Nigerian Journal of Paediatrics

C:\Users\Pelroxltd\Desktop\NJP website\2019\v46n2\2Accuracy of mid upper arm circumference in detection of obesity among school children in Yenagoa City South south region of Nigeria.pdf

Niger J Paediatr 2019; 46 (2): 48 -

ORIGINAL

Okosun OA

CC – BY

Accuracy of mid upper arm

Akinbami FO

Orimadegun AE

circumference in detection of

Tunde-Oremodu II

obesity among school children

in Yenagoa City, South-south

region of Nigeria

DOI:http://dx.doi.org/10.4314/njp.v46i2.2

Accepted: 3rd May 2019

Abstract: The search for alterna-

Under the ROC (AUC) was also

tive method that is easier and less

determined to assess MUAC’s

Akinbami FO

(

)

cumbersome than body mass in-

ability to correctly identify obesity.

Department of Paediatrics,

dex (BMI), for identification of

Results : MUAC correlated posi-

Niger Delta University, Yenagoa,

obese individuals has been contro-

tively with BMI and age, the cut-

Bayelsa State, Nigeria

versially discussed in recent lit-

off values increased with age in

Email: kaytoks@yahoo.com

erature. This study was carried

both boys and girls. When com-

out to determine the accuracy of

pared with BMI, using sex and age

Ofure AO, Tunde-Oremodu II

Mid Upper Arm Circumference

group specific cut-off for MUAC,

Department of Paediatrics,

(MUAC) compared to BMI.

the estimated specificities were

Federal Medical Centre, Yenagoa,

Method : We recruited 920 chil-

relatively higher than sensitivities

Bayelsa State, Nigeria

dren aged 5 – 18 years from pri-

in all age groups. However, the

mary and secondary schools in

best performance of MUAC for

Orimadegun AE

Yenagoa in the Niger Delta region

detection of obesity was recorded

Institute of Child Health,

of Nigeria using a multistage ran-

for girls (AUC = 0.94, 96% CI =

College of Medicine, University of

dom sampling technique. Weight,

0.89, 0.99) and boys (AUC = 0.89,

Ibadan, Nigeria

height and MUAC of the children

95% CI = 0.78, 0.99) in age group

were measured using standard

10-14 years. Similarly, the NPV

methods. We calculated BMI,

were higher than the PPV.

defined obesity as BMI-for-age z-

Conclusion : The MUAC showed

scores >2 and the corresponding

remarkably high accuracy for diag-

cut-off values of MUAC for de-

nostic and screening use among

fining obesity were determined.

children aged 10-14 years but in-

Sensitivity, specificity, negative

consistent results in other age

predictive values (NPV) and posi-

groups.

tive predictive values (PPV) of

MUAC were determined using

Keywords : body mass index, mid

BMI as the gold standard. Area

-upper arm circumference, obesity

Introduction

(MUAC) because it is less affected by the localised ac-

cumulation of excess fluid than BMI and it does not

require height measurement.

14-16

Obesity among children and adolescents is rapidly be-

coming a global epidemic.

1,2

The estimated prevalence

Nigeria is on the verge of experiencing an increase in

of overweight/obesity in children aged 5-17 years

childhood obesity, possibly because the rapidly chang-

worldwide is 10%, with variation of over 30% in Amer-

ing economy and population lifestyle tends towards

ica to <2% in sub-Saharan Africa. In Nigeria, the na-

those of developed countries. A potentially more

tional prevalence is 1.1%, ranging from 0% to 37.2% in

friendly, faster and easier method of screening for over-

different regions of the country.

5-9

Due to the complica-

weight and obesity among young people such as the

tions of childhood obesity, it is important to identify

MUAC needs to be validated. This study was conducted

children at risk, but unfortunately information on the

to evaluate the accuracy of MUAC compared to BMI in

problem is scarce in Nigeria.

determining overweight and obesity. It was also in-

The body mass index (BMI) is the main proxy indicator

tended to determine MUAC cut-off values for defining

of body fatness in both children and adults,

10,11

but its

obesity among school children in Yenagoa, Nigeria's

use requires equipment and calculations, thus limiting its

south-south region.

use in resource poor settings and necessitating the need

for a more friendly and reliable method.

12,13

One such

method may be the mid-upper arm circumference

Methods

difficult or problem area. All students were invited to a

Study design, settings and population

meeting in a large hall during the data collection. We

explained the purpose of the study to them and assured

A cross-sectional design was adopted in this study.

them of the confidentiality of any voluntary information.

