obesity reviews doi: 10.1111/j.1467-789X.2011.00920.x
Etiology and Pathophysiology/Obesity Prevention
A review of the association between obesity and
cognitive function across the lifespan: implications
for novel approaches to prevention and treatment obr_920 740..755
E. Smith1,2, P. Hay3,4, L. Campbell5,6 and J. N. Trollor1,2
1
Brain and Ageing Research Program, School Summary
of Psychiatry, University of New South Wales, Recent research suggests that increased adiposity is associated with poor cogni-
Sydney, New South Wales, Australia; tive performance, independently of associated medical conditions. The evidence
2
Department of Developmental Disability regarding this relationship is examined in this review article. A relatively consis-
Neuropsychiatry, School of Psychiatry, tent finding across the lifespan is that obesity is associated with cognitive deficits,
University of New South Wales, Sydney, New especially in executive function, in children, adolescents and adults. However, as
South Wales, Australia; 3School of Medicine, illustrated by contradictory studies, the relationship between obesity and cogni-
University of Western Sydney, Sydney, New tion is uncertain in the elderly, partly because of inaccuracy of body mass index as
South Wales, Australia; 4School of Medicine, a measure of adiposity as body composition changes with aging. This review
James Cook University, Sydney, New South further discusses whether obesity is a cause or a consequence of these cognitive
Wales, Australia; 5Garvan Institute of Medical deficits, acknowledging the possible bidirectional relationship. The possible effects
Research, Darlinghurst, New South Wales, of increased adiposity on the brain are summarized. Our investigations suggest
Australia; 6Department of Endocrinology, St that weight gain results, at least in part, from a neurological predisposition
Vincent’s Hospital, Darlinghurst, New South characterized by reduced executive function, and in turn obesity itself has a
Wales, Australia compounding negative impact on the brain via mechanisms currently attributed to
low-grade systemic inflammation, elevated lipids and/or insulin resistance. The
Received 2 March 2011; revised 9 July 2011; possible role of cognitive remediation treatment strategies to prevent and/or treat
accepted 10 July 2011 obesity is discussed.
Address for correspondence: Dr E Smith, Keywords: Adiposity, body mass index, cognition, cognitive impairment, obesity.
Department of Developmental Disability
Neuropsychiatry, 34 Botany Street, University obesity reviews (2011) 12, 740–755
of New South Wales, Sydney, NSW 2031,
Australia. E-mail: evelyn.smith@unsw.edu.au
fed a high-fat diet have also been shown to be cognitively
Introduction
impaired compared with those fed normal chow (9–13). In
Obesity, defined by accumulation of excess adipose tissue, humans, recent investigations across the lifespan have
has become a worldwide epidemic with dramatic conse- examined whether obese individuals have cognitive deficits
quences for health because of its association with increased compared with their normal-weight counterparts. Child-
heart disease, hypertension, diabetes, stroke and cancer hood and adult obesity is increasing dramatically and if
(1,2). While some of these medical comorbidities are them- obese individuals show early impairment of cognitive per-
selves associated with adverse cognitive effects (3–6), recent formance this could lead to a larger and even more devas-
research suggests that adiposity has a specific association tating epidemic: that of earlier onset dementia. Midlife
with cognitive function. Obese rats have been shown to obesity is already a risk factor for dementia in the elderly
perform worse on learning and memory tasks compared (14–18). While there is a focus on improving cognition and
with rats of normal weight (7,8). Notably, non-obese rats function in the elderly, in an obese individual there may be
740 © 2011 The Authors
obesity reviews © 2011 International Association for the Study of Obesity 12, 740–755
,obesity reviews Obesity and cognitive function across lifespan E. Smith et al. 741
greater impediments to improvement. With the substantial Cognitive function refers to the processing, integration,
increase in obesity, a significant decline in cognitive perfor- storage and retrieval of information. Cognitive function
mance could have further functional implications, not only includes perception, attention, memory and executive func-
for the individual but for society in general. tion. Executive function, the most consistent deficit found in
The evidence regarding the relationship between obesity the obese, encompasses a diverse range of cognitive pro-
and cognition across the lifespan is reviewed in this article. cesses facilitating initiation, planning, regulation, sequenc-
We first summarize and discuss the findings of the associa- ing and achievement of complex goal-oriented behaviour
tion in childhood and adolescence (age 4 to 18 years), in and thought (22,23), which in turn may impact on eating
adulthood (age 19–65 years) and in the old (age 65–99 behaviour. Executive function has been usually measured by
years). There is also evidence that poor performance in tasks such as the Wisconsin Sorting task, the Stroop task or
cognitive function in children predicts future increase in the Trail Making test, which measure switching and cogni-
body mass index (BMI), suggesting a bidirectional relation- tive flexibility, and the Iowa gambling task, which measures
ship. The effects of weight loss and high-fat diets on cog- inhibition and rule acquisition. However, working memory
nition will be briefly discussed. In addition, this review will has also been suggested to be part of executive function (24).
examine the effects of obesity on the brain via neuroimag- Cognitive dysfunction or impairment has been usually esti-
ing methods, with mention of potential biological mecha- mated by statistically significant deficits in test performance
nisms for this association. compared with healthy controls or normative data.
