ORIGINAL

Niger J Paed 2013; 40 (2): 179 –183

Triceps skin fold thickness as a

measure of body fat in Nigerian

adolescents

Ahmad MM

Ahmed H

Airede KI

DOI:http://dx.doi.org/10.4314/njp.v40i2,15

Accepted: 26th November 2012

Abstract Background: Skin fold

thickness (SFT) at selected areas

offers a simple method of subcuta-

neous fat assessment and provides a

good estimate of obesity and body

fat distribution. The triceps SFT has

been shown to be one of the best

and most popular sites for SFT

measurement in children.

Objective: To assess the body fat of

school adolescents and to compare

the performance of triceps SFT

with Bioelectrical impedance

method in the detection of over-

weight (OW) and obesity (OB)

among the subjects

UK Ltd; 2004) were used respec-

tively to measure the triceps SFT

and body fat content (%) according

to the manufacturer’s instructions.

Results: The mean triceps SFT val-

ues were 8.9mm (±4.7) for males

and 12.9mm (±4.6) for the females

Ahmad MM

Ahmed H

(

)

Department of Paediatrics

Usmanu Danfodiyo University

Teaching Hospital, Sokoto

Tel: +2348033691126

(

8

±

p<0.001). Mean % BF values were

.2 ± 4.1% for the males, and 20.0

6.8% for the females (p<0.001).

E-mail: murisbn@yahoo.com

Airede KI

The triceps SFT gave a prevalence

of overweight of 2.5%, while that

of obesity was 0.8%. With the BIA

method, 2.5% of the subjects were

classified as overweight and 1.7%

as obese.

Department of Paediatrics,

Faculty of Clinical Sciences University

of Abuja, Gwagwalada, FCT Nigeria.

Methods: The study was cross sec-

tional; involving secondary school

students within Sokoto metropolis.

Subjects were selected by a multi-

stage random sampling method.

Harpenden skin fold caliper

Conclusion: Triceps SFT remain a

fair surrogate for the assessment of

adiposity, the component of over-

weight that leads to pathology.

Key words: Skin fold thickness,

bioelectrical impedance analysis,

body fat, obesity, adolescents

(

ASSIST Creative Resources Ltd,

LL13 9UG, UK) and Tanita Body

fat scale (model UM-030, Tanita,

Introduction

Skin fold thickness has been used to estimat₁e body fat-

1

ness and this has gained more popularity. There are

Nutrition transition is increasingly₁ evident in middle

income and low income countries. Together with re-

duced energy expenditure and sedentary life styles; this

has contributed to rises in the incidence of obesity and

different recognized areas for the measurement of SFT,

which includes the triceps SFT, biceps, subscapular,

suprailiac, abdominal (flank), chest and thigh

(quadriceps) skin folds. If only one₁s₂_,k₁₃in fold is meas-

1

non communicable diseases. Skin fold thickness (SFT)

ured, it is usually the triceps SFT.

Therefore, the

at selected areas offers a simple method of subcutaneous

fat assessment and provi₂d_,₃es a good estimate of obesity

and body fat distribution. Skin fold thickness measure-

ments are a well established means of assessing the

thickness of subcutaneous fa_-₆t at all ages; including in-

triceps area was chosen to assess body fat in our sub-

jects. In addition, this site is easy to expose and is more

convenient to the subjects (₁p₂articularly adolescents and

adults) than the other areas. It has also been shown to

give better results for o₁₃besity screening in adolescents

compared to other sites.

4

fancy and neonatal periods. Direct assessment of adi-

posity, the component of overweight that leads to pa-

thology_,₈, represent a significant advance over body mass

A number of equations have been proposed whereby

SFT can be used to predict total body fat from density-

derive₅d_,₁₄e_-₁s₆timate, both for adults, children and adoles-

7

index. The measurement is relatively easy, fast, non-

3

,4

invasive and requires little equipment. It does not re-

9

quire a high degree of technical skill, although a simple

cents.

