Examining the relationship between obesity and cognitive function: A systematic literature review

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Obesity Research & Clinical Practice (2014) xxx, xxx.e1—xxx.e21

REVIEW

Examining the relationship between obesity and cognitive function: A systematic literature review Christina Prickett a,b, Leah Brennan c,b,∗, Rene Stolwyk a a

School of Psychological Sciences, Faculty of Medicine, Nursing & Health Sciences, Monash University, Australia b Centre for Obesity Research and Education, Faculty of Medicine, Nursing & Health Sciences, Monash University, Australia c School of Psychology, Faculty of Arts and Sciences, Australian Catholic University, Australia Received 17 January 2014 ; received in revised form 5 May 2014; accepted 18 May 2014

KEYWORDS Obesity; Body mass index; Cognition; Systematic review

Summary The increasing prevalence of both obesity and dementia is a significant public health concern, especially as recent research demonstrates a significant relationship between these conditions. However, while there is evidence of an obesity—dementia relationship, the effect of obesity on cognitive function in adults, independent of obesity related comorbidities, remains ambiguous. Furthermore, research is yet to systematically compare evidence for domain specific cognitive deficits in obese adults. A systematic literature review was conducted to assess evidence for domain specific cognitive deficits in obese (BMI > 30 kg/m2 ) adults (18—65 years of age) and whether these studies have been able to determine an independent relationship between obesity and cognition over and above relevant comorbidities. Seventeen articles were identified. The literature revealed impairments in obese adults across almost all cognitive domains investigated (e.g. complex attention, verbal and visual memory, decision making). However, numerous methodological limitations were identified which need to be considered in interpretations and conclusions regarding an independent effect. While cognitive impairments in obese adults are evident, as a result of these methodological limitations there is currently insufficient evidence to indicate a reliable and valid independent association between obesity and cognitive impairment in mid-life adults. Further research

∗

Corresponding author at: School of Psychology, Australian Catholic University, 115 Victoria Parade/Locked Bag 4115, Melbourne, VIC 3065, Australia. Tel.: +61 3 9953 3662; fax: +61 3 9953 3205. E-mail addresses: [email protected], [email protected] (L. Brennan). http://dx.doi.org/10.1016/j.orcp.2014.05.001 1871-403X/© 2014 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e2

C. Prickett et al. addressing key methodological limitations (e.g. application of relevant exclusions and control variables, use of appropriate comparison groups and measures) is recommended in order to improve understanding of the relationship between mid-life obesity and cognition. Such research will inform the development of appropriate approaches to identification, prevention and treatment of cognitive decline in obese adults. © 2014 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.

Contents Introduction.................................................................................................. Methods...................................................................................................... Search strategy .......................................................................................... Study selection .......................................................................................... Data extraction.......................................................................................... Design considerations/limitations........................................................................ Results ....................................................................................................... Description of studies.................................................................................... Assessment of cognitive function ........................................................................ Obesity and current cognitive function in obese adults .................................................. General cognitive performance.................................................................... Intellectual functioning ........................................................................... Psychomotor performance and speed.............................................................. Orientation and attention ......................................................................... Visual and space perception....................................................................... Visual construction ................................................................................ Memory ........................................................................................... Language.......................................................................................... Executive function ................................................................................ Discussion .................................................................................................... Summary of findings ..................................................................................... Limitations of existing literature ........................................................................ Consideration of confounds relevant to obesity and cognitive function ............................ Employment of appropriate study designs ......................................................... Use of appropriate comparison groups............................................................. Assessment of all neuropsychological domains..................................................... Use of quality and appropriate assessment tools .................................................. Publication bias ................................................................................... Future directions ........................................................................................ Strengths and limitations of this review ................................................................. Conclusions .................................................................................................. References ...................................................................................................

Introduction Obesity is a significant public health concern in Australia, with the proportions of males and females with obesity increasing from 9% to 19% and 10% to 17% respectively between 1989—90 and 2004—05 [1]. It is predicted that this will increase to 38.4% of men and 20.2% of women by 2022 [2]. This is concerning given that obesity is a significant risk

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

factor for conditions such as cardiovascular disease, diabetes, osteoarthritis, various forms of cancer [3] and depression [4]. Many of these obesity-related comorbidities including hypertension [5], elevated triglycerides [6,7] and type 2 diabetes [8] have been associated with cognitive impairment and increased risk of dementia. Moreover, numerous studies have reported mid-life obesity to be a significant risk factor for later-life dementia (including Alzheimer’s

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Cognitive function in obese adults disease and vascular dementia) [9—12], independent of comorbidities such as hypertension and diabetes [13—15]. This indication that obesity has an independent contribution to dementia risk has significant implications given the increased prevalence of obesity [16] and the ageing population in Australia [17]. Recent research suggests that obesity in middleage may be associated not only with an increased risk of later-life dementia, but also with reduced cognitive performance in middle-age [18]. These findings suggest obesity may impact upon cognition prior to any later life dementia-related cognitive decline. If obesity does impact on cognition in mid-life, this could potentially compromise a person’s memory, attention or decision-making abilities leading to direct occupational and social functioning impairment. Furthermore, if obesity is compromising cognition in mid-life, obesity intervention (e.g. weight-loss) may attenuate later-life dementia risk. Consequently, interest in this area of research has increased, culminating in a recent literature review of the relationship between obesity and cognitive function across the lifespan [19]. This review provided a valuable summary of the research to date examining obesity and cognition in childhood, mid-life and older adulthood. With regards to mid-life obesity, the review concluded that there was significant evidence for cognitive dysfunction in obese adults (19—65 years old), with deficits most consistently evident in the areas of executive functioning [19]. Given preliminary research provides support for a relationship between obesity and cognitive impairment, interest now turns to potential underlying mechanisms of this relationship. As noted above, it is well established that comorbidities of obesity such as depression, hypertension, dyslipidaemia, and type 2 diabetes are associated with cognitive impairment. However, emerging animal and human research have purported a potential independent contribution of obesity (i.e. high adiposity) to cognitive impairment via a range of potential mechanisms including impaired cerebral metabolism [20], elevated leptin [21,22] inflammation [19] and neuronal degradation [19]. Nevertheless, while numerous previous studies have consistently reported reduced cognitive function in obese populations, studies have been inconsistent with delineating the effects of obesity versus obesity-related comorbidities, particularly depression and CVD variables. Current research suggests higher rates of depression in obese populations [4,23] and given depression is also known to be associated with cognitive dysfunction [25,26], it should be a consideration in such research. This

xxx.e3 delineation is important to improve our understanding of how obesity specifically compromises cognitive function and will be crucial in determining which interventions might be most appropriate (i.e. whether targeting depression, CVD variables or weight loss will be most effective). The review described above [19] concluded that mid-life obesity was associated with mid-life deficits in domains of language, motor and memory performance, with deficits most consistently evident in the area of executive function. However, it is noted that the above review did not directly compare cognitive domains across studies and did not account for the quantity and quality of research conducted in various cognitive domains. Thus it is possible that this finding of selective executive dysfunction within obese populations reflects the weight of research conducted in this domain compared to other cognitive domains. A domainspecific review of this literature would enable a clearer understanding of the cognitive profile of obese adults. To summarise, the growing body of literature, including a recent review on the topic [19], have indicated increasing evidence for cognitive dysfunction in mid-life obese individuals, specifically in the area of executive functioning. Despite such research, we are yet to determine the relative influence of obesity given other related comorbidities (e.g. depression) are also known to influence cognitive function. Furthermore, a domain specific review of the cognition and obesity literature has not yet been undertaken. Such knowledge would assist in the appropriate targeting of interventions to reduce such deficits or assist with weight-loss. Consequently, the aims of the current review are twofold. One, to systematically examine the literature to determine the current evidence for domain specific cognitive deficits in obese (BMI ≥ 30 kg/m2 ) mid-life adults (18—65 years). Two, to ascertain whether the relationship between obesity and cognitive function occurs independently of factors known to be associated with both obesity and cognitive dysfunction (i.e. depression, CVD risk factors).

Methods Search strategy This systematic literature review was conducted and reported in line with the current Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) statement [27]. Articles were identified through PsychInfo, MedLine, CINAHL,

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e4

C. Prickett et al. 1.

Obes* or overweight or over weight or BMI or body mass index or waist hip ratio or body weight or Waist circumference or WC or Body fat distribution or weight or body mass or adipos* or Metabolic syndrome or insulin resistance syndrome or syndrome x

2.

Obesity/ or overweight/ or body mass index/ or metabolic syndrome/

3.

1 OR 2

4.

Cognit* or neuropsych* or neurocognit* or memor*

5.