School children aged 5 to 18 years were investigated for

The selected participants received consent forms con-

obesity using anthropometric measurements and at the

taining detailed explanations of the study and process of

same time documented their socio-demographic charac-

voluntary consent to be signed by their parents / caregiv-

teristics. Participants were recruited during school hours,

ers and returned the following day. Furthermore, before

minimal physical examination of each child was per-

the questionnaire was given, we informed the students

formed in a dedicated cubicle in each of the schools and

that their participation was voluntary and obtained ac-

strict confidentiality was observed. The study was con-

cent from each person. Each student was given time to

ducted in the petroleum-rich town of Yenagoa, Nigeria's

think through and decide on their participation. The in-

Niger-Delta region during the months of June to Sep-

vestigators were available to give the necessary clarifi-

tember 2015. Yenagoa is a Local Government Area

cations to anyone who asked for more information on

with an estimated total population of 459,693, of which

any part of the questionnaire.

about 48.3% were children (2015 projection from 2006

census). The adult population are mainly civil servants,

Anthropometric measurements

traders, and sustenance farmers. According to the 2013

demographic health survey report, the net school atten-

The investigators measured the weight, height and

dance rate was 78.1%. The choice of this study area

MUAC using standard methods with the help of re-

search assistants.

14,20

was informed by the anecdotal observation of rapidly

The weight and height were meas-

changing economy, proliferation of fast foods restau-

ured at the nearest 0.1 kg and 0.1 cm respectively, while

rants and increasing transportation vehicles, and rela-

the MUAC was measured at the nearest 0.1 cm with a

tively few youth friendly recreational centres for play or

non-stretchable tape on bare skin at the midpoint be-

physical exercises.

tween the tip of the olecranon and the acromion process.

A battery powered Seca 872 digital floor scale (Seca,

Sample size calculation

Inc., Columbia, MD, USA) was used for weight meas-

urements while height was measured using a standard

stadiometer. BMI (kg/m ) was calculated as weight (kg)

The study sample size was calculated using formula for

divided by the square of the height (m ). Obesity was

estimating single proportion, an assumed prevalence of

childhood obesity of 18.0% reported in two previous

defined by the World Health Organization (WHO) 2007

publications,

9,18

a margin of error ±5% margin of error,

BMI-for-age reference, z score >+2.0.

21,22

95% level of confidence and non-response rate of 5%.

The calculated number of children required for each of

Ethical considerations

the 4 categories of schools (public primary, private pri-

mary, public secondary and private secondary) was 218

Ethical approval was obtained from the State Universal

giving the minimum total sample size as 872.

Basic Education Board, State Senior Secondary School

Board, and the Ethics Committee of the Federal Medical

Sampling technique

Centre, Yenagoa. Parents signed the informed consent

forms, and assent was obtained from the participants.

A three-stage random sampling method was used to se-

lect two educational zones (Okolobiri and Yenagoa

Statistical analysis

town), schools (16 primary and 16 secondary schools in

each zone) and a total of 920 participants out of 91,238

The z scores for the calculated BMI were generated us-

ing the WHO Anthro Plus 2007 software, exported to

in the school registers, respectively. The number of par-

ticipants from each school was based on the proportion

MS Excel and added to other corresponding anthropom-

of the school population in each of the public primary

etric measures in IBM SPSS statistics version 20 (SPSS

schools, private primary schools, public secondary

Inc., Chicago, IL, USA). Comparisons were made be-

schools, and private secondary schools, respectively, out

tween groups using the Chi square test and the Student t-

of the overall pupil population.

test, while the Pearson’s correlation coefficient was used

to determine the strength of association between two

Data collection

variables. A p value less than 0.05 was considered as

statistically significant. For each of the three categories

The interviewer-administered questionnaire used for this

of 5-9 years, 10-14 years and 15-18 years, the different

study was designed by the investigators, items were

MUAC (cm) cut-off values for obesity corresponding to

adapted from the 2013 USA Youth Risk Behaviour Sur-

BMI-for-age z score > + 2.0 were determined using the

veillance instrument. The questionnaire had four sec-

point of interception of the graphical plots of sensitivity

tions: socio-demographic characteristics, parent and

and specificity against MUAC. Thereafter, the Receiver

family information, child health status, and anthropom-

Operating Characteristic (ROC) curves were used to test

etric measurements. It was pre-tested in a rural primary

MUAC’s ability to correctly identify obese children

and secondary school, as well as an urban primary and

using the BMI as the gold standard as described by Han

and colleagues.

secondary school, and this helped modify any noticed

The sensitivity and specificity of

MUAC as a screening tool for obesity were also calcu-

value (NPV) for different age groups and gender were as

lated for all cut-off points to find the optimal cut-off

shown in Table 3. Table 3 also shows that for both ages

values.