Methods Results
PubMed, Medline and PsycInfo were searched for words
Cognitive deficits in obese children and
such as ‘obesity’, ‘adiposity’, ‘overweight’, ‘body mass
adolescents (aged 4–18 years)
index’, ‘weight’, ‘waist circumference’ or ‘waist to hip
ratio’ paired with ‘cognition’, ‘cognitive function’ or ‘cog- Nine cross-sectional studies were found which examined
nitive performance’. Studies that used the Mini-mental obesity and cognitive function in children and adolescents
state examination (19) were excluded because (i) it is of (see Table 1). Some studies were population-based whereas
relevance as a screening measure only in the elderly; (ii) it is others reflected a clinically obese population. Obesity
not a sensitive or specific measure of integrity for particular was defined using standard growth curves (weight above
cognitive domains and (iii) the score is affected by educa- the 95th percentile classified as obese) and compared with
tion and cultural factors (20). Any additional articles normal weight. Alternatively, using a linear regression
meeting inclusion criteria but not identified by search were model, an increasing BMI was associated with lower cog-
crosschecked. Articles were scanned to ensure that the asso- nitive function, suggesting a linear relationship. All cross-
ciation between obesity and cognition was examined. sectional articles were published within the last 5 years,
In this review, we selected only articles that measured highlighting this as a developing area. Findings have been
both obesity and cognition, with primary aim to identify relatively consistent, with eight of the nine studies showing
the relationship between these two constructs. That is, significantly poor cognitive indices in obese individuals
other articles examining obesity with another condition, compared with those of normal weight.
such as hypertension, were not included in this review Cognitive deficits in obese children and adolescents were
(unless hypertension was included as a covariate). The lit- found most consistently in tests of executive function (25–
erature on metabolic syndrome was also excluded for four 28), while some studies also found deficits in short-term
reasons (i) the clinical construct metabolic syndrome has memory (29) (an aspect of executive function (24)), global
multiple definitions with different criteria and obesity is the functioning (30,31) and verbal abilities (32) compared with
main factor we are interrogating; (ii) a review of cognitive normal-weight adolescents. It has been suggested that
dysfunction in metabolic syndrome was published in 2007 impairment in verbal abilities in young age could lead to
(21) (although most studies included in the present review impairment in executive function (33). The study that did
were published after 2007); (iii) metabolic syndrome not report an association between high weight and poor
cannot be diagnosed in children under 10 years old and (iv) cognitive function used standard growth curve of BMI in a
metabolic syndrome and its relationship to cognition have regression model to predict cognitive function (34). This
usually been examined in the middle aged or elderly. study was limited in that only 10% of participants were
In these studies obesity was defined primarily by BMI overweight. It is likely that the negative effects of adiposity
(kg m-2) of 30 or above in adults (unless otherwise stated), on cognition in children and adolescents are only detected
and above the 95th percentile for age and height in chil- over a threshold, i.e. only in the obese.
dren. High waist-to-hip ratio or high waist circumference Probably one of the best studies showing the clear asso-
was also used as a measure of adiposity in a few studies. ciation between elevated BMI and cognitive dysfunction in
© 2011 The Authors
obesity reviews © 2011 International Association for the Study of Obesity 12, 740–755
, 742 Obesity and cognitive function across lifespan
Table 1 Nine cross-sectional studies of the association between obesity and cognition in children and adolescents (4–19 years old)
Reference How was obesity Study population Controls Cognitive tests Covariates/exclusions Results
measured? and ascertainment
Azurmendi BMI as a continuum 60 boys and 69 girls Multiple regression Matrices, verbal abilities, None Negative relationship between BMI and verbal
et al. (2005) aged 5 years analyses affective labelling, display abilities, especially evident in girls
(32) – school children rules, false beliefs
Miller et al. >150% of ideal body 12 early onset 21 Normal-weight Woodcock–Johnson test of Sex Overall cognitive ability score was significantly lower
(2006) (30) weight morbidly obese (EMO siblings; 17 cognitive abilities and in the EMO compared with control siblings, but
individuals aged 4–22 Prader–Willi achievement, overall cognitive significantly higher compared with Prader–Willi
years [mean: 10.9 Syndrome ability score syndrome. EMO subjects also had lower thinking
E. Smith et al.
years]) ability, cognitive efficiency, phonemic awareness &
– clinic setting working memory (executive function) than controls.
obesity reviews © 2011 International Association for the Study of Obesity 12, 740–755
Cserje’si et al. BMI = 27.16 kg m-2 24 (12 obese) boys, Age matched Digit Span memory task, None Obese children performed worse on the Wisconsin
(2007) (27) mean age 12 years normal-weight Raven’s progressive matrices, sorting card test (executive function) and D2
– school children boys semantic verbal fluency, D2 attention endurance task despite similar intelligence
attention endurance test, and memory capacity to that of the control group
Wisconsin sorting card
Mond et al. BMI > 19.8 kg m-2 9,415 children aged Non-obese Standardized test to assess Age, nationality, year of An association between obesity and impairment in
(2007) (26) for boys and 4.4–8.6 years (4.3% of motor development, recruitment, location and gross motor skills was found in the entire sample
BMI > 19.7 kg m-2 for girls and 2.4% of boys development of speech, duration of kindergarten after controlling for covariates. When separated by
girls were obese) cognitive development, and gender, obese women were impaired in their ability
– population-based psychosocial development to focus attention continuously during the
examination (executive function) compared with
normal-weight girls.
Gunstad et al. Standard BMI growth 478 (45 overweight) Compared with Digit Span backward, Trial Age and education No associations were found between BMI and
(2008) (34) curves mean children/adolescents other weight Making Test B, Verbal recall; Exclusion: mental illness or cognitive function
BMI = 26.21 kg m-2 aged 12.46 years groups Animal fluency; finger tapping family history of mental
(range 6–19 years) illness, neurological injury and
obesity reviews
– population-based any significant medical
condition
© 2011 The Authors
