meta-regression analysis were aerobic capacity Overall, 11 studies with 303 total participants were included in
(p = 0.02) and duration of exercise (p = 0.02). the final analysis, which revealed that programs of aerobic
A 3rd meta-analysis was then published by Dinoff et al.20 It exercise (median 12 h/week, lasting 4 24 weeks) were effec-
aimed at exploring the effects of long-term exercise interven- tive in significantly increasing resting BDNF values
tions on circulating BDNF values in patients with major (SMD = 0.84; 95%CI: 0.47 1.20; p < 0.001). In line with
depressive disorders, diagnosed according to the Diagnostic these findings, Hirsch et al.23 recently carried out another
Physical exercise, neurotrophins, and cognitive function 77

meta-analysis that included 4 pre-experimental studies and 2 of cholesterol and saturated fats. Neither form of the 2 inter-
randomized controlled trials with a total of 100 patients who ventions involving physical exercise was effective in signifi-
had idiopathic Parkinson’s disease. Overall, physical exercise cantly modifying the circulating values of NGF (p = 0.74).
interventions lasting 28 90 days were effective in signifi- Bonini et al.28 carried out a cross-sectional study including
cantly increasing circulating BDNF concentration in these 96 pre-Olympic athletes and 49 matched healthy sedentary
patients (SMD = 2.06; 95%CI: 1.36 2.76; p < 0.001). controls. The authors reported that the concentration of circu-
Another systematic review was recently published by lating NGF in athletes was nearly twice that of the controls
Enette et al.,24 who analyzed the impact of interval and contin- (368 § 776 pg/mL vs. 174 § 484 pg/mL; p < 0.001).
uous aerobic training on BDNF concentration in subjects aged Bansi et al.29 carried out a randomized controlled clinical
60 years or older. Overall, 14 randomized or nonrandomized trial involving 60 patients with multiple sclerosis who engaged
intervention studies with a total of 988 participants were in 30 min of controlled cycling daily at 60% of maximal aero-
included in the final analysis. The resting BDNF concentration bic capacity for 3 weeks performed on a cycle ergometer or an
was found to be significantly increased among participants in aquatic bike (pool depth, 130 cm). Interestingly, the baseline
4 of the 5 (80%) studies that applied aerobic interval training BDNF concentration increased by 23% (p = 0.046) after water
protocols and among participants in 6 of the 9 (67%) studies cycling but not after land cycling (p = 0.699), whereas the
that applied continuous aerobic training. baseline NGF values did not change after either form of exer-
cise (p = 0.087 after water cycling and p = 0.721 after land
cycling).
2.2. Other neurotrophins
Kim30 carried out a cross-sectional study that included 22
Unlike the large volume of data available for BDNF male adolescents aged between 14 and 18 years. These sub-
(Table 2), a rather limited number of studies have evaluated jects were divided in 2 groups: those who did not perform any
the influence of physical exercise on other neurotrophins. An regular exercise program for more than 1 year (control group,
electronic search conducted in Scopus and PubMed using n = 9) and those with more than 3 years of training who regu-
“nerve growth factor” AND “exercise” OR “sport” as search larly exercised 18 h/week (trained group, n = 13). Although
parameters identified a total number of 371 documents, only the concentration of NGF in the trained group was slightly
12 of which involved humans and investigated the effect of higher than in the control group (6.7 § 2.8 ng/mL vs. 5.0 §
physical exercise on this neurotrophin. The findings from these 2.2 ng/mL), the difference did not reach statistical significance
studies are briefly presented and discussed in this section. (p > 0.05).
The 1st study that investigated the effect of exercise on cir- Azali Alamadari and Choobineh31 carried out a study
culating NGF values was published by Gold et al.25 in 2003. involving 30 subjects (mean age, 58 § 5 years) with metabolic
The authors studied 25 patients with multiple sclerosis and 20 syndrome, who were randomized to no exercise (n = 14) or an
healthy matched controls who were asked to perform a cycle 8-week aerobic training program encompassing 3 sessions/
ergometry test for 30 min at 60% of maximal aerobic capacity. week of exercise at 50% 60% of maximal aerobic capacity
Although the immediate postexercise values of BDNF were (n = 16). Interestingly, a significant increase in resting circulat-
significantly increased by approximately 30% over the base- ing values of both BDNF (+4%; p < 0.05) and NGF (+48%;
line in both patients and controls (cumulative effect, p = 0.03), p < 0.05) was noted in the exercise group.
