Does cognitive training work? (For Whom? For What?)

The grow­ing field of cog­ni­tive train­ing (one of the tools for brain fit­ness) can appear very con­fus­ing as the media keeps report­ing con­tra­dic­to­ry claims. These claims are often based on press releas­es, with­out a deep­er eval­u­a­tion of the sci­en­tif­ic evidence.

Let’s take a cou­ple of recent exam­ples, in suc­ces­sive days:

It does­n’t work!” type of headline:
Reuters (Feb. 10, 2009)  For­mal brain exer­cise won’t help healthy seniors: research
Healthy old­er peo­ple should­n’t both­er spend­ing mon­ey on com­put­er games and web­sites promis­ing to ward off men­tal decline, the author of a review of sci­en­tif­ic evi­dence for the ben­e­fits of these “brain exer­cise” pro­grams says.

It works! type of headline:
Sci­enceDai­ly (Feb. 11, 2009)  “Com­put­er Exer­cis­es Improve Mem­o­ry And Atten­tion, Study Suggests”
Accord­ing to the researchers, par­tic­i­pants who used the Brain Fit­ness Pro­gram also scored as well as those ten years younger, on aver­age, on mem­o­ry and atten­tion tests for which they did not train.

So, does struc­tured brain exer­cise / cog­ni­tive train­ing work or not?

The prob­lem may in fact reside in ask­ing this very ques­tion in the first place, as Alvaro point­ed out a while ago in his arti­cle Alzheimer’s Dis­ease: too seri­ous to play with head­lines.

We need a more nuanced set of questions.

Why? Because:
1. Cog­ni­tion is made of sev­er­al dif­fer­ent abil­i­ties (work­ing mem­o­ry, atten­tion, exec­u­tive func­tions such as deci­sion-mak­ing, etc)
2. Avail­able train­ing pro­grams do not all train the same abilities
3. Users of train­ing pro­grams do not all have the same needs or goals
4. We need to dif­fer­en­ti­ate between enhanc­ing cog­ni­tive func­tions and delay­ing the onset of cog­ni­tive deficits such as Alzheimer’s.

Let’s illus­trate these points, by ana­lyz­ing briefly the very same study that allowed Reuters Health to claim that “For­mal “brain exer­cise” won’t help healthy seniors”:

1 + 2: Con­trary to what you may believe if you only read the Reuters head­line, the review study cit­ed did find 10 high-qual­i­ty stud­ies in which brain exer­cis­es suc­cess­ful­ly improved tar­get­ed func­tions in healthy old­er adults. Which sug­gests that brain exer­cis­es “work” when used to boost spe­cif­ic cog­ni­tive skills, and that effects last longer than the train­ing itself.

Train­ing pro­grams do not all train the same abil­i­ties. If I need to train my exec­u­tive func­tions and use a pro­gram that trains basic audi­to­ry skills, I may well con­clude that this pro­gram does not “work”. But this pro­gram may work for some­body who needs help in audi­to­ry pro­cess­ing areas.

3 + 4- Why a per­son uses a train­ing pro­gram makes a big dif­fer­ence in the assess­ment of whether this pro­gram works or not.

As men­tioned above, a per­son who reads the review ask­ing whether using a train­ing pro­gram may enhance some cog­ni­tive skills, will prob­a­bly con­clude that a good num­ber of pro­grams work — age itself is not an obstacle.

Fur­ther, the review shows that in the 4 stud­ies that tracked ben­e­fits a few months to a few years after the train­ing, ben­e­fits (or sig­nif­i­cant part of them) remained. So it is not that the report did­n’t find any study that showed that the pro­grams work beyond the train­ing peri­od itself; what it not­ed, and jus­ti­fi­ably so, was the very small num­ber of long-term stud­ies to start with!
The authors of the review also acknowl­edge that evi­dence sup­ports the val­ue of cog­ni­tive train­ing with peo­ple who have mild cog­ni­tive impair­ment and even Alzheimer’s Dis­ease (before exclud­ing such evi­dence from their review in order to focus exclu­sive­ly on healthy adults).

Now, if some­one in Oba­ma’s health­care team reads the review ask­ing, Is there enough evi­dence to invest $5 bil­lion today in buy­ing prod­ucts and dis­trib­ute them to the pub­lic at large in order to pre­vent the upcom­ing Alzheimer’s Dis­ease prob­lem?, it is much hard­er to decide that the pro­grams work for the good rea­son that sci­en­tists do not know yet whether this is an effi­cient and sus­tained way to do so. It would be arguably be pre­ma­ture to invest those $5 bil­lion in buy­ing prod­ucts today.

Let’s now con­trast in more depth those 2 recent stud­ies and try to under­stand what they mean and what they do not mean.