These equations are population-based, and

training is required to standardize the measurement. A

properly trained₁₀individual can attain a precision of

within 5% easily.

they relate the sum of two or more SFT measurements to

the body density. For this reason, these methods are usu-

ally not effective for a population that differs substan-

tially from the original reference population, due to

1

80

cross-population diffe₁r₇ences in the parameters that are

used in the equations.

the coeducational schools, students were stratified into

groups of boys and girls, for each selected class, to al-

low for equal gender representation.

Bioelectrical impedance analysis (BIA) on the other

hand is a simple, portable, non-invasive, safe and highly

acceptable to patient technique that measures total-body

electrical conductivity by electrical impedance, t₁h₀ereby

providing an assessment of the body composition.

Trice₉ps skin fold thickness percentile values by Frisan-

1

cho and percentage body fat (%BF) cut-off values ac-

8

cording to McCarthy et al. were used to classify sub-

jects as overweight or obese. Overweight and obesity

were respectively defined as triceps SFT and %BF val-

ues at or greater than 85 to <95 and ≥95 percentiles

Since the pathology associated with obesity is driven by

the excess fat mass, the ideal assessment tool should

directly assess adiposity. Many of the available tools

th

for the age and gender.

1

that can do that are complex and expensive. Bioelectri-

cal impedance is one of the simple and cheap tools de-

veloped, that can distinguish fat and lean tissue mass. It

also offer the advantage of increased speed, ease of

measureme₈_,n₁₈t as well as high inter- and intra-observer

Ethical Consideration

Ethical approval for the study was obtained from the

Ethical Committee of Usmanu Danfodiyo University

Teaching Hospital, Sokoto, Ministry of Education

Sokoto State, and the affected school authorities. In-

formed consent was also obtained from parents/

guardians of day school students.

reliability.

Therefore, BIA was employed (in the ab-

sence of gold standard methods of body composition

analysis - hydrodensitometry and/or dual energy X-ray

absorptiometry) to compared the performance of SFT

method in assessing body fatness.

Data collection and measurements

Questionnaires were used to record each subject’s bio-

data including ethnicity, school, class level and parenta₂₀l

socioeconomic class (SEC) according to Oyedeji.

Three (3) trained research assistants (all graduates) and

author administered the study questionnaires. However,

all measurements were carried out by the first author

(AMM) so as to avoid inter-observer errors.

Objective

To compare the performance of triceps SFT with Bio-

electrical impedance analysis (in the absence of gold

standard methods of body composition analysis) in the

assessment of body fat among school adolescents.

Harpenden skin fold caliper (ASSIST Creative Re-

sources Ltd, LL13 9UG, UK) and Tanita Body fat scale

(

model UM-030, Tanita, UK Ltd; 2004) were used re-

spectively to measure the triceps SFT and body fat con-

tent (%) according to the manufacturer’s instructions.

Subjects’ heights were also measured using stadiometer

scale (Seca 213, UK). The height in centimeter and age

in years were first entered into the digital BIA scale and

the appropriate gender option selected, for the assess-

ment of percentage body fat (%BF) by BIA. Body

weight and %BF were simultaneously measured as the

subjects bare feet make pressure contact with the elec-

trodes and the digital scale. Fat mass was derived from

the percentage body fat (%BF) and body weight as fol-

lows: FM = %BF × Body weight (kg).

Subjects and Methods

This was a cross-secti_to_hnal study condu_tc_hted over a six-

week period (from 10 February to 25 March, 2010).

Three hundred and sixty adolescent students aged 10 to

1

8 years were enrolled.

The subjects were drawn from both public and private

schools within Sokoto metropolis, by a multistage ran-

dom sampling as follows: There were 32 secondary

schools within the metropolis (from the Statistics Unit of

the State’s Ministry of Education). Of these schools,

twenty-one (21) were public schools while eleven (11)

were privately owned.

The schools were grouped into private and public groups

based on the 3 local government areas (LGAs) within

the metropolis. A total of six (6) secondary schools were

selected for the study, two from each of the 3 LGAs (a

public; and a private school). For each group (public or

private) in an LGA, the names of the schools were writ-

ten on pieces of paper which were folded and mixed up.