CBT or cognitive behavioral therapy or cognitive behavioural therapy or cognitive behaviour therapy or cognitive behavior therapy or cognitive therapy

6.

4 NOT 5

7.

Cognition/ or Memory/

8.

6 OR 7

9.

3 AND 8

10. Limit search Age 18-65 11. Limit to English Language, Human studies

Figure 1 Example search strategy from MedLine database.

EMBASE, Web of Knowledge and Cochrane corporation electronic databases using terms related to obesity (e.g. ‘obesity’, ‘body mass index’, ‘waist hip ratio’, ‘adiposity’) and cognitive domains (e.g. ‘cognition’, ‘neuropsychology’, ‘neurocognition’, ‘memory’). See Fig. 1 for an example of the search strategy. The search was limited to studies of humans that were published in English. Reference lists of articles included in this review and other relevant reviews were also considered to identify any articles overlooked in the electronic search.

Data extraction Data relating to the concurrent measurement of obesity and cognitive function in adults 18—65 were extracted including: (a) author, year and country of study; (b) BMI and age; (c) study design; (d) study population (e.g., clinical, community sample); (e) comparison group (e.g. normative or non-obese comparison group); (f) control for confounding variables (matching, covariates or exclusions); (g) cognitive measures used; (h) cognitive domain assessed; and (i) reported outcome (significant, non-significant), see Table 1.

Study selection Articles were included if they assessed cognitive functioning in obese adults and met the following inclusion criteria (a) male and/or female participants between the ages of 18—65; (b) a mean BMI ≥ 30 kg/m2 ; (c) concurrent objective measurement of obesity and cognitive function; (d) employed a normative or non-obese comparison group; (e) published in a peer reviewed journal; and (f) written in English. Cognition was to be assessed using tests designed to measure general or specific domains of cognitive function. The above criteria were used to identify potential relevant titles and abstracts from the search results yielded. If abstracts suggested the paper may meet the inclusion criteria, the full text was obtained and evaluated. Ambiguous studies were discussed amongst authors to determine whether they met criteria for inclusion. The papers that met the inclusion criteria were included in the final qualitative analysis (Fig. 2).

Design considerations/limitations In order to address the two aforementioned aims, each study was summarised based on whether key methodological criteria were met. To address the first aim regarding a general relationship between obesity and cognition, studies were reviewed regarding whether they address generally accepted considerations in neuropsychological research (See Table 2): (a) sample size; (b) appropriate exclusions (e.g., considered factors such as neurological history, major psychiatric history, major substance abuse); (c) control group (non-obese comparison group); (d) utilised measures which have undergone a scale development process, and have evidence of reliability and validity. To address the second aim regarding an independent relationship between obesity, studies were reviewed according to whether they controlled or matched for variables relevant specifically when investigating this relationship (see Table 2): (e) age; (f)

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Obese group: age (years)

Study design

Study population (clinical or community)

Sample size and comparison group

Controlled/matched/excluded Cognitive measures variables

Cognitive domain measured

Reported outcomes (significant, non-significant)

Ariza et al. (2012) [36]

 = 38.30 ± 7.59

 = 31.80 ± 6.51

Cross-sectional

Community — Population based

Healthy weight: N = 42

Excluded: History of neurological or psychiatric disorder, possible presence of anxiety or depression (as indicated by the HADS), Alcohol or drug abuse (SCID), obesity related disorders (e.g. thyroid dysfunction), diabetes, hypertension and cognitive impairment Matched: Age, education, HADS raw score (measures of overall anxiety and depression), gender Controlled: None reported

Letter-Number Sequencing

Working memory

NS (p = 0.179)

Symbol Digit Modalities Test

Psychomotor performance and speed

NS (p = 0.215)

Spain

Obese: N = 42

Boeka and Lokken (2008) [31] United States

 = 51.18

 = 41 ± 8.76

Cross-sectional

Clinical — Individuals seeking bariatric surgery

Trail Making Test (B and B-A) Controlled Oral Word Association Test Stroop Colour Word Test Wisconsin Card Sorting Test

Complex attention

NS (p = 0.44 and 0.324)

Verbal fluency

NS (p = 0.691)

Inhibition Concept formation and set shifting

NS (p = 0.403) NS (p = 0.869)

Excluded: None reported

WRAT-3 (reading subtest)

Reading ability

Descriptive

Obese group: N = 68

Matched: Age, gender, education (when normative data allowed) Controlled: Results re-analysed excluding those with reported history of learning problems, head trauma, substance abuse, seizure disorder or CVA/TIA. Did not control for medical co-morbidities hypertension, sleep apnea and diabetes, but instead did further analyses comparing individuals with and without these conditions to see if there were significant differences to indicate whether these conditions had an effect on cognitive performance

WAIS-III (similarities, block design, digit span and digit symbol) Rey Complex Figure Task (copy, 3 min delay, 30 min delay)

Intellectual function

Descriptive

Visual construction and visual memory

Sig (copy and delay; p < 0.001)

Trail Making Test Wisconsin Card Sorting Test Controlled Oral Word Association Test

Complex attention Concept formation and set shifting Verbal fluency

NS (p > 0.005) Sig (p < 0.001)

Animal Naming Task California verbal learning test WMS-III (logical memory subtest)

Verbal fluency Verbal memory

Obese individuals performed better than norms (p < 0.001) NS (p = 0.04) NS (p > .005)

Verbal memory

NS (p = .05)

xxx.e5

Normative sample

ARTICLE IN PRESS

Obese group: body mass index

ORCP-376; No. of Pages 21

Association between obesity and cognitive function in adults 18—65 years.

Author, year, country

Cognitive function in obese adults

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Table 1

Obese group: age (years)

Study design

Study population (clinical or community)

Sample size and comparison group

Controlled/matched/excluded Cognitive measures variables

Cognitive domain measured

Reported outcomes (significant, non-significant)

Chelune et al. (1986) [37]

 = 44.3 (converted mean BMI from weight and height info provided)

 = 32.7 ± 7.5

Cross-sectional

Clinical — Outpatients seeking gastroplasty

Normative sample

Excluded: History of endogenous causes contributing to their obesity (i.e. genetic or endocrine causes) Matched: None (probably matched on age, gender for normative comparison but not reported) Controlled: None

WAIS-R

Intellectual function

Scores normally distributed in similar pattern to normal distribution

Trail Making Test (A)

Complex attention

Substantial number of patients fall in impaired range

Trail Making Test (B)

Complex attention

Category test

Concept formation and set shifting

Substantial number of patients fall in impaired range Substantial number of patients fall in impaired range

Rey Auditory verbal learning test (adapted from) WAIS (version not indicated; Digit Symbol Substitution Subtest) Selective attention test (derived from Sternberg test, 2 subtests)

Verbal memory

Sig (p < 0.001)

Psychomotor performance and speed Complex attention

Sig (p < 0.001)

Obese group: N = 44

United States

Cournot et al. (2006) [30]

Quintiles of BMI were used (quintile 5 obese group:  = 30.5 ± 2.8)

Grouped by age 32—62 years

Cross-sectional and prospective (only using cross-sectional section)

Community — Population based

France

Five groups divided into quintiles of BMI

Excluded: None

Obese group: N = 444

Matched: Age, gender

Controlled: Age, sex, educational level, blood pressure, diabetes, physical activity, region of residence, perceived health score Davis et al. (2010) [38]

Obese Binge Eating Disorder:  = 35.7 ± 9.0

Obese Binge Eating Disorder:  = 34.3 ± 6.5

Canada

Obese control:  = 38.6 ± 7.1

Obese control:  = 35.2 ± 6.7

Etou et al. (1989) [39]

 = 34.6 ± 1.1

 = 34.2 ± 2.3 (SEM)

A case [double] control (normal weight and obese) design

Community sample

Healthy weight: N = 71

Obese Binge Eating Disorder: N = 65

Obese controls: N = 73 Cross-sectional

Clinical — Outpatient obesity clinic

Obese group: N = 13

Healthy weight: N = 13

Delayed free recall

Verbal memory

Sig (p < 0.001)

Decision making

Sig (p < 0.044), did not remain significant when education added to the model

Delay discounting task (computerised)

Decision making

Sig (p < 0.001), did not remain significant when education added to the model

Excluded: No neurological signs, orthopaedic deficits or uncorrected disturbances of vision or hearing in any subject Matched: Age, height, IQ and education

Tap test

Psychomotor performance and speed

Sig (p < 0.01)

Transfer co-ordination test

Sig (p < 0.05)

Controlled: None

Transverse speed test

Psychomotor performance and speed Psychomotor performance and speed Time estimation Time estimation

