The Effects of Nonpharmacological Interventions on ...

Special Article Psychother Psychosom 2010;79:6–19 DOI: 10.1159/000254901

Received: September 30, 2008 Accepted after revision: April 23, 2009 Published online: November 4, 2008

The Effects of Nonpharmacological Interventions on Subjective Memory Complaints: A Systematic Review and Meta-Analysis B. Metternich a D. Kosch b L. Kriston c M. Härter c M. Hüll b a

Epilepsy Center, Department of Neurosurgery, b Center for Geriatrics and Gerontology Freiburg, University Hospital Freiburg, Freiburg im Breisgau, c Center for Psychosocial Medicine, Institute and Policlinic for Medical Psychology, Hamburg University Hospital, Hamburg, Germany

Key Words Systematic review ⴢ Meta-analysis ⴢ Subjective memory complaint ⴢ Memory training ⴢ Cognitive restructuring ⴢ Healthy older adults ⴢ Functional memory complaint

Abstract Background: Subjective memory complaints (SMC) in the absence of psychiatric or neurological disorders are common among older adults. Although increasing numbers of individuals are suffering from SMC, research into interventions alleviating SMC is sparse. The present systematic review was conducted in order to present a clearer picture of the quality, quantity and outcomes of SMC intervention research. Methods: PubMed and PsychInfo were searched with a comprehensive search string delivering 4,496 hits. Abstracts and – if applicable – full texts of these studies were screened by independent raters according to predefined criteria. Fourteen relevant studies were included in the review. The categories of intervention were: conventional memory training (MT; n = 8), expectancy change (EC: cognitive restructuring, psychoeducation, etc.; n = 5), combined interventions (n = 5), physical training (n = 1), and physical and mental training combined (n = 2). The outcomes were: subjective memory measures, objective memory, depressive symptoms and well-being. Results: EC, followed by com-

© 2009 S. Karger AG, Basel 0033–3190/10/0791–0006$26.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/pps

bined interventions, was most efficient in influencing subjective memory. MT or physical and mental training combined were not efficient. On objective memory, MT was the only efficient intervention. No effects were found on depressive symptoms and well-being, with low numbers of studies in these comparisons. Conclusion: EC seems to be most efficient in influencing SMC, whereas in objective memory MT is the only effective intervention. Depressive symptoms or well-being do not seem to be influenced significantly by any of the interventions. To date, the number of high-quality publications on SMC interventions is relatively low. More randomized controlled trials including measures of depressive symptoms and well-being are needed. Copyright © 2009 S. Karger AG, Basel

Introduction

Memory complaint has become a topic of interest for researchers of different backgrounds in recent years. The known relationship between mood and memory has been investigated, e.g. by Antikainen et al. [1], who found that a successful treatment of depressive symptoms helps alleviate memory complaint. A great deal of research has focused on memory complaint as a possible prodrome of Alzheimer’s disease [2–4]. Yet another perspective comes Dipl.-Psych. Birgitta Metternich Zentrum für Geriatrie und Gerontologie Freiburg Lehener Str. 88, DE–79106 Freiburg (Germany) Tel. +49 761 270 5422, Fax +49 761 270 7089 E-Mail [email protected]

from psychosomatic research: memory complaint without known organic impairment could be a disorder in its own right. Various authors have suggested that subjective memory complaints (SMC) in this sense could be viewed as a type of somatoform disorder [5–8]. Thus, it is important to clarify what is meant by the term ‘subjective memory complaints’. In some studies the term ‘subjective’ refers to a complaint voiced by the subject of investigation. However, objective memory impairment is not explicitly excluded. Typically, in such studies, dementia poses an exclusion criterion. Nevertheless, patients are included if their score on the Mini-Mental Status is as low as 23 points (e.g. Willlis et al. [9]). This does, of course, not exclude impaired performance on tests of declarative verbal memory, which may be the cause of these SMC. In this context, SMC is a construct not very different from mild cognitive impairment [10]. In the present review, however, SMC is defined as a condition with no known organic basis and no objective memory impairment, i.e. subjective reports of limited everyday memory performance are contrasted with a lack of pathological findings in neuropsychological tests of memory. Persons with SMC in this sense can experience disruption of daily occupational and social activity, which can consequently become an important secondary stress factor. A vicious circle may be set in motion: the patients’ perceived memory lapses could initially be caused by psychological stress [7, 8]. If memory problems are noticed, they can lead to anger and fear. Ultimately this leads to heightened stress levels, which in turn increase the risk of further memory lapses. This type of memory complaint is common in the general population. Commissaris et al. [11] found that 34– 41% of the general Dutch population aged between 40 and 65 had memory complaints and that the frequency of memory complaint seems to increase with age. Blazer et al. [12] report memory complaints in over 50% of a population aged 65 years and older. In memory clinics the percentage of patients with memory complaints without organic or major psychiatric disorders varies between 6 and 12% [6, 13, 14]. As SMC is a common phenomenon and may precipitate anxiety, embarrassment, fear of dementia and repeated medical investigations, interventions directed at alleviating SMC need to be designed and evaluated. In 1997, Floyd and Scogin [15] published a review on the effects of memory training (MT) on subjective memory functioning in older adults. They concluded: ‘Given the prevalence of memory complaints among older adults and their motivation for seeking improvement, the sub-

jective measures of memory performance have clinical significance’ [15, p. 151]. Therefore, the aim of the present systematic review is to identify relevant interventions and evaluate their efficacy in treating SMC. Apart from this efficacy evaluation, another important goal of the present study was to provide an up-to-date overview of the relevant literature in this research field. Rather than only focusing on MT interventions, all types of nonpharmacological interventions, e.g. psychoeducational programs, MT and physical training, were included. However, interventions based on the intake of medication or any type of substance (e.g. cranberry juice, ginkgo) were not included in this review. The most important research questions guiding this work were the following: (1) How many randomised controlled trials evaluating interventions with regard to SMC have been published? (2) How effective are these interventions in comparison with wait-list (WL), placebo (P) or each other in influencing subjective memory (primary outcome)? (3) How effective are these interventions in comparison with WL, P or each other in influencing objective memory measures and measures of depressive symptoms and general psychological well-being (secondary outcomes)?

