Published Ahead of Print on January 9, 2019 as 10.1212/WNL.

0000000000006879
ARTICLE

Association of body mass index and waist-to-hip

ratio with brain structure
UK Biobank study
Mark Hamer, PhD, and G. David Batty, DSc Correspondence

®
Dr. Hamer
Neurology 2019;92:e1-e7. doi:10.1212/WNL.0000000000006879 m.hamer@lboro.ac.uk

Abstract
Objective
To examine the association of body mass index (BMI) and waist-to-hip ratio (WHR) with brain
volume.

Methods
We used cross-sectional data from the UK Biobank study (n = 9,652, age 55.4 ± 7.5 years, 47.9%
men). Measures included BMI, WHR, and total fat mass as ascertained from bioimpedance.
Brain images were produced with structural MRI.

Results
After adjustment for a range of covariates, higher levels of all obesity measures were related to
lower gray matter volume: BMI per 1 SD (β coefficient −4,113, 95% confidence interval [CI]
−4,862 to −3,364), WHR (β coefficient −4,272, 95% CI −5,280 to −3,264), and fat mass
(β coefficient −4,590, 95% CI −5,386 to −3,793). The combination of overall obesity (BMI
≥30 kg/m2) and central obesity (WHR >0.85 for women, >0.90 for men) was associated with
the lowest gray matter compared with that in lean adults. In hypothesis-free testing with
a Bonferroni correction, obesity was also related to various regional brain volumes, including
caudate, putamen, pallidum, and nucleus accumbens. No associations between obesity and
white matter were apparent.

Conclusion
The combination of heightened BMI and WHR may be an important risk factor for gray matter
atrophy.

From School Sport (M.H.), Exercise & Health Sciences, Loughborough University; and Department of Epidemiology and Public Health (M.H., G.D.B.), University College London, UK.

Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

Copyright © 2019 American Academy of Neurology e1

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Glossary
BMI = body mass index; CI = confidence interval; WHR = wait-to-hip ratio.

In the absence of effective treatment modalities, the primary The aim of this study was to examine the joint associations of
prevention of neurodegenerative diseases, including de- BMI and WHR with brain structure using cross-sectional data
mentia, has gathered much research interest1–3 but remains from a large-scale population based imaging study of >9,000
poorly understood. Obesity was associated with lower rates of adults.
dementia in a large-scale study of >2 million adults,4 although
cohort studies with extended follow-up have shown null
findings5 or even the reverse gradient.6 That weight loss is Methods
common in the preclinical phase (up to a decade before di-
Participants
agnosis) of dementia6,7 may go some way to explaining these
Participants 40 to 69 years of age were recruited in 2006
apparently paradoxical findings.
to 2010 as part of the UK Biobank study and attended 1
of 22 clinical assessment centers in England, Wales, and
The mechanisms underlying the association between obesity
Scotland.18
and neurodegenerative diseases are not well known. Struc-
tural alterations in gray and white matter have been linked to Standard protocol approvals, registrations,
episodic memory decline and dementia risk.8 Various small- and patient consents
scale imaging studies9–16 have shown that higher levels of Ethics approval was provided by the National Health Service,
body mass index (BMI) and waist-to-hip ratio (WHR) are National Research Ethics Service (reference 11/NW/0382).
linked to lower gray matter volume. Although studies have Participants provided written informed consent.
examined BMI and WHR (as a marker of central obesity)
separately, there has been no investigation of the joint Obesity measures
effects, which may be important given the existing evidence Body weight and fat mass were collected with a Tanita
in relation to other disease outcomes.17 Some data have (Tokyo, Japan) BC418MA body composition analyzer using
suggested greater risk of cardiovascular disease in partic- bioimpedance.19 Nurses measured standing height using
ipants with both elevated BMI and WHR,17 which may be of a Seca (Hamburg, Germany) height measure with the head
relevance to neurodegenerative disorders given the apparent positioned in the Frankfort plane. BMI was calculated with
vascular origin. the standard formula (weight in kilograms divided by height