One school was picked at random, from each of the two

groups and was subsequently enrolled.

Results

Of the three hundred and sixty subjects studied, one

hundred and ninety eight (55%) belonged to the Hausa

ethnic group, 48 (13.3%) were Fulanis, 40 (11.1%) were

Yorubas, 33 (9.2%) were Ibos and 41 (11.4%) were

from other minor ethnic groups. The predominant group

2

were the Hausas (X =9.341, df=4, p=0.05). The age and

At the school level, sixty (60) students were selected per

school (10 at each class level) by systematic random

sampling. Where there was more than one arm per class

level, one of the arms was selected by balloting. Among

gender distribution of the subjects is depicted in fig 1

1

81

Fig 1: Age and gender distribution of the subjects

Table 3: Mean % Body fat ± SD according to age and gender

Males

BF (%)

Females

BF (%)

+

Age

(yr)

n

Mean (± SD)

n

Mean (± SD)

t

P

1

0

1

2

3

4

5

6

7

8

3

5

12.53 ± 0.71

9.66 ± 3.48

9.05 ± 2.94

9.80 ± 4.62

8.39 ± 2.20

8.26 ± 5.35

7.29 ± 4.58

7.55 ± 3.53

5.94± 3.00

2

8

17.10 ± 1.84

15.05 ± 7.38

-3.194

-1.528

-4.057

-5.970 <0.001

-8.535 <0. 001

-7.020 <0. 001

-10.782 <0. 001

-5.907 <0. 001

-5.810 <0. 001

19.817 <0.001

0.193

0.201

0.002

13

28

24

35

32

16

19 17.36 ± 7.01

18 20.22 ± 5.57

30 20.40 ± 5.83

31 22.28 ± 6.24

30 21.65 ± 4.72

23 15.93 ± 6.88

19 22.44 ± 9.67

Total 180

8.23 ± 4.10 180 19.97 ± 6.85

+

%

BF= percentage body fat; SD= standard deviation

The % BF values were generally lower in our subjec₈ts

compared to the values reported by McCarthy et al.

among children in the UK. Females showed more

progressive increase in %BF with age as shown in Fig 2.

Age group 10 years had the least number of participating

subjects (1.4%), as majority of pupils in this age group

may still be in the primary school. Age 16 years had the

highest number of subjects (18% of study population).

The mean ages of the male and female subjects were 15

Fig 2: Comparison of the mean values of %BF as found by

McCarthy et al. with those of the present study.

2

8

(

±2.50) and 14.8 (±2.58) years, respectively (X =0.8095,

df=2, p=0.667).

Table 1: Distribution of subjects according to type of school

and socio-economic class

SEC

Public schools Private schools Total (%)

Upper

22

65

87 (24.2)

Middle

Lower

79

92

23

171 (47.5)

102 (28.3)

Total

180

360 (100)

2

X =52.986, df=2, p<0.0001

SEC= socioeconomic class

Middle SEC formed the predominant group in the study

population (47.5%), with a fair distribution of the mid-

dle class subjects between public and private schools.

McCM= mean %BF for males in McCarthy study; McCF= mean for %

BF females in McCarthy study; STUDM= mean %BF for males in

present study; STUDF= mean %BF for females in present study

The mean triceps SFT values were 8.9mm (±4.7) for

males and 12.9mm (±4.6) for the females (p<0.001).

Mean % BF values were 8.2 ± 4.1% for the males, and

Table 4: Prevalence of overweight and obesity based on the

two methods of assessment

2

0.0 ± 6.8% for the females (p<0.001). Tables 2 and 3

respectively depict the mean triceps SFT and % BF val-

ues for the specific age groups based on gender.

