Background

Aging, the brain, and cognition

Throughout the lifespan, the human organism undergoes considerable changes. As a consequence of aging, the structure and function of organic systems (i.e., brain) can be negatively affected, which in turn can converge in a decline of individual capabilities (e.g., cognition). In this regard, in recent years, evidence has shown that the hippocampus [1,2,3,4] and the grey matter in the frontal lobe [1,2,3, 5,6,7,8,9,10,11,12] are affected by age-related shrinking. In contrast, the grey matter volume of other brain structures such as the parietal and occipital cortices have been reported to change slightly with increasing age [1, 5, 8], whereas a severe decline in white matter volume of the prefrontal cortex (PFC) is most pronounced in the very oldest [1, 8, 9, 13, 14]. These age-related changes in brain structure [15, 16] are assumed to play major roles in the worsening of cognition functions, such as processing speed and memory [17,18,19,20]. In fact, in older adults, it was observed that a decrease in hippocampal volume is associated with worsening of memory performance [21,22,23]. Conversely, an increase in hippocampal volume after a yearlong aerobic training intervention was associated with memory improvements [24]. These findings suggest that the preservation of brain structures (e.g., hippocampus) is important to ensure the proper functioning of cognitive processes (e.g., memory). Similar to the relationship of brain structure and cognition, it is assumed that changes in brain function (e.g., brain activation during a cognitive task) contribute to changes in cognition [16, 25,26,27]. Such an intertwined relationship between brain activation and cognition is underpinned by the findings linking activation of the PFC to behavioral performance in executive function tasks [28,29,30,31], in visuomotor tasks [32], or in working memory tasks [33,34,35]. Currently, several hypotheses exist that aim to explain age-related alterations in brain activation and cognition [16, 25,26,27]. For instance, the HAROLD model predicts that there is hemispheric asymmetry reduction in older adults in the PFC during the execution of memory tasks [27, 36]. In the compensation-related utilization of the neural circuits hypothesis (CRUNCH), it is postulated that adults will recruit more brain regions (mainly the PFC) as the task load increases and that older adults need to recruit these brain regions at lower levels of cognitive load than younger adults (e.g., during working memory tasks) [26, 37,38,39]. In the Scaffolding Theory of Aging and Cognition (STAC), it is postulated that increased brain activity with age, especially in the PFC, is a compensatory mechanism caused by reorganization of the brain in response to the age-related decline in neural structures and neural functioning [16, 39, 40]. To date, none of these hypotheses satisfactorily explain the observed age-related changes in brain function [41], but all of these hypotheses emphasize the important role of the PFC in age-related functional brain changes. It is well recognized in the literature that physical exercises [28,29,30, 42, 43] and physical training [44,45,46,47] lead to positive changes in cognitive performance (e.g., executive functions) and brain activation patterns. Furthermore, the changes in brain activation patterns (i.e., shown by higher levels of oxygenated hemoglobin in brain regions) are associated with cognitive performance improvements [28,29,30, 47], which illustrate the important role of physical interventions in preserving cognition and brain health.

In summary, distinct cognitive functions (e.g., memory) are negatively affected, and substantial changes in brain structure (e.g., shrinkage of hippocampus) and brain function (e.g., compensatory brain activation; i.e., of PFC) occur as consequences of “normal” aging. Notably, regular engagement in physical exercise is a valuable strategy to counteract age-related decline in brain and cognition [48,49,50,51,52].

Aging, muscular system, and cognition

There is solid evidence in the literature that muscle mass (sarcopenia) [53,54,55,56,57] and muscular strength (dynapenia) [53, 57,58,59], which constitute the ability to produce muscular force and power [60], decline gradually as a function of age. Notably, the age-related decrease in muscular strength was noticed to be more pronounced than the decrease in muscle mass [61,62,63]. Moreover, the decline in maximum muscular strength is more serious in the lower limbs than in upper limbs [62, 64,65,66,67]. In general, it was observed that the age-related loss in, for instance, maximum isokinetic hip/leg extensor strength is rather minimal until the fifth decade of life but accelerates considerably thereafter [58, 68,69,70]. Potential reasons for the pronounced decline in muscular strength are the reduction in cross-sectional area of the muscle fibers [64, 71] as well as the loss of muscle fibers and motor units [55, 56, 58, 61, 72, 73]. However, appropriate levels of muscular strength are needed for independent and healthy living. For instance, an appropriate level of muscular strength in the muscles of the lower limbs (e.g., hip and leg extensors) is required to ensure proper function for engaging in activities of daily living (e.g., balance and gait) [74, 75]. Hence, it is not surprising that a decline in isokinetic muscular strength in leg extensors is associated with reduced mobility [76,77,78] and increased risk of mortality [77, 79, 80].