e2 Neurology | Volume 92, Number 6 | February 5, 2019 Neurology.org/N

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
in meters squared). To account for the fact that taller indi- (i.e., large weight loss may be a marker of disease onset). In
viduals tend to have more fat and lean mass, height-adjusted addition, we removed 1 participant with self-reported
indices were used to create a fat index score (created by diagnosis of dementia or cognitive impairment to avoid
dividing fat mass in kilograms by height in meters1.2).20 reverse causation. We modeled the associations between
Waist and hip circumferences were measured with a Seca various measures of obesity (per 1-SD increase) and brain
200 measuring tape using standard procedures. Participants structure (total gray and white volume) using multiple
were excluded from bioimpedance measures if they were linear regression. β Coefficients were adjusted for age, sex,
pregnant, using a pacemaker, wheelchair-bound, an ampu- smoking, alcohol consumption, physical activity, educa-
tee, unable to grip the handles of the Tanita analyzer, unable tion, heart disease, hypertension, and depression. We
to stand, wearing a plaster cast, or unwilling to remove their tested for nonlinearity by fitting a squared term for each
shoes. obesity measure. We selectively examined effect modifi-
cation by fitting interaction terms for age (3 categories)
Structural MRI and sex only. In doing so, the joint effects of overall (BMI ≥
In 2014, 100,000 participants from the original UK Biobank 30 kg/m2) and central (WHR >0.8 for women, >0.9 for
sample were invited back for brain, heart, and body imag- men) obesity on brain structure were examined by creating
ing.21 Approximately 10,000 were scanned between 2014 6 groups (reference category BMI <25 kg/m2 without
and 2016; processed data were used in the present analyses. central obesity, BMI <25 kg/m2 with central obesity,
Total gray and white volumes were measured with structural BMI 25–<30 kg/m2 without central obesity, BMI
MRI across 3 imaging centers that were equipped with 25–<30 kg/m2 with central obesity, BMI ≥30 kg/m2
identical scanners (Siemens Skyra 3T running VD13A SP4 without central obesity, BMI ≥30 kg/m2 with central
with a Siemens 32-channel RF receive head coil, Munich, obesity). Analyses were performed with SPSS version 22
Germany). The MRI protocols have been described in detail (SPSS Inc, Chicago, IL).
elsewhere.21 For each scan, Siemens auto-align software
determined the field-of-view, which aligns a scout scan to an Data availability statement
atlas. If auto-align failed, the radiographer set the alignment. All bona fide researchers can apply to use the UK Biobank
Structural images were acquired with straight sagittal ori- resource for health-related research that is in the public in-
entation (i.e., with the field of view aligned to the scanner terest (ukbiobank.ac.uk/register-apply/).
axes), with a resolution of 1 × 1 × 1 mm and a field of view of
208 × 256 × 256 matrix, over a duration of 5 minutes, and
with 1-mm isotropic resolution using a 3-dimensional Results
magnetization-prepared rapid-acquisition gradient echo.21
Information on structural image segmentation and data The flow of study participants into the analytic sample is
normalization is provided in detail elsewhere.22 Publicly depicted in figure 1. The analytical sample comprised 9,652
available image processing tools were used to process the healthy adults (age 55.3 ± 7.5 years, 47.9% men). Obesity was
data largely taken from FSL (the FMRIB Software Library). apparent in 18.7% of the sample. As anticipated, obese people
The output of the standard biobank processing pipeline were less likely to have a degree and to be physically active;
was used for the present analyses. Data were normalized for they also had a higher prevalence of heart disease and hy-
head size. pertension (table 1).