the concentration of NGF remained virtually unchanged in Cho et al.32 studied 36 obese middle-aged Korean women
both groups (cumulative effect, p = 0.09). (mean age, 55 § 3 years) who were randomized to no interven-
Schulz et al.26 studied 15 patients with multiple sclerosis tion, an 8-week aerobic exercise program (3 sessions/week of
who were engaged in an 8-week aerobic bicycle training pro- 40-min treadmill running at 70% of heart rate reserve), and an
gram (30 min of cycling at 75% of maximal watts, twice a 8-week aerobic exercise program (same as before) plus cranial
week). The resting concentration of BDNF was found to be electrotherapy stimulation. The resting concentration of
increased after the exercise training period (+17%), whereas BDNF was found to be significantly increased after the 2 exer-
the concentration of NGF did not substantially change. Similar cise intervention modalities (+17% after exercise alone and
results were seen after a 30-min endurance test, with BNDF +15% after exercise plus cranial electrotherapy stimulation),
concentration increasing by approximately 36% and NGF val- as were the resting values of NGF (+10% after exercise alone
ues remaining unvaried. and +9% after exercise plus cranial electrotherapy stimula-
In an ensuing investigation, Rokling-Andersen et al.27 con- tion). The same authors carried out another intervention study
ducted a randomized, controlled, 2 £ 2 trial including 188 par- based on a 40-min treadmill exercise at 70% heart rate reserve
ticipants in the Oslo Diet and Exercise Study. These 3 times/week for 8 weeks in 10 obese (mean age, 23 § 2 years)
participants were randomized to 4 different 1-year protocols: a and nonobese (mean age, 23 § 2 years) young men.33 This
no intervention protocol (n = 37), a diet protocol (n = 45), an exercise intervention was effective in increasing the resting
exercise protocol (n = 48), and a diet plus exercise protocol concentration of BDNF in both obese and nonobese subjects
(n = 58). The exercise protocol consisted of 60 min of aerobic (1% 20%; p = 0.013), although the circulating levels of NGF
training (e.g., fast walking or jogging), 3 times weekly, remained mostly unchanged in both categories of subjects
whereas the dietary intervention increased participants’ intake (p = 0.196). These authors also performed another study
of fish, vegetables, and fiber-rich products and reduced intake including 15 healthy men who were asked to engage in
78 G. Lippi et al.

treadmill running under low-, moderate-, and high-intensity of NT-4 (r = 0.59; p = 0.034), but not with that of NT-3, nor
conditions.34 In accordance with previous findings, the con- were the postexercise changes of NT-3 and NT-4 significantly
centration of BDNF displayed an exercise-volume-dependent associated.
immediate increase (+6%, +23%, and +27% after low-, moder-
ate-, and high-intensity exercise, respectively), with values
3. Discussion
returning to the baseline after 60 min. A similar exercise vol-
ume-dependent kinetics was noted for NGF values (+12%, Taken together, the currently available data seemingly con-
+19%, and +23% immediately after low-, moderate-, and firm the existence of a positive relationship between physical
high-intensity exercise, respectively), with values returning to exercise and circulating BDNF levels, both in the short and
the baseline after 60 min. long term, and also support the beneficial impact of training
In a subsequent study, Mokhtarzade et al.35 randomly programs for amplifying the acute BDNF response (Table 2).
assigned 66 patients with multiple sclerosis to no exercise or Notably, an accurate assessment of postexercise variation of
an 8-week exercise intervention based on cycling at BDNF changes requires an appropriate blood sample collec-
60% 70% of peak power 3 days/week. Although the exercise tion and should be corrected for hemoconcentration.38
intervention was effective in significantly increasing BDNF Although both aerobic exercise and resistance training seem to
values in normal weight (n = 33; p = 0.001) but not in obese be effective in acutely increasing BDNF values, the variation
patients (n = 33; p > 0.05), the circulating levels of NGF promoted by the former type of exercise seems greater. Con-
remained similar to pre-exercise intervention levels both in troversial information has been reported for exercise intensity
normal weight and obese patients (p > 0.05). and duration. Although most published studies seem to rule
Moradi et al.36 measured NGF in 48 autistic children aged out a clear impact of intensity on BDNF concentration, exer-
6 9 years, who were equally assigned to 4 different 2-month cise lasting for 30 min or more seems necessary to generate a
protocols, encompassing no intervention, perceptual-motor significant postexercise increase in this molecule (Table 2).