Overweight

OW)

Method used n (%)

Obese

(OB)

n (%)

Combined

(OW+OB)

n (%)

(

Table 2: Mean triceps skin fold thickness ± SD according to

age and gender

%BF

SFT

9 (2.5)

6 (1.7)

3 (0.8)

15 (4.2%)

12 (3.3%)

Males

*SFT (mm)

Mean (± SD)

Females

*SFT (mm)

Mean (± SD)

2

X = 0.765, p= 0.682

%

BF= %body fat; SFT= skin fold thickness

Age

(

yr)

n

t

P

Based on the BIA (% body fat) as criteria, 9 (2.5%) of

1

0

1

2

3

4

5

6

7

8

3

5

5.80 ± 1.00

7.0 0± 1.02

2

8

8.6 0± 2.55

10.00 ± 6.38

10.58 ±3.98

12.38 ± 4.37

13.14 ± 4.08

14.16 ± 5.05

13.44 ± 3.58

12.77 ± 4.12

14.70 ± 5.51

-3.500

-0.644

-2.772

-2.942

-6.731

-3.697

-6.544

-5.088

-3.715

-8.571

0.177

0.555

0.017

0.009

<0.001

0.001

<0.001

0.002

th

the subjects were classified as overweight (85 to <95

percentile) and 6 (1.7%), were classified as obese (≥95

th

13 6.79 ± 1.92

28 7.93 ± 4.09

24 6.83 ± 1.45

24 7.82 ± 4.86

35 7.25 ± 2.85

32 7.57 ± 2.61

16 7.80 ± 2.92

19

18

30

31

30

23

19

percentile). With the use of triceps SFT, Nine (2.5%) of

the subjects were overweight and 3 (0.8%) were obese.

For the obesity prevalence however, higher values were

recorded with the use of BIA (1.7%), and lower (0.8%)

with the use of triceps SFT. There were some overlaps

between the methods, in the classification of subjects as

overweight or obese (BIA and SFT simultaneously

Total 180 8.86 ± 4.66

180 12.95 ± 4.61

<0.001

*

SFT= skin fold thickness; SD= standard deviation

1

82

classified six subjects as overweight and 3 as obese). No

subject was singly classified as overweight or obese by

either the BIA or SFT method alone, without being clas-

sified as either OW or obese by the other method.

triceps skin folds and other measu₃_,r₂e₁s of body fatness

have been reported in the literature.

The mean percentage body fat (%BF) for boys in this

study was found to be highest at 10years (though the

number of subjects in this age group was very small).

The %BF in boys continued to fluctuate down wards

with increasing₈age. This finding is consistent with that

Table 5: Correlation coefficients (r) between SFT and BIA

methods

Gender

N

r

p- value

1

of Sung et al. in Hong Kong, which showed an in-

Males

180

0.69

0.81

<0.001

<0. 001

crease in %BF in boys from age 8, peaked at age 11 and

subsequently maintained linear values (leveled off) from

age 14years.

Females

The correlation between the two methods was better (r value

approaching 1) among the females, compared to the males.

In contrast to the %BF values in boys, the mean %BF

values in girls increased steadily and significantly with

increase in age. This observed increase in %BF with age

in the females is in agreement w₁i₈th the reports by Owa

Table 6: Two-way analysis of variance (ANOVA) for the

assessed parameters in relation to age and gender

2

and Adejiyugbe, Sung et al. as well as that of

Parameter

F-value

for age

F-value for

gender

p-value

for age

p-value

for gender

8

McCarthy et al. Fat gain has been shown to occur in

both boys and girls early in adolescence, but it ceases

and may even temporarily reverse in boys₃, while it con-

SFT

2.86

0.36

93.70

57.61

0.079

0.917

<0.001

2

%

Body fat (BF)

tinues throughout adolescence in girls. This is as a

result of the effect of sex hormones induced sexual di-

morphism. The females lay down fat as a natural part of

the ontogeny of their sexual and reproductive physiol-

ogy, whereas the males₈_,g₁₂a_,i₂n₁ proportionately more mus-

cle mass rather than fat.

Fat mass (FM)

1.33

36.63

0.349

<0.001

*SFT, %BF and fat mass (FM) varied significantly with gender, being

higher in the females than males (p< 0.001). These parameters did not

vary significantly with age (p>0.05).

A significant gender difference in correlation (r) be-

tween the 2 methods of body fat assessment was noted.