Covariates Higher levels of all obesity indicators were associated with

Covariates were chosen a priori on the basis of previous lower gray matter volume after adjustment for covariates
evidence.4–6 During the clinic visit, data were collected (table 2). No associations with white matter were observed at
via self-report for age, sex, smoking history, frequency of
alcohol intake (daily or almost daily, 1–2 times a week
or monthly, never or almost never), educational attain-
Figure 1 Flowchart describing sample selection
ment (college/degree, A-level, O-level, CSEs or equiva-
lent, National Vocational Qualifications/Higher National
Diploma or equivalent, other professional qualifica-
tion, none), physical activity, and self-reported physician-
diagnosed heart disease, hypertension, and major
depression.

Statistical analysis
In all our analyses, underweight participants (BMI <18.5
kg/m2, n = 41) were removed because this subsample was
too small to analyze in its own right and inclusion within
the healthy weight category may have introduced bias

Neurology.org/N Neurology | Volume 92, Number 6 | February 5, 2019 e3

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Table 1 Characteristics of sample according to obesity Table 2 β Coefficients for the association between
(n = 9,652) markers of adiposity and brain volume
(n = 9,652)
BMI category
White matter
Normal Overweight
Gray matter volume volume
(18.5 ≤ (25–29.99 Obese
25 kg/m2) kg/m2) (≥30 kg/m2) β (95% CI) β (95% CI)

People, n 3,680 4,167 1805 BMI −4,113 (−4,862 to −3,364) a

−177 (−985 to 630)

Age at examination 54.7 (7.5) 56.0 (7.5) 55.0 (7.3) WHR −4,272 (−5,280 to −3,264) a
294 (−790 to 1,379)
(mean, SD), y
Fat indexb −4,590 (−5,386 to −3,793)a −367 (−1,226 to 493)
Sex, %

Women 63.7 42.2 49.9 Abbreviations: BMI = body mass index; WHR = waist-to-hip ratio.
Coefficients reflect a 1-SD increase in obesity marker and are adjusted for
age, sex, smoking, vigorous physical activity, alcohol, education, major de-
Men 36.6 57.6 50.1
pression, heart disease, and hypertension.
a
p < 0.001.
Educational attainment, 48.4 42.4 38.1 b
Calculated from impedance data; total fat mass (kg)/height (m).
% degree/college

Smoking, % current 6.4 7.0 6.1

Alcohol intake, % daily 22.9 23.6 19.2 Second, data were available on “fluid intelligence” (a task
Frequency vigorous with 13 logic/reasoning-type questions and a 2-minute time
physical activity, % limit23) in a subsample of participants (n = 3,477). We
None 29.0 32.7 43.2
conducted a sensitivity analysis on obesity and gray matter
volume with additional adjustment for fluid intelligence
1 or 2 times per week 34.1 33.3 29.6
score in this subsample. Again, results remained largely un-
≥3 times per week 36.9 33.9 27.2 changed: 1-SD unit increase in BMI (β = −3,292, 95%
CI −4,560 to −2025), WHR (β = −3,661, 95% CI −5,322
Heart disease, % 1.9 3.6 4.8
to −2000), and fat index (β = −3,448; 95% CI −4,790 to
Hypertension, % 11.7 21.6 31.7 −2,107).
Diabetes mellitus, % 1.1 2.6 7.2
Third, we examined the combined influences of BMI
Major depression, % 2.6 2.5 2.8
and WHR on brain volume. Within normal and overweight
Whole-brain gray 804,047 ± 789,089 ± 787,577 ± BMI categories, there were no differences in gray matter
matter, mm3 47,524 46,979 49,573
volume between participants with and those without central
Whole-brain white 711,042 ± 711,459 ± 710,707 ± obesity. Within obese participants (BMI ≥30 kg/m2) with
matter, mm3 41,074 41,327 39,831
central obesity (present in 72%), however, there was evi-
Abbreviations: BMI = body mass index.
dence of lower gray matter volume compared with those
who were not categorized as centrally obese (β = −4,496,
95% CI, −8,820 to −172, p = 0.04) (figure 3). These dif-
ferences were marginally attenuated with further adjustment
conventional levels of statistical significance. The linear nature for total body fat percentage (β = −3,907, 95% CI, −8,246 to
of the BMI–gray matter association was similar across age 432, p = 0.078).
categories (figure 2) (p for interaction = 0.50) and sex (p for
interaction = 0.23). Lastly, an exploratory analysis was conducted to examine
associations between obesity and 7 specific brain region
Next, we conducted a series of sensitivity analyses. First, volumes using partial correlations. These tests were hy-
diabetes mellitus is likely to be on the intermediate pathway pothesis free, and, owing to their frequency, we applied
between obesity and brain atrophy. To test its role, we ad- a Bonferroni correction. Obesity was associated with a lower
ditionally controlled for a diagnosis of diabetes mellitus volume of the caudate (only WHR), putamen (only BMI
(sample prevalence 2.9%). Participants with diabetes melli- and total fat mass), pallidum, and nucleus accumbens
tus demonstrated lower gray matter volume (β = −14,200, regions (p < 0.001) (table 3).
95% confidence interval [CI] −18,595 to −9,804) compared
to those without the condition. When we added diabetes
mellitus to the models featured in table 2, the associations
between BMI and gray matter (β = −3,847, 95% CI −4,600 to
Discussion
−3,093) and WHR and gray matter (β = −3,942, 95% CI The aim of the present paper was to examine associations
−4,954 to −2,930) were, in fact, only partially attenu- between obesity and brain volumes, taking advantage of
ated (≈6%–8%). a considerably larger study population than in previous work.