activities combined with music, vitamin D supplementation Interestingly, the potential role of anaerobic exercise has been
alone, and perceptual-motor activities combined with music highlighted by Schiffer et al.,39 who showed that blood lactate
and vitamin D supplementation. As expected, no variation in generation may interplay with BDNF blood concentrations,
NGF levels was observed in the no intervention group, whereby intravenous infusions of approximately 4 mol/L of
whereas the postintervention concentration of this neurotro- sodium lactate solution were effective in acutely increasing
phin gradually increased from approximately 3% in the per- blood BDNF values by approximately 1.4-fold. Finally,
ceptual-motor activities combined with music group to although the impact of exercise on BDNF concentration has
approximately 30% in the vitamin D supplementation group, occasionally been found to be higher in men than in women,
up to approximately 70% in the perceptual-motor activities no significant associations have been reported based on age or
combined with music and vitamin D supplementation group. body mass index. Surprisingly, no effect of physical exercise
As regards the other 2 neurotrophins, NT-3 and NT-4, a on circulating BDNF values was found in patients with major
combined Scopus and PubMed search using the keywords depressive disorders, whereas in those with neurologic disor-
“neurotrophin-3” OR “neurotrophin-4” AND “exercise” OR ders the effect was globally comparable with that observed in
“sport” identified 93 documents for NT-3 and 48 documents the general population (Table 2).
for NT-4, but only one of these was an experimental study Less solid and clear information can be garnered from the
involving human subjects. limited number of studies that investigated the impact of phys-
Domınguez-Sanchez et al.37 carried out a randomized, par- ical exercise on circulating values of the 3 other major neuro-
allel group clinical study including 51 physically inactive trophins (NGF, NT-3, and NT-4). Although definitive
obese men (mean age, 24 § 3 years) who were randomized to conclusions cannot be made, the current evidence suggests
no exercise, high-intensity exercise (4 £ 4-min treadmill run- that—unlike BDNF—the impact of physical exercise on circu-
ning at 85% 95% of maximum heart rate (HRmax) alternated lating values of NGF remains largely uncertain. Significant
with 4-min recovery at 75% 85% of HRmax), resistance train- increases were observed in the several interventional
ing (12 15 repetitions/set at 50% 70% of HRmax alternated studies,28,31,32,34,36 although the results of the remaining stud-
with 1-min recovery), or combined high-intensity and resis- ies do not support the existence of any significant interaction.
tance exercise. Blood samples were then drawn before exercise In the only interventional study that evaluated the postexercise
and immediately after (i.e., within 1 min) each exercise proto- variation of both NT-3 and NT-4, the concentration of these 2
col. Interestingly, BDNF values acutely increased after high- biomarkers increased after resistance and combined training.37
intensity exercise (+7%) and combined training (+12%), but Thus, given that physical exercise is capable of both acutely
were not modified by resistance training. The values of NT-3 and chronically increasing the circulating values of some neu-
and NT-4 were both found to be increased after resistance rotrophins, especially NBDF, understanding how this phenom-
training (+13% and +7%, respectively) and combined training enon would translate into tangible benefits in humans becomes
(+12% and +9%, respectively), whereas their relative increase essential. Some studies have combined the assessment of
(+13% and +1%, respectively) did not reach statistical signifi- BDNF variation and neurocognitive improvement after exer-
cance after high-intensity exercise. Notably, the relative varia- cise. Kimhy et al.40 carried out a single-blind randomized clin-
tion of BNDF was found to be marginally associated with that ical study that included 33 patients with schizophrenia who
Physical exercise, neurotrophins, and cognitive function 79

were randomized either to a 12-week aerobic whole body exer- phenomenon over time. Interestingly, exercise duration but
cise program (n = 16) or to standard psychiatric treatment not its intensity was found to be associated with BDNF varia-
(n = 17). At the end of the study, the concentration of BDNF tions,18 thus supporting the hypothesis that regular perfor-
increased by 11% in the aerobic exercise group compared to mance of light exercise may be sufficient to trigger beneficial
2% in the cohort that received standard psychiatric treatment. effects on BDNF. This concept has been recently confirmed
Neurocognitive functioning was also substantially improved in by Park et al.,45 who assessed the effects of an exercise inter-
the aerobic exercise group, and approximately 15% of this var- vention based on 20 min of daily low- to moderate-intensity
iation was explained by postaerobic exercise BDNF variation. gardening (i.e., approximately 3.5 metabolic equivalents,
In a subsequent study, Heisz et al.41 explored the impact of involving digging, fertilizing, raking, planting, cleaning, and
physical exercise and cognitive training on neurotrophic fac- watering) in 41 elderly subjects (mean age, 77 § 6 years).
