Body Mass Index Does Not Affect Grooved Pegboard Performance in Healthy South African Adults

doi:10.4236/psych.2013.41004

Paper Menu >>

Journal Menu >>

Psychology

2013. Vol.4, No.1, 33-37

Published Online January 2013 in SciRes (http://www.scirp.org/journal/psych) http://dx.doi.org/10.4236/psych.2013.41004

Body Mass Index Does Not Affect Grooved Pegboard

Performance in Healthy South African Adults

Charles H. van Wijk

Private Practice, Simon’s Town, South Africa

Email: [email protected]

Received October 26th, 2012; revised November 25th, 2012; accepted December 22nd, 2012

Obesity has been associated with poorer performance on the Grooved Pegboard (GP) among healthy older

adults. The GP is widely used in South Africa, among others for the assessment of HIV Associated Neu-

rocognitive Disorders. Obesity is growing among the younger adult population in South Africa, which is

the group also most at risk for HIV. It is not clear what the interaction between body mass and GP per-

formance would be among a group of healthy younger adults. This study investigated whether body mass

might affect fine psychomotor skills. A sample of 850 healthy adults (20 - 49 years) completed the GP

and had their Body Mass Index (BMI) calculated. The relationship between GP and BMI was examined

using ANOVA and correlation coefficients. The expected gender differences in GP performance found

elsewhere were demonstrated in this sample. No significant interactions between BMI categories and GP

times were found, and no significant correlations between BMI continuous scores and GP times were

found either. In spite of the presence of a wide weight spectrum among the participants and the absence of

any history of known medical disease, the lack of significant BMI-GP interactions suggest that the effect

of BMI may generally be discounted when interpreting GP results.

Keywords: Body Mass Index; BMI; Gender Differences; Grooved Pegboard; Psychomotor Performance;

HIV

Introduction

The Grooved Pegboard (GP) is a manipulative dexterity task

that can be used for neuropsychological assessment (Lezak,

Howieson, & Loring, 2004). The HIV epidemic has stimulated

renewed interest in the GP because of its potential to identify

HIV associated neurocognitive disorders (HAND). It is a com-

mon neuropsychological instrument used in HAND (Grant,

2008), which can differentiate between the HIV statuses of

asymptomatic patients in Sub-Saharan Africa (Moshani, 2009;

Sacktor et al., 2005). In particular the GP non-dominant hand

score is sensitive for HIV-associated neuropsychological im-

pairment (Carey et al., 2004; Davis, Skolasky, Selnes, Burgess,

& McArthur, 2002), and the GP non-dominant hand test has

been established to detect signs of HIV dementia (cf. Sacktor et

al., 2005). Completion time has been related to stage of HIV

disease (Heaton, Grant, Butters, White, Kirson, & Atkinson,

1995), while declining times have been linked to future pro-

gression to HIV dementia (Selnes, Galai, McArthur, & Cohen,

1997). The GP is also sensitive to improvement in neuropsy-

chological performance in HIV+ individuals receiving HAART

(Joska, Gouse, Paul, Stein, & Flisher, 2010).

A large percentage of asymptomatic HIV infected persons

show mild neurocognitive difficulties (Heaton et al., 1995), and

recent South African (SA) figures indicated that 17% - 23.5%

of HIV patients display cognitive impairment (Ganasen, Fin-

cham, Smit, Seedat, & Stein, 2008; Joska et al., 2010). HIV

prevalence for SA adults (15 - 49) was estimated at 16.9% in

2008. SA is home to the world’s largest population of people

living with HIV (5.7 million) (UNAIDS, 2009).

To meaningfully use the GP non-dominant hand test to

screen for HAND, performance of HIV individuals need to be

compared to age and education adjusted peer means (Sactor et

al., 2005). There are many factors that could influence GP per-

formance in healthy adults. Obesity is one example of this, and

is a growing concern for SA healthcare. SA studies found that

between 24% and 27% of women are overweight, with a further

30% to 54% obese, while figures for men indicate that between

15% and 22% are overweight, with a further 7% to 19% obese

(Malhotra et al., 2008; Puoane et al., 2002). Recent studies

indicate that obesity seems to start at an increasingly young age,

with about 10% of women obese by the age of 24 years

(Puoane et al., 2002, Reddy, Panday, Swart, Jinabhai, Amosun,

& James, 2003).