e4 Neurology | Volume 92, Number 6 | February 5, 2019 Neurology.org/N

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Figure 2 Association between body mass index and gray matter volume relative to age

Data are presented as means adjusted for age, sex,

smoking, vigorous physical activity, alcohol, edu-
cation, major depression, heart disease, and
hypertension.

Our main finding was that people with obesity, as ascertained disease.26 Epidemiological studies of mortality risk have
with BMI and WHR, had a lower gray matter volume. BMI tested whether WHR can provide additional predictive
and WHR appeared to have additive effects only in obese utility over and above BMI, although results have been
(BMI ≥30 kg/m2) participants, although associations mixed.17,25,27,28 The present data suggest that the combi-
appeared to be driven partly by total body fat percentage. For nation of high BMI and high WHR is associated with
the first time, we also found apparent associations of obesity greater gray matter atrophy. Visceral fat is thought to be
with specific brain regions, relationships that need detailed a major site for inflammatory cytokine production and has
replication with other datasets. been linked to other vascular risk factors (hypertension,
diabetes mellitus)29 that may be important mechanisms in
The association between obesity and health outcomes has brain atrophy.13,30,31 Associations between obesity and gray
been controversial,24,25 and this might be partly explained matter volume were only partly explained by diabetes
by specific health effects of different fat depots. BMI is mellitus in the present study. In contrast, subcutaneous fat
thought to be more reflective of fat stored peripherally, in the hips and legs has been linked to healthier metabolic
whereas WHR is an indicator of fat located viscerally and profiles,32 which may provide partial support for the con-
potentially considered a greater risk factor for heart cept of metabolically healthy obesity. Indeed, our data
suggested that obese participants (BMI ≥30 kg/m2) with-
out central obesity had a gray matter volume similar to that
of overweight participants.
Figure 3 Body mass index and waist-to-hip in relation to
gray matter brain volume Hippocampal atrophy is thought to be particularly relevant
to the etiology of neurodegenerative diseases such as Alz-
heimer disease,33 although we did not observe consistent
associations with obesity. Previous work has hypothesized
obesity–gray matter associations specifically in areas in-
volved in behavioral control, reward processing (e.g., the
prefrontal cortex in the frontal lobe or striatum with caudate
nucleus, globus pallidus, and putamen), homeostasis (hy-
pothalamus), and motor control (cerebellum and gyrus
precentralis). These areas could conceivably be linked to
obesogenic behavior such as appetite and satiety regula-
tion.15 The present results confirm earlier work15 by dem-
onstrating associations between obesity and smaller volumes
in some of these specific areas of the brain (i.e., caudate,
putamen). Structural brain abnormalities that disrupt ap-
Data are presented as means adjusted for age, sex, smoking, vigorous petite regulation/reward could precede the development of
physical activity, alcohol, education, major depression, heart disease, and
hypertension. obesity. Although brain imaging data were collected after
measures of obesity, these were essentially cross-sectional