tors and neurocognitive functioning in 95 healthy young adults The authors reported that the circulating levels of BDNF sig-
(mean age, 21 § 3 years; 58 women and 37 men). The partici- nificantly increased by approximately 8% (p = 0.038) imme-
pants were randomized to no intervention, 6 weeks of exercise diately after this low to moderate level of physical activity.
training (20 min of high-intensity interval training 3 times/ Given the evidence that physical exercise has a favorable
week), or 6 weeks of the previously mentioned exercise train- impact on BDNF concentration in patients with neurologic dis-
ing combined with cognitive training. Individuals with greater orders, it is reasonable to suggest that regular performance of
aerobic fitness improvements in response to exercise training aerobic exercise should be an active part of a rehabilitation
showed greater increases in BDNF levels (r = 0.25; p < 0.05), program in these patients, thus modulating the interplay of
as well as better high-interference memory performance as a neurotrophins, especially BDNF, with the pathogenesis of
result of the combined exercise and cognitive training com- these conditions. This conclusion is supported by results of 2
pared with exercise alone (r = 0.42; p < 0.05). More recently, very recent meta-analyses showing that BDNF concentration
Leckie et al.42 explored the effect of aerobic training on execu- was significantly reduced in patients with Parkinson’s disease
tive function and BDNF levels in 90 older adults (mean age, (SMD = 1.03; 95%CI: 1.83 to 0.23; p = 0.012)46 and
67 § 6 years; 59 women and 33 men). The subjects were Alzheimer’s disease (Hedges’ g = 0.339; 95%CI: 0.572 to
randomized either to a no exercise protocol or to an interven- 0.106; p = 0.004)47 compared to healthy populations.
tion involving 1 year of walking (3 days/week for 40 min at
60% 75% of HRmax reserve). In members of the walking 4. Conclusion
group aged more than 65 years, the serum concentration of
Several lines of evidence now attest to the idea that physical
BDNF significantly increased by approximately 25%, and this
exercise is essential for brain health and that the repeated or
variation was significantly associated with improvement in
chronic release of neurotrophins—especially BDNF—triggered
executive function and task-switch performance. Encouraging
by motor activities may be a key mediator of this effect. There-
results were also published by Hakansson et al.,43 who
fore, promotion of adequate volumes and intensities of physical
recruited 19 healthy older adults 65 85 years of age. All of
exercise (i.e., approximately 3 months of moderate-intensity
them performed a single session of 3 different types of inter-
aerobic exercise, with 2 3 sessions/week lasting not less than
vention: 35 min of whole body aerobic exercise at moderate
30 min) represents an inexpensive and safe strategy for boosting
intensity, 35 min of cognitive training, and 35 min of mindful-
BDNF release that may preserve or restore cognitive function.
ness practice. Blood samples for BDNF analysis were col-
Additional studies should be conducted that further explore the
lected immediately before, immediately after, and at 20 min
interplay between physical exercise, neurotrophins, and cogni-
and 60 min after each exercise. In agreement with previous
tive function.
findings, physical exercise generated an immediate and signifi-
cant increase in circulating BDNF values (+17%; p = 0.004),
Acknowledgment
although no significant changes were observed after cognitive
training and mindfulness practice. Notably, the postexercise Fabian Sanchis-Gomar is supported by a postdoctoral con-
BDNF variation was significantly associated with working tract granted by Subprograma Atraccio de Talent - Contractes
memory function (r = 0.89; p = 0.02). Postdoctorals de la Universitat de Valencia.
