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
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