However, there is growing evidence that an appropriate level of muscular strength is also linked to brain health and functioning (e.g., cognitive functions). In this regard, it has been reported in the literature that higher levels of isokinetic strength of the M. quadriceps femoris are linked to better performance in general cognitive abilities (operationalized by Mini-Mental State Examination [MMSE]) [81] and to better performance in executive functions [82, 83]. This link is further reinforced by the findings that higher leg power [84] and higher whole-body muscle strength [85] are associated with higher scores in standardized cognitive test batteries. Furthermore, higher handgrip strength is linked to higher scores in general cognitive abilities (e.g., operationalized by MMSE) [86, 87] and to higher scores in standardized cognitive test batteries [88,89,90]. Moreover, it was observed that gains in dynamic muscular strength (assessed by one repetition maximum in different resistance exercises) after 6 months of progressive resistance training mediate improvements in global cognitive performance (according to the Alzheimer’s Disease Assessment Scale–cognitive subscale) [91]. Similar to the previously mentioned finding, it was reported that changes in isokinetic knee extension and knee flexion torques after 3 months of progressive resistance training mediate improvements in executive functions [92]. Notably, a meta-analysis did not observe a correlation between muscle size and cognition [93] but reported that both muscle function (e.g., muscular strength) and muscle structure (e.g., muscle size) were linked to brain structure [93].

Taken together, during aging processes, a substantial decline in muscular strength, especially in lower limb muscles, occurs, and accumulating evidence suggests that lower muscular strengths are linked to poorer cognitive performance. Hence, resistance (strength) exercises (a single bout of resistance exercise, also referred to as acute exercise) and resistance (strength) training (more than one resistance exercise session, also referred to as chronic exercise; see also section ‘Data extraction’) seem to be promising activities to ensure the preservation of physical functioning and cognitive functions with aging.

Resistance exercises, resistance training, brain, and cognition

One physical intervention strategy that is frequently recommended to counteract the age-related deterioration of both physical functioning and cognition is the continuous and regular execution of resistance exercises and/or resistance training [94,95,96,97,98,99,100,101,102,103,104,105,106]. There is solid evidence in the form of systematic reviews and meta-analyses indicating that resistance exercises and resistance training (for distinction, see section ‘Data extraction’) have substantial benefits for specific domains of cognitive functions (e.g., executive functions) [105, 107,108,109,110,111], but the underlying neurobiological mechanisms of resistance exercise-induced improvements in cognitive functions are not yet fully understood [107, 110].

As shown in Fig. 1, cognitive improvements in response to resistance exercises and/or resistance training are based on changes on multiple levels of analysis [112, 113]. At the first level, molecular and cellular changes occur, which are summarized in the “neurotrophic hypothesis” [114,115,116,117]. The “neurotrophic hypothesis” claims that in response to physical exercises (e.g., resistance exercises), a pronounced release of distinct neurochemicals occurs (e.g., brain-derived neurotrophic factor [BDNF]) [114,115,116,117]. The pronounced release of specific neurochemicals triggers complex neurobiological processes evoking functional and/or structural brain changes that facilitate, at best, improvements in cognitive functions [24, 50, 114, 118,119,120]. With regard to the molecular and cellular levels, a systematic review summarized the evidence of resistance exercise and resistance training-induced changes in the release of several myokines (e.g., BDNF) and highlighted their positive effects on cognitive functions [121]. However, with respect to functional and structural brain changes and socioemotional changes (see Level 2 and Level 3 in Fig. 1), knowledge about resistance exercise and/or resistance training-induced changes is still relatively scarce, and the available literature has not yet been systematically pooled. In particular, the pooling of available evidence regarding functional and structural brain changes is needed because the brain may act as a mediator for the effect of resistance exercises and/or resistance training on cognition [112, 122]. Such a systematic pooling of available evidence is needed to provide evidence-based recommendations for individualized exercise prescriptions [123,124,125]. Because resistance exercises and/or resistance training is a promising strategy that could “hit many birds with one stone” (i.e., simultaneously counteracting different types of physical and brain-related health problems), the objective of this systematic review is to provide an overview of resistance exercise and/or resistance training-induced functional and/or structural brain changes that are related to changes in cognitive functions.