These converging findings, along with evidence that the
brain is the major contributor of exercise-induced circulating Authors’ contributions
BDNF,44 convincingly support the hypothesis that acute or
GL and FSG reviewed the literature and participated in
chronic postexercise increases in BDNF translates into an
drafting, writing, and revising this review manuscript; CM
improvement of some executive and cognitive functions, thus
revised the manuscript and provided meaningful edits and
carrying substantial clinical implications. First, both disease- and
comments. All authors have read and approved the final ver-
age-related impairment in cognitive function may be attenuated
sion of the manuscript, and agree with the order of presentation
by participation in exercise programs targeting the neurotrophin
of the authors.
pathways (i.e., moderate-intensity aerobic exercise lasting
for 30 min). Repeated training is also advisable, since each
Competing interests
session of exercise seems to be effective in generating BDNF
production, thus amplifying the magnitude of this The authors declare that they have no competing interests.
80 G. Lippi et al.

References systematic review and meta-analysis. Neural Plast 2017;2017: 4716197.

doi:10.1155/2017/4716197.
1. Pareja-Galeano H, Mayero S, Sanchis-Gomar F. Exercise, neuroplasticity, 23. Hirsch MA, van Wegen EEH, Newman MA, Heyn PC. Exercise-induced
and growth factors in adolescence. In: Farooqui TFaAA, editor. Diet and increase in brain-derived neurotrophic factor in human Parkinson’s dis-
exercise in cognitive function and neurological diseases. Hoboken, NJ: ease: a systematic review and meta-analysis. Transl Neurodegener
Wiley Blackwell; 2015.p.323–37. 2018;7:7. doi:10.1186/s40035-018-0112-1.
2. Bothwell M. Recent advances in understanding neurotrophin signaling. 24. Enette L, Vogel T, Fanon JL, Lang PO. Effect of interval and continuous
F1000Res 2016;5. pii: F1000 Faculty Rev-1885. doi:10.12688/ aerobic training on basal serum and plasma brain-derived neurotrophic
f1000research.8434.1. factor values in seniors: a systematic review of intervention studies. Reju-
3. Vilar M, Mira H. Regulation of neurogenesis by neurotrophins during venation Res 2017;20:473–83.
adulthood: expected and unexpected roles. Front Neurosci 2016;10:26. 25. Gold SM, Schulz KH, Hartmann S, Mladek M, Lang UE, Hellweg R, et al.
doi:10.3389/fnins.2016.00026. Basal serum levels and reactivity of nerve growth factor and brain-derived
4. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development neurotrophic factor to standardized acute exercise in multiple sclerosis
and function. Annu Rev Neurosci 2001;24:677–736. and controls. J Neuroimmunol 2003;138:99–105.
5. Lu B, Nagappan G, Lu Y. BDNF and synaptic plasticity, cognitive func- 26. Schulz KH, Gold SM, Witte J, Bartsch K, Lang UE, Hellweg R, et al.
tion, and dysfunction. Handb Exp Pharmacol 2014;220:223–50. Impact of aerobic training on immune-endocrine parameters, neurotrophic
6. Pareja-Galeano H, Brioche T, Sanchis-Gomar F, Montal A, Jovani C, factors, quality of life and coordinative function in multiple sclerosis.
Martinez-Costa C, et al. Impact of exercise training on neuroplasticity- J Neurol Sci 2004;225:11–8.
related growth factors in adolescents. J Musculoskelet Neuronal Interact 27. Rokling-Andersen MH, Reseland JE, Veierod MB, Anderssen SA,
2013;13:368–71. Jacobs Jr DR, Urdal P, et al. Effects of long-term exercise and diet
7. Warburton DER, Bredin SSD. Health benefits of physical activity: a sys- intervention on plasma adipokine concentrations. Am J Clin Nutr
tematic review of current systematic reviews. Curr Opin Cardiol 2007;86:1293–301.