The agreement between MUAC and BMI as

and genders, the accuracy levels of MUAC for identify-

method of identifying obesity was assessed using Kappa

ing obesity, using the area under the curve (AUC), was

statistics according to the scale suggested by Altman.

0.72 (95% CI = 0.59, 0.84), and 0.91 (0.86, 0.97) in the

Altman proposed Kappa of less than 0.20 as poor agree-

age groups 5-9 years and 10-14 years, respectively (p

ment, 0.20 to 0.40 as fair agreement, 0.40 to 0.60 as

<0.001). However, these levels of accuracy were not

moderate agreement, 0.60 to 0.80 as good agreement

demonstrated when the same analysis was done sepa-

and 0.80 to 1.00 as very good agreement.

[25]

Correct

rately for males and females. Whilst the MUAC signifi-

classification means that both MUAC and BMI identi-

cantly detected obesity among female participants in all

fied the child as having obesity or not, while misclassifi-

the three age categories (AUC >0.78), among males,

cation indicates that there is discordance between the

MUAC only showed significant accuracy for detection

two methods.

of obesity in the age group 10-14 years (AUC = 0.89).

The MUAC cut-off values for obesity were 20.8 cm,

24.8 cm, and 27.8 cm in girls aged 5-9 years, 10-14

Results

years and 15-18 years respectively, while the cut-offs in

Characteristics of the study population

boys aged 5-9 years, 10-14 years and 15-18 years were

21.8 cm, 25.4 cm and 27.8 cm, respectively. For the

There were 920 participants, 403 (43.8%) males and 517

MUAC cut-offs that showed significant AUC, the sensi-

(56.2%) females.

The age of the study participants

tivities were 82.0% in boys aged 10-14 years; 68.0%,

ranged from 5 to 18 years (overall mean age = 11.7±3.0

92.0% and 86.0% in girls aged 5-9 years, 10-14 years

years). Table 1 shows the mean age and anthropometric

and 15-18 years, respectively. Similarly, the specificity

measurements by gender of the participants. The ages of

was 88.0% in boys aged 10-14 years and approximately

the boys and girls were similar (p = 0.088). The mean

96.0%, 84.0% and 75.0% in girls aged 5-9 years, 10-14

values of weight, MUAC, BMI, and height were signifi-

years and 15-18 years, respectively. The positive and

cantly higher in girls than in boys (p < 0.05).

negative predictive values for each cut-off point are also

shown in Table 3.

Table 1: Anthropometric measurements of male and female

participants

Table 2: Relationship between mid-upper-arm circumference

Variables

Male

Female

and other anthropometric variables by gender

Age (years)

11.6±3.1

11.9±2.9

0.088

Male

Female

Variables

Weight (cm)

39.2±13.7

43.2±14.5

<0.001

MUAC (cm)

21.7±3.8

23.1±4.9

<0.001

Weight (cm)

0.765

<0.001

0.737

<0.001

BMI (kg/m )

19.2±4.0

20.5±4.8

<0.001

Height (cm)

0.469

<0.001

0.347

<0.001

Height-for-age z-score

-0.93±1.36

-0.61±1.42

0.001

Weight-for-age z-score

0.557

<0.001

0.626

<0.001

Weight-for-age z-score

-0.14±1.43

0.08±1.24

0.016

Height-for-age z-score

0.416

<0.001

0.344

<0.001

Height (cm)

140.9±16.9

143.4±15.2

0.017

BMI-for-age z-score

0.589

<0.001

0.624

<0.001

BMI-for-age z-score

0.52±1.34

0.63±1.26

0.215

r = correlation coefficient

*p value < 0.001 in all cases

Prevalence of obesity and correlation between MUAC

and other indices

According to the WHO recommended cut-offs for detec-

tion of obesity using z-scores, the prevalence rate of

obesity in boys and girls was 8.0% and 9.7%, respec-

tively. The prevalence of obesity was not significantly

different between girls and boys (p = 0.436). Table 2

presents the Pearson correlation coefficients between

MUAC, and anthropometric parameters for boys and

girls. MUAC showed a strong positive correlation with

BMI (p < 0.001). In both boys and girls, all the anthro-

pometrics including BMI show positive correlation with

MUAC (p < 0.001). However, the strongest correlation

was shown between weight and MUAC while the corre-

lations between MUAC and height and its z-score were

relatively weak.