Neurology.org/N Neurology | Volume 92, Number 6 | February 5, 2019 e5

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Table 3 Partial correlations between measures of adiposity and specific brain volumes
Thalamus Caudate Putamen Pallidum Hippocampus Amygdala Nucleus accumbens

BMI −0.015 −0.024 −0.04a −0.04a 0.015 0.016 −0.035a

p Value 0.127 0.018 <0.001 <0.001 0.14 0.12 <0.001

a a a
WHR −0.036 −0.047 −0.025 −0.050 −0.024 −0.011 −0.031

p Value <0.001 <0.001 0.014 <0.001 0.017 0.28 0.002

Fat indexb −0.012 −0.023 −0.04a −0.05a 0.003 0.012 −0.037a

p Value 0.25 0.022 <0.001 <0.001 0.75 0.26 <0.001

Abbreviations: BMI = body mass index; WHR = waist-to-hip ratio.

Correlations adjusted for age, sex, smoking, vigorous physical activity, alcohol, education, major depression, heart disease, and hypertension.
a
Statistically significant after Bonferroni correction.
b
Calculated from impedance data; total fat mass (kg)/height (m)1.2.

analyses; therefore, we are unable to speculate on the di-

rection of the association. Appendix 1 Author contributions

Name Location Role Contribution

Our study of course has its limitations. Chronic disease was
based on self-report of physician diagnosis, although previous Mark Loughborough Author Design and conceptualized
work has demonstrated the validity of this approach.34 As in Hamer, PhD University, UK study; analyzed the data;
drafted the manuscript for
any observational study, the possibility of residual con- intellectual content
founding cannot be excluded. Only ≈5% of the target pop-
G. David University Author Drafted the manuscript for
ulation took part in UK Biobank,35 and study members Batty, DSc College intellectual content
typically showed more favorable risk profiles than the non- London, UK