Body Mass Index (BMI) is calculated from a person’s weight

and height, and is an effective method for population assess-

ment of overweight and obesity (Centers for Disease Control

and Prevention, 2010). Its main importance lies in the relation-

ship between body weight and disease and death (World Health

Organisation, 1995), with overweight and obese individuals at

increased risk for many diseases and health conditions (Na-

tional Institutes of Health, 1998). The negative health cones-

quences associated with increased BMI in SA have been well

described (Joubert, Norman, Bradshaw, Goedecke, Steyn, &

Puoane, 2007). BMI serves two functions, firstly, to indicate

risk for various diseases, and secondly as a general indication

of body fatness.

Obesity has been associated with poorer cognitive function in

several studies (Elias, Elias, Sullivan, Wolf, & D’Agostino,

2003, 2005; Kilander, Nyman, Boberg, & Lithell, 1997; Wald-

stein & Katzel, 2006). Obesity is associated with lower memory

and executive function in late middle-aged and elderly men (but

C. H. VAN WIJK

not women), independent of other common cardiovascular dis-

ease (CVD) risk factors (Elias et al., 2003, 2005), and further

related to diminished performance on tests of motor speed and

manual dexterity, independent of other components of the

metabolic syndrome and several other potential confounding

factors, among healthy older adults (Waldstein & Katzel, 2006).

Interactions between BMI and Grooved Pegboard (GP) scores

in older adults were significant, with individuals with greater

BMI performing most poorly on the GP (Waldstein & Katzel,

2006).

The mechanism for this is not well understood. Waldstein &

Katzel (2006) hypothesised that central obesity has been asso-

ciated with various neuroendocrine disturbances (Bjorntorp &

Rosmond, 2000) that have also been associated with enhanced

sympathetic nervous system activity (Bjorntorp & Rosmond,

2000; Ren, 2004) and that may promote structural brain abnor-

malities including silent cerebrovascular disease and stroke

(Everson, Lynch, Kaplan, Lakka, Sivenius, & Salonen, 2001;

Waldstein, Siegel, Lefkowitz, Maier, Pelletier-Brown, & Obu-

chowski, 2004). Obesity has also been associated with en-

hanced pro-inflammatory factors (Toni, Malaguti, Castorina,

Roti, & Lechan, 2004), which have been shown to exert nega-

tive effects on cognitive function (Yaffe, Lindquist, Penninx,

Simonsick, Pahor, & Kritchevsky, 2003).

With the interaction of BMI and neurocognitive performance

in healthy older adults established, it is not clear whether the

same patterns will hold for healthy younger adults. It could be

hypothesised that shorter exposure to the neurotoxic effects of

CVD factors, hormonal abnormalities, or inflammatory factors

may leave younger brains less affected by the disease related

aspects of obesity. However, in an environment of reduced

neurotoxic risk (i.e. youth), BMI might still influence GP per-

formance through, for example, clumsy fingers, and thus con-

found the results of this psychomotor task.

In SA the effect of BMI on GP performance is of interest

because of its potential to identify HAND. The demographic

curve of HIV is skewed towards younger people, creating a

younger risk group for HAND, which is the same demographic

group increasingly at risk for obesity. While obesity is not usu-

ally associated with HIV positive status, it is widespread within

the general population. Thus, in order to use the GP non-

dominant hand test with HIV positive persons, the effect of

BMI on GP performance among healthy individuals needs to be

clarified first, in order to determine the usefulness of peer

norms. If BMI does affect GP performance, accurate peer

norms for HIV positive persons may need to control for the

confounding effects of BMI.

This study thus set out to explore whether excessive body

mass might affect fine motor skills, in particular GP perform-

ance, among healthy younger adults. This could potentially

have implications for the interpretation of neuropsychological

scores, in particular for the screening of HIV associated neuro-

cognitive decline.

Methods

Sample

This study used a convenience sample, recruiting participants

through an occupational health surveillance program. This al-

lowed for the measurement of BMI without inconveniencing

participants. Individuals were excluded from the study if they

had a history of neurological, psychiatric or cardio-vascular

disorders, were HIV positive, had any physical impediments

that could affect motor performance, or were not adequately

proficient in English to understand the GP instructions or an-

swer the health questionnaire. Due to the recruitment channel,

all participants were employed at the time.