MUAC cut-offs and its abilities to accurately define

obesity

The cut-offs for MUAC, its specificity, sensitivity, posi-

tive predictive value (PPV) and negative predictive

Fig 1: Receiver Operating Curves for MUAC as predictor of BMI-for-age >2.0 z-score among children

Table 3: Area under the ROC curves, optimal cut-off values, sensitivities, and specificities for mid-upper-arm circumference asso-

ciated with obesity in children

True

Age (years)

Prevalence

Cut-off

AUC

Sensitivity

Specificity

PPV

NPV

(%)

(cm)

(95% CI)

All children

5-9

205

17.0

22.1

0.72 (0.59, 0.84)

<0.001

0.47 (0.30, 0.65)

0.95 (0.91, 0.98)

0.67 (0.45, 0.84)

0.90 (0.85, 0.94)

10-14

526

8.0

25.6

0.91 (0.86, 0.97)

<0.001

0.88 (0.73, 0.95)

0.86 (0.82, 0.89)

0.35 (0.26, 0.45)

0.99 (0.97, 1.00)

15-18

189

4.0

27.9

0.73 (0.49, 0.96)

0.032

0.75 (0.36, 0.96)

0.76 (0.69, 0.82)

0.12 (0.05, 0.25)

0.99 (0.94, 1.00)

Male

5-9

15.0

21.8

0.58 (0.37, 0.78)

0.351

0.20 (0.05, 0.49)

0.95 (0.87, 0.98)

0.43 (0.12, 0.80)

0.87 (0.77, 0.93)

10-14

232

7.0

25.4

0.89 (0.78, 0.99)

<0.001

0.82 (0.56, 0.95)

0.88 (0.83, 0.92)

0.35 (0.21, 0.52)

0.98 (0.95, 1.00)

15-18

1.0

27.8

0.36 (0.22, 0.50)

0.071

0.00 (0.11, 0.95)

0.81 (0.70, 0.88)

0.00 (0.01, 0.24)

0.99 (0.91, 1.00)

Female

5-9

108

18.0

20.8

0.83 (0.69, 0.97)

<0.001

0.68 (0.43, 0.86)

0.96 (0.88, 0.99)

0.76 (0.50, 0.92)

0.93 (0.86, 0.97)

10-14

294

8.0

24.8

0.94 (0.89, 0.99)

0.024

0.92 (0.72, 0.99)

0.84 (0.79, 0.88)

0.34 (0.23, 0.47)

0.99 (0.97, 1.00)

15-18

115

6.0

27.8

0.78 (0.54, 0.99)

0.012

0.86 (0.42, 0.99)

0.75 (0.66, 0.83)

0.18 (0.08, 0.36)

0.99 (0.92, 1.00)

AUC Z= area under the curve; CI = confidence interval; PPV = Positive predictive value; NPV = Negative predictive value

Agreement between MUAC and BMI

Table 4: Agreement between MUAC with BMI in the identifi-

cation of obesity among children

The agreement between MUAC and BMI in classifying

Correctly

Misclassi-

Kappa statistic

children aged 5-18 years in the identification of weight-

Age group

identified by

fied as by

value

MUAC, n

for-height categories were as shown in Table 4. For all

MUAC, n(%)

(%)

participants, irrespective of gender and age, the Kappa

All children

776 (84.3)

114 (15.7)

0.369

<0.001

statistics (0.369) suggested a significant but fair agree-

Male Children

ment between MUAC and BMI. This level of agreement

All Boys

341 (84.6)

62 (15.4)

0.276

<0.001

after stratifying the participants by gender revealed a

5-9

91 (83.5)

16 (16.5)

0.193

0.037

relatively higher Kappa statistics for females (0.422)

10-14

203 (87.5)

29 (12.5)

0.433

<0.001

compared with 0.276 for males suggesting a moderate

15-18

51 (77.0)

17 (22.6)

-0.261

0.587

agreement for female and only fair agreement for male.