2017;32:541–56. 28. Bonini M, Fioretti D, Sargentini V, Del Giacco S, Rinaldi M, Tranquilli C,
8. Vina J, Sanchis-Gomar F, Martinez-Bello V, Gomez-Cabrera MC. Exer- et al. Increased nerve growth factor serum levels in top athletes. Clin J
cise acts as a drug; the pharmacological benefits of exercise. Br J Pharma- Sport Med 2013;23:228–31.
col 2012;167:1–12. 29. Bansi J, Bloch W, Gamper U, Kesselring J. Training in MS: influence of
9. Pareja-Galeano H, Mayero S, Perales M, Garatachea N, Santos-Lozano A, two different endurance training protocols (aquatic versus overland) on
Fiuza-Luces C, et al. Biological rationale for regular physical exercise as cytokine and neurotrophin concentrations during three week randomized
an effective intervention for the prevention and treatment of depressive controlled trial. Mult Scler 2013;19:613–21.
disorders. Curr Pharm Des 2016;22:3764–75. 30. Kim YI. The impact of exercise training on basal BDNF in athletic adoles-
10. Ploughman M. Exercise is brain food: the effects of physical activity on cents. J Phys Ther Sci 2016;28:3066–9.
cognitive function. Dev Neurorehabil 2008;11:236–40. 31. Azali Alamadari K, Choobineh S. Integrated effects of aerobic training on
11. Lee MC, Byun K, Kim JS, Lee H, Kim K. Trends in exercise neurosci- metabolic risk factors, circulatory neurotrophins, testosterone and cortisol
ence: raising demand for brain fitness. J Exerc Rehabil 2019;15:176–9. in midlife males with metabolic syndrome. Med Sport 2016;69:228–39.
12. El-Sayes J, Harasym D, Turco CV, Locke MB, Nelson AJ. Exercise- 32. Cho SY, So WY, Roh HT. Effects of aerobic exercise training and cranial
induced neuroplasticity: a mechanistic model and prospects for promoting electrotherapy stimulation on the stress-related hormone, the neurotrophic
plasticity. Neuroscientist 2019;25:65–85. factor, and mood states in obese middle-aged women: a pilot clinical trial.
13. Pareja-Galeano H, Sanchis-Gomar F, Mayero S. Autism spectrum disor- Salud Mental 2016;39:249–56.
ders: possible implications of BDNF modulation through epigenetics. 33. Roh HT, So WY. The effects of aerobic exercise training on oxidant-
Acta Psychiatr Scand 2013;128:97. doi:10.1111/acps.12071. antioxidant balance, neurotrophic factor levels, and blood-brain bar-
14. Premack D. Human and animal cognition: continuity and discontinuity. rier function in obese and non-obese men. J Sport Health Sci
Proc Natl Acad Sci U S A 2007;104:13861–7. 2017;6:447–53.
15. Sherwood CC, Subiaul F, Zawidzki TW. A natural history of the human 34. Roh HT, Cho SY, Yoon HG, So WY. Effect of exercise intensity on neu-
mind: tracing evolutionary changes in brain and cognition. J Anat rotrophic factors and blood-brain barrier permeability induced by oxida-
2008;212:426–54. tive-nitrosative stress in male college students. Int J Sport Nutr Exerc
16. Szuhany KL, Bugatti M, Otto MW. A meta-analytic review of the effects Metab 2017;27:239–46.
of exercise on brain-derived neurotrophic factor. J Psychiatr Res 35. Mokhtarzade M, Motl R, Negaresh R, Zimmer P, Khodadoost M, Baker
2015;60:56–64. JS, et al. Exercise-induced changes in neurotrophic factors and markers of
17. Dinoff A, Herrmann N, Swardfager W, Liu CS, Sherman C, Chan S, et al. blood-brain barrier permeability are moderated by weight status in multi-
The effect of exercise training on resting concentrations of peripheral ple sclerosis. Neuropeptides 2018;70:93–100.
brain-derived neurotrophic factor (BDNF): a meta-analysis. PLoS One 36. Moradi H, Sohrabi M, Taheri H, Khodashenas E, Movahedi A. The effects
2016;11: e0163037. doi:10.1371/journal.pone.0163037. of different combinations of perceptual-motor exercises, music, and vita-
18. Dinoff A, Herrmann N, Swardfager W, Lanctot KL. The effect of acute min D supplementation on the nerve growth factor in children with high-
exercise on blood concentrations of brain-derived neurotrophic factor in functioning autism. Complement Ther Clin Pract 2018;31:139–45.