The Effects of Nonpharmacological Interventions on SMC

Psychother Psychosom 2010;79:6–19

Methods Methods and results are reported in accordance with the Quality of Reporting of Meta-Analyses statement [16]. Identification and Selection of Studies A key word search of MedLine and PsycInfo formed the basis of our article selection. The key words used were: memory AND (training OR intervention OR therapy) AND (anxiety OR perceived OR self-assessment OR self-report OR self-reported OR self-efficacy OR confidence OR complainer OR complaint OR complaints OR complained OR subjective OR belief OR beliefs). In addition, relevant secondary literature was manually scanned, and references to other publications in identified studies were followed up. Research submitted and accepted for publication was included as well, if the authors were made aware of it. Screening and Data Extraction Abstracts and full texts were screened and extracted by 2 independent raters. Congruence was over 99%. Disagreements were resolved by discussion. Only in instances when both raters were unsure whether to include a study and this could not be fully solved by discussion was a third rater’s opinion requested. In this way, all inconsistencies could be solved to the full satisfaction of the involved raters. Inclusion and Exclusion Criteria All nondrug interventions were included. They did not have to target SMC explicitly or primarily, but subjective memory per-

7

formance had to be among the outcome measures. Only peer-reviewed journal articles published in English, Dutch, German or French were included. Furthermore, an abstract and a full-text publication had to be available. Only randomized controlled trials were included. The criteria are listed in table 1. Additionally, in order to be included in the meta-analyses, the studies had to report sufficient data [i.e., mean values, standard deviations (SD) or SE, all divided by intervention group] or provide such on request. According to criterion E1, studies which recruited participants for MT were also included if the participants were not expressly asked for the presence of memory complaints. It can be assumed that the main reason for wishing to participate in MT is the feeling that one’s memory should improve or should be prevented from deteriorating. A study by Verhaeghen et al. [17] showed that out of 129 participants 79% participated because they felt their memory deteriorating or were afraid thereof, wanted to take precautions or to refresh their memory. Fourteen percent wanted to learn about their memory, and only 7% had other reasons. Outcome Measures Subjective Memory. For subjective memory – being the main focus of the present review – we included all measures referring to subjective memory assessment. This covered measures of subjective memory capacity, subjective change in memory functioning, memory-related feelings of anxiety and memory-related control beliefs, but not predictions of performance on selected memory tasks. We also excluded measures of strategy use, as they do not represent subjective memory assessments. The included measures were chosen according to the following criteria. Validated and commonly used measures with representative norms were given priority over measures that were less commonly used. Comprehensive measures (i.e. including aspects of subjective memory capacity and memory-related anxiety, etc.) were preferred over measures assessing only a single aspect of subjective memory (e.g. ‘memory decline index’). The included measures were: the Metamemory in Adulthood Questionnaire (MIA) [18], the Memory Controllability Inventory (MCI) [19], the Memory Functioning Questionnaire (MFQ) [20], a 44-item SMC questionnaire by Zarit [21], a 50-item subjective memory questionnaire regarding everyday memory problems [22], the 10-item Metamemory questionnaire assessing e.g. frequency and seriousness of memory failures [23], a memory diary recording how many times they remembered or forgot items in a checklist of everyday situations [24] and the memory item from the Subjective Quality of Life Profile (see also ‘Well-Being’) [25]. All of these measures contain questions referring to 1 or more of the following: self-rated memory capacity, frequency of forgetting, memory-related anxiety, self-rated change in one’s memory functioning, memory selfefficacy [26], memory-related achievement motivation and memory-related control beliefs. Based on validation data and comprehensiveness of the measures we judged the MIA to be the best measure, followed by the MCI and the MFQ. Although reliability and internal consistency were somewhat inferior for the MCI, we chose this measure over the MFQ because of its greater comprehensiveness. Thus, in studies reporting more than 1 of these 3 measures the best of the used measures was included in our results. This procedure resulted in selecting the MIA for 4 studies, the MCI for 5 and the MFQ for 3. The remaining 4 studies used

8


Table 1. Inclusion and exclusion criteria

Inclusion E1 Studies including subjects with SMC or the desire to improve their memory performance E2 Randomized controlled trials E3 Studies published in English, French, Dutch or German E4 At least one of the outcome measures refers to subjective memory Exclusion A1 Studies exclusively investigating groups of patients with objective memory deficits (e.g. dementia, MCI, AAMI), a relevant neuropsychiatric disorder, traumatic brain injury or cognitive impairment in the course of cytostatic treatment A2 Studies investigating subjects the majority of whom show impaired performance on tests of declarative memory (>1 SD below the norm population’s mean) A3 Studies not evaluating objective memory performance with at least one validated test

less common or self-constructed measures of lower quality. Nevertheless, these were included to allow an assessment of subjective memory in the absence of higher-quality measures. Objective Memory. For objective memory, we tried to include mostly declarative verbal long-term memory measures. This was possible in all but 1 study, where a verbal-visual task (Face-Name Task, Andrewes et al. [24]) was the only available long-term memory measure. The memory measures were chosen according to the following criterion: validated and commonly used measures with representative norms were preferred over laboratory tasks and measures that were less commonly used. The included measures were: the California Verbal Learning Test [27, 28], the Hopkins Verbal Learning Test [29], other list recall tasks [30–32], the Visual Verbal Learning Test [33], the logical memory subtest of the Wechsler Memory Scale [34], the Buschke Selective Reminding Test [35], the Guild Memory Test [36] and the Face-Name Task used by Andrewes et al. [24]. Depressive Symptoms. The included measures were: The Profile of Moods States [37], the Geriatric Depression Scale [38], the Depression Adjective Check List (DACL) [39] and the Zung Self-Rating Depression Scale [40]. Well-Being. The included measures were: the Symptom Checklist-90-R [41], the Well-Being Questionnaire [42] as well as the well-being-item of the Subjective Quality of Life Profile [25]. Data Synthesis For all comparisons standardized mean differences of change scores (from baseline to posttreatment) were calculated. Standardization was necessary because of the large diversity of instruments measuring the same outcome. Hedge’s adjusted g was applied due to the small sample size in some trials. This measure is similar to Cohen’s d but includes an adjustment to correct for the ‘small sample bias’ [43]. As in most cases only results for per-protocol analyses were reported, the numbers of randomized patients were small and selective dropout might be present in some cases, comparability of baseline characteristics could not always be war-