As the general HIV prevalence figures are available for the

ages up to 49, the same maximum age limit was used for this

study. In SA, both BMI profiles (Malhotra et al., 2008) and GP

performance (Van Wijk, 2012) differs across gender, and

women completed the GP faster than men. Further, the associa-

tion of BMI and GP performance also differed along gender

lines among older people, and it was thus decided to treat the

data for women and men separately.

Eight hundred and fifty volunteers between the ages of 20

and 49 years (mean age = 32 ± 8) completed the GP on an indi-

vidual basis. All participants had at least 8 years of formal edu-

cation (mean years = 12 ± 1). The sample comprised 307

women (36.1%) and 543 men (63.9%), and was drawn from all

SA language groups and provinces of origin.

The female group had a mean age of 28 years (±7), and a

mean of 12 years (±1) of education. Of the women, 67% were

between 19 and 29 years, 27% were between 30 and 39 years,

and 6% were between 40 and 49 years, while 61% were Black,

20% were Coloured, 3% were Indian, and 16% were White.

The male group had a mean age of 35 years (±8), and a mean of

12 (±1) years of education. Of the men, 33% were between 19

and 29 years, 36% were between 30 and 39 years, and 31%

were between 40 and 49 years, while 41% were Black, 33%

were Coloured, 5% were Indian, and 21% were White. The

women were significantly younger than the men (p < 0.01).

Instruments

1) Grooved Pegboard. The GP is a manipulative dexterity

test (Lafayette Instrument Company, 1989), which measures

psychomotor speed, fine motor control, and rapid visual-motor

coordination (Mitrushina, Boone, Razan, & D’Elia, 2005).

Performance is highly dependent on psychomotor speed (Lezak

et al., 2004). Scores represe nt time in seconds required to com-

plete the matrix with each hand, with higher scores reflecting a

lower level of performance. The non-dominant hand is often

considered more sensitive for psychomotor slowing. The task

was administered according to the standard instruction set, as

described in the manual (Lafayette Instrument Company, 1989).

One trial each was allowed with first the dominant hand (GPd),

and then the non-dominant hand (GPn).

2) Anthropometric measurement. Participants were measured

while wearing light clothes without shoes or jackets. Measure-

ments were done on a Secca scale, and took place under the

supervision of a dietician. The scale’s automatic BMI calcula-

tion feature was used; height had to be entered manually, and

was rounded to the nearest centimetre for this purpose. BMI

was computed as weight (in kilograms) divided by the square of

the height (in meters). The following WHO (1995) categories

were used: underweight (BMI < 18.5), healthy weight (BMI

18.5 to 24.9), overweight (BMI 25.0 to 29.9), and obese (BMI

> 30).

3) Health questionnaire. Participants completed a self-report

health questionnaire, developed for this study, inquiring about

neurological or psychiatric history, and history of cardiovascu-

lar disease.

C. H. VAN WIJK

Procedure

All participants completed an informed consent form. They

did their BMI measurement as part of their health surveillance

program, and did the GP individually, usually at the end of their

health screening. The study was approved by the Surgeon Gen-

eral’s Health Research Ethics Committee.

Data Analysis

Some previous studies used dichotomous variables to index

the presence of obesity (Elias et al., 2003, 2005), while others

used continuous measures of obesity relating to cognition

(Waldstein & Katzel, 2006). This present study will employ

both. Firstly, the relationship between the WHO’s BMI catego-

ries and GP timed scores will be calculated with ANOVA. Post

hoc analysis will use Tukey HSD tests. The calculations will be

done separately for each gender group. Secondly, to explore

further the ability of BMI scores to predict GP performance,

correlation coefficients will be calculated for BMI raw scores

and GP timed scores, again separately for each gender group.

As there was no significant time-score difference between race

groups, the results of the total group will be used for analysis

here. All analyses were done using STATISTICA 7.

Results

BMI profile for the sample: None of the participants fell into

the underweight category. In the female sample, 37% were of

healthy weight, 34% were overweight, and 29% were obese. In

the male sample, 30% were of healthy weight, 38% were over-

weight, and 32% were obese. The gender difference may be the

result of the gendered age profiles, as there were more older

men and more younger women in the group.

GP performance of the sample: Women completed the GPd

in a mean time of 60.9 seconds (±8.0), while the men did it in a

mean time of 65.8 seconds (±9.2). The difference (using t-tests

for independent samples) was significant (p < 0.001). Women

completed the GPn in a mean time of 66.3 seconds (±9.4),

while the men did it in a mean time of 70.8 seconds (±11.2).