healthy adults: a meta-analysis. Eur J Neurosci 2017;46:1635–46. 37. Domınguez-Sanchez MA, Bustos-Cruz RH, Velasco-Orjuela GP, Quin-
19. Feter N, Alt R, Dias MG, Rombaldi AJ. How do different physical exer- tero AP, Tordecilla-Sanders A, Correa-Bautista JE, et al. Acute effects of
cise parameters modulate brain-derived neurotrophic factor in healthy and high intensity, resistance, or combined protocol on the increase of level of
non-healthy adults? A systematic review, meta-analysis and meta-regres- neurotrophic factors in physically inactive overweight adults: the Brainfit
sion. Sci Sports 2019;34:293–304. Study. Front Physiol 2018;9:741. doi:10.3389/fphys.2018.00741.
20. Dinoff A, Herrmann N, Swardfager W, Gallagher D, Lanctot KL. The 38. Pareja-Galeano H, Alis R, Sanchis-Gomar F, Cabo H, Cortell-Ballester J,
effect of exercise on resting concentrations of peripheral brain-derived Gomez-Cabrera MC, et al. Methodological considerations to determine
neurotrophic factor (BDNF) in major depressive disorder: a meta-analysis. the effect of exercise on brain-derived neurotrophic factor levels. Clin
J Psychiatr Res 2018;105:123–31. Biochem 2015;48:162–6.
21. Kurebayashi Y, Otaki J. Does physical exercise increase brain-derived 39. Schiffer T, Schulte S, Sperlich B, Achtzehn S, Fricke H, Struder HK. Lac-
neurotrophic factor in major depressive disorder? A meta-analysis. Psy- tate infusion at rest increases BDNF blood concentration in humans. Neu-
chiatr Danub 2018;30:129–35. rosci Lett 2011;488:234–7.
22. Mackay CP, Kuys SS, Brauer SG. The effect of aerobic exercise on brain- 40. Kimhy D, Vakhrusheva J, Bartels MN, Armstrong HF, Ballon JS, Khan S,
derived neurotrophic factor in people with neurological disorders: a et al. The impact of aerobic exercise on brain-derived neurotrophic factor
Physical exercise, neurotrophins, and cognitive function 81

and neurocognition in individuals with schizophrenia: a single-blind, ran- 44. Hill T, Polk JD. BDNF, endurance activity, and mechanisms underlying
domized clinical trial. Schizophr Bull 2015;41:859–68. the evolution of hominin brains. Am J Phys Anthropol 2019;168(Suppl.
41. Heisz JJ, Clark IB, Bonin K, Paolucci EM, Michalsk Bi, Becker S, et al. 67):S47–62. doi:10.1002/ajpa.23762.
The effects of physical exercise and cognitive training on memory and 45. Park SA, Lee AY, Park HG, Lee WL. Benefits of gardening activities for
neurotrophic factors. J Cogn Neurosci 2017;29:1895–907. cognitive function according to measurement of brain nerve growth factor
42. Leckie RL, Oberlin LE, Voss MW, Prakash RS, Szabo-Reed A, Chad- levels. Int J Environ Res Public Health 2019;16. pii: E760. doi:10.3390/
dock-Heyman L, et al. BDNF mediates improvements in executive func- ijerph16050760.
tion following a 1-year exercise intervention. Front Hum Neurosci 46. Jiang L, Zhang H, Wang C, Ming F, Shi X, Yang M. Serum level of brain-
2014;8:985. doi:10.3389/fnhum.2014.00985. derived neurotrophic factor in Parkinson’s disease: a meta-analysis. Prog
43. Hakansson K, Ledreux A, Daffner K, Terjestam Y, Bergman P, Carlsson Neuropsychopharmacol Biol Psychiatry 2019;88:168–74.
R, et al. BDNF responses in healthy older persons to 35 minutes of physi- 47. Qin XY, Cao C, Cawley NX, Liu TT, Yuan J, Loh YP, et al. Decreased
cal exercise, cognitive training, and mindfulness: associations with work- peripheral brain-derived neurotrophic factor levels in Alzheimer’s dis-
ing memory function. J Alzheimers Dis 2017;55:645–57. ease: a meta-analysis study (N=7277). Mol Psychiatry 2017;22:312–20.