Metternich /Kosch /Kriston /Härter /Hüll

First screening (abstracts) 4,496 studies Exclusion of 4,421 studies because of 1 or more exclusion criteria

Second screening (fulltext screening) 75 studies

14 studies included in review

Exclusion of 61 studies (see table 1 for details) reason

n

%

E1

10

16.4

E2

22

36.1

E4

21

34.4 24.6

A1

15

A2

25

41.0

A3

21

34.4

Other reasons 3 4.9 (see text) % and n add up to >100%, as in most cases >1 reason led to exclusion

1 study not included in any meta-analysis, due to missing data

13 studies included in meta-analyses

Fig. 1. Flow diagram of study inclusion process.

ranted. Thus, the probability of selection bias was rather high, and the analysis of change scores rather than posttreatment results was preferred. If not reported by the primary authors, change scores and their corresponding SD were calculated according to the Cochrane Handbook for Systematic Reviews of Interventions [44]. Missing correlations between baseline and posttreatment scores were imputed using test-retest reliability coefficients of the applied instruments. Examined active interventions were compared pairwise with examined control interventions and with each other. Due to clinical and methodological diversity, individual effect sizes were summarized across studies using a random effects model [45]. Statistical heterogeneity was examined using the I2 statistic, which describes the percentage of the variability in effect estimates that is not due to sampling error (chance) [46]. All analyses were performed using Review Manager 5.0 (Cochrane Collaboration, Copenhagen).


Results

Included Studies The search produced 4,496 potentially relevant articles (after duplicate elimination). The abstracts of these articles were screened, and 74 studies were identified as potentially relevant in this first screening. The full-text publications of these 74 studies were then included in the second screening. Fourteen relevant publications were included in the final pool. Thirteen were included in the meta-analyses across different outcomes, 10 in the analyses for the primary outcome subjective memory (fig. 1). The most common reason was the presence of objective memory impairment in a majority of the subjects. This result strengthens the claim to provide a stricter defPsychother Psychosom 2010;79:6–19

9

Table 2. Overview of included studies: design and sample description Study

Assessment times

Duration of intervention

Comparison groups

Andrewes et al. [24], 1996

baseline post: 1 month

4 weeks

Best et al. [51], 1992

baseline about 2 weeks, post: 2–3 weeks 1.5–2 h/week

Caprio-Prevette baseline and Fry [52], post: 10 weeks 1996

10 weeks 2 h/week

Cases

Relevant measures

1. Memory handbook (MT) 2. Booklet with 3 mnemonic strategies (P)

40

SM: memory diarya OM: face-name test

1. 2. 3. 4.

49

SM: subjective memory complaint questionnaire OM: WMS logical memory D: depression adjective checklist

Expectancy change group (EC) Memory training (MT) Art discussion control group (P) WL

1. Cognitive restructuring (EC) 2. Memory training (MT)

117

SM: MIA OM: guild memory test D: GDS SM: SQLPa OM: WMS logical memory Well-being: SQLPa

Fabre et al. [53], baseline 8 weeks 1999 post: 2 months PT: 2 h/week MT: 1.5 h/week CPT: 3.5 h/week

1. 2. 3. 4.

Aerobic training (PT) Mental training (MT) Combination of both (CPT) WL

32

Lachman et al. [54], 1992

baseline CR: 2 days post: 1.5 weeks MT: 3 days CRMT: 5 days PG: 3 days

1. 2. 3. 4. 5.

Cognitive restructuring (EC) Memory training (MT) Combination of both (CT) Unguided practice (P) WL

105

Levy [58], 1996

baseline post: same day

0.5 SD at baseline. d No mention of blinded assessment. e Subjective memory assessment only after intervention, not at baseline. f Result tables with mean values but no SD. g Statistics unclear. h Not for objective memory measures. i Memory assessment mostly done by computer. j No objective memory outcome. k ITT analysis done. l Number of participants varies among the different tests and assessment times. m Table only for objective memory tests.

effects for MT versus WL and P. A significant effect for CT over WL (p = 0,047), but not versus P, was observed. There was a weak trend towards significance for CT over MT (p = 0.098). There was no significant effect for combined physical and mental training versus WL. There were no data for comparisons regarding PT.

a significant effect for EC versus MT, but not for MT versus WL, P or CT. There were no data for other comparisons. Well-Being. On well-being there was no significant effect for EC versus WL nor for MT versus WL/P. There were no data for other comparisons.

Meta-Analyses for Secondary Outcomes For the secondary outcome parameters only some of the studies could be included in the meta-analyses, as many studies did not report (sufficient) data, especially on depressive symptoms and well-being. Objective Memory. On objective memory there were no significant effects for EC versus WL, P, CT or MT. There were significant effects for MT versus WL and P, as well as MT versus CT; nor were there any for MT versus PT or CPT, for CT versus WL/P and for PT versus WL or combined physical and mental training versus WL/P. Depressive Symptoms. On depressive symptoms there was no significant effect for EC versus WL, but a trend towards significance for EC over P (p = 0.055). There was

Side Effects The only study reporting side effects is Levy [58]. Some subjects complained about eye problems after having been exposed to the visual priming conditions. In most other analyses, side effects were not mentioned. However, it is unlikely that side effects have been systematically investigated in these studies.