responders. Thus, the issue of selection is likely to be much

more serious in the present study. Highly select groups of the
References
population, however, tend to reveal the same risk factor– 1. Beydoun MA, Beydoun HA, Wang Y. Obesity and central obesity as risk factors for
disease associations as those seen in the general population.36 incident dementia and its subtypes: a systematic review and meta-analysis. Obes Rev
2008;9:204–218.
On balance, therefore, while the prevalence of selected char- 2. Anstey KJ, Cherbuin N, Budge M, Young J. Body mass index in midlife and late-life as
acteristics will differ in the UK Biobank, risk factor–health a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev 2011;12:
e426–e437.
outcome associations should not. 3. Emmerzaal TL, Kiliaan AJ, Gustafson DR. 2003-2013: A decade of body mass index,
Alzheimer’s disease, and dementia. J Alzheimers Dis 2015;43:739–755.
4. Qizilbash N, Gregson J, Johnson ME, et al. BMI and risk of dementia in two million
Obesity was associated with lower gray matter brain volumes. people over two decades: a retrospective cohort study. Lancet Diabetes Endocrinol
It is unclear whether structural brain abnormalities drive 2015;3:431–436.
5. Batty GD, Galobardes B, Starr JM, Jeffreys M, Davey Smith G, Russ TC. Examining if
obesity or whether obesity induces changes in gray matter being overweight really confers protection against dementia: sixty-four year follow-up
volume that play a mechanistic role in future risk of of participants in the Glasgow University Alumni Cohort Study. J Negat Results
Biomed 2016;15:19.
neurodegeneration. 6. Singh-Manoux A, Dugravot A, Shipley M, et al. Obesity trajectories and risk of
dementia: 28 years of follow-up in the Whitehall II Study. Alzheimers Dement 2018;
Study funding 14:178–186.
7. Knopman DS, Edland SD, Cha RH, Petersen RC, Rocca WA. Incident dementia in
This research has been conducted using the UK Biobank women is preceded by weight loss by at least a decade. Neurology 2007;69:739–746.
Resource under application 23,476. M.H. acknowledges 8. Nyberg L. Functional brain imaging of episodic memory decline in ageing. J Intern
Med 2017;281:65–74.
support from the National Institute for Health Research 9. Pannacciulli N, Del Parigi A, Chen K, Le DS, Reiman EM, Tataranni PA. Brain
Leicester Biomedical Research Centre. G.D.B. is supported abnormalities in human obesity: a voxel-based morphometric study. Neuroimage
2006;31:1419–1425.
by the UK Medical Research Council (MR/P023444/1) and 10. Raji CA, Ho AJ, Parikshak NN. Brain structure and obesity. Hum Brain Mapp 2010;
the US National Institute on Aging (1R56AG052519-01, 31:353–364.
11. Yokum S, Ng J, Stice E. Relation of regional gray and white matter volumes to current
1R01AG052519-01A1). BMI and future increases in BMI: a prospective MRI study. Int J Obes (Lond) 2012;
36:656–664.
12. Debette S, Beiser A, Hoffmann U, et al. Visceral fat is associated with lower brain
Disclosure volume in healthy middle-aged adults. Ann Neurol 2010;68:136–144.
The authors report no disclosures relevant to the manuscript. 13. Debette S, Seshadri S, Beiser A, et al. Midlife vascular risk factor exposure accelerates
structural brain aging and cognitive decline. Neurology 2011;77:461–468.
Go to Neurology.org/N for full disclosures. 14. Kurth F, Levitt JG, Phillips OR, et al. Relationships between gray matter, body mass
index, and waist circumference in healthy adults. Hum Brain Mapp 2013;34:
Publication history 1737–1746.
15. Janowitz D, Wittfeld K, Terock J, et al. Association between waist circumference and
Received by Neurology May 23, 2018. Accepted in final form December gray matter volume in 2344 individuals from two adult community-based samples.
6, 2018. Neuroimage 2015;122:149–157.