Female children

However, further stratification of participants by age

All Girls

435 (84.1)

82 (15.9)

0.422

<0.001

revealed the highest levels of agreement (moderate)

5-9

98 (90.7)

10 (9.3)

0.667

<0.001

were recorded among children in age 10-14 years among

10-14

250 (85.0)

44 (15.0)

0.433

<0.001

boys (0.433) and 5-9 years among girls (0.667).

15-18

87 (75.6)

28 (24.4)

0.222

<0.001

Discussion

shows some promise as a reliable method for detection

of obesity among children population irrespective of

This study investigated the use of MUAC in the detec-

geographical location. However, many studies have

tion of obesity among Nigerian children aged 5-18 years

shown that body size, which varies by geographical

using the BMI-for-age reference, z score >+2.0 as the

boundaries and ethnicity, is an essential determinant of

gold standard.

21,22

The MUAC strongly correlated posi-

MUAC in children, and so it is necessary to consider

tively with BMI and age in this study. The MUAC cut-

population specific characteristics such as average

off values increase with age in both boys and girls.

height to determine whether MUAC is appropriate or

not.

29,30

Thus, MUAC can accurately identify obesity in all Ni-

gerian girls aged 5-18 years but only among boys in the

ages 10-14 years. In these age categories, the areas un-

The results of our data show that, when compared to

der the ROC curve were consistent with robust diagnos-

BMI, the sensitivities of MUAC in screening for obesity

tic performance and indicated that measurement of

in children were sufficiently good but lower than those

obtained in other studies. Craig etal observed sensitiv-

MUAC has a good ability to identify children with or

without elevated BMI. To our knowledge, this study

ity and specificity to be generally high (76 – 97%) in

provides the first MUAC cut-off values for Nigerian

their study of 5 – 14 year old South African children.

Similarly, de Almeida et al obtained a sensitivity and

children aged 5-18 years.

specificity of 76.5% and 77.9% respectively in 1 – 5

year old Brazilian children. Lu et al

One of the potential benefits of using MUAC for detec-

also obtained a

tion of obesity is that its measurement is not influenced

sensitivity of 83.8 – 94.5% and 82.5 – 90.2% in Han

by respiratory movements and postprandial abdominal

girls and boys respectively, with a specificity of 81.7 –

distension as in the case of waist circumference. There-

94.0% and 89.0 – 95.7% in girls and boys respectively.

fore, MUAC may be a good alternative and reliable in-

Overall, it is evident from the results of this study that

dex for obesity among girls. Our findings agree with the

specificities and negative predictive values were re-

report by Chomtho et al

which showed that MUAC

markably greater than sensitivities and positive predic-

correlated strongly with fat mass but weakly with fat-

tive values, respectively, in all age groups. The high

free mass. In that study, the MUAC value explained

specificities imply that the fraction of those without the

63% of variability in total fat mass and only 16% of

obesity correctly identified as negative by the MUAC

variability in total fat-free mass in healthy children. A

was higher than the fraction of those with the obesity

number of direct measurements of body-fat content and

correctly identified as positive by the MUAC. Therefore,

its distribution such as dual X-ray absorptiometry can be

MUAC would be a better tool for diagnostic use than

used to accurately measure degree of obesity, but many

screening. Similarly, the fractions of people with

of such methods are neither practical nor affordable to

MUAC indicating obesity who actually have the condi-

people in resource-limited countries like Nigeria. Thus,

tion as shown by the positive predictive values were

the use of MUAC provides a cheaper and easy to use

lower than those people who actually did not have the

alternative method in such settings.

condition detected by MUAC as shown by negative pre-

dictive values. The relatively lower positive predictive

The MUAC cut-offs that produced good accuracy for

values obtained for MUAC is not surprising as studies

the different age groups were relatively higher than

have shown that when the prevalence of the disease is

those reported by Lu et al who reported values of 18.9

low the predictive value of a positive test will also be

low.