The difference was again significant (p < 0.001).

When using ANOVA, women did not display any significant

differences between the BMI categories and GPd time scores

(F2,304 = 1.13; p = 0.3) or the GPn time scores (F2,304 = 0.95; p =

0.4). Men did not display any significant differences between

the BMI categories and GPd time scores (F2,540 = 1.3; p = 0.3)

and GPn time scores (F2,540 = 2.1; p = 0.1) either.

A small stepwise progression of obesity and performance

was observed in both the women and men’s groups, but never

achieved significance. In general, participants who were obese

posted slightly longer times than those who were overweight,

who in turn posted slightly longer times than those in the

healthy weight category (see Table 1).

When correlation coefficients were calculated for BMI raw

scores and GP time scores, no significant correlations were

found for both GPd and GPn for either gender or the total group

(see Table 2).

Discussion

This study demonstrated the expected gender difference in

GP performance reported elsewhere (Bryden & Roy, 2005;

Schmidt, Oliveira, Rocha, & Abreu-Villaca, 2000). The results

Table 1.

Means and standard deviations o f GP times across weight categories.

Women

N GPd GPn

mean SD mean SD

Healthy

weight 113 60.0 7.5 65.3 9.0

Over weight 106 61.1 8.5 66.8 9.0

Obese 88 61.8 8.0 66.9 10.8

Men

N GPd GPn

mean SD mean SD

Healthy

weight 162 65.0 8.8 69.4 10.3

Over weight 209 65.8 9.4 71.3 10.9

Obese 172 66.6 9.4 71.7 12.4

Table 2.

Correlation coefficients between BMI and GP scores.

N GPd GPn

Total group 850 0.04 0.05

Women 307 0.00 0.04

Men 543 0.04 0.04

further found no significant interaction between BMI categories

and GP times, nor significant interactions between BMI scores

and GP times.

This stands in apparent contrast to earlier studies that found

obesity associated with poorer performance on the GP (Wald-

stein & Katzel, 2006), among older adults.

The findings differed from previous studies for a number of

reasons: Firstly, the age distribution in the female group was

skewed towards younger women. As age is associated with

weigh (Puoane et al., 2002), the weight distribution was thus

also skewed towards lower body mass. Age is also associated

with GP performance (Heaton, Ryan, Grant, & Matthews,

1996), and the total skewed distribution could have influenced

possible effects of BMI on GP performance. Secondly, as the

sample was comprised of younger adults, the results may sug-

gest that lesser exposure to the neurotoxic effects of CVD,

neuroendocrine disorders, and so forth might indeed lessen the

effect of such factors on psychomotor performance. This would

give support to current hypotheses regarding the mechanism of

the obesity/cognitive performance interaction (Elias et al., 2003;

Waldstein & Katzel, 2006).

This study has a number of limitations. Most notably, the

women sample did not reflect the normal population distribu-

tion in either age or BMI categories. Further to this, there was

no objective control (e.g. an examination by a physician) of

possible CVD history or risk. As unknown or unreported CVD

risk factors could have been present, the results need to be in-

terpreted with caution. Lastly, the difference in composition

(age and BMI categories) of the gender sub-samples precludes

C. H. VAN WIJK

direct comparisons of women and men’s results.

Future studies need to include participants that are more re-

flective of the general population distribution (with regard to

age, gender, employment status, education, and so forth), as

well as external (vs self-reported) controls for medical and

lifestyle risk factors (e.g. hypertension, diabetes, exercise and

nutrition).

To conclude: in the presence of a wide weight spectrum

among the participants, and the absence of history of known

medical risk factors, the lack of significant BMI-GP interaction

suggests that there is no real evidence of body mass signifi-

cantly affecting GP performance among younger adults.

Slowed performance on the GP (in the presence of elevated

body mass) would thus probably be due to neurological disease

processes, rather than body fat percentage.

Thus, the effect of BMI may generally be discounted when

interpreting GP results until further corroboration of the inter-

action has been reported.

Acknowledgements

I wish to acknowledge Chesray Hans-Arendse, Sadia Edross,

Marilize Willers and Ruby Muller who assisted in the admini-

stration of the GP, and Aileen van der Spuy who did the BMI

measurements.