Papp, Walsh, & Sny­der (2009), report­ed in Reuters as “train­ing does­n’t work”, con­duct­ed a meta-analy­sis of the stud­ies pub­lished after 1992 that used cog­ni­tive train­ing with healthy old­er adults. They includ­ed only high-qual­i­ty well-con­duct­ed stud­ies (ran­dom­ized con­trolled tri­als pub­lished in peer-reviewed jour­nals), which left 10 stud­ies in the analysis.

The results con­sis­tent­ly sug­gest­ed that cog­ni­tive skills can be trained, no mat­ter the age group, on a short-term basis. How­ev­er it is not always the case that the train­ing ben­e­fits trans­fer to untrained tasks, even when these are cog­ni­tive­ly close to the trained tasks.

Only 4 stud­ies assessed long-term train­ing ben­e­fits. They showed that small ben­e­fits were main­tained over time after the train­ing (dur­ing a peri­od of sev­er­al months for 2 stud­ies and up to 5 years for one study).

The authors con­clud­ed that there was almost no evi­dence for long-term ben­e­fits BECAUSE most of the stud­ies do not assess long-term benefits.

What this study means:

  • Short-term improve­ments can be obtained for some spe­cif­ic cog­ni­tive skills when using a com­put­er­ized train­ing pro­gram. These improve­ments last longer than the train­ing itself.
  • Stud­ies that assess long-term ben­e­fits of cog­ni­tive train­ing (i.e., delay in onset of demen­tia) are rare; thus the evi­dence for long-term ben­e­fits is scant. We need more research.

What this study does not mean:

  • Cog­ni­tive train­ing CAN post­pone the emer­gence of Alzheimer’s.
  • Cog­ni­tive train­ing CANNOT post­pone the emer­gence of Alzheimer’s.
  • We sim­ply don’t know yet! We need more research track­ing the direct impact of cog­ni­tive train­ing over the long haul.

Now, let’s review the oth­er study, report­ed in Sci­ence Daily.

Smith and col­leagues (2009), “train­ing works!”, report­ed the results of a large ran­dom­ized, con­trolled, dou­ble-blind study test­ing the short-term effects of a com­put­er­ized train­ing pro­gram (Posit Sci­ence clas­sic pro­gram). The IMPACT study involved 487 healthy adults, aged 65 and old­er, for an amount of 40h of train­ing (1h per day, 5 days per week for 8 weeks). Par­tic­i­pants either used a brain train­ing pro­gram or watch edu­ca­tion­al DVD fol­lowed by quizzes (con­trol group). The pro­gram includes 6 exer­cis­es designed to improve the speed and accu­ra­cy of audi­to­ry infor­ma­tion processing.

Par­tic­i­pants who used the train­ing pro­gram showed improve­ment in most of the tests used to assess their audi­to­ry mem­o­ry per­for­mance. Such improve­ment was not shown in the con­trol group.

What this study means:

  • Short-term improve­ments can be obtained for some spe­cif­ic cog­ni­tive skills by using a com­put­er­ized train­ing program.
  • These improve­ments can gen­er­al­ize from the trained tasks to untrained tasks that are cog­ni­tive­ly close.

What this study does not mean:

  • Com­put­er­ized train­ing pro­grams can post­pone the emer­gence of dementia.
  • Com­put­er­ized train­ing pro­grams can­not post­pone the emer­gence of dementia.
  • This train­ing gen­er­al­izes to every impor­tant cog­ni­tive skill one would like to main­tain as we age, or that one’s brain get’s 10 years younger.
  • All train­ing pro­grams will show ben­e­fits for every­body: Train­ing ben­e­fits do not seem to trans­fer to tasks that are not cog­ni­tive­ly close to the trained tasks. Thus one needs to under­stand what tool to use — which is why Sharp­Brains released this 10-Ques­tion Eval­u­a­tion Check­list to help con­sumers and pro­fes­sion­als make informed decisions.


What those 2 recent stud­ies say and imply

  • Cog­ni­tive train­ing can help healthy adults improve spe­cif­ic cog­ni­tive skills, and improve­ments seem to last longer than the train­ing itself (Willis et al., 2006; Smith et al., 2009).
  • Cog­ni­tive train­ing can help adults in the ear­ly stages of cog­ni­tive impair­ment and demen­tia improve some cog­ni­tive skills (Sitzer et al, 2006)
  • One needs to make informed deci­sions. Sharp­Brains’ Eval­u­a­tion Check­list may prove useful.

What nei­ther study says or implies

  • Whether cog­ni­tive train­ing can post­pone the emer­gence of demen­tia: More long-term stud­ies are need­ed. (We know that men­tal­ly stim­u­lat­ing activ­i­ties can help build a Cog­ni­tive Reserve and delay symp­toms of Alzheimer’s Dis­ease, but that evi­dence is not based on ran­dom­ized clin­i­cal tri­als like the ones dis­cussed above).