NS (significance level not reported)

Time judgement estimation Time judgement reproduction

Sig (p < 0.05)

Sig (p < 0.05)

C. Prickett et al.

Japan

Sig (p < 0.001)

Iowa gambling task (computerised)

Excluded: Serious medical condition, were not fluent in English, were pregnant (or had recently given birth), and were being treated for (or had a history of) any psychiatric disorder including eating disorders and substance abuse Matched: Participants were not matched for education, normal weight controls significantly younger Controlled: Education

ARTICLE IN PRESS

Obese group: body mass index

ORCP-376; No. of Pages 21

Author, year, country

xxx.e6

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Table 1 (Continued)

Obese group: age (years)

Study design

Study population (clinical or community)

Sample size and comparison group

Controlled/matched/excluded Cognitive measures variables

Cognitive domain measured

Reported outcomes (significant, non-significant)

Fagundo et al. (2012) [41]

 = 39.8 ± 7.4

 = 40.5 ± 11.1

Cross-sectional

Clinical — Various hospital samples

Healthy weight: N = 137

Excluded: History chronic medical condition, neurological condition, head trauma with loss of consciousness >2 min, learning disability, intellectual disability, use of psychoactive medications or drugs, co-morbid binge eating disorder diagnosis Matched: Not matched on age and education as there were significant differences between groups Controlled: Age and education

Wisconsin Card Sorting Test

Concept formation and set shifting

Sig (p < 0.001)

Stroop colour and word test

Inhibition

Sig (p < 0.001)

Iowa gambling task

Decision making

Sig (p < 0.001)

Clock Drawing Task

Visual construction

NS (p > 0.05)

Trail Making Test A and B (combined score)

Complex attention

Sig (OR: 3.77; p < 0.0061)

MMSE

General cognitive performance

NS (p = 0.17)

Spain

Obese: N = 52 Fergenbaum et al. (2009) [32]

≥30

Range = 19—65 years

Cross-sectional

Community — Canadian First Nations population

Canada

Comparison groups classified as ‘‘impaired’’ and ‘‘non-impaired’’ and odds ratios calculated for obesity

Total N = 207

Excluded: When examining the effects of dyslipidaemia, obesity, metabolic syndrome, and insulin resistance, those having diabetes or fasting plasma glucose levels ≥7 mmol/l (e.g., undiagnosed diabetes) were excluded Matched: None Controlled: Age, Sex, hypertension, CVD, diabetes, insulin resistance, smoking

Gonzales et al. (2010) [29]

 = 34.3 ± 3.5

 = 48.5 ± 8.6

Cross-sectional

Community sample

Healthy weight: N = 9

xxx.e7

Excluded: History of coronary artery disease, angina pectoris, myocardial infarctions, heart failure, cardiac surgery, history of neurological disease (e.g., stroke, parkinsons, clinically significant TBI), major psychiatric illness (e.g., bipolar, schizophrenia), substance abuse (e.g., diagnosed abuse, and/or previous hospitalisation for substance abuse), metabolic disorder (e.g., diabetes, thyroid disorder), smoking (within last 2 years), or MRI contraindications, excluded if fasting blood glucose >126 mg/dl

ARTICLE IN PRESS

Obese group: body mass index

ORCP-376; No. of Pages 21

Author, year, country

Cognitive function in obese adults

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Table 1 (Continued)

Obese group: body mass index

Obese group: age (years)

Study design

Study population (clinical or community)

Controlled/matched/excluded Cognitive measures variables

Cognitive domain measured

Reported outcomes (significant, non-significant)

WASI FSIQ

Intellectual function

NS (p = 0.07)

Overweight: N = 11

Matched: Age, years of education, SBP, DBP, fasting blood concentrations of glucose, LDL-cholesterol and triglycerides Controlled: None

California verbal learning test II (delayed recall) Rey Complex Figure task (delayed recall) WAIS-III (digit-span subtest) Controlled Oral Word Association Test Trail Making Test A and B Verbal n-back test

Verbal memory

NS (p = 0.39)

Visual memory

NS (p = 0.48)

Working memory

NS (p = 0.79)

United States

Obese: N = 12

Gunstad et al. (2006) [33]

≥30

Range = 21—50

Cross-sectional

Community sample

Healthy weight: N = 194

United States

Overweight: N = 106 Obese: N = 43 Halkjaer et al. (2003) [42]

United States of America

NS (p = 0.37) NS (p = 0.91 and 0.43) NS (p = 0.711)

Spot the real word test (computerised task)

Estimated verbal intellectual function

NS (p = 0.34)

12 word list (total recall, delayed recall, recognition)

Verbal memory

Sig (p = 0.015; 0.015 and 0.036)

Borge Priens Prove

Intellectual function

Non-obese control group significantly higher median intelligence score. No statistics reported.

 = 19 (range = 18—24)

Cross-sectional and prospective (only using cross-sectional section)

Community — Population based (Danish Military)

Healthy weight controls: N = 883

Obese group: N = 907

Excluded: Chronic diseases, Diseases sequelae, Handicaps, Mental retardation that requires institutionalisation Matched: Age Controlled: None

 = 49.24 ± 32.32

 = 42.19 ± 9.9

Cross-sectional

Clinical — Individuals seeking bariatric surgery

Normative sample

Excluded: None

WRAT-4 (reading subtest)

Reading ability

Descriptive

Obese group: N = 169

Matched: Age and education matched normative data Controlled: Scores below 10 on Beck Depression Inventory suggest minimal or no depression present in sample

WAIS-III (similarities, block design, digit span, digit symbol)

Intellectual function

NS (p = 0.08)

Wisconsin Card Sorting Test Rey Complex Figure task (copy only)

Concept formation and set-shifting Visual construction

Sig (p < 0.001) Sig (p < 0.001)

C. Prickett et al.

 = 32.5 ± 6.8

Denmark Lokken et al. (2010) [34]

Excluded: Medical conditions known to impact cognitive functioning including: Neurological disorders, head injury, CVD, diabetes, hypertension, thyroid disease, history of significant psychiatric illness including ADHD, Schizophrenia, Bipolar disorder, alcohol and drug use disorders. or a family history of the above conditions Matched: No group differences for age, estimated IQ, education, gender, depression, anxiety, stress Controlled: Age

Verbal fluency Complex attention Working memory

ARTICLE IN PRESS

Sample size and comparison group

ORCP-376; No. of Pages 21

Author, year, country

xxx.e8

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Table 1 (Continued)

Obese group: age (years)

Study design

Study population (clinical or community)

Sample size and comparison group

Controlled/matched/excluded Cognitive measures variables

Cognitive domain measured

Reported outcomes (significant, non-significant)

Nederkoorn et al. (2006) [40] The Netherlands

 = 39.0 ± 5.3

 = 40.9 (6.6)

Cross-sectional

Community — Population based

Healthy weight: N = 28

Excluded: Men, no other exclusions Matched: Age Controlled: None

Stop signal task

Inhibition

NS (p = 0.18)

Delay discounting task

Decision making

NS (p = 0.3)

Pignatti et al. (2006) [35]

 = 42.17 ± 6.00

 = 43.4 ± 8.13

Cross-sectional

Clinical — Hospital obesity clinic

Healthy weight: N = 20

Excluded: History of psychiatric disorders, alcohol or drug abuse or sexual addictions Matched: Age, education and IQ Controlled: Age, education, gender

Gambling task (computerised)

Decision making

Sig (p < 0.04)

Healthy weight: N = 2123

Excluded: None

Borge Priens Prove

Intellectual function

Intelligence score all reduced in obese group (p < 0.001). Within education levels, no significant group differences on intelligence score (p > 0.1)

Obese: N = 1143

Matched: None Controlled: Compared within education levels Excluded: Illnesses or handicaps that warranted rejection from the service (obesity not one of these illnesses); Mental retardation requiring institutionalisation Matched: Age Controlled: Time and place of examination

Borge Priens Prove

Intellectual function

Sig (p < 0.0001)

Excluded: Smoking, substance or gambling problem, eating disorder (based on eating disorders diagnostic scale; EDDS), a serious health condition, current or previous experience of hallucinations or delusions, taking medication that could affect thinking or emotion, controls ever having a BMI >30 Matched: Age Controlled: Subsequent analyses indicated that the delay discounting difference between obese and control women was not related to differences in IQ, age or income

Delay discounting (computerised)

Decision making

Sig (women only; p < 0.02)

Obese: N = 31

Italy Obese: N = 20 Sorensen and Sonne-Holm (1985) [44]

≥31

18

Cross-sectional

Community — Population based (Danish Military)

Denmark

Sorensen et al. (1982) [43]

≥31

Range = 18—21

Cross-sectional

Community — Population based (Danish Military)

Healthy weight: N = 2719

Denmark Obese: N = 1806 Weller et al. (2008) [28]

Male:  = 35.4 ± 4.8

Male:  = 19.2 ± 1.3

Female:  = 38.4 ± 6.6

Female:  = 19.6 ± 2.9 (range = 18—50 years)

Cross-sectional

Community — University sample

Healthy weight: N = 47

United States of America Obese: N = 48

NS (men only; p > 0.05)

ARTICLE IN PRESS

Obese group: body mass index

ORCP-376; No. of Pages 21

Author, year, country

Cognitive function in obese adults xxx.e9

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Table 1 (Continued)

ORCP-376; No. of Pages 21

xxx.e10

Key methodological considerations of studies.