REFERENCES

Bjorntorp, P., & Rosmond, R. (2000). Neuroendocrine abnormalities in

visceral obesity. International Journal of Obesity Related Metabolic

Disorders, 24, S80-S85. doi:10.1038/sj.ijo.0801285

Bryden, P. J., & Roy, E. A. (2005). A new method of administering the

grooved pegboard test: Performance as a function of handedness and

sex. Brain and Cognition, 58, 258-268.

doi:10.1016/j.bandc.2004.12.004

Carey, C. L., Woo ds, S. P., Rippeth, J. D., Gonzalez, R ., Moore, D. J.,

Marcotte, T. D. et al. (2004). Initial validation of a screening battery

for the detection of HIV-associated cognitive impairment. Clinical

Neuropsychology, 18, 234-248. doi:10.1080/13854040490501448

Centers for Disease Control and Prevention (2010). About BMI for

adults. URL (last checked 12 October 20 12 ).

http://www.cdc.gov/healthyweight/assessing/bmi/adult_BMI/index.h

tml

Davis, H. F., Skolasky, R. L., Selnes, O. A., Burgess, D. M., &

McArthur, J. C. (2002). Assessing HIV-associated dementia: Modi-

fied HIV dementia scale versus the grooved pegboard. The AIDS

Reader, 12, 29-38.

Elias, M. F., Elias, P. K., Sullivan, L. M., Wolf, P. A., & D’Agostino, R.

B. (2003). Lower cognitive function in the presence of obesity and

hypertension: The Framingham heart study. International Journal of

Obesity, 27, 260-268. doi:10.1038/sj.ijo.802225

Elias, M. F., Elias, P. K., Sullivan, L. M., Wolf, P. A., & D’Agostino, R.

B. (2005). Obesity, diabetes and cognitive deficit: The Framingham

heart study. Neurobiology of Aging, 26, S11-S16.

doi:10.1016/j.neurobiolaging.2005.08.019

Everson, S. A., Lynch, J. W., Kaplan, G. A., Lakka, T. A., Sivenius, J.,

& Salonen, J. (2001). Stress-induced blood pressure reactivity and

incident stroke in middle-aged men. Stroke, 32, 1263-1270.

doi:10.1161/01.STR.32.6.1263

Ganasen, K. A., Fincham, D., Smit, J., Seedat, S., & Stein, D. (2008).

Utility of the HIV dementia scale (HDS) in identifying HIV demen-

tia in a South African sample. Journal of the Neurological Sciences,

269, 62-64. doi:10.1016/j.jns.2007.12.027

Grant, I. (2008). Neurocognitive disturbances in HIV. International

Review of Psychiatry, 20, 33-47. doi:10.1080/09540260701877894

Heaton, R. K., Grant, I., Butters, N., White, D. A., Kirson, D., & At-

kinson, J. H. (1995). The HNRC 500: Neuropsychology of HIV in-

fection at different disease stages. Journal of the International

Neuropsychological Socie ty , 1, 231-251.

doi:10.1017/S1355617700000230

Heaton, R. K., Ryan, L., Grant, I., & Matthews, C. G. (1996). Demo-

graphic influences on neuropsychological test performance. In I.

Grant, & K. M. Adams (Eds.), Neuropsychological assessment of

neuropsychiatric disorders (2nd ed., pp. 141-163). New York: Ox-

ford University Press.

Joska, J. A., Gouse, H., Paul, R. H., Stein, D. J., & Flisher, A. J. (2010).

Does highly active antiretroviral therapy improve neurocognitive

function? A systematic review. Neurovirology, 16, 101-114.

doi:10.3109/13550281003682513

Joubert, J., Norman, R., Bradshaw, D., Goedecke, J. H., Steyn, N. P., &

Puoane, T. (2007). Estimating the burden of disease attributable to

excess body weight in South Africa in 2000. South African Medical

Journal, 97, 683-690.

Kilander, L., Nyman, H., Boberg, M., & Lithell, H. (1997). Cognitive

function, vascular risk factors and education: A cross-sectional study

based on a cohort of 70-year-old men. Journal of Internal Medicine,

242, 313-321. doi:10.1046/j.1365-2796.1997.00196.x

Lafayette Instrument Company (1989). Grooved Pegboard: Owner’s

Manual. Lafayette, IN: Lafayette Instrument Company.

Lezak, M. D., Howieson, D. B., & Loring, D. W. (2004). Neuropsy-

chological Assessm ent (4th ed.). New York: Oxford University Press.