- Papp, Walsh, & Sny­der. (2009). Imme­di­ate and delayed effects of cog­ni­tive inter­ven­tions in healthy elder­ly: A review of cur­rent lit­er­a­ture and future direc­tions. Alzheimer’s & Demen­tia, 50–60.

- Sitzer, Twamley, & Jeste (2006). Cog­ni­tive train­ing in Alzheimer’s Dis­ease: A meta-analy­sis of the lit­er­a­ture. Acta Psy­chi­a­tr Scand, 114, 75–90.

- Smith et al. A Cog­ni­tive train­ing pro­gram designed based on prin­ci­ples of brain plas­tic­i­ty: Results from the Improve­ment in Mem­o­ry with Plas­tic­i­ty-based Adap­tive Cog­ni­tive Train­ing Study. Jour­nal of the Amer­i­can Geri­atrics Soci­ety, April 2009.

- Willis, S. L., Tennst­edt, S. L., Mar­siske, M., Ball, K., Elias, J., Koep­ke, K. M., Mor­ris, J. N., Rebok, G. W. Unverza­gt, F. W. Stod­dard, A. M., & Wright, E. (2006). Long-term effects of cog­ni­tive train­ing on every­day func­tion­al out­comes in old­er adults. Jour­nal of the Amer­i­can Med­ical Asso­ci­a­tion, 296(23), 2805–2814.

Pascale MichelonPas­cale Mich­e­lon, Ph. D., is Sharp­Brains’ Research Man­ag­er for Edu­ca­tion­al Projects. Dr. Mich­e­lon has a Ph.D. in Cog­ni­tive Psy­chol­o­gy and has worked as a Research Sci­en­tist at Wash­ing­ton Uni­ver­si­ty in Saint Louis, in the Psy­chol­o­gy Depart­ment. She con­duct­ed sev­er­al research projects to under­stand how the brain makes use of visu­al infor­ma­tion and mem­o­rizes facts. She is now an Adjunct Fac­ul­ty at Wash­ing­ton University.


  1. Mike on February 26, 2009 at 12:56

    Thanks for pro­vid­ing this objec­tive view­point on the state of cog­ni­tive train­ing research. Edu­cat­ed and sober arti­cles are hard to find these days!

  2. abdo on February 27, 2009 at 6:44

    thanks soo much …

  3. Maria on February 27, 2009 at 11:15

    Thanks for your respon­si­ble insight. Every­one must research one-on-one brain train­ing. Human to human brain training.

  4. Nick Almond on March 2, 2009 at 3:07

    This is a real­ly good arti­cle and it address­es two major issues in the research area. The ques­tion of who can ben­e­fit from cog­ni­tive train­ing is very impor­tant. It is also direct­ly linked to what cog­ni­tive func­tion­ing needs improv­ing. I am sur­prised that you did not men­tion metacog­ni­tion. Our research has demon­strat­ed that cog­ni­tive activ­i­ty can increase metacog­ni­tion in indi­vid­u­als who have a low num­ber of years in edu­ca­tion or who do not take part in as many cog­ni­tive activities. 

    I must add a note of cau­tion in the con­clu­sions. The major­i­ty of research has used a between-sub­jects design and our research has demon­strat­ed how small indi­vid­u­als dif­fer­ence can play a large role in the effec­tive­ness of a cog­ni­tive inter­ven­tion pro­gramme. You are defi­nate­ly right when you say that more research is need­ed, but these must use a dif­fer­ent exper­i­men­tal approach.

  5. Michael Patterson on March 2, 2009 at 6:53

    This arti­cle, and your web site, pro­vide a won­der­ful ser­vice in help­ing the gen­er­al pub­lic under­stand how to read beyond the head­lines and the mar­ket­ing hyper­bole. Keep up the good work.

  6. coolmathaz on March 7, 2009 at 12:51

    Although neur­al plas­tic­i­ty is rel­a­tive­ly new in terms of tar­get­ed treat­ment pro­grams the jury is in and it most cer­tain­ly does work. Neur­al train­ing pro­grams have doc­u­ment­ed suc­cess in ani­mals and humans, from all spec­trums old and young.

About SharpBrains

SHARPBRAINS is an independent think-tank and consulting firm providing services at the frontier of applied neuroscience, health, leadership and innovation.
SHARPBRAINS es un think-tank y consultoría independiente proporcionando servicios para la neurociencia aplicada, salud, liderazgo e innovación.

Top Articles on Brain Health and Neuroplasticity

Top 10 Brain Teasers and Illusions


Subscribe to our e-newsletter

* indicates required

Got the book?