Study

Obesity-related comorbid variables

Sample size

Matched/controlled

N = 84 N = 68 N = 44 N = 2223 N = 209 N = 26 N = 189 N = 207 N = 32 N = 343 N = 1790 N = 169 N = 59 N = 40 N = 3266 N = 4525 N = 95

Adequate exclusions ×

× × × ×

×

Control group × × × × × × × × × × × × × ×

Standardised measures × × × × × × × × × × × ×

Age

Education/SES

CVD risk factors

Depression

Total

× × × ×

× × × × × × ×

×

×

7 3 3 5 3 3 5 4 6 5 2 4 3 3 3 3 4

× × × × × × × × × ×

× × × × × ×

×

× ×

× ×

ARTICLE IN PRESS

Ariza et al. (2012) [36] Boeka and Lokken (2008) [31] Chelune et al. (1986) [37] Cournot et al. (2006) [30] Davis et al. (2010) [38] Etou et al. (1989) [39] Fagundo et al. (2012) [41] Fergenbaum et al. (2009) [32] Gonzales et al. (2010) [29] Gunstad et al. (2006) [33] Halkjaer et al. (2003) [42] Lokken et al. (2010) [34] Nederkoorn et al. (2006) [40] Pignatti et al. (2006) [35] Sorensen and Sonne-Holm (1985) [44] Sorensen et al. (1982) [43] Weller et al. (2008) [28]

General neuropsychological research considerations

C. Prickett et al.

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

Table 2

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Identification

Cognitive function in obese adults

xxx.e11

Records identified through database searching (n = 10355)

Additional records identified through other sources (n =14)

Eligibility

Screening

Records after duplicates removed (n = 8154)

Records screened (n =8154)

Full-text articles assessed for eligibility (n = 154)

Records excluded (n = 8000)

Full-text articles excluded, with reasons e.g. participants not obese (BMI<30 kg/m2) (n = 137)

Included

Studies included in qualitative synthesis (n = 17)

Studies included in quantitative synthesis (meta-analysis) (n = 0) as insufficient data to conduct meta-analysis

Figure 2 Flowchart of literature search performed.

education/socioeconomic status; (g) cardiovascular disease risk factors; and (h) depression.

reference lists). Subsequent to full-text review, 17 articles were found to meet the inclusion criteria and were included in the final analysis, see Fig. 2.

Results Description of studies The literature search yielded 8154 articles following the exclusion of 2215 duplicates. One hundred and fifty four articles were considered following screening of titles and abstracts, with 14 additional articles identified via additional methods (e.g.,

Participant characteristics: Subject ages ranged from 18 to 65, with mean age ranging from 19.2 [28] to 48.5 [29] years old between studies. Average BMI ranged from 30.5 kg/m2 [30] to 51.18 kg/m2

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e12 [31], where averages were reported. Three of the studies included male participants only [32—34]; four studies assessed women only [35—38], while ten studies included both male and female participants [28—31,39—44]. Study design and comparison groups: Twelve studies were cross-sectional in design; four had cross-sectional and prospective components, with one case—control design. Three used normative comparison groups, thirteen studies utilised normal weight and/or overweight comparison groups, and one study used an arbitrary ‘‘impaired’’ cutoff point to compare number of obese individuals in cognitively impaired and non-cognitively impaired groups.

Assessment of cognitive function Cognitive function was assessed using a variety of measures. In total 33 different measures of cognitive function were used and 13 different cognitive functions were assessed. Table 3 provides a summary of domains, assessment tools and abbreviations. Different studies classified cognitive tests and domains differently. Therefore, for the purpose of this review, cognitive functions are grouped into categories as classified by globally recognised references within clinical neuropsychology including Lezak, Howieson and Loring [45] and Strauss, Sherman and Spreen [46]. These categories include: general cognitive performance, intellectual functioning, psychomotor performance and speed, orientation and attention, visual and space perception, visual construction, memory, language, and executive function.

Obesity and current cognitive function in obese adults General cognitive performance The term general cognitive performance will be used in this paper to describe studies that utilised brief cognitive function screening instruments. Only one study utilised such a measure with obese adults [29]. No significant differences were found between groups (normal, overweight, and obese) on Mini Mental State Examination (MMSE) performance. The MMSE was designed as a screening tool for dementia and is thus not sensitive to mild cognitive deficits, particularly attention and executive dysfunction [47]. Accordingly, there appears to be no evidence to support impaired general cognitive performance in obese adults.

C. Prickett et al. Intellectual functioning Intellectual function is a construct purporting to measure hypothesised global cognitive ability [45]. Intellectual function in obese adults was assessed in seven studies [29,32—34,40,41,44]. Three studies utilising data from the Danish draft board between 1956 and 1977 reported a significant relationship between intellectual function and obesity [32—34]. Four studies, two of which utilised normative comparisons [41,44], found no significant differences in general intelligence between obese individuals and normative or control comparisons [29,40,41,44]. Of the three studies finding a significant relationship, two of these did not control for the effect of education [32,33]. The one study that did, found no group intelligence differences when participants were compared within the same education levels [34]. None of these studies controlled for mood or CVD variables [32—34]. Of the studies finding a non-significant relationship, all matched or controlled for education and age, one considered CVD risk variables [29], and two considered depression [40,41]. Despite finding no significant differences between (obese, overweight and healthy weight) groups; one of these studies found a non-significant trend (p = .07) towards lower scores in the obese group [29]. This study matched the groups on age, education, and CVD variables, making it the most controlled of the studies investigating obesity and intelligence. This trend is noteworthy given this study had one of the smallest sample sizes and limited statistical power (N = 32) [29]. Overall, there is evidence of a relationship between obesity and intellectual function, with three studies finding a significant relationship. However, given the inconsistency in consideration of important comorbidities, in cannot be concluded that this effect is fully accounted for by obesity. Psychomotor performance and speed Psychomotor performance is considered the coordination of a sensory or cognitive activity and motor performance [46]. One study assessed psychomotor abilities using a tap test, transfer co-ordination tests and a transverse speed test [36]. Two other studies utilised measures of processing speed (Symbol Digit Modalities Test [43]; and WAIS Digit Symbol Substitution Subtest [30]). Results from the psychomotor performance tasks indicated significantly reduced performance in obese individuals [36]. On measures of processing speed, one study failed to find a difference on Symbol Digit Modalities Test performance between obese and healthy control groups [43]. The other study found scores on the WAIS Digit Symbol Substitution Subtest were significantly poorer for the obese group compared

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Cognitive function in obese adults Table 3

xxx.e13

Domains assessed, neuropsychological assessment tools and abbreviations.

Domain

Measure

Abbreviation

Complex attention

Selective attention test (derived from Sternberg test, 2 subtests) Trail Making Test

TMT

Concept formation and set-shifting

Category test Wisconsin Card Sorting Test

Decision making

Delay discounting task Gambling task (computerised) Iowa gambling task (computerised)

General cognitive performance

Mini Mental Status Examination

Inhibition

Stop signal task Stroop Colour Word Test

Intellectual function

Borge Priens Prove Spot the real word test Wechsler Adult Intelligence Scale-III (similarities, block design, digit span and digit symbol) Wechsler Adult Intelligence Scale-Revised Wechsler Abbreviated Scale of Intelligence

Psychomotor performance and speed

WCST

MMSE

WAIS-III WAIS-R WASI

Symbol Digit Modalities Test

Wechsler Adult Intelligence Scale (version not indicated; Digit Symbol Substitution Subtest) Tap test Transfer co-ordination test Transverse speed test

WAIS

Time estimation

Time judgement estimation Time judgement reproduction

Verbal fluency

Animal Naming Task Controlled Oral Word Association Test

ANT COWAT

Verbal memory

12 word list (total recall, delayed recall, recognition) California Verbal Learning Test Rey Auditory Verbal Learning Test (adapted from) Wechsler Memory Scale-III (logical memory subtest)

WMS-III

Clock Drawing Task Rey Complex Figure task (copy only)

RCF

Visual memory

Rey Complex Figure task (delayed recall)

RCF

Working memory

Letter-Number Sequencing Verbal n-back test Wechsler Adult Intelligence Scale-III (digit-span subtest)

Visual construction

with lower weight groups (quintiles, based on BMI) [30]. Of the three studies investigating this domain, two found significant deficits in performance in obese individuals, providing evidence for a relationship between obesity and psychomotor performance and speed. All studies controlled or

WAIS-III

matched for age and education, two for CVD risk factors [30,43] and one for depression [43]. The study employing the most controls for comorbidities did not find a significant difference, thus there does not appear to be sufficient evidence that is relationship is fully accounted for by obesity.

