Table 3.

Main characteristics of selected studies.

Study	Design  Duration  Long-term follow- up  n	Intervention	Result neurobiological outcomes	Result behavioural outcomes
Healthy	The elderly
Valenzuela et al. (2003) [18]	RCT  5 weeks  No follow-up  n = 20	Method of loci	Increased creatine and choline in hippocampus	Improvement in reproduction memory task  No effect on depression or anxiety scores
Nyberg et al. (2003) [30]	Prospective cohort  1 day  No follow-up  n = 24	Method of loci	Acquisition: no group differences   Use phase: adults and improved elderly increased activity in intervention related areas	8 of 16 older persons no improvement in memory task (the unimproved elderly). 8 of 16 older persons and all 8 adults improved in memory task
Small et al. (2006) [20]	RCT  14 days  No follow-up  n = 17	Multicomponent health promotion	Intervention group: decreased activity   prefrontal cortex	Better verbal fluency in intervention group  no significant effect on memory task and subjective memory ability
Erickson et al. (2007) [21]	RCT  2-3 weeks  No follow-up  n = 65	Attentional training	Improvement in dual task performance is correlated with increased activity in left ventral prefrontal cortex and decreased activity in the dorsolateral prefrontal cortex	Both reaction time and accuracy improved most in the dual task intervention group  Young and old adults showed the same degree of performance improvement after intervention.
Brehmer et al. (2011) [22]	RCT  5 weeks  No follow-up  n = 24	Adaptive working memory training. Control group: low level fixed working memory training	All participants decreased brain activity  Participants who profit the most showed the largest decreases in intervention related brain areas and largest increase in caudate	Both groups Improved in span board backward, digit span backward, PASAT, RAVLT  No intervention related performance gains for in scanner task  Intervention group showed training related improvement in span board backward task and PASAT compared to controls
Belleville et al. (2014) [23]	RCT  3 weeks  No follow-up  n = 46	Attentional training	Better performance in single tasking correlated with decreased activity in right inferior and middle frontal gyrus  In the strategic control of attention condition, a better ability to modulate attention according to task instruction correlated with increased activity in Brodmann area 10	All intervention groups improved in reaction time and accuracy for alphanumeric task, no effect for visual detection task  Both dual task conditions better performance dual tasking compared to single task intervention group  Strategic control of attention condition significant effect of task instruction. They were able to modify attention according to instruction.
Mild	Cognitive	Impairment
Clare et al. (2009) [31]	Single case study  8 weeks  No follow-up  n = 1	Goal oriented cognitive intervention	Encoding: increased activity in intervention related brain areas, decreased activity higher visual areas, and frontal areas Recognition: increased activity in intervention related brain areas, decreased activity higher visual areas, and frontal areas	Better subjective memory performance, memory satisfaction
Hampstead et al. (2011) [32]	Multiple single cases  2 weeks  No follow-up  n = 6	Face-name association learning	Encoding: increased activity in default network  Effective connectivity changes: increased connectivity	Significant improvement in performance  trained and untrained memory task
Belleville et al. (2011) [33]	Case control   6 weeks  No follow-up  n = 15	Episodic memory training	Encoding healthy elderly: decreased activity in brain areas related to intervention. Encoding MCI: increased activity in brain areas related to intervention   Retrieval healthy elderly and MCI: increased activity new brain areas and accumulated activity in specialized areas both related to intervention.	Both groups improved on a memory test
Rosen et al. (2011) [16]	RCT  2 months  No follow-up  n = 12	Auditory processing training	Increased activity hippocampus in intervention group decreased activity hippocampus in control group	Intervention group improved in memory test and training tasks
Hampstead et al. (2012) [24]	RCT   2 weeks  No follow-up  n = 34	Mnemonic training	Encoding MCI: increased activity left hippocampal body  Retrieval MCI: increased activity hippocampal body and tail bilaterally  Healthy controls: decreased activity right hippocampal body	MCI group and healthy controls improved in encoding and retrieving trained object locations. No intervention effect for untrained object locations.
Dementia
Heiss et al. (1994) [25]	RCT  6 months  No follow-up  n = 80	(1) Social support  (2) Cognitive training  (3) Cognitive training  + pyritinol  (4) Cognitive training + phosphatidylserine	EEG: increased global power gr 3 + 4 Decreased delta power gr 4  PET: significant correlation MMSE score and glucose metabolism tempero-parietal cortex.   Gr. 4 increased activity primary visual cortex during intervention, but not at the end of the intervention	Gr 4 more responders and significant higher scores on orientation than gr 1 + 2 in week 8 + 16. At the end of the intervention (6 months) there were no differences.
Nagaya et al. (2005) [34]	Within subjects   Unknown  No follow-up  n = 11	Recreational rehabilitation	Responders: decreased activity frontal regions   Non responders: decreased activity all regions	Responders: improved 3 MMSE points
Tanaka et al. (2007) [35]	Single case  2 months  Follow-up for 6 months  n = 1	Reminiscence	increased activity frontal areas, postcingulate gyrus, and precuneus	Improvement in cognition, vitality, volition, and daily life activities
Förster et al. (2011) [27]	RCT  6 months  No follow-up  n = 36	Multipurpose	MCI controls: decreased activity in brain regions typically impaired in AD  MCI intervention: no declines  AD controls: decreased activity in brain regions typically impaired in AD   AD intervention: decreased activity in 2 clusters; lingual gyrus, left inferior temporal gyrus	No changes
Akanuma et al. (2011) [26]	RCT   3 months  No follow-up  n = 24	Reminiscence with reality orientation	Intervention group: increased activity anterior cingulate bilateral, left inferior temporal cortex. Correlation between increased act anterior cingulate and improvement social and communication scales brse	Intervention group: improved in social and communication scales (brse)
Spironelli et al. (2013) [36]	Observational study   5 weeks  No follow-up  n = 11	Cognitive training	The amplitudes of the recognition potential (negative potential) were significantly increased on the left sides of posterior regions for high frequency words	A marginally significant improvement on the verbal reasoning score No significant treatment effect on the dementia screening tests, four out of five cognitive tasks, and the independent functioning questionnaires.
van Paasschen et al. (2013) [28]	RCT  8 weeks  No follow-up  n = 19	Tailored cognitive rehabilitation	Intervention group   Encoding: no significant changes. Recognition: significant increased activity bilateral prefrontal areas and the bilateral insula   Control  group   Recognition: decreased activity in bilateral prefrontal areas and the bilateral insula	Intervention group improved on satisfaction and performance of individual goals (COPM)   No treatment effect on the in scanner face-name association task
Baglio et al. (2014) [29]	RCT  10 weeks  Follow-up for 22 weeks  n = 60, n = 30 Follow-up	Multidimensional stimulation program	Postintervention: increased activity bilateral superior temporal gyrus (right > left), right lentiform nucleus, and thalamus	Intervention group showed clinical relevant improvement in NPI, language, and memory scales of ADAS-cog  No significant change in functional status or physical well-being  Long-term follow-up: improvement in the language and memory scales of the ADAS-cog is preserved

Legenda: 

qol: quality of life.

MMSE: minimental state examination.

ADAS cog: dementia screening test.

BRSE: scale for social and communication skills.

GDS: geriatric depression scale.