Discussion

The results of the present systematic review show that conventional MT is not efficient in reducing SMC. On subjective memory conventional memory training inter13

Table 4. Results of the meta-analyses for all comparisons and outcomes

Study/outcome

Effect estimate

Expectancy change vs. wait list Subjective memory Best et al. [51], 1992 Lachman et al. [54], 1992 Metternich et al. [57], 2008 Objective memory Best et al. [51], 1992 Lachman et al. [54], 1992 Depressive symptoms Best et al. [51], 1992 Well-being Metternich et al. [57], 2008

0.6315 0.7100 0.6700 0.5300 0.0205 0.0000 0.0300 0.3800 0.3800 –0.0400 –0.0400

Expectancy change vs. placebo Subjective memory Best et al. [51], 1992 Lachman et al. [54], 1992 Objective memory Best et al. [51], 1992 Lachman et al. [54], 1992 Depressive symptoms Best et al. [51], 1992

0.6853 0.8400 0.5800 0.0635 0.2800 –0.0900 0.7500 0.7500

Expectancy change vs. memory training Subjective memory 0.9172 Best et al. [51], 1992 0.7600 Caprio-Prevette and Fry [52], 1996 1.9100 Lachman et al. [54], 1992 0.0400 Objective memory 0.0383 Best et al. [51], 1992 –0.4500 Caprio-Prevette and Fry [52], 1996 0.4700 Lachman et al. [54], 1992 –0.1200 Depressive symptoms 1.3020 Best et al. [51], 1992 0.7900 Caprio-Prevette and Fry [52], 1996 1.6900 Expectancy change vs. combined memory training Subjective memory –0.3600 Lachman et al. [54], 1992 –0.3600 Objective memory 0.4800 Lachman et al. [54], 1992 0.4800 Memory training vs. wait list Subjective memory Best et al. [51], 1992 Lachman et al. [54], 1992 Scogin et al. [32], 1985 Scogin et al. [50], 1992 Objective memory Best et al. [51], 1992 Fabre et al. [53], 1999 Lachman et al. [54], 1992 Scogin et al. [32], 1985 Scogin et al. [50], 1992 Depressive symptoms Best et al. [51], 1992 Scogin et al. [32], 1985 Well-being Scogin et al. [50], 1992

14

0.2194 0.0300 0.6800 0.09 –0.0600 0.4903 0.5700 0.6700 0.1600 0.7000 0.5200 0.0226 –0.2900 0.1500 0.3600 0.3600


SE

0.4900 0.3200 0.3800 0.4700 0.3200 0.4700 0.3600

0.4000 0.3300 0.3800 0.3200 0.3900

0.4000 0.2300 0.3000 0.3900 0.1900 0.2900 0.4000 0.2200

0.3100 0.3000

0.4600 0.3200 0.3000 0.3600 0.4700 0.5200 0.3100 0.3100 0.3600 0.4700 0.3000 0.3600

CI start

CI end

I2 (%)

Z

p

0.2023 –0.2504 0.0428 –0.2148 –0.4979 –0.9212 –0.5972 –0.5412 –0.5412 –0.7456 –0.7456

1.0607 1.6704 1.2972 1.2748 0.5389 0.9212 0.6572 1.3012 1.3012 0.6656 0.6656

0.0000

2.8839

0.0039

0.0000

0.0775

0.9382

N/A

0.8085

0.4188

N/A

0.1111

0.9115

0.1864 0.0560 –0.0668 –0.4162 –0.4648 –0.7172 –0.0144 –0.0144

1.1842 1.6240 1.2268 0.5433 1.0248 0.5372 1.5144 1.5144

0.0000

2.6922

0.0071

0.0000

0.2595

0.7953

N/A

1.9231

0.0545

–0.3086 –0.0240 1.4592 –0.5480 –0.5145 –1.2144 0.0976 –0.6884 0.4284 0.0060 1.2588

2.1430 1.5440 2.3608 0.6280 0.5911 0.3144 0.8424 0.4484 2.1756 1.5740 2.1212

92.1737

1.4665

0.1425

66.7054

0.1357

0.8920

74.2716

2.9212

0.0035

–0.9676 –0.9676 –0.1080 –0.1080

0.2476 0.2476 1.0680 1.0680

N/A

1.1613

0.2455

N/A

1.6000

0.1096

–0.1220 –0.8716 0.0528 –0.4980 –0.7656 0.1670 –0.3512 –0.3492 –0.4476 0.0924 –0.1856 –0.4730 –1.2112 –0.4380 –0.3456 –0.3456

0.5608 0.9316 1.3072 0.6780 0.6456 0.8136 1.4912 1.6892 0.7676 1.3076 1.2256 0.5183 0.6312 0.7380 1.0656 1.0656

0.9811

1.2596

0.2078

0.0000

2.9723

0.003

0.0000

0.0895

0.9287

N/A

1.0000

0.3173


Table 4 (continued)

Study/outcome

CI end

I2 (%)

Z

p

–0.0737 –0.6052 –0.0272 –0.6044 0.0570 0.0144 0.1656 –0.5676 –0.4840 –0.6352 –0.6352 –0.7944 –0.7944

0.7324 0.8452 1.2272 0.9244 0.8720 1.4256 1.6944 0.6476 1.0840 0.8152 0.8152 0.7344 0.7344

0.0000

1.6016

0.1092

24.3780

2.2341

0.0255

N/A

0.2432

0.8078

N/A

0.0769

0.9387

0.0715 0.1980 0.3548 1.0424 1.1884 1.2848 0.1548 0.1548

0.0000

1.6563

0.0977

0.0000

2.5617

0.0104

N/A

1.5526

0.1205

0.3800

–0.8515 –0.9780 –1.1348 0.1387 0.0516 –0.2048 –1.3348 –1.3348

0.6400 0.6400

–0.3792 –0.3792

1.6592 1.6592

N/A

1.2308

0.2184

0.5200

Memory training vs. physical-cognitive training Objective memory 0.1800 Fabre et al. [53], 1999 0.1800