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e14 Orientation and attention Time estimation. Time estimation describes the ability of a subject to estimate a passage of time [45], with impaired performance on such tasks associated with cognitive deficits seen in amnesia and Alzheimer’s disease [48]. Only one study investigated time judgement, estimation and reproduction in obese individuals [36], finding obese individuals’ significantly underestimated time compared to healthy weight controls. No significant differences between groups on a time judgement reproduction tasks were found. Consequently, there appears to be evidence for a relationship between obesity and time estimation, but not time judgement, in one study. Adequate exclusion criteria were not employed, nor variables such as depression or CVD risk factors controlled for, meaning there is currently no evidence for an independent relationship between obesity and time estimation performance in obese adults. Complex attention. Six studies investigated complex attention performance in obese adults [29—31,39,43,44]. One study used subtests of the Sternberg test to assess selective attention [30], while five studies employed the Trail Making Test (TMT) as a measure of attentional switching [29,31,39,43,44]. Results of two subtests of the Sternberg test [49] (used to derive a selective attention score) found obese individuals performed significantly worse than those in lower BMI quintiles [30]. Five studies utilised the Trail Making Test (TMT) as a measure of switching attention [29,31,39,43,44]. Two of these studies demonstrated significantly reduced performance in obese adults compared to comparison groups [39,44]; while three studies did not find any significant differences [29,31,43]. Of the three studies indicating a significant relationship with complex attention, all controlled or matched for age [30,39,44], two for education [30,44] and CVD risk factors [30,39] and none for depression. Notably, one of these studies reported no statistical analyses meaning little can be concluded from this study [44]. The three studies that indicated no impairment in obese complex attention performance each controlled or matched for age and education [29,31,43]. Only one of these studies controlled for numerous CVD risk factors [29] and only one considered depression [43]. Each study used relatively small samples (total samples ranging from N = 32 to N = 84) and therefore nonsignificant results may be consequent to a lack of statistical power. Overall, three studies indicated impaired complex attention performance in obese adults, and three demonstrated no significant findings. Consequently, evidence is equivocal regarding

C. Prickett et al. a relationship between obesity and complex attention and there is insufficient evidence to conclude an independent relationship at this time. Working memory. Working memory is the ability to retain information over the short term and perform mental operations on these contents [46]. Two studies employed three separate measures of working memory (WAIS-III digit span subtest, Verbal n-back and Letter-Number Sequencing) [29,43]. Specifically, no significant differences were found between three weight groups (obese, overweight and normal weight control) on the WAIS-III digit span subtest total score (forwards and backwards) [29]. Accuracy and reaction time performance on the verbal n-back task did not differ significantly between the three weight groups [29]. The relatively small sample size utilised in this study (N = 32) may have influenced the ability to detect an effect [29]. Letter-Number Sequencing also failed to find a difference between obese and healthy control groups, while matching and excluding for important variables including age, education, depression and cardiovascular factors [43]. Combined, these findings provide no evidence for impairments working memory performance in obese individuals. Accordingly, there is no evidence for an independent link between obesity and working memory performance. Visual and space perception Visual and space perception involves organising and interpreting sensory information to enable an understanding of the environment. Tasks of perception generally require minimal or no physical manipulation of testing materials [45]. According to Lezak, Howieson and Loring classifications [45], no studies have assessed this cognitive domain. Visual construction Constructional abilities combine motor response, perceptual and spatial abilities [45]. Three studies investigated construction abilities [31,39,41]. The first, using the Clock Drawing Test, found BMI was not associated with impaired performance [39]. This is unremarkable given the likely ceiling effect for this test in non-demented individuals [50]. The copy portion of the Rey Complex Figure (RCF) was used in two studies; both reporting significant differences between groups [31,41]. Of the studies finding significant differences on the RCF, both controlled for age and education [31,41] while neither controlled for CVD factors and one controlled for depression [41]. Overall, two studies provide evidence of a relationship between obesity and reduced visual construction performance. However, certain comorbidities were not controlled

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Cognitive function in obese adults for, meaning there is currently insufficient evidence for an independent link between obesity and visual construction. Memory Visual memory. Visual memory is the process by which one encodes, stores, and retrieves visual information [46]. Two studies assessed visual memory using the recall administration of the RCF task [29,31]. One study found no differences between normal weight, overweight and obese individuals using this measure [29]. While this study was adequately controlled, its sample size was small (N = 32), limiting power to detect potential between group differences. The other study found that obese individuals had significantly poorer mean performance compared to normative data on the RCF recall tasks [31], however, these participants were found to have reduced performance on the copy portion of this task compared to the normative sample, which indicates limited encoding of the figure and compromises its use as a measure of delayed visual memory. Overall given these conflicting findings, evidence is equivocal for a relationship between visual memory performance in obese adults. Furthermore, the available research provides no evidence of an independent relationship between visual memory and obesity. Verbal memory. Verbal memory is the process of encoding, storing and retrieving verbal information [46]. Assessment of verbal memory function was conducted in four studies [29—31,40]. Two of these studies reported a significant difference in memory performance (learning and delayed recall) between obese individuals and non-obese control comparisons on word list learning tasks [30,40]. The other two studies found no differences in verbal memory performance of obese adults in comparison to the respective comparison groups (healthy weight and overweight [29] and normative data [31]). Both of the studies obtaining significant findings matched and controlled for age [30,40], while one matched for education [40] and the other controlled for it [30]. One controlled for depression [40]; and one controlled for cardiovascular factors [30]. Of the studies demonstrating a non-significant finding both controlled for age and education [29,31], one controlled for CVD factors [29] and neither for mood factors. One of these studies reported a two point mean difference between obese and healthy controls [29], indicating the small sample (N = 32) may have meant the study was inadequately powered to detect an effect. Overall, given the conflicting findings results are equivocal for an association between mid-life obesity and current verbal memory performance. Accordingly,

xxx.e15 evidence is insufficient to conclude an independent relationship. Language Verbal fluency. Verbal fluency refers to the ease and quantity of spontaneous speech production under semantic and phonemic constraints [45,46]. Three studies assessed phonemic verbal fluency using the COWAT [29,31,43], with one of these studies also measuring semantic fluency using the Animal Naming Task (ANT) [31]. Two studies did not find any significant difference in COWAT performance compared to control comparisons, after matching, controlling or excluding for a number of important confounding variables [29,43]. One of these studies had a small sample size (N = 32) [29]. Interestingly, the third study did find a significant difference using the COWAT; with obese individuals outperforming normative data [31]. This study had a larger sample size (N = 68), but did not control for variables including cardiovascular and mood factors. This study also used the ANT, with no significant differences found between groups [31]. Therefore, there appears to be no evidence for impaired semantic or phonemic verbal fluency in obese adults and no evidence that there is a relationship independent of comorbidities. Furthermore, these higher order verbal fluency skills appear to be the only verbal or language functions assessed in this literature. No studies appear to have assessed comprehension or nominal functions. Executive function Executive function is an umbrella term used to describe a complex set of cognitive processes necessary for responding adaptively in novel situations [46]. This may include a variety of higher-order tasks including problem-solving, planning and judgement. There are issues with defining ‘‘executive function’’ and there is little consensus in the literature about what tasks constitute measures of specific aspects of executive function [46]. As such, for the purpose of this review, executive functions have been divided into sub-classifications to allow for comparisons between tasks with similar underlying processes. The texts of Lezak, Howieson and Loring [45] and Strauss, Sherman and Spreen [46] were again used to aid in classification of the follow measures. Concept formation and set-shifting. Concept formation and set-shifting encapsulate tasks requiring the ability to use abstraction, flexibility and novel problem solving to form concepts and shift back and forth between mental sets in response to the environment [45,46]. Five studies [31,38,41,43,44] assessed these abilities, with four of these using the