– 23.4cm for 7 – 12 year old Han children, and those

reported by Craig et al among South African children

who

reported

values

18.3cm/18.9cm

and

One major issue that can limit the generalisation of our

18.4cm/18.6cm for 5 – 9 year olds girls and boys respec-

finding to the entire Niger Delta region of Nigeria is

tively, and 22.5cm/22.8cm and 22.2cm/23.2cm for 10 –

geographical restriction of participants to only residents

14 year old girls and boys respectively. These differ-

of Yenagoa Local Government Area. However, the most

ences in the MUAC cut-offs suggest that children in

recent demographic health survey conducted in the re-

these three populations had remarkably different arm

gion showed that the characteristics of the population in

different localities might not be remarkably different.

[4]

muscle mass. Another possible explanation for the dif-

ferences in cut-offs may be the variation in prevalence

Thus, we speculate that variation in the prevalence of

of obesity (defined by using BMI). A recent systematic

obesity and the ability of MUAC to detect it may not be

review showed that sensitivity and specificity of a test

considerably different across the Niger Delta region.

often vary with disease prevalence and this effect is

Another limitation is the lack of data on pre- school chil-

likely to be the result of mechanisms, such as patient

dren. The challenges of collecting accurate data on an-

spectrum and diagnostic cut-off.

thropometry in addition to the fact that MUAC vary

The fact that AUC were excellent for boys aged 10-14

considerably even within the same ages being a period

years and girls in all age groups (AUC >0.78) agree with

of rapid growth, make the use MUAC alone less de-

pendable compared to actual estimation of BMI. None-

the report by Lu et al.

for Han children (0.934 –

0.975) and Craig et al

among South African children

theless, we suggest that the feasibility of including pre-

(0.96 and 0.90 in girls and boys in the 5 – 9 year old age

school children and use of MUAC method be considered

group; 0.94 and 0.97 in girls and boys in the 10 – 14 year

in future studies.

old age group). The implication of this is that MUAC

References

16. de Almeida CA, Del Ciampo

1. Ng M, Fleming T, Robinson M,

Ansa V, Odigwe C, Anah M.

LA, Ricco RG, Silva Jr SM,

Thomson B, Graetz N, Mar-

Profile of body mass index and

Naves RB, Pina JF. Assess-

gono C , et al. Global, regional,

obesity in Nigerian children

ment of mid-upper arm cir-

and national prevalence of

and adolescents. Nigerian Jour-

cumference as a method for

overweight and obesity in chil-

nal of Medicine: journal of the

obesity screening in preschool

dren and adults during 1980 –

National Association of Resi-

children. Jornal de pediatria

2013: a systematic analysis for

dent Doctors of Nigeria

2003;79: 455-460.

the Global Burden of Disease

2001;10: 78-80.

17. National Population Commis-

Study 2013. Lancet 2014;384:

9. Akesode F, Ajibode H. Preva-

sion. 2006 Population Census.

766-781.

lence of obesity among Nige-

National Bureau of Statistics:

2. Popkin BM, Adair LS, Ng SW.

rian school children. Soc Sci

Abuja, Nigeria, 2006.

Global nutrition transition and

Med 1983;17: 107-111.

18. Owa JA, Adejuyigbe O. Fat

the pandemic of obesity in de-

10. Centre for Disease Control and

mass, fat mass percentage,

veloping countries. Nutr Rev

Prevention. Body mass index:

body mass index, and mid-

2012;70: 3-21.

Considerations for practitio-

upper arm circumference in a

3. Lobstein T, Baur L, Uauy R,

ners. United States of America:

healthy population of Nigerian

Iaso International Obesity

Department of Health and Hu-

children. J Trop Pediatr

TaskForce. Obesity in children

man Science, Centre for Dis-

1997;43: 13-19.

and young people: a crisis in

ease Control and Prevention;

19. Kann L, Kinchen S, Shanklin

public health. Obes Rev 2004;5

2010 [cited 01 March, 2015].

SL, Flint KH, Hawkins J, Har-

Suppl 1: 4-104.

Available from: www.cdc.gov/

ris WA , et al. Youth risk be-

4. National Population Commis-

obesity/downloads/

havior surveillance — United

sion (NPC) [Nigeria], ICF In-

bmiforpactitioners.pdf.

States, 2013. 2014.

ternational. Nigeria Demo-

11. Sweeting HN. Measurement

20. Organization WH. Physical

graphic And Health Survey

and definitions of obesity in

status: The use of and inter-

2013. National Population

childhood and adolescence: a

pretation of anthropometry,

Commission, Federal Republic

field guide for the uninitiated.

Report of a WHO Expert

of Nigeria and ICF Interna-

Nutr J 2007;6: 32.