–0.8000 –0.8000

1.1600 1.1600

N/A

0.3600

0.7188

0.5000

0.0085 0.3432 0.0072 –0.3526 –1.0872 0.0476 0.1916

1.2201 1.6368 0.7128 0.8789 0.1672 0.7924 1.3284

64.3991

1.9874

0.0469

76.5188

0.8375

0.4023

1.2253 1.5664 0.4528 0.5942 0.0272 0.5328

76.8750

1.1296

0.2587

77.8435

0.4262

0.6699

0.3200 0.1800

–0.3293 0.2336 –0.2528 –0.9244 –1.2272 –0.1728 –0.7900 –0.7900

1.1700 1.1700

N/A

0.3800

0.7039

0.5000

1.0604 1.0604 1.1104 1.1104

N/A

0.2041

0.8383

N/A

0.3061

0.7595

0.4900

–0.8604 –0.8604 –0.8104 –0.8104 –1.4296 –1.4296

0.5696 0.5696

N/A

0.8431

0.3992

0.5100

Memory training vs. placebo Subjective memory Best et al. [51], 1992 Lachman et al. [54], 1992 Scogin et al. [50], 1992 Objective memory Andrewes et al. [24], 1996 Best et al. [51], 1992 Lachman et al. [54], 1992 Scogin et al. [50], 1992 Depressive symptoms Best et al. [51], 1992 Well-being Scogin et al. [50], 1992

Effect estimate 0.3294 0.1200 0.6000 0.1600 0.4645 0.7200 0.9300 0.0400 0.3000 0.0900 0.0900 –0.0300 –0.0300

Memory training vs. combined memory training Subjective memory –0.3900 Lachman et al. [54], 1992 –0.3900 Rasmusson et al. [55], 1999 –0.3900 Objective memory 0.5906 Lachman et al. [54], 1992 0.6200 Rasmusson et al. [55], 1999 0.5400 Depressive symptoms –0.5900 Rasmusson et al. [55], 1999 –0.5900 Memory training vs. physical training Objective memory Fabre et al. [54], 1999

Combined memory training vs. wait list Subjective memory 0.6143 Lachman et al. [54], 1992 0.9900 Valentijn et al. [59], 2005 0.3600 Objective memory 0.2631 Lachman et al. [54], 1992 –0.4600 Valentijn et al. [59], 2005 0.4200 Wolters et al. [31], 1996 0.7600 Combined memory training vs. placebo Subjective memory Lachman et al. [54], 1992 Mohs et al. [60], 1998 Objective memory Lachman et al. [54], 1992 Mohs et al. [60], 1998

0.4480 0.9000 0.1000 –0.1651 –0.6000 0.1800

Physical training vs. wait list Objective memory Fabre et al. [53], 1999

0.1900 0.1900

Physical-cognitive training vs. wait list Subjective memory Small et al. [61], 2006 Objective memory Small 2006

0.1000 0.1000 0.1500 0.1500

Physical training vs. physical-cognitive training Objective memory –0.4300 Fabre et al. [53], 1999 –0.4300


SE

0.3700 0.3200 0.3900 0.3600 0.3900 0.3100 0.4000 0.3700 0.3900

0.3000 0.3800 0.2900 0.3800

0.3300 0.1800 0.3200 0.1900 0.2900

0.3400 0.1800

0.4900

CI start


15

ventions did not show significant effects compared to P, WL or other active treatment control conditions. A combination of physical and mental training was not successful in influencing SMC either. Lack of data did not allow any comparisons for physical training alone. However, EC interventions were more successful. They showed significant effects compared to WL and P control conditions (ES = 0.63 and 0.69, respectively). They were not statistically significantly superior to other active treatment conditions, i.e. combined MT or conventional MT. However, for the latter comparison, the pooled effect size was high (ES = 0.92) and in favor of EC, but the heterogeneity of effects was too high to produce a significant result. The combined intervention (conventional MT plus EC) was also somewhat successful in affecting SMC in a positive way. There were significant effects compared to WL (ES = 0.61) but not P control conditions. Combined MT was not as successful in comparison with EC. However, there was a trend of significance in comparison with conventional MT (ES = 0.39). The present results are fairly congruent with those of Floyd and Scogin’s [15] review. Therefore, it is possible that – as Floyd and Scogin suggested – EC is the more important component, when it comes to influencing SMC, and that contrarily to their original findings, conventional MT has no additional positive effect at all. Floyd and Scogin [15] identified 3 studies investigating EC interventions and 5 investigating combined interventions. In the present review, the numbers were: 5 EC interventions and 5 combined interventions. Our inclusion criteria were stricter than those of Floyd and Scogin, and a great number of studies were excluded because of a lack of randomization. The small number of studies is unfortunate, as it reduces the power of the meta-analyses. However, the congruence between our and Floyd and Scogin’s [15] results lends a greater credibility to the present findings. The results regarding objective memory functioning followed an utterly different pattern than those for the subjective measures. The conventional MT interventions had a positive effect compared to P and WL control groups (ES = 0.41 and 0.46, respectively). There were no significant effects of conventional MT over EC, physical training or combined physical and mental training. However, there was only 1 study included in each of those comparisons. There was a significant effect of conventional MT versus combined MT (ES = 0.59). Neither EC nor combined MT or (combined) physical training had any significant positive effects on objective memory. This result does not come unexpected. Many studies have 16