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e16 Wisconsin Card Sorting Test (WCST) [31,38,41,43] and one the category test [44]. In the study employing the category test 50% of obese individuals exceeded the cut off indicative of impairment [44], however, no variables were matched or controlled for. Of the studies employing the WCST, three indicated a significant difference in obese performance compared to comparisons [31,38,41]. One study found no significant deficits in obese performance on the WCST relative to comparisons [43]. Of the three studies indicating impaired performance on WCST, all controlled for the effects of age and education [31,38,41], while none considered CVD factors and only one considered depression [41]. Interestingly, the only study that found no significant deficits on the WCST relative to comparisons controlled, matched or excluded for age, education, gender, depression and cardiovascular variables and had an adequate sample (N = 42 in each group) [43]. Overall, four of the five studies indicate deficits in this domain. While these studies provide evidence of a relationship, given variability in methodological controls utilised, research thus far cannot provide conclusions regarding an independent relationship. Decision making and delay discounting. Decision making refers to the ability to select an option or action out of several alternatives, while modifying decisions based on new information available [46]. Three studies investigated decision making performance with a computerised gambling task requiring participants to make decisions about long and short term rewards based on feedback received. Significant differences in decision making performance were seen in each of these three studies compared to comparisons [35,38,42]. Of these studies, the study that matched for education demonstrated a significant difference in decision making performance between obese and normal weight control participants [42]. One study did not match for education, but when education was added to the model in one study, group differences did not remain significant [35]. Group differences remained significant in the other study that controlled for education [38]. No studies controlled for CVD factors or depression. Overall, all three studies provide evidence of deficits in decision making. However, given the variability in the comorbidities considered, research cannot yet conclude whether there is an independent relationship. Delay discounting, one component which underlies decision-making [51], is considered a measure of an individual’s ability to delay smaller more immediate rewards for delayed larger rewards based on available information and feedback [37]. None of these tasks involved decisions around health

C. Prickett et al. behaviours. Three studies employed various versions of delay discounting tasks [28,35,37]. One study found no differences [37], one significant differences (that were washed out following the addition of education to the model) [35], and one a significant difference in obese women, but not men in delay discounting performance compared to normal weight controls [28]. Along with the variability of these results, the comorbid variables considered by these studies was varied with only two controlling for education/SES [28,35]. None of these studies controlled for CVD variables or depression. Overall, the conflicting findings of these studies highlight the equivocal evidence regarding delay discounting performance in obese adults. Furthermore, the variability in the comorbidities considered between studies suggests an independent relationship cannot yet be established. Inhibition. Inhibition refers to the ability to make desired responses while being able to suppress a habitual response when required [46]. One study investigated the response inhibition performance in obese individuals using a stop signal task [37]. No significant deficit was apparent in obese individuals compared to normal weight controls in the community sample of women matched on age. No other matched or control variables were reported [37]. Two studies employed the Stroop Colour Word Test (interference outcome measure) as a measure of inhibitory control [38,43]. One reported that obese adults demonstrated worse performance on this task compared to healthy controls [38]. The other study found no significant difference on this measure despite adequate controls and sample size [43]. Overall, given one study provides evidence of obese related deficits in inhibition and two other studies conflict this finding, evidence is equivocal regarding a deficits in this domain. Accordingly, there is insufficient evidence for an independent link between obesity and inhibition.

Discussion This review aimed to (1) systematically determine whether there is evidence for domain specific cognitive deficits in mid-life obese adults; (2) assess whether these deficits are independent of obesity-related variables and comorbidities (e.g. age, education, mood, CVD factors).

Summary of findings Seventeen articles were reviewed which assessed obesity and cognitive function across a variety of

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Cognitive function in obese adults cognitive domains. Before consideration of obesity related comorbidities, findings provide evidence that mid-life obese individuals exhibit cognitive problems in the following domains: intellectual functioning, psychomotor performance and speed, visual construction, concept formation and set shifting, and decision making. Limited evidence was available for the domain of time estimation. Available evidence was equivocal for a relationship between obesity and visual memory, verbal memory, complex attention, delay discounting and inhibition. There was no evidence of deficits in the areas of general cognitive performance, time judgement, working memory and verbal fluency. While these results support the conclusions from a previous review that found consistent evidence that mid-life obesity was associated with impaired cognitive function, specifically in the area of executive functioning [19], it has also identified evidence for a relationship between obesity and intellectual functioning, psychomotor performance and speed, and visual construction. Once obesity related comorbidities were considered, in order to delineate the independent effect of obesity on cognition, numerous methodological inconsistencies between papers were revealed. Following consideration of such variables there were no domains where there was strong evidence of an independent relationship between obesity and cognition. However, among the following domains there was some evidence of an independent relationship, albeit insufficient to draw sound and meaningful conclusions: psychomotor performance and speed, visual construction, verbal memory, concept formation and set shifting, decision making, delay discounting and inhibition. Among the following domains there was no available evidence to support an independent relationship between obesity and cognitive function: general cognitive performance, intellectual function, time estimation, working memory, visual memory and verbal fluency. Consequently, overall there is currently not enough evidence to indicate a reliable and valid independent association between obesity and cognitive impairment in mid-life adults. Given current research only provides limited evidence of the independent contribution of obesity, further investigation is warranted. Emerging research proposes a range of plausible mechanisms that might contribute to an independent effect of obesity of cognitive function including structural brain changes [52,53], impaired cerebral metabolism [20], elevated leptin [21,22] and inflammation [19]. Independent associations between obesity and structural brain changes including greater brain atrophy [52,53], decreased

xxx.e17 grey matter volumes [53,54], and increased white matter hyperintensities [55] have been demonstrated. Furthermore, a negative relationship between prefrontal cerebral metabolism and BMI has been identified [20]. Leptin, a hormone contained in body fat, and thus found in greater levels in obese individuals, has been associated with impaired cognitive performance [21,22]. While inflammatory proteins (e.g. c-reactive protein), found more commonly in obese individuals, have also been associated with lower cognitive scores among obese females [56]. Inflammatory markers have also been linked with reduced total brain volume [57] providing further evidence that neuronal degradation might be implicated in the obesity—cognition relationship. Nevertheless, while such plausible mechanisms suggest a possible independent effect of obesity on cognitive function, the results of this review indicate research is yet to confirm this relationship. Consequently, future research into the specific impact of obesity on cognitive function in mid-life is warranted. Such information will be crucial in informing further research regarding potential mechanisms underlying this relationship and will be important in important for developing appropriately targeted interventions. There are a number of novel strengths within our review methodology that support the reliability and validity of our findings. First, this review directly compared between specific domains of cognitive function to enable more direct comparisons between studies, along with a more detailed reflection of how such methodological issues may influence interpretation of results. Second, as this paper is specifically addressing mid-life, strict inclusion criteria were used to minimise variability in findings which may be attributable to the aforementioned confounding variables (e.g. age). However, these stringent inclusion criteria resulted in only seven articles that were common to this and the previous review. This present review includes an additional ten studies that were not examined by the previous review paper.

Limitations of existing literature This current review provides some important insights into the strengths and weaknesses of the cognitive performance and obesity literature. This information is valuable in informing the design of future studies in this area. Table 2 demonstrates the extent to which various methodological issues and confounding variables were considered by each of the reviewed studies. While some of these criteria are considered necessary across

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e18 all neuropsychological research, some are specifically relevant to obesity. Some variables were considered consistently across studies (e.g. age, education); however, others were rarely considered (e.g. depression, CVD risk factors). As can be seen in Table 2, only one of the included studies fulfilled all of the aforementioned design considerations criteria considered necessary to investigate an independent link between obesity and cognition. Consideration of confounds relevant to obesity and cognitive function The results of this review demonstrate the limited extent to which studies matched or controlled for relevant confounds. Many of these important considerations (e.g. age, education, depression and CVD risk variables) have been independently associated with cognitive deficits and obesity [58—62] and thus must be considered by studies investigating the relationship between cognitive function and obesity in mid-life. Employment of appropriate study designs The results of this review demonstrate considerable variability between studies in the employment of appropriate exclusion criteria, with few studies screening participants for variables known to be associated with cognitive deficits (e.g. neurological diagnoses, substance use and head injury). As studies have not done this consistently it is possible that some of the reported effects of obesity on cognitive function may have been attenuated had these potential confounds been considered. Conversely, it is also possible that the small sample sizes seen in the majority of these studies may have masked important effects. Use of appropriate comparison groups This review also reveals appropriate comparisons were not consistently employed across studies. Three of the included seventeen studies utilised normative data instead of control comparison groups [31,41,44]. All three of these studies reported significant findings of executive function deficits. Although the use of normative data instead of control groups does not entirely discount the findings of these studies, the use of such data is not considered best practice in neuropsychological research. Different tests use different normative samples which results in multiple comparison groups being utilised within single studies. This is problematic given the characteristics of the different comparison groups (e.g. age, gender, nationality) will vary. Furthermore, as weight data of normative samples are often not available, it is likely that such samples contain both overweight