Fig. 1
figure 1

Schematic illustration of the objective of the present systematic review and the levels of analysis. ‘a’ indicates that the brain could be regarded as an outcome, a mediator or a predictor [122]. ‘b’ indicates several possibilities for how structural and functional brain changes, socioemotional changes, and cognitive changes are intertwined [112]. ERP: event-related potentials; FDG-PET: F-2-deoxy-D-glucose (FDG) positron-emissions tomography (PET); GMV: grey matter volume; LTM: long-term memory; STM: short-term memory; WMV: white matter volume

Full size image

Methods

Search strategy and process

In accordance with the guidelines for systematic reviews [126], two independent researchers conducted a systematic literature search on the 25th of April 2019 across the following six electronic databases (applied specifications): PubMed (all fields), Scopus (title, abstract, keywords), Web of Science (title), PsycInfo (all text), SportDiscus (abstract), and the Cochrane Library (title, abstract, keywords; trials). The following terms were used as search strings:

  • “strength exercise” OR “strength training” OR “resistance exercise” OR “resistance training” OR “weight exercise” OR “weight training” OR “weight lifting” OR “weight bearing” OR “elastic band” OR toning OR calisthenics OR “functional training”

AND

  • mental OR neuropsychological OR brain OR cogniti* OR neurocogni* OR executive OR attention OR memory OR “response time” OR “reaction time” OR accuracy OR error OR inhibition OR visual OR spatial OR visuospatial OR processing OR recall OR learning OR language OR oddball OR “task switching” OR “problem solving” OR Flanker OR Stroop OR Sternberg OR “Trail Making” OR “Tower of London” OR “Tower of Hanoi” OR “Wisconsin Card Sorting” OR “Simon task”

AND

  • cortex OR hemodynamic OR oxygenation OR “grey matter” OR “gray matter” OR “white matter” OR “brain volume” OR plasticity OR neuroelectric OR electrophysiological OR “P 300” OR “P 3” OR “event-related potentials” OR ERP OR Alpha OR Beta OR Gamma OR Theta OR NIR OR fNIRS OR “functional near-infrared spectroscopy” OR “near-infrared spectroscopy” OR “functional near-infrared spectroscopic” OR “optical imaging system” OR “optical topography” OR fMRI OR MRI OR “MR imaging” OR “magnetic resonance imaging” OR EEG OR electroencephalography OR electrocorticography OR MEG OR magnetoencephalography OR PET OR “positron emission tomography”

Afterwards, the results of the systematic search were loaded into a citation manager (Citavi 6.3), which was used for further analyses and for removing duplicates (see Fig. 2).

Fig. 2
figure 2

Flow chart with information about the search, screening, and selection processes that led to the identification of relevant articles included in this systematic review

Full size image

Inclusion and exclusion criteria

Screening for relevant studies was conducted using the established PICOS-principle [126, 127]. The acronym “PICOS” stands for participants (P), intervention (I), comparisons (C), outcomes (O), and study design (S) [126, 127]. The following inclusion and exclusion criteria were used: (P) we applied no restrictions and included all age groups regardless of pathologies; (I) only studies involving resistance exercises and/or resistance training were included; (C) in this systematic literature search, no specific restrictions were used; (O) studies considered relevant assessed functional brain changes and/or structural brain changes related to cognitive changes; (S) interventional or cross-sectional studies.

As shown in Fig. 3, 46 studies were excluded after full text screening because they did not meet our inclusion criteria. Eight studies were excluded because they only assessed functional or structural brain changes but did not measure cognitive performance [128,129,130,131,132,133,134,135]. Vice versa, 38 studies were excluded because they solely measured changes in cognitive performance without quantifying functional or structural brain changes [81, 91, 136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171].