shown that MT can influence memory test performance (see also Verhaeghen et al. [62]). Perhaps, larger effects could have been predicted. On the other hand, the present review explicitly excluded studies investigating participants with objective memory impairment. So, possibly this group of participants did not have ample room for improvement, as their performance on memory tests was already in the normal range. The results regarding objective memory functioning also need to be viewed in comparison to the results concerning subjective memory. The lack of a relation of the results in these 2 different categories of measures can be interpreted as a genuine effect of EC and combined MT interventions on subjective memory rather than a secondary effect via the improvement of objective memory functioning. It seems that objective and subjective memory can be influenced relatively independently and that different types of intervention are suited to influence either subjective or objective memory functioning. The measures of general psychological well-being and depressive symptoms could also potentially influence subjective memory functioning, as various studies have shown correlations between subjective memory and affective measures [6, 7]. In the present review, apart from EC over conventional MT, none of the interventions had any reliable effects on either depression or well-being. However, the number of studies included in these analyses was low, sometimes down to 1 study per comparison. Nevertheless, this is congruent with the notion forwarded by Floyd and Scogin [15] that interventions focused on alleviating SMC can have a direct effect on subjective memory which is not mediated by an improvement in depressive symptomatology. However, the present review did not include patient samples with major depression. The possibility of individuals suffering from minor depression, which also seems to be associated with cognitive complaints [63], cannot be excluded. However, more than 70% of the included studies explicitly excluded individuals with major depression. So, this lack of a relationship might only be valid for healthy elderly samples, as mood improvement followed by the reduction in SMC has been reported in depressed patients [1]. The relatively low number of randomized controlled trials for the treatment of SMC is an important result of the present review. With 14 publications the number of studies investigating SMC in healthy elderly populations applying a randomized controlled design is not very high. This result is somewhat surprising, as Floyd and Scogin already showed in 1997 that more studies invesMetternich /Kosch /Kriston /Härter /Hüll

tigating SMC are needed. Furthermore, they emphasized the importance of including expectancy modification components in new interventions for SMC. The low numbers of high-quality studies including such components show that this suggestion does not seem to have been taken up by many researchers. One reason for this could be the subjects’ expectations. Especially ambitious, anxious and rigid personalities might be prone to interpret normal age-related changes of memory in the light of imminent dramatic decline in functioning. People with SMC often seem to prefer MT or over-the-counter medication over psychological interventions such as cognitive restructuring. This might be because they find it difficult to accept their memory is normal compared to their age group. These clients are unlikely to feel the need for a psychological intervention targeting their attitudes, beliefs and expectations regarding their own memory. However, this is a good argument for combined interventions. It might be easier to reach people with SMC if they are offered MT along with expectancy modification. The main limitation of the findings is probably the relatively low number of included studies in some comparisons. However, most analyses yielded clear (statistically and clinically significant) results without considerable heterogeneity. The methodological and especially the publication quality of the studies was not generally high, with a few exceptions, e.g. lack of blinded assessments could have potentially influenced the results regarding objective memory measures but not the subjective measures. Due to the report of per-protocol rather than intention-to-treat results by primary authors the possibility of pooled effect sizes being slightly biased cannot be ruled out [44]. The main advantages of the statistical methods that were applied in the present study are the correction for small sample bias, the adjustment for baseline differences in the outcomes and the fitting of a random-effects model that accounts for heterogeneity between the results of primary studies. Some results are heterogeneous, e.g. the comparison EC versus MT. One out of three studies [54] found no effect on subjective memory. One possible explanation could be the comparison groups’ difference in number and duration of sessions. The MT subjects received more sessions than the EC subjects. The overall treatment duration was longer for the MT group, which could have caused a bias favoring MT in this case. Other cases of heterogeneity could be due to differences in design and study quality (e.g. lack of blinded assessment).

The heterogeneity of the P treatments poses a potential problem to the interpretation of the results. Baskin et al. [64] found that inadequately designed P treatments (i.e. P structurally different from the active treatment) produced larger effect sizes in a meta-analysis than adequately designed ones. None of the 5 P treatments investigated in the included studies fulfilled all of the criteria for structural equivalence suggested by Baskin et al. [64]. The 2 studies with the most inadequate P [24, 50] solely investigated conventional MT versus P. Here, the effects of MT could have been overestimated. The generalizability of the findings might be somewhat restricted as all of the included studies were conducted in ‘Western’ societies (USA, France, Germany, the Netherlands, Australia). Therefore, it is doubtful whether these results can be generalized to societies with different cultural backgrounds. In all studies, the mean age was over 50 years. Most commonly, the participants were aged between 60 and 80 years. Therefore, the results might not be the same for younger populations. Most participants were female, with the proportions ranging between two thirds and three quarters of female participants across the studies with few exceptions. This also has to be taken into consideration if one tries to generalize the findings to populations with higher proportions of males. All in all, the present review provides evidence that interventions containing expectancy modification components alone, and lesser so in combination with mnemonics, are effective when it comes to influencing SMC, whereas conventional MT and physically oriented interventions do not seem to be effective. Furthermore, these effects do not appear to be secondary to an improvement in objective memory functioning, depressive symptoms or general psychological well-being. However, the small number of included studies limits the interpretation of some of the results, especially those regarding well-being and depressive symptoms. Therefore, future intervention studies on SMC should include measures of depressive symptoms and psychological well-being. As demographic change occurs in Western countries, more importance should be placed on psychological conditions affecting the elderly. In many elderly SMC sufferers the complaints reach a critical level, warranting their labeling as a disorder [7, 8]. These patients seek medical help, as their everyday memory problems cause psychosocial stress and often fear of dementia. A mere feedback, reassuring them of their normal test results, has not proven effective in alleviating their complaints [8]. A recent review has shown that SMC is related to lower quality of



17

life [65]. It is important to understand SMC, in order to be able to treat them or even prevent them from occurring in the first place. Older people might be more prone to suffer from SMC because they cannot evade negative age stereotypes [58]. Therefore, apart from developing and evaluating interventions directed at alleviating SMC, it might also be necessary to make society as a whole more aware of negative age-related stereotypes and their potential consequences.

Conflict of Interest The authors have been supported by grants from the German Federal Ministry of Education and Research (Kompetenznetz Demenzen 01 GI 0420), the German Federal Ministry of Health (Leuchtturmprojekt Demenzen) and have been involved in clinical research with anti-dementia drugs supported by the pharmaceutical industry. The authors have served as experts for the Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen. No funding has been received in the context of the present research project.