C. Prickett et al. and obese individuals. This is problematic when obesity is the variable of interest and conclusions about the obesity—cognition relationship are being made based on these group comparisons. Assessment of all neuropsychological domains Despite the broad range of cognitive areas assessed, some domains (e.g. visual and space perception) have not yet been investigated in mid-life obese adults. Additionally, only a small number of studies have investigated each domain of cognitive function. This highlights the need for broad assessment of cognitive domains in mid-life obese adults. Despite the weight of the evidence supporting deficits in the area of executive function, given the scarcity of research in other areas, all domains should continue to be investigated. Furthermore, it is acknowledged that inconsistencies between studies in classification of domains, including a lack of consensus regarding what tests measure what constructs, further compounds difficulties in making meaningful comparisons. Use of quality and appropriate assessment tools The extent to which studies considered issues relevant to measurement selection in this population varied significantly between studies. There was considerable variety in measurement tools selected between studies making comparison difficult. Furthermore, some studies continue to employ measures that should be considered inappropriate for this population (e.g. MMSE). Additionally, neuropsychological test outcome scores were not always interpreted in a consistent and recommended manner (e.g. RCF). Publication bias In addition to these limitations there is also a likely publication bias due to the challenges in publishing non-significant results [63]. The conclusions of this review may have been different if such a bias towards the publication of studies that reject the null hypothesis is evident in this research area.

Future directions The results of this review highlight the need for more methodologically sound research to further understanding the relationship between mid-life obesity and cognition. While there is considerable evidence for reduced cognitive function in obese populations, it remains to be seen whether this relationship is independent of other obesityrelated comorbidities. This information is crucial in determining the potential mechanisms underlying this relationship, along with identifying the

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Cognitive function in obese adults Table 4

xxx.e19

Design considerations for future research.

Design considerations

Checklist

Consideration of confounds relevant to obesity and cognitive function

Match or control for age, education and/or socioeconomic status, cardiovascular disease risk factors and depression

Employ appropriate study designs Exclude participants with a neurological history, major psychiatric history, major substance abuse history Adequate power and sample size Use appropriate comparison groups Control group (non-obese comparison group) Assess all neuropsychological domains Use quality and appropriate assessment tools

Assessing cognitive domains as informed by the research Utilise measures that have undergone a scale development process and have evidence of reliability and validity Select tools with consideration of sensitivity to mild deficits, ceiling effects and potential practice effects

most appropriate interventions. Such interventions may include weight loss (and determining whether this may attenuate later life dementia risk) or targeting such deficits to determine if this might assist in the weight-loss process. Table 4 provides a checklist to inform future research about key design considerations relevant to delineating this relationship. Incorporating these recommendations in future research will assist in understanding the complex nature of the relationship between obesity and cognitive function.

Strengths and limitations of this review To our knowledge this is the only systematic review to have examined domain specific weight-related cognitive deficits in mid-life obese adults, while assessing for evidence of an association independent of obesity related co-morbidities. This review included studies that concurrently measured obesity (BMI ≥ 30 kg/m2 ) and cognitive function in adults (18—65 years). These stringent inclusion criteria are a strength of the current review as they eliminate variability in the included studies, giving more certainty to the applicability of findings to the group of interest (i.e. mid-life obese adults). However, these stringent inclusion criteria also result in the exclusion of studies that fall outside these parameters but may contribute to understanding of the relationships of interest. This strict age criteria resulted in prospective studies being excluded as follow-ups of such studies often fell outside

this aforementioned age criteria. Furthermore, the limited number and varying methodologies across studies, precluded valid meta-analysis and evaluation of effect sizes in accordance with the PRISMA guidelines [64]. Nevertheless, most importantly the current review highlights the importance of research investigating this relationship and the need for numerous issues to be considered in future research in order to determine the exact nature, potential mechanisms, and intervention targets of cognitive deficits in this population.

Conclusions Findings from this review indicate evidence for cognitive impairment in obese individuals. However, critical evaluation of this research suggests that while numerous studies report cognitive impairment in obese individuals, conclusions regarding an independent relationship between these variables are currently limited given inconsistencies in consideration of the impact of obesity related comorbidities. This review highlights the need for more methodologically sound research to be conducted, considering and controlling for a number of key methodological issues. Such research will facilitate a clearer theoretical understanding of the potential mechanisms underlying the relationship between obesity and cognitive function in midlife and is essential to ensuring the development of appropriate and effective interventions. This is

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

xxx.e20 particularly important given the potentially damaging impact of such deficits on occupational and social functioning in obese individuals.

References

C. Prickett et al.

[20]

[21] [22]

[1] More than half of adults are overweight and the numbers are increasing. Australian Bureau of Statistics; 2008. [2] Schneider H, Dietrich ES, Venetz WP. Trends and stabilization up to 2022 in overweight and obesity in Switzerland, comparison to France, UK, US and Australia. Int J Environ Res Public Health 2010;7(2):460—72. [3] Obesity and Overweight [Fact Sheet]. World Health Organisation; 2011. [4] 2007 National Survey of Mental Health and Wellbeing. Australian Bureau of Statistics; 2008. [5] Qiu C, Winblad B, Fratiglioni L. The age-dependent relation of blood pressure to cognitive function and dementia. Lancet Neurol 2005;4(8):487—99. [6] Panza F, D’Introno A, Colacicco AM, Capurso C, Pichichero G, Capurso SA, et al. Lipid metabolism in cognitive decline and dementia. Brain Res Rev 2006;51(2):275—92. [7] Whitmer RA, Sidney S, Selby J, Johnston SC, Yaffe K. Midlife cardiovascular risk factors and risk of dementia in late life. Neurology 2005;64(2):277—81. [8] Grimley Evans J, Areosa Sastre A. Effect of the treatment of type II diabetes mellitus on the development of cognitive impairment and dementia. Cochrane Database Syst Rev 2003, http://dx.doi.org/10.1002/14651858.CD003804. [9] Whitmer RA, Gunderson EP, Barrett-Connor E, Quesenberry Jr CP, Yaffe K. Obesity in middle age and future risk of dementia: a 27 year longitudinal population based study. BMJ (Clin Res Ed) 2005;330(7504):1360. [10] Gustafson D. Adiposity indices and dementia. Lancet Neurol 2006;5(8):713—20. [11] Rosengren A, Skoog I, Gustafson D, Wilhelmsen L. Body mass index, other cardiovascular risk factors, and hospitalization for dementia. Arch Intern Med 2005;165(3):321—6. [12] Whitmer RA, Gunderson EP, Quesenberry Jr CP, Zhou J, Yaffe K. Body mass index in midlife and risk of Alzheimer disease and vascular dementia. Curr Alzheimer Res 2007;4(2):103—9. [13] Whitmer RA, Gustafson DR, Barrett-Connor E, Haan MN, Gunderson EP, Yaffe K. Central obesity and increased risk of dementia more than three decades later. Neurology 2008;71(14):1057—64. [14] Anstey KJ, Cherbuin N, Budge M, Young J. Body mass index in midlife and late-life as a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev 2011;12(5):426—37. [15] Beydoun MA, Beydoun HA, Wang Y. Obesity and central obesity as risk factors for incident dementia and its subtypes: a systematic review and meta-analysis. Obes Rev 2008;9(3):204—18. [16] Australian Health Survey: First Results, 2011—12. Australian Bureau of Statistics; 2012. [17] Measures of Australia’s Progress, 2010. Australian Bureau of Statistics; 2010. [18] Gunstad J, Paul RH, Cohen RA, Tate DF, Spitznagel MB, Gordon E. Elevated body mass index is associated with executive dysfunction in otherwise healthy adults. Compr Psychiatry 2007;48(1):57—61. [19] Smith E, Hay P, Campbell L, Trollor JN. A review of the association between obesity and cognitive function across the