Fig. 3
figure 3

Analysis of the risk of bias in the included studies in accordance with the Cochrane Collaboration guidelines. This figure was created using Review Manager [172]. A “green plus” indicates a low risk of bias, a “yellow question mark” indicates an unclear risk of bias, and a “red minus” denotes a high risk of bias

Full size image

Data extraction

We extracted information about the first author, year of publication, population characteristics including age, gender, cognitive status, exercise characteristics (e.g., muscle action, loading and volume, rest period between sets/between exercises, repetition velocity, frequency, resistance exercise selection), cognitive testing (e.g., tested cognitive domain, administration after exercise cessation), and functional and structural brain data. The extraction of information followed the recommendations of Hecksteden et al. [173].

Prior to presentation of the findings, it is necessary to clarify the different terms used in the field of exercise cognition. ‘Physical activity’ is defined as any muscle-induced bodily movements that increase energy expenditure from 1.0 to 1.5 MET [174, 175]. Hence, physical activity covers a wide range of acute and chronic physical activities (e.g., from housework to resistance exercises/resistance training). Specific forms of structured, planned, and regularly (chronically) conducted physical activities aiming to increase individual capabilities in a certain fitness domain are referred to as ‘training’ or ‘chronic (repetitive) exercises’ [174, 176,177,178]. Single sessions of physical activities (exercises) are referred to as ‘an acute (single) bout of physical activities’ or ‘physical exercises’ [174, 179, 180]. In this article, we use the term ‘resistance training’ when more than two exercise sessions were conducted. Consequently, a single session of resistance exercises is referred to as ‘a single (acute) bout of resistance exercises’ and/or ‘resistance exercises’. Furthermore, we use ‘exercise prescription’ as an umbrella term to denote exercise (e.g., load for an exercise) and training variables (e.g., frequency).

Risk of bias assessment

Two evaluators independently performed the risk of bias assessment using the Cochrane Collaboration’s Risk of Bias tool [181]. The Cochrane Collaboration’s Risk of Bias tool evaluates the methodological quality of a study by rating the risk of bias in distinct criteria (see Figure 3) as being ‘low’, ‘high’, or ‘unclear’ [181]. Any discrepancies in the ratings of the risk of bias were resolved by a discussion among the two evaluators or/and the consultation of the third author of the review. The risk of bias assessment is summarized in Fig. 3.

Results

Risk of bias

As shown in Fig. 3, the results regarding the judgment of risk of bias are heterogeneous. In the domains of sequence generation, allocation concealment, blinding of participants and personnel, and blinding of outcome assessment, the majority of studies were rated as low risk of bias or unclear risk of bias. The reviewed studies were judged as having an unclear risk of bias in those domains because procedures were not described in sufficient detail (e.g., method of random sequence generation). In the domains of incomplete outcome data, selective reporting, and other bias, most studies were judged as having a low risk of bias.

Participants’ characteristics and study design

In the reviewed studies, the effect of resistance exercises and/or resistance training on cognition and the brain was investigated in different cohorts, including healthy young adults [43, 182, 183], healthy older adults [44, 45, 184,185,186,187,188], older adults with mild cognitive impairment [188,189,190,191], older adults in an early stage of dementia [192], and individuals with multiple sclerosis [193]. Detailed information about participant characteristics (e.g., age, height, body mass) is provided in Table 1.

Table 1 Overview of the population characteristics and resistance exercises and/or resistance training characteristics of the reviewed studies
Full size table

Regarding the study design, almost all studies could be classified as interventional and as randomized controlled trials [43,44,45, 183,184,185,186, 188,189,190, 195, 197].

Additionally, three resistance exercise studies [43, 182, 183, 195] accounted for circadian variability as a possible moderating factor.

Resistance exercise characteristics

In four studies investigating the acute effects of single resistance exercise sessions on cognitive performance and on functional neuroelectric or hemodynamic brain processes, the exercise sessions lasted approximately 30 min [183] or 40 min [43, 182, 195].