e6 Neurology | Volume 92, Number 6 | February 5, 2019 Neurology.org/N

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
16. Taki Y, Kinomura S, Sato K, et al. Relationship between body mass index and gray 26. Emdin CA, Khera AV, Natarajan P, et al. Genetic association of waist-to-hip ratio with
matter volume in 1,428 healthy individuals. Obesity 2008;16:119–124. cardiometabolic traits, type 2 diabetes, and coronary heart disease. JAMA 2017;317:
17. Hamer M, O’Donovan G, Stensel D, Stamatakis E. Normal-weight central obesity and 626–634.
risk for mortality. Ann Intern Med 2017;166:917–918. 27. Wormser D, Wormser D, Kaptoge S, et al. Separate and combined associations of
18. Sudlow C, Gallacher J, Allen N, et al. UK Biobank: an open access resource for body-mass index and abdominal adiposity with cardiovascular disease: collaborative
identifying the causes of a wide range of complex diseases of middle and old age. PLoS analysis of 58 prospective studies. Lancet 2011;377:1085–1095.
Med 2015;12:e1001779. 28. Czernichow S, Kengne AP, Stamatakis E, Hamer M, Batty GD. Body mass index, waist
19. Ortega LS, Bradbury KE, Cross AJ, Morris JS, Gunter MJ, Murphy N. A prospective circumference and waist-hip ratio: which is the better discriminator of cardiovascular
investigation of body size, body fat composition and colorectal cancer risk in the UK disease mortality risk? Evidence from an individual-participant meta-analysis of 82 864
Biobank. Sci Rep 2017;7:17807. participants from nine cohort studies. Obes Rev 2011;12:680–687.
20. Bann D, Kuh D, Wills AK, Adams J, Brage S, Cooper R. Physical activity across 29. Despres JP. Intra-abdominal obesity: an untreated risk factor for type 2 diabetes and
adulthood in relation to fat and lean body mass in early old age: findings from the cardiovascular disease. J Endocrinol Invest 2006;29(suppl):77–82.
Medical Research Council National Survey of Health and Development, 1946-2010. 30. Korf ES, van Straaten EC, de Leeuw FE, et al. Diabetes mellitus, hypertension
Am J Epidemiol 2014;179:1197–1207. and medial temporal lobe atrophy: the LADIS study. Diabet Med 2007;24:
21. Miller KL, Alfaro-Almagro F, Bangerter NK, et al. Multimodal population brain 166–171.
imaging in the UK Biobank prospective epidemiological study. Nat Neurosci 2016;19: 31. den Heijer T, Vermeer SE, van Dijk EJ, et al. Type 2 diabetes and atrophy of medial
1523–1536. temporal lobe structures on brain MRI. Diabetologia 2003;46:1604–1610.
22. Alfaro-Almagro F, Jenkinson M, Bangerter NK, et al. Image processing and quality 32. Manolopoulos KN, Karpe F, Frayn KN. Gluteofemoral body fat as a determinant of
control for the first 10,000 brain imaging datasets from UK Biobank. Neuroimage metabolic health. Int J Obes (Lond) 2010;34:949–959.
2018;166:400–424. 33. Karas GB, Burton EJ, Rombouts SA, et al. A comprehensive study of gray matter loss
23. Lyall DM, Cullen B, Allerhand M, et al. Cognitive test scores in UK Biobank: data in patients with Alzheimer’s disease using optimized voxel-based morphometry.
reduction in 480,416 participants and longitudinal stability in 20,346 participants. Neuroimage 2003;18:895–907.
PLoS One 2016;11:e0154222. 34. Pierce MB, Zaninotto P, Steel N, Mindell J. Undiagnosed diabetes: data from the
24. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality with English Longitudinal Study of Ageing. Diabet Med 2009;26:679–685.
overweight and obesity using standard body mass index categories: a systematic re- 35. Fry A, Littlejohns TJ, Sudlow C, et al. Comparison of sociodemographic and health-
view and meta-analysis. JAMA 2013;309:71–82. related characteristics of UK Biobank participants with those of the general pop-
25. Sahakyan KR, Somers VK, Rodriguez-Escudero JP, et al. Normal-weight central ulation. Am J Epidemiol 2017;186:1026–1034.
obesity: implications for total and cardiovascular mortality. Ann Intern Med 2015; 36. Batty GD, Shipley M, Tabák A, et al. Generalizability of occupational cohort study
163:827–835. findings. Epidemiology 2014;25:932–933.

Neurology.org/N Neurology | Volume 92, Number 6 | February 5, 2019 e7

Copyright ª 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Association of body mass index and waist-to-hip ratio with brain structure: UK
Biobank study
Mark Hamer and G. David Batty
Neurology published online January 9, 2019
DOI 10.1212/WNL.0000000000006879

This information is current as of January 9, 2019

Updated Information & including high resolution figures, can be found at:
Services http://n.neurology.org/content/early/2019/01/09/WNL.0000000000006
879.full
Permissions & Licensing Information about reproducing this article in parts (figures,tables) or in
its entirety can be found online at:
http://www.neurology.org/about/about_the_journal#permissions
Reprints Information about ordering reprints can be found online:
http://n.neurology.org/subscribers/advertise

Neurology ® is the official journal of the American Academy of Neurology. Published continuously since
1951, it is now a weekly with 48 issues per year. Copyright © 2019 American Academy of Neurology. All
rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.