The correlation between the methods were generally

better (r value approaching 1) among the females

Discussion

(

ence was statistically significant.

r=0.81), compared to the males (r=0.69). This differ-

The mean triceps SFT values in our subjects were gener-

ally lower among the males and did not vary uniformly

with increasing age in the males, unlike in the females

where the values increased steadily with age. This dif-

ference was statistically significant.

Highest prevalence of overweight/obesity in female sub-

jects was recorded among the age group 18years, fol-

lowed by those aged 15 and 16 years. This finding is

2

consistent with that of Akesode et al. who demonstrated

A similar study by Akesode and Ajibode in Abeokuta,

that highest frequencies of obesity and overweight in

females occurred in age groups 18 and 17 years respec-

tively. In contrast, the highest prevalence of obesity/

overweight among the male subjects in this study was

found in the age groups 13 and 15 years.

South Western Nigeria, also did not show significant

change in the SFT values with age, among their male

subjects, as corroborated by our findings. However, the

mean triceps SFT values obtained in our study were

2

higher than those reported by Akesode and Ajibode

among school students aged 6 to 19 years. The latter

study was carried out about three decades ago, and the

age limit of the study groups differ (6 to 19years against

When both overweight and obese subjects were com-

bined in relation to their socioeconomic classification,

only 26.3% of the overweight/obese subjects belonged

to the upper SEC, 42.1% belonged to the middle socio-

economic class, whereas, the remaining 31.6% came

from the lower SEC. However, this distribution was not

statistically significant. A negative association between

lower SEC and obesity may be related to lack of aware-

ness of the problems of obesity (ignorance) as well as

excessive eating of cheaper, high calorie diet by the

lower SEC groups, who are usually poorer and less well

informed members of the community. However, this

study did not look at the dietary habits and other risk

factors for obesity among the study subjects.

1

0 to 18 years in the present study). The higher SFT

values in the present study may indicate a gradual in-

crease in body fat content in these children over time,

which may connote some secular trend in the body fat

content as a result of changes in lifestyle and dietary

habits.

Conversely, the mean triceps SFT values in the present

th

study were generally lower as compared to the 50 per-

centile values₉obtained from age- and gender matched

1

U.S children. Factors such as demography, lifestyles

and socioeconomic differences may be responsible for

the lower figures seen in our subjects compared to the

‘

western children.’ These differences in anthropometric

dimensions among age groups and gender as well as the

1

83

Conclusion

Conflict of interest: None

Funding: None

Triceps skin fold thickness remains a fair surrogate for

body fat estimation as compared to BIA, in the assess-

ment of overweight and obesity among adolescent sub-

jects. The body fat content, in terms of SFT, % BF and

fat mass varied significantly with gender (but not with

age), being higher in female subjects than the males.

Based on the prevalence of overweight and obesity in

this study, it seemed that obesity is not yet a major

health problem in the study area. However, there is need

for continuous vigilance so that early detection and

prompt intervention can be made.

Limitation of the study

•

The eating habits and activity levels of the subjects

were not assessed in relation to the body fat

The ‘gold standard’ methods of body composition

analysis (hydrodensitometry and/or dual energy X-

ray absorptiometry) could not be used as standards

for comparison due to non availability and lack of

expertise.

References

9

.

Speiser PW, Rudolf MCJ, Anhalt

H, et al. On behalf of the obesity

consensus working group. Consen-

sus statement: Childhood obesity.

J Clin Endocrinol Metab 2005;

90:1871-87.

17. Deurenberg P and Deurenberg-

Yap M. Validation of skinfold

thickness and hand-held imped-

ance measurements for estimation

of body fat percentage among

Singaporean Chinese, Malay and

Indian subjects. Asia Pacific J Clin

Nutr 2002;11:1-7.

18. Sung RYT, Lau P, Yu CW, et al.

Measurement of body fat using leg

to leg bioimpedance. Arch Dis

Child 2001; 85:263-7.

19. Frisancho AR. Triceps skin fold

and upper arm muscle size norms

for assessment of nutritional status.