Studies on the effects of resistance training on cognition and functional and/or structural brain changes involved groups who trained 1 day [45, 184,185,186], 2 days [45, 184,185,186, 188,189,190, 193, 197], or 3 days per week [44, 187, 191]. Exercise sessions in the resistance training studies lasted 30 min [44], 40 min [191], 60 min [45, 184,185,186,187,188,189, 197] or 90 min [190]. The regimes were conducted for 9 weeks [194], 10 weeks [192], 12 weeks [188], 16 weeks [44, 191], 24 weeks [193], 26 weeks [190, 197], 48 weeks [187], or 52 weeks [45, 184,185,186, 189]. In most of the resistance training studies reviewed, the exercise sessions were conducted in supervised classes [44, 45, 184,185,186,187, 189,190,191, 193, 197]. Furthermore, in most of the reviewed studies, participants were asked to perform two or three sets during the exercise sessions with a minimum of six and a maximum of ten repetitions of upper and lower body exercises at a load ranging from 50 to 92% of 1RM (one repetition maximum) using free weights and/or machines (for a detailed overview, see Table 1).

Main findings

Functional brain changes and cognition

Hemodynamic functional brain changes and cognition

With regard to an acute bout of resistance exercises, in healthy young adults, a decrease in tissue oxygenation index in the left prefrontal cortex during the Stoop test and improved behavioral performance (i.e., faster reaction time and higher number of solved items in neutral condition) was observed after a single bout of high-intensity resistance exercise [43].

With regard to resistance training, after a 16-week intervention with healthy older adults, oxygenated hemoglobin and total hemoglobin were lowered in the left prefrontal cortex during the Stroop task (Stroop interference effect, posttest compared with pretest), while cognitive task performance (i.e., reaction time) was improved [44]. At the end of 52 weeks of resistance training, older adults who had conducted resistance exercises twice a week exhibited better performance in tasks of executive functions (i.e., Stroop test) than those who had performed balance and toning exercises [45]. Furthermore, in the same study, the hemodynamic response during the incongruent flanker condition was increased in the left anterior insula and the left lateral orbitofrontal cortex, whereas the hemodynamic response during the congruent flanker condition decreased in the same areas [45].

In older individuals with mild cognitive impairment (MCI), the right lingual and occipital-fusiform gyri and the right frontal pole exhibited increased activation during the associative memory test after a twice-weekly performed resistance training lasting for 52 weeks when compared with older individuals conducting balance and toning exercises in this time period [189]. Furthermore, in this study, a positive correlation between increased hemodynamic activity in the right lingual gyrus and improved associative memory performance was observed [189]. After 26 weeks of resistance training, decreased resting-state functional connectivity of the PCFC with the left inferior temporal lobe and the anterior cingulate cortex and between the HIPFC and the right inferior temporal lobe was observed in older adults with MCI [190]. In the same study, an increase in resting-state functional connectivity between the HIPFC and the right middle frontal lobe was evident in older adults with MCI in the resistance training group [190].

Neuroelectric functional brain changes and cognition

With regard to an acute bout of resistance exercises, cognitive performance was improved in younger adults [182, 183] and older adults with MCI [195]. After exercising in younger adults, an increase in the P3 amplitude during a Go/No-Go task combined with the Eriksen Flanker paradigm was observed [182], and in older adults with MCI, the P3 amplitude across all electrode positions (except Pz) during the Eriksen Flanker task was larger posttest compared with pretest [195]. Furthermore, in younger adults, a time-dependent and condition-dependent increase in P3 amplitude (obtained during the Stroop task) was observed [183]. In incongruent trials, larger P3 amplitudes were observed 30 min and 40 min after exercise cessation, whereas in congruent trials, larger P3 amplitudes were observed 10 min and 40 min after exercise cessation [183]. However, in the same study, no statistically significant differences between the resistance exercise group and the loadless movement group were observed [183]. Additionally, larger P3 amplitudes were associated with lower serum cortisol levels after an acute bout of resistance exercise in younger adults [182].