References 1 Antikainen R, Hanninen T, Honkalampi K, Hintikka J, Koivumaa-Honkanen H, Tanskanen A, Viinamaki H: Mood improvement reduces memory complaints in depressed patients. Eur Arch Psychiatry Clin Neurosci 2001;251:6–11. 2 O’Brien, John T, et al: Do subjective memory complaints precede dementia? A three-year follow-up of patients with supposed ‘benign’ senescent forgetfulness’. Int J Geriatr Psychiatry 1992;7:481–486. 3 Jorm AF, Christensen H, Korten AE, et al: Memory complaints as a precursor of memory impairment in older people: a longitudinal analysis over 7–8 years. Psychol Med 2001;31:441–449. 4 Glodzik-Sobanska L, Reisberg B, De Santi S, et al: Subjective memory complaints: presence, severity and future outcome in normal older subjects. Dement Geriatr Cogn Disord 2007;24:177–184. 5 Ponds RWHM, Schmidt AJM, de Lugt M, et al: De angst om te vergeten: behandeling van functionele geheugenklachten. Tijdschr Psychiatr 1995;37:62–68. 6 Berrios GE, Markova IS, Girala N: Functional memory complaints: hypochondria and disorganization; in Berrios GE, Hodges JR (eds): Memory Disorders in Psychiatric Practice. Cambridge, Cambridge University Press, 2000, pp 384–399. 7 Metternich B, Schmidtke K, Hüll M: How are memory complaints in functional memory disorder related to measures of affect, metamemory, and cognition? J Psychosom Res 2009;66:435–444. 8 Schmidtke K, Pohlmann S, Metternich B: The syndrome of functional memory disorder: definition, etiology and natural course. Am J Geriatr Psychiatry 2008;16:981–988. 9 Willis SL, Tennstedt SL, Marsiske M, Ball K, Elias J, et al: Long-term effects of cognitive training on everyday functional outcomes in older adults. JAMA 2006;296:2805–2814. 10 Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings JL, DeKosky ST: Practice parameter: early detection of dementia: mild cognitive impairment (an evidence-based review). Neurology 2001; 56:1133–1142.

18

11 Commissaris CJAM, Ponds RWHM, Jolles J: Subjective forgetfulness in a normal Dutch population: possibilities for health education and other interventions. Patient Educ Couns 1998;34:25–32. 12 Blazer DG, Hays JC, Fillenbaum GG, Gold DT: Memory complaint as a predictor of cognitive decline. J Aging Health 1997; 9: 171– 184. 13 Schmidtke K, Metternich B, Hüll M: Funktionelle Gedächtnis- und Konzentrationsstörungen. Neurogeriatrie 2007;4:79–84. 14 Weiner MF, Bruhn M, Svetlik D, Tintner R, Horn J: Experiences with depression in a dementia clinic. J Clin Psychiatry 1991;52:234– 238. 15 Floyd M, Scogin F: Effects of memory training on the subjective memory functioning and mental health of older adults: a metaanalysis. Psychol Aging 1997; 12:150–161. 16 Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF: Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of Reporting of Meta-Analyses. Lancet 1999;354:1896–1900. 17 Verhaeghen P, Van Ranst N, Marcoen A: Memory training in the community: evaluations by participants and effects on metamemory (preview). Educ Gerontol 1993;19:525–534. 18 Dixon RA, Hultsch DF, Hertzog C: The Metamemory in Adulthood (MIA) questionnaire. Psychopharmacol Bull 1988; 24: 671–688. 19 Lachman ME, Bandura M, Weaver SL, Elliott E: Assessing memory control beliefs: the memory controllability inventory. Aging Cogn 1995;2:67–84. 20 Gilewski MJ, Zelinski EM, Schaie KW: The Memory Functioning Questionnaire for assessment of memory complaints in adulthood and old age. Psychol Aging 1990; 5: 482–490. 21 Zarit Sh: Affective correlates of self-reports about memory of older people. Meet Gerontol Soc, Washington, 1979.


22 Deelman BG, Saan RJ: Memory deficits: assessments and recovery; in Deelman BG, Saan BJ, vanZomeren AH (eds): Traumatic Brain Injury. Clinical, Social and Rehabilitation Aspects. Lisse, Swets & Zeitlinger, 1990. 23 Zelinski EM, Gilewski MJ, Thompson LW: Do laboratory tests relate to self-assessment of memory ability in the young and old?; in Poon LW, Fozard JL, Cermak LS, Arenberg D, Thompson LW (eds): New Directions in Memory and Aging. Proceedings of the George A. Talland Memorial Conference. Hillsdale, Erlbaum, 1980, pp 519–544. 24 Andrewes DG, Kinsella G, Murphy M: Using a memory handbook to improve everyday memory in community-dwelling older adults with memory complaints. Exp Aging Res 1996;22:305–322. 25 Gerin P, Dazord A, Boissel JP, Chifflet R: Quality of life assessment in therapeutic trials: rationale for and presentation of a more appropriate instrument. Fundam Clin Pharmacol 1992;6:263–276. 26 Hertzog C, Hultsch D, Dixon R: Evidence for the convergent validity of two self-report questionnaires. Dev Psychol 1989; 25: 687– 700. 27 Delis DC, Kramer JH, Kaplan E, Ober BA: California Verbal Learning Test – Adult Version. New York, Psychological Corporation, 1986. 28 Delis DC, McKee R, Massman PJ, Kramer JH, Kaplan E, Gettman D: Alternate form of the California Verbal Learning Test: development and reliability. Clin Neuropsychol 1991;5:154–163. 29 Brandt J: The Hopkins Verbal Learning Test: development of a new memory test with six equivalent forms. Clin Neuropsychol 1991;5: 125–142. 30 Hertzog C, Dixon RA, Hultsch DF: Relationships between metamemory, memory predictions, and memory task performance in adults. Psychol Aging 1990; 5:215–227. 31 Wolters G, Theunissen I, Bemelmans KJ, van der Does AJW, Spinhoven Ph: Immediate and intermediate-term effectiveness of a memory training program for the elderly. J Cogn Rehabil 1996;14:16–22.