Am J Clin Nutr 1974; 27:1052-7.

20. Oyedeji GA. Socio-economic and

cultural background of hospital-

ized children in Ilesha. Nig J Paed

1985; 12(4):111-7.

1

.

Chopra M, Galbraith S and Darn-

ton-Hill I. A global response to a

global problem: the epidemic of

overnutrition. World Health Or-

ganization 2002; 80:952-8.

2

3

4

.

Akesode FA and Ajibode HA.

Prevalence of obesity among Nige-

rian school children in Abeokuta.

Soc Sci Med 1983; 17:107-11.

Himes JH, Roche AF and Webb P.

Fat areas as estimates of total body

fat. Am J Clin Nutr 1980; 33:2093-

10. Lukaski HC. Methods for the as-

sessment of human body composi-

tion: traditional and new. Am J

Clin Nutr 1987; 46:537-56.

11. David SG, George AB, Bauer M,

Kaplan K, Gemayel N. Skin fold

thickness measurements in obese

subjects. Am J Clin Nutr 1990;

51:571-7.

12. Poskitt EME. Nutrition in child-

hood. In: Hendrickse RG, Barr

DGD, Matthews TS, editors. Pae-

diatrics in the tropics. Oxford

Blackwell scientific publications

1991; 90-118.

13. Sardinha LB, Going SB, Teixeira

PJ and Lohman TG. Receiver

operating characteristic analysis of

body mass index, triceps skin fold

thickness and arm girth for obesity

screening in children and adoles-

cents. Am J Clin Nutr 70:1090-5.

14. Reilly JJ, Wilson J, Durnin JVGA.

Determination of body composi-

tion from skin fold thickness: a

validation study. Arch Dis Child

1995; 73:305-10.

15. LJerome Brandon. Comparison of

existing skin folds equations for

estimating body fat in African

American and white women. Am J

Clin Nutr 1998; 67:1155-61.

16. William WW, Janice ES, Nancy

FB, et al. Estimating body fat in

African American and white ado-

lescent girls: a comparison of skin

folds thickness equations with a 4-

compartment criterion model. Am

J Clin Nutr 2000; 72:348-54.

1

00.

Rodríguez G, Samper MP, Oliva-

res JL, et al. Skin fold measure-

ment at birth: Sex and anthropo-

metric influence. Arch Dis Child

Fetal Neonatal 2005; 90:F273-5.

Dauncey MJ, Gillian G and Doug-

las G. Assessment of total body fat

in infancy from skin fold thickness

measurements. Arch Dis Child

5

6

.

21. Krebs NF, Primak LE. Normal

childhood nutrition and its disor-

ders. In: Hay WW, Levin MJ,

Sondheimer JM, Deterding RR

(editors). Current Diagnosis &

1

977; 52:223-7.

James ES, Gatchina MF and Karen

AF. Evaluation methods for intrau-

terine growth using neonatal fat

stores instead of birth weight as

outcome measures: Fetal and neo-

natal measurements correlated

with neonatal skin fold thick-

th

Treatment in Paediatrics, 18 ed.

McGraw-Hill companies 2007;

283-314.

22. Owa JA and Adejuyigbe O. Fat

Mass, Fat Mass Percentage, Body

Mass Index and Mid-upper Arm

Circumference in a Healthy Popu-

lation of Nigerian Children. J Trop

Pediatr 1997; 43:13-9.

23. Lobstein T, Baur L, Uauy R. Obe-

sity in children and young people:

a crisis in public health. Obesity

Reviews 2004; 5:4-85

nesses. J Clin Ultrasound 1990;

1

8:9-14.

7

.

Mei Z, Grummer-Strawn LM,

Pietrobelli A, et al. Validity of

body mass index compared with

other body composition screening

indices for the assessment of body

fatness in children and adolescents.

Am J Clin Nutr 2002; 75:978-85.

McCarthy HD, Cole TJ, Fry T,

Jebb SA and Prentice AM. Paedi-

atric Highlight. Body fat reference

curves for children. Int J Obes

8

.

2

006; 30:598-602.