[23]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

lifespan: Implications for novel approaches to prevention and treatment. Obes Rev 2011;12(9):740—55. Volkow ND, Wang G-J, Telang F, Fowler JS, Goldstein RZ, Alia-Klein N, et al. Inverse association between BMI and prefrontal metabolic activity in healthy adults. Obesity (Silver Spring, Md) 2009;17(1):60—5. Harvey J. Leptin regulation of neuronal excitability and cognitive function. Curr Opin Pharmacol 2007;7(6):643—7. Gunstad J, Spitznagel MB, Keary TA, Glickman E, Alexander T, Karrer J, et al. Serum leptin levels are associated with cognitive function in older adults. Brain Res 2008;1230:233—6. Friedman KE, Ashmore JA, Applegate KL. Recent experiences of weight-based stigmatization in a weight loss surgery population: psychological and behavioral correlates. Obesity (Silver Spring) 2008;16(Suppl. 2):S69—74. Austin M-P, Mitchell P, Goodwin GM. Cognitive deficits in depression: possible implications for functional neuropathology. Br J Psychiatry 2001;178:200—6. Castaneda AE, Tuulio-Henriksson A, Marttunen M, Suvisaari J, Lonnqvist J. A review on cognitive impairments in depressive and anxiety disorders with a focus on young adults. J Affect Disord 2008;106(1—2):1—27. Moher D, Liberati A, Tetzlaff J, Altman DG, Group TP. Preferred reporting items for systematic reviews and metaanalyses: the PRISMA statement. PLoS Med 2009;6(7):1—5. Weller RE, Cook Iii EW, Avsar KB, Cox JE. Obese women show greater delay discounting than healthy-weight women. Appetite 2008;51(3):563—9. Gonzales MM, Tarumi T, Miles SC, Tanaka H, Shah F, Haley AP. Insulin sensitivity as a mediator of the relationship between BMI and working memory-related brain activation. Obesity 2010;18(11):2131—7. Cournot M, Marquie J, Ansiau D, Martinaud C, Fonds H, Ferrieres J, et al. Relation between body mass index and cognitive function in healthy middle-aged men and women. Neurology 2006;67(7):1208—14. Boeka AG, Lokken KL. Neuropsychological performance of a clinical sample of extremely obese individuals. Arch Clin Neuropsychol 2008;23(4):467—74. Halkjaer J, Holst C, Sorensen TIA. Intelligence test score and educational level in relation to BMI changes and obesity. Obes Res 2003;11(10):1238—45. Sorensen TI, Sonne-Holm S, Christensen U, Kreiner S. Reduced intellectual performance in extreme overweight. Hum Biol 1982;54(4):765—75. Sørensen TIA, Sonne-Holm S. Intelligence test performance in obesity in relation to educational attainment and parental social class. J Biosoc Sci 1985;17(4):379—87. Davis C, Patte K, Curtis C, Reid C. Immediate pleasures and future consequences. A neuropsychological study of binge eating and obesity. Appetite 2010;54(1):208—13. Etou H, Sakata T, Fujimoto K, Kurata K, Terada K, Fukagawa K, et al. Characteristics of psychomotor performance and time cognition in moderately obese patients. Physiol Behav 1989;45(5):985—8. Nederkoorn C, Smulders FTY, Havermans RC, Roefs A, Jansen A. Impulsivity in obese women. Appetite 2006;47(2):253—6. Fagundo AB, de la Torre R, Jiménez-Murcia S, Agüera Z, Granero R, Tárrega S, et al. Executive functions profile in extreme eating/weight conditions: from anorexia nervosa to obesity. PLoS ONE 2012;7(8):e43382. Fergenbaum JH, Bruce S, Lou W, Hanley AJ, Greenwood C, Young T. Obesity and lowered cognitive performance in a Canadian First Nations population. Obesity 2009;17(10):1957—63.

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001

ORCP-376; No. of Pages 21

ARTICLE IN PRESS

Cognitive function in obese adults

xxx.e21

[40] Gunstad J, Paul RH, Cohen RA, Tate DF, Gordon E. Obesity is associated with memory deficits in young and middle-aged adults. Eat Weight Disord 2006;11(1):e15—9. [41] Lokken KL, Boeka AG, Yellumahanthi K, Wesley M, Clements RH. Cognitive performance of morbidly obese patients seeking bariatric surgery. Am Surg 2010;76(1):55—9. [42] Pignatti R, Bertella L, Albani G, Mauro A, Molinari E, Semenza C. Decision-making in obesity: a study using the Gambling Task. Eat Weight Disord 2006;11(3):126—32. [43] Ariza M, Garolera M, Jurado MA, Garcia-Garcia I, Hernan I, Sanchez-Garre C, et al. Dopamine genes (DRD2/ANKK1TaqA1 and DRD4-7R) and executive function: their interaction with obesity. PLoS ONE 2012;7(7):e41482. [44] Chelune GJ, Ortega D, Linton JC, Boustany M. Personality and cognitive findings among patients electing gastroplasty for morbid obesity. Int J Eat Disord 1986;5(4):701—12. [45] Lezak MD, Howieson DB, Loring DW, editors. Neuropsychological assessment. 4th ed. New York: Oxford University Press; 2004. [46] Strauss E, Sherman EMS, Spreen O. A compendium of neuropsychological tests: administration, norms and commentary. 3rd ed. New York: Oxford Univeristy Press; 2006. [47] Filley CM, Davis KA, Schmitz SP, Stears JC, Heaton RK, Kelly J, et al. Neuropsychological performance and magnetic resonance imaging in Alzheimer’s Disease and Normal Aging. Cogn Behav Neurol 1989;2(2):81—91. [48] Nichelli P, Venneri A, Molinari M, Tavani F, Grafman J. Precision and accuracy of subjective time estimation in different memory disorders. Brain Res Cogn Brain Res 1993;1:87—93. [49] Sternberg S. Memory scanning: new findings and current controversies. Q J Exp Psychol 1975;27:1—32. [50] Lam LC, Chiu HF, Ng KO, Chan C, Chan WF, Li SW, et al. Clock-face drawing, reading and setting tests in the screening of dementia in Chinese elderly adults. J Gerontol B Psychol Sci Soc Sci 1998;53(6):353—7. [51] da Matta A, Gonc ¸alves FL, Bizarro L. Delay discounting: concepts and measures. Psychol Neurosci 2012;5(2):135—46. [52] Ward MA, Carlsson CM, Trivedi MA, Sager MA, Johnson SC. The effect of body mass index on global brain volume in middle-aged adults: a cross sectional study. BMC Neurol 2005;5:23. [53] Gunstad J, Paul RH, Cohen RA, Tate DF, Spitznagel MB, Grieve S, et al. Relationship between body mass

[54]

[55] [56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64]

index and brain volume in healthy adults. Int J Neurosci 2008;118(11):1582—93. Taki Y, Kinomura S, Sato K, Inoue K, Goto R, Okada K, et al. Relationship between body mass index and gray matter volume in 1,428 healthy individuals. Obesity (Silver Spring) 2008;16(1):119—24. Jagust W, Harvey D, Mungas D, Haan M. Central obesity and the aging brain. Arch Neurol 2005;62(10):1545—8. Sweat V, Starr V, Bruehl H, Arentoft A, Tirsi A, Javier E, et al. C-reactive protein is linked to lower cognitive performance in overweight and obese women. Inflammation 2008;31(3):198—207. Jefferson AL, Massaro JM, Wolf PA, Seshadri S, Au R, Vasan RS, et al. Inflammatory biomarkers are associated with total brain volume: the Framingham Heart Study. Neurology 2007;68(13):1032—8. Waldstein S, Katzel L. Blood pressure, obesity and cognitive function: the answers lie in the literature. Int J Obes 2007;31(7):1187—8. Bruehl H, Wolf OT, Sweat V, Tirsi A, Richardson S, Convit A. Modifiers of cognitive function and brain structure in middle-aged and elderly individuals with type 2 diabetes mellitus. Brain Res 2009;1280:186—94. McDermott LM, Ebmeier KP. A meta-analysis of depression severity and cognitive function. J Affect Disord 2009;119(1—3):1—8. Waldstein SR. Hypertension and neuropsychological function: a lifespan perspective. Exp Aging Res 1995;21(4):321—52. Knopman DS, Mosley TH, Catellier DJ, Coker LH. Fourteenyear longitudinal study of vascular risk factors, APOE genotype, and cognition: the ARIC MRI study. Alzheimers Dement 2009;5(3):207—14. Song F, Parekh-Bhurke S, Hooper L, Loke YK, Ryder JJ, Sutton AJ, et al. Extent of publication bias in different categories of research cohorts: a meta-analysis of empirical studies. BMC Med Res Methodol 2009;9:79. Liberati A, Altman DG, Tetzlaff J, Mulrow C, GÃtzsche PC, Ioannidis JPA, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009;151(4):W65—94.

Available online at www.sciencedirect.com

ScienceDirect

Please cite this article in press as: Prickett C, et al. Examining the relationship between obesity and cognitive function: A systematic literature review. Obes Res Clin Pract (2014), http://dx.doi.org/10.1016/j.orcp.2014.05.001