Committee. 1995.

tional Rockville, Maryland,

12. Bhurosy T, Jeewon R. Pitfalls

21. WHO Multicentre Growth

USA: Abuja, Nigeria and

of using body mass index

Reference Study Group. WHO

Rockville, Maryland, USA:,

(BMI) in assessment of obesity

Child Growth Standards based

2014, pp 165-171.

risk. Curr Res Nutr Food Sci

on length/height, weight and

5. Sabageh AO, Ogunfowokan

2013;1: 71.

age. Acta Paediatr Suppl

AA, Ojofeitimi EO. Obesity

13. Lu Q, Wang R, Lou D-H, Ma C

2006;450: 76-85.

and body image discrepancy

-M, Liu X-L, Yin F-Z. Mid-

22. de Onis M. Growth curves for

among school adolescents in Ile

upper-arm circumference and

school age children and ado-

-Ife, Osun State, Nigeria. Pak J

arm-to-height ratio in evalua-

lescents. Indian Pediatr

Nutr 2013;12: 377-381.

tion of overweight and obesity

2009;46: 463-465.

Ahmadu BU, Usiju NM, Ibra-

in Han children. Pediatr Neo-

23. Organization WH. AnthroPlus

him A, Adiel AA, Tumba D,

natol 2014;55: 14-19.

2007: software for the global

Rimamchika M , et al. Linger-

14. Jeyakumar A, Ghugre P, Gad-

application of the WHO Ref-

ing hunger among primary

have S. Mid-Upper-Arm Cir-

erence 2007 for 5-19 years to

school pupils residing in rural

cumference (MUAC) as a Sim-

monitor the growth of school-

areas of Borno State, North-

ple measure to assess the nutri-

age children and adolescents.

Eastern Nigeria: implication for

tional status of adolescent girls

Geneva, Switzerland: World

education and food supplemen-

as compared with BMI. Infant

Health Organization; 2007.

tation programs. Glo Adv Res J

Child Adolesc Nutr 2013;5: 22-

2009.

Food Sci Technol 2012;1: 93-

25.

24. Han TS, van Leer EM, Seidell

97.

15. Mazıcıoğlu MM, Hatipoğlu N,

JC, Lean ME. Waist circum-

Senbanjo I, Adejuyigbe E.

Öztürk A, Çiçek B, Üstünbaş

ference as a screening tool for

Prevalence of overweight and

HB, Kurtoğlu S. Waist circum-

cardiovascular risk factors:

obesity in Nigerian preschool

ference and mid − upper arm

evaluation of receiver operat-

children. Nutr Health 2007;18:

circumference in evaluation of

ing characteristics (ROC).

391-399.

obesity in children aged be-

Obes Res 1996;4: 533-547.

tween 6 and 17 years. J Clini-

25. Altman DG. Practical statis-

cal Res Pediatric Endocrinol-

tics for medical research.

ogy 2010;2: 144.

Chapman and Hall: London,

England, 1997.

26. Chomtho S, Fewtrell MS, Jaffe

28. Leeflang MM, Rutjes AW,

30. Arnold R. The arm circumfer-

A, Williams JE, Wells JC.

Reitsma JB, Hooft L, Bossuyt

ence as a public health index

Evaluation of arm anthropome-

PM. Variation of a test’s sensi-

of protein-calorie malnutrition

try for assessing pediatric body

tivity and specificity with dis-

of early childhood. XVIII. The

composition: evidence from

ease prevalence. Canadian

QUAC stick: a field measure

healthy and sick children. Pedi-

Medical Association Journal

used by the Quaker Service

atr Res 2006;59: 860-865.

2013;185: E537-E544.

Team in Nigeria. J Trop 3-

27. Craig E, Bland R, Ndirangu J,

29. Joseph B, Rebello A, Kullu P,

247.

Reilly JJ. Use of mid-upper

Raj VD. Prevalence of malnu-

31. Parikh R, Mathai A, Parikh S,

arm circumference for deter-

trition in rural Karnataka, South

Chandra Sekhar G, Thomas R.

mining overweight and overfat-

India: a comparison of anthro-

Understanding and using sen-

ness in children and adoles-

pometric indicators. J Health

sitivity, specificity and predic-

cents. Arch Dis Child 2014;99:

Popul Nutr 2002;20: 239-244.

tive values. Indian J Ophthal-

763-766.

mol 2008;56: 45-50.