With regard to resistance training, after 9 weeks of training (three times per week), the elderly participants showed a significant decrease in N1 latencies at the Fz and Cz positions during an auditory task, whereas the N1-P2, P2-N2 and N2-P3 amplitudes (at Fz) and the N1-P2 amplitude (at Cz) increased [194]. In comparison to both an aerobic training group and an inactive control group, the resistance training group showed  a greater absolute reduction in P2 and N2 latencies and larger absolute increase in N1-P2, P2-N2, and N2-P3 amplitudes [194]. Furthermore, after 10 weeks of resistance training in healthy older adults and in older adults at an early stage of dementia, a decrease in beta asymmetry, a decrease in N200 A asymmetry, and an increase in theta asymmetry was observed [192]. The decrease in N200 A asymmetry was significantly negatively correlated with improvements in the Fuld immediate recall score and the Fuld delayed recall score, while the increase in delta asymmetry was significantly positively correlated with a better Fuld delayed recall score [192]. After resistance training with elastic bands for 12 weeks, healthy older adults showed a decrease in relative theta power at P3 and P4, but their cognitive measures remained unchanged [188]. However, in the same study, exercising older adults with MCI exhibited significantly higher scores in the digit span backward test than their non-exercising counterparts [188]. Furthermore, from pre- to posttest, theta power at F3 increased and alpha power at T3 decreased in exercising older adults with MCI [188]. After 16 weeks of resistance training in older adults with amnestic MCI, larger P3 amplitudes during a task-switching paradigm were observed [191]. Furthermore, in the same study, decreased reaction times (i.e., in the non-switching condition and in the switching condition) and higher accuracy rates (i.e., in the pure condition, in the non-switching condition, and in the switching condition) were noticed in the resistance training group and the aerobic training group when the posttest was compared with the pretest [191]. Additionally, in the resistance training group, a positive correlation between changes in serum levels of insulin-like growth factor 1 (IGF-1) and P3 amplitudes (measured during switching condition) and a negative correlation between serum levels of tumor necrosis factor-alpha and accuracy rates in the switching condition were observed, which both barely failed to attain statistical significance [191]. In another study, 48 weeks of resistance training led to superior cognitive performance (i.e., reaction time) as well as to larger P3a and P3b amplitudes in an oddball task [187]. Moreover, serum IGF-1 concentrations increased and were correlated with faster reaction times and larger P3b amplitudes only in the resistance group [187].

Structural brain changes and cognition

After resistance training performed once or twice weekly for 52 weeks, compared with older adults conducting balance and toning exercises, older adults in the resistance training groups exhibited (i) an increased performance in Stroop test [186], (ii) a reduction in whole brain volume [186], (iii) a lower volume of cortical white matter atrophy [184], and (iv) a lower degree of cortical white matter lesions [185]. In older female adults with probable MCI, resistance training over 26 weeks did not led to significant changes in hippocampal volume [197]. In another study, older adults with MCI resistance training performed twice a week for 26 weeks exhibited improved ADAS-Cog scores (global cognition assessed with Alzheimer’s Disease Assessment Scale) and increased the cortical thickness of grey matter in the posterior cingulate gyrus [190]. Moreover, the increase in grey matter thickness was negatively correlated with ADAS-Cog scores, indicating better cognitive performance [190]. In individuals with multiple sclerosis (MS), resistance training lasting 24 weeks led to an increase in cortical thickness in the anterior cingulate sulcus and gyrus, the temporal pole, the inferior temporal sulcus, and the orbital H-shaped sulcus [193]. The increased thickness in the temporal pole was significantly negatively correlated with lower scores on the Expanded Disability Status Scale (i.e., lower disability) [193]. More detailed information on the main findings is provided in Table 2.

Table 2 Overview of the characteristics of cognitive testing and the main outcomes of the reviewed studies
Full size table

Discussion

Risk of bias

In general, our results regarding the source of the risk of bias are somewhat heterogeneous (see Fig. 3); nevertheless, the overall quality of the majority of the reviewed studies can be regarded as sufficiently high. However, the risk of bias could be further minimized by proper planning of the study, which would strengthen the plausibility of observed effects. To ensure and enhance the study quality, it appears imperative that future studies report their procedures in sufficient detail (e.g., exercise and training variables) and pay attention to established guidelines such as the CONSORT statement [202] or the STROBE statement [203].

Selection of participants and study design

The reviewed studies were conducted with healthy young adults, healthy older adults, or older adults with MCI or beginning dementia. Therefore, our knowledge about the effect of resistance exercises and/or resistance training on cognitive functions is limited to these cohorts, and further investigations with other cohorts are required. In particular, older adults with sarcopenia are a key group because there is a high prevalence (ranging from 1 to 33%) of this condition in various older populations [204], which poses major economic costs to the welfare system [205]. Sarcopenia comprises the age-related loss of muscle mass [206,207,208,209,210] but in the literature the term has often been (incorrectly) extended to the age-related loss of muscle function (e.g., muscle strength) [210,211,212,213,214,215,216,217,218,219]. The latter one should be referred to as dynapenia which encompasses the age-related loss of muscle function (e.g., loss of muscular strength and power) [209,210,211, 220]. However, age-related muscular changes (e.g., sarcopenia) could also lead to a decline in cognitive performance [221, 222]. Hence, older adults with sarcopenia and/or dynapenia may profit in two ways (physically and cognitively) from resistance exercises/resistance training.