Malhotra, R., H oyo , C., Østbye, T., Hughes, G., Schwartz, D., Tsolekile,

L. et al. (2008). Determinants of obesity in an urban township of

South Africa. South African Journal of Clinical Nutrition, 21, 315-

320.

Mitrushina, M. N., Boone, K. B., Razan, J., & D’Elia, L. F. (2005).

Handbook of normative data for neuropsyc ho l ogical assessment (2nd

ed.). New York: Oxford University Press.

Moshani, M. L. (2009). The exploration of neuropsychological disor-

ders in HAART-naïve young adults in South Africa. Unpublished

Honours Thesis, Cape Town: University of Cape Town. URL (last

checked 12 October 2012).

http://web.uct.ac.za/depts/psychology/postgraduate/Hons2009Project

s/Nomakhawuta.Moshani.pdf

National Institutes of Health (1998). Clinical guidelines on the identifi-

cation, evaluation, and treatment of overweight and obesity in adults.

URL (last checked 12 October 2012) .

http://www.nhlbi.nih.gov/guidelines/obesity/ob_home.htm

Puoane, T., Steyn, K., Bradshaw, D., Lau bsch er, R., Fou rie, J., La mbert,

V. et al. (2002). Obesity in South Africa: The South African demo-

graphic and health survey. Obesity Research, 10, 1038-1048.

doi:10.1038/oby.2002.141

Reddy, S. P., Panday, S., Swart, D., Jinabhai, C. C., Amosun, S. L., &

James, S. (2003). Umthenthe uhlaba usamila—The South African

youth risk behaviour survey 2002. Cape Town: South African Medi-

cal Research Council. URL (last checked 12 O c t ob e r 2012).

http://www.mrc.ac.za/healthpromotion/reports.htm

Ren, J. (2004). Leptin and hyperleptinemia: From friend to foe for

cardiovascular function. Journal of Endocrinology, 18, 1-10.

doi:10.1677/joe.0.1810001

Sacktor, N. C., Wong, M., Nakasujja, N., Skolaskya, R. L., Selnes, O.

A., Musisi. S. et al. (2005). The international HIV dementia scale: A

new rapid screening test for HIV d ementia. AIDS, 19, 1367-1374.

Schmidt, S. L., Oliveira, R. M., Rocha, F. R., & Abreu-Villaca, Y.

(2000). Influences of handedness and gender on the grooved peg-

board test. Brain a nd Co gn it ion, 44, 445-454.

doi:10.1006/brcg.1999.1204

Selnes, O. A., Galai, N., McArthur, J. C., & Cohen, S. (1997). HIV

infection and cognition in intravenous drug users: Longterm follow-

up. Neurology, 48, 223-230. doi:10.1212/WNL.48.1.223

Toni, R., Malaguti, A., Castorina, S., Roti, E., & Lechan, R. M. (2004).

New paradigms in neuroendocrinology: Relationships between obe-

sity, systemic inflammation and the neuroendocrine system. Journal

of Endocrinologic al Investigation, 27, 182-186.

UNAIDS (2009). AIDS epidemic update: November 2009. Geneva:

World Health Organisation.

Van Wijk, C. H. (2012). Self-reported generalised anxiety and psy-

chomotor test performance in healthy South Africans. South African

C. H. VAN WIJK

Journal of Psychology, 42, 7-14.

Waldstein, S. R., & Katzel, L. I. (2006). Interactive relations of central

versus total obesity and blood pressure to cognitive function. Inter-

national Journal of Obesity, 30, 201-207. doi:10.1038/sj.ijo.0803114

Waldstein, S. R., Siegel, E. L., Lefkowitz, D., Maier, K. J., Pelletier-

Brown, J. R., & Obuchowski, A. M. (2004). Stress-induced blood

pressure reactivity and silent cerebrovascular disease. Stroke, 35,

1294-1298. doi:10.1161/01.STR.0000127774.43890.5b

World Health Organization (1995). Physical status: The use and inter-

pretation of anthropometry. WHO Technical Report Series. Geneva:

World Health Organization.

Yaffe, K., Lindquist, K., Penninx, B., Simonsick, E., Pahor, M., &

Kritchevsky, S. (2003). Inflammatory markers and cognition in well-

functioning African-American and white elders. Neurology, 61, 76-

80. doi:10.1212/01.WNL.0000073620.42047.D7