32 Scogin F, Storandt M, Lott L: Memory skills training, memory complaints, and depression in older adults. J Gerontol 1985;40:562– 568. 33 Brand N, Jolles J: Learning and retrieval rate of words presented auditorily and visually. J Gen Psychol 1985; 112:201–210. 34 Wechsler D: A standardized memory scale for clinical use. J Psychol 1945;19:87–95. 35 Buschke H, Fuld PA: Evaluating storage, retention and retrieval in disordered memory and learning. Neurology 1974; 24: 1019– 1025. 36 Gilbert JG, Levee RF, Catalano FL: Guild Memory Test Manual. Bloomfield, Unico National Mental Health Research Foundation, 1981. 37 Spielberger C: Profile of Mood Scale (POMS). San Diego, Educational and Industrial Testing Service, 1971. 38 Sheikh JI, Yesavage JA: Geriatric Depression Scale (GDS): recent evidence and development of a shorter version; in Brink TL (ed): Clinical Gerontology. A Guide to Assessment and Intervention. New York, Haworth Press, 1986, pp 165–173. 39 Lubin B: Depression Adjective Checklists. San Diego Educational and Industrial Testing Service, 1967. 40 Zung WW: A self-rating depression scale. Arch Gen Psychiatry 1965;12:63–70. 41 Derogatis LR: SCL-90-R, Administration, Scoring and Procedures Manual-I for the R(evised) Version. Baltimore, Johns Hopkins University School of Medicine, 1977. 42 Tempelman CJJ: Welbevinden bij ouderen (dissertation). Groningen, University of Groningen, 1987. 43 Hedges LV, Olkin I: Statistical Methods for Meta-Analysis. San Diego, Academic Press, 1985. 44 Higgins JPT, Green S (eds): Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.0. The Cochrane Collaboration, 2008. 45 DerSimonian R, Laird N: Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–188.


46 Higgins JP, Thompson SG, Deeks JJ, Altman DG: Measuring inconsistency in meta-analyses. Br Med J 2003;327:557–560. 47 Rebok GW, Rasmusson DX, Brandt J: Prospects for computerized memory training in normal elderly: effects of practice on explicit and implicit memory tasks. Appl Cogn Psychol 1996;10:211–223. 48 Rebok GW, Rasmusson DX, Bylsma FW, Brandt J: Memory improvement tapes: how effective for elderly adults? Aging Neuropsychol Cogn 1997;4:304–311. 49 Scogin F, Prohaska M, Weeks E: The comparative efficacy of self-taught and group memory training for older adults. J Clin Geropsychol 1998; 4:301–314. 50 Scogin F, Prohaska M: The efficacy of selftaught memory training for communitydwelling older adults. Educ Gerontol 1992; 18:751–766. 51 Best DL, Hamlett KW, Davis SW: Memory complaint and memory performance in the elderly: the effects of memory-skills training and EC. Appl Cogn Psychol 1992; 6: 405– 416. 52 Caprio-Prevette MD, Fry PS: Memory enhancement program for community-based older adults: development and evaluation. Exp Aging Res 1996;22:281–303. 53 Fabre C, Massé-Biron J, Chamari K, Varray A, Mucci P, Préfaut Ch: Evaluation of quality of life in elderly healthy subjects after aerobic and/or mental training. Arch Gerontol Geriatr 1999;28:9–22. 54 Lachman ME, Weaver SL, Bandura M, Elliott E, Lewkowicz CJ: Improving memory and control beliefs through cognitive restructuring and self-generated strategies. J Gerontol 1992;47:293–299. 55 Rasmusson DX, Rebok GW, Bylsma FW, Brandt J: Effects of three types of memory training in normal elderly. Aging Neuropsychol Cogn 1999;6:56–66. 56 Metternich B, Schmidtke K, Dykierek P, Härter M, Hüll M: Konzeption und Erprobung eines gruppentherapeutischen Behandlungskonzepts für die funktionelle Gedächtnis- und Konzentrationsstörung. Psychother Psychosom Med Psychol 2009 (E-pub ahead of print).

57 Metternich B, Schmidtke K, Dykierek P, Hüll M: A pilot group therapy for functional memory disorder. Psychother Psychosom 2008;77:259–260. 58 Levy B: Improving memory in old age through implicit self-stereotyping. J Pers Soc Psychol 1996;71:1092–1107. 59 Valentijn SA, van Hooren SA, Bosma H, Touw DM, Jolles J, van Boxtel MP, Ponds RW: The effect of two types of memory training on subjective and objective memory performance in healthy individuals aged 55 years and older: a randomized controlled trial. Patient Educ Couns 2005; 57:106–114. 60 Mohs RC, Ashman TA, Jantzen K, Albert M, Brandt J, Gordon B, Rasmusson X, Grossman M, Jacobs D, Stern Y: A study of the efficacy of a comprehensive memory enhancement program in healthy elderly persons. J Psychiatr Res 1998;77:183–195. 61 Small GW, Silverman DHS, Siddarth P, Ercoli LM, Miller KJ, Lavretsky H, Wright BC, Bookheimer SY, Barrio JR, Phelps ME: Effects of a 14-day healthy longevity lifestyle program on cognition and brain function. Am J Geriatr Psychiatry 2006;14:538–545. 62 Verhaeghen P, Marcoen A, Goossens L: Improving memory performance in the aged through mnemonic training: a meta-analytic study. Psychol Aging 1992; 7:242–251. 63 Howland RH, Schettler PJ, Rapaport MH, Mischoulon D, Schneider T, Fasiczka A, Delrahiem K, Maddux R, Lightfoot M, Nierenberg AA: Clinical features and functioning of patients with minor depression. Psychother Psychosom 2008;77:384–389. 64 Baskin TW, Tierney SC, Minami T, Wampold BE: Establishing specificity in psychotherapy: a meta-analysis of structural equivalence of placebo controls. J Consult Clin Psychol 2003;71:973–979. 65 Mol M, Carpay M, Ramakers I, Rozendaal N, Verhey F, Jolles J: The effect of perceived forgetfulness on quality of life in older adults; a qualitative review. Int J Geriatr Psychiatry 2007;22:393–400.


19