In the terms of study design, in future resistance exercise and/or resistance training studies, moderator variables such as gender [223,224,225,226] or genotype [227, 228], which may influence the effectiveness of the resistance exercise and/or resistance training, should be considered and analyzed. The assessment and analysis of moderators may help provide a better understanding of the observed inter-individual variability regarding the effect of physical exercise (e.g., resistance training) on the brain and on cognitive functions and help to foster the optimization of physical exercise interventions [125]. Furthermore, chronobiological factors (such as circadian variability) should be considered since they affect muscular adaptions in response to resistance exercises [229,230,231,232] and affect cognitive performance [233,234,235]. However, hemodynamic responses are reported to be relatively unaffected by, for instance, circadian variability [236].

Moreover, larger cohorts and longer intervention intervals could be beneficial (especially in [f] MRI studies) for increasing the external validity and for adaptation processes to manifest [237]. In addition, concerning cognitive testing, it seems advisable to use standardized sets of cognitive tests or to employ the latent variable approach (create an unobserved [latent] variable for a distinct set of cognitive tests) [238]. In this context, the ‘human baseline hypothesis’ should be considered, which claims that the baseline values of strength (e.g., grip strength, knee extensor strength) assessed prior to resistance training and/or after a detraining period are a more appropriate indicator of health outcomes than the training-related increase in strength values [239].

With regard to upcoming cross-sectional studies, neuroimaging methods (e.g., fNIRS, see [179]) should be employed as they help to better understand the association between superior cognitive performance (e.g., in global cognitive abilities) and superior muscular performance previously operationalized by (i) hand grip strength [86, 88, 89], (ii) isokinetic quadriceps strength [82, 83], (iii) leg power [84], or (iv) whole-body muscular strength [85].

Functional brain changes and cognition in response to resistance exercises or resistance training

Hemodynamic functional brain changes and cognition

Currently, only a few studies have investigated the influence of resistance exercises and/or resistance training on functional brain parameters in healthy adults during standardized cognitive tasks. However, regardless of whether resistance exercises were conducted as an acute bout [43] or over a period of 16 weeks [44], proxies of cortical activation in the prefrontal cortex during the Stroop test were found to be decreased. In another resistance training study (52 weeks), a decrease in brain activation was observed exclusively during the relatively easy task condition, whereas increased activation was found in the more difficult task condition [45]. These observations stand in contrast to the findings of acute aerobic exercise studies [28, 29, 43] and aerobic training studies [44], in which, in general, increased activation of prefrontal areas during cognitive testing was observed after exercising [180]. Notably, similar to the findings of most aerobic exercise or aerobic training studies, the reviewed resistance exercise and/or resistance training studies also reported improved cognitive functions [43,44,45]. Hence, decreases in the applied proxies of neuronal activity might indicate more efficient processing or automatization of cognitive processes. Moreover, it is likely that the decrease in brain activation in response to resistance exercises and/or resistance training is related to neurobiological mechanisms different from those induced by aerobic exercises or aerobic training [107, 223, 240]. Future studies are urgently needed to investigate the underlying neurobiological mechanisms of different types of acute physical exercises (e.g., resistance exercises vs. aerobic exercises) and chronic physical training (e.g., resistance training vs. aerobic training). Analysis of the neurobiological changes in response to different physical exercise/training interventions will also contribute to a better understanding of the functional changes in the brain. In this regard, Liu-Ambrose et al. [45] noticed that after the completion of a 52-week long resistance training program, functional brain activations in the left anterior insula extending from the lateral orbital frontal cortex and in the anterior portion of the left middle temporal gyrus during execution of a cognitive task were altered [45]. The left anterior insula, for instance, plays an important role in successful performance in response inhibition tasks [241], which may be based on their involvement in (i) the stop** ability [242], (ii) the assurance of general task accuracy [242], and (iii) maintaining a stable task set control [243, 244]. The left middle temporal gyrus is especially activated in complex Go−/No-Go situations [245]. However, in contrast, in comparable aerobic training, higher task-related activation in prefrontal areas and parietal cortices and decreased activation of the anterior cingulate cortex was observed [246]. Parietal areas [247] and prefrontal areas [248, 249] are involved in a variety of cognitive processes, among them attention [250, 251]. In particular, the parietal areas [252, 253] and the prefrontal areas [254, 255] are strongly involved in selective attention and the frontoparietal network in maintaining and manipulating task-relevant information in working memory [243]. In the context of attentional processes, the anterior cingulate cortex is also an important structure because it allocates attentional resources based on the recruitment of task-appropriate processing centers [256]. Moreover, the anterior cingulate cortex is activated in conflict processing where erroneous responses are highly probable [257,258,259,260]. Taken together, resistance training might be beneficial for cognitive processes that aim to avoid unwanted responses (e.g., maintaining stable task set control and increased stop efficacy), whereas aerobic exercises may enhance cognitive processes such as selective attention (e.g., maintaining task-relevant information) [45]. Further research is needed to verify this assumption.

The positive effect of resistance training on brain health is also underpinned by findings of Nagamatsu et al. [189], who observed higher cortical activation during an associative memory task in older individuals with MCI after they had undergone long-term resistance training (52 weeks). Moreover, this higher cortical activity was positively correlated with improvements in cognitive performance [189]. Another mechanism through which resistance training may ensure or/and improve brain health in MCI may be related to the modulation of functional connectivity. It was observed that (i) the resting-state functional connectivity between posterior cingulate cortex and other brain regions is generally decreased in individuals with MCI [261,262,393,394].

Recommendations for future studies

  • Based on the available evidence derived from the reviewed studies and other recommendations [107], resistance exercises and/or resistance training aiming to enhance cognitive functions and evoke positive functional and structural brain changes should be designed to induce muscle hypertrophy.

  • Future studies are needed to investigate the influence of the adjustment of different resistance exercise variables (e.g., load, number of sets, training frequency, training duration) on functional and structural brain changes in conjunction with cognitive functions.

  • To understand the time-course of functional and structural brain changes, neuroimaging should be performed at several time points after an acute bout of resistance exercise or during the resistance training intervention.

  • The inclusion of further cohorts (e.g., older individuals with sarcopenia and/or dynapenia) is needed to verify whether resistance exercise-induced improvements also occur in such needy cohorts and how this is related to functional and structural brain changes.

  • Interventional studies (or cross-sectional studies) investigating the relationship of resistance exercises (or strength, muscle function/structure) and cognition should utilize different neuroimaging methods during standardized cognitive testing and assess neurochemical substances (e.g., neurotransmitters, neurotrophic factors) to elucidate underlying neurobiological mechanisms.

  • Bed rest studies, which reported a worsening of executive functions [395,396,397], profound brain changes [397,398,399], and a decrease in muscle mass as well as muscle strength [400,401,402,403,404,405,406,407,408], could be an interesting model to study the relationship between the muscular system, functional and structural brain changes, and cognition.

Conclusions

In summary, resistance exercises and resistance training are powerful physical intervention strategies to induce meaningful functional brain changes, especially in the frontal lobe, which are accompanied by improvements in executive functions. Furthermore, based on the studies reviewed, resistance training leads to lower white matter atrophy and lower volumes of white matter lesions. However, given the small number of available studies that have mostly been part of greater study projects (Brain Power Study and SMART [Study of Mental and Resistance Training]), further research investigating the influence of an acute bout of resistance exercise and chronic resistance training on cognition and the underlying neurobiological mechanisms (e.g., functional and/or structural brain changes) is needed. This future research should also focus on the effects of systematically manipulating exercise and training variables (dose-response relationship) and further including specific cohorts with the greatest need (e.g., older individuals with sarcopenia and/or dynapenia). Most importantly, engaging regularly in resistance exercises and/or resistance training across the whole lifespan appears to be imperative for ensuring physical and brain health because muscular weakness in the early years of life (e.g., adolescence) has been shown to be associated with disability in later life (e.g., after 30 years) [409] and even 4 weeks of detraining (being physical inactive) completely reversed the physical and cognitive improvements of 22-week resistance training in older adults [410]. Hence, to summarize in a metaphorical sense: “May the force be with you across your lifespan.”