Comparison of factor loadings for anthropometric and physiometric measures among type 2 diabetic males, pre and postmenopausal females in North Indian Punjabi population

doi:10.4236/ns.2010.27093

Paper Menu >>

Journal Menu >>

Vol.2, No.7, 741-747 (2010) Natural Science

http://dx.doi.org/10.4236/ns.2010.27093

Comparison of factor loadings for anthropometric and

physiometric measures among type 2 diabetic males,

pre- and post-menopausal females in North Indian

Punjabi population

Badaruddoza*, Basanti Barna, Amarjit Singh Bhanwer

Department of Human Genetics, Guru Nanak Dev University, Amritsar, India; *Corresponding Author: [email protected]

Received 10 April 2010; revised 18 May 2010; accepted 23 May 2010.

ABSTRACT

Background: The objective of the present study

was to compare the relationship of anthropom-

etric and physiometric characteristics using

principal component factor analysis among

three groups of type 2 diabetic subjects such as

males, pre- and post-menopausal females in

North Indian Punjabi population. Method: A total

of 349 type 2 diabetic subjects (males 157; fe-

males 192; 88 pre- and 104 post-menopausal)

were ascertained for the present study. Different

anthropometric and physiometric measure-

ments were taken. Principal component factor

analysis (PCFA) was applied to identify the

components which are more close to type 2

diabetes among the three groups. Results:

PCFA revealed five uncorrelated components

which explained 79% of the total variance

among diabetic males and six unrelated com-

ponents which explained 78% of the total vari-

ance among pre- and post-menopausal females.

The important two factors could be identified as

central obesity (factor 1) and blood pressure

(factor 2) among these three groups. Conclu-

sion: Higher clustering of obesity and blood

pressures were found in diabetic males as com-

pared to pre- and post-menopausal diabetic

females in North Indian Punjabi population

whereas, waist to hip ratio (WHR) has maximum

loading in post-menopausal females as com-

pared to others.

Keywords: Factor Analysis; Blood Pressure; Type

2 Diabetes; Anthropometry; Punjabi Population

1. INTRODUCTION

The relationships between type 2 diabetes mellitus

(T2DM), anthropometric variables and blood pressures

are statistically complex [1,2]. Strong inter-correlation

between anthropometric and physiometric variables cre-

ates complexities in the analysis and interpretation of

independent associations of these variables with the de-

velopment of type 2 diabetes. Principal Component

Factor Analysis (PCFA) is the technique to reduce a

large number of variables to a smaller number of factors

which are more closely associated with antecedent [2-4].

The objective of the present study using principal com-

ponent factor analysis is to compare the relationship

between anthropometric and physiometric components

with diabetic males, pre- and post-meno-pausal females

in North Indian Punjabi population. The attention has

also been given to find out which factors can be used as

significant predictors of T2DM.

2. MATERIALS AND METHODS

Present study was conducted at the different clinical

centres such as Heart Station and Diabetic Clinic, A.P.

Hospital and Heart Care Centre, Diabetic Clinic and

Research Institute in Amritsar district in the state of

Punjab among Punjabi population. Punjabi population

may be defined as similar genotype groupings and ag-

gregate of similar cultural practices, life style pattern,

social influence and similar ethnic characteristics with

Punjabi language speaking and at least reside in Punjab

for the last 20 years. A total of 349 type 2 diabetic indi-

viduals participated in the baseline examination for the

present study which occurred from October 2008 to

September 2009. Among total individuals 157 and 192

are males and females respectively whereas, among fe-

males 88 and 104 are pre- and post-menopausal. All par-

ticipants provided written informed consent.

2.1. Anthropometric Measurements

Actual age and age on the onset of the disease were re-

corded from the subject’s health card provided by the

Badaruddoza et al. / Natural Science 2 (2010) 741-747

742

clinical centres. Height, weight, circumferences of waist

(WC), hip (HC), arm (AC) and Calf (CC) and four skin

fold thickness (biceps, triceps, sub-scapular and su-

pra-iliac) were taken by female author on each individ-

ual using standard anthropometric techniques and tools

[5,6]. Height and weight were measured to the nearest

0.5 cm and 0.1 kg respectively. Body mass index (BMI)

was calculated for an estimate of overall adiposity using

the formula: BMI = weight (kg)/height (m2). Waist and

Hip circumference (WC and HC) for an estimate of cen-

tral obesity [7] were measured to the nearest 0.5 cm with

a steel tape. Waist to hip ratio (WHR) was calculated

using the standard formula: WHR = WC (cm)/HC (cm).

A Lange skinfold calliper was used to measure the skin-

folds to the nearest 0.2 mm. Two subsequent measure-

ments were taken and averages were used in the analysis.

2.2. Physiometric Measurements

Left arm blood pressures (first phase systolic and fifth

phase diastolic) were taken from each participant with

standard mercury sphygmomanometer after a 5 min rest.

The average of the two subsequent measurements was

used for analysis. All efforts were made to minimize the

factors which affect the blood pressure like anxiety, fear,

stress, laughing and recent activity [8]. Mean arterial

blood pressure (MBP) was calculated for each of the two

readings taken for SBP and DBP by using the formula:

MBP = SBP ＋ (SBP-DBP)/3 [9]. The radial artery at

the wrist was used to count the pulse. It was counted

over one minute. The difference of SBP and DBP was

used as pulse pressure.

3. STATISTICAL ANALYSIS

Descriptive statistics such as means, standard deviations

and coefficient of skewness were calculated for all vari-

ables. All statistical analysis including factor analyses

were conducted by SPSS (Statistical Package for Social

Sciences, version 17.0, SPSS Inc. USA). Each of the

anthropometric and physiometric variables is highly

inter-correlated with each others and creates a methodo-

logical problem for analysis the data. PCFA is used when

variables are highly correlated and this multivariate sta-

tistical tool able to reduce a large number of inter-corre-

lated variables to a smaller number of principal compo-

nents which account for most of the variance in the data

[10,11]. Factor analysis has done on the basis of correla-

tion matrix which helps to understand the amount of

association between the variables, factor extraction and

orthogonal rotation to make factors easily interpretable.

Hence, PCFA was used to extract uncorrelated factors

and varimax rotation, which is an orthogonal rotation in

which the factors are assumed to act independently, was

used in the present study. Factor loadings were equiva-

lent to the correlation coefficients between the variables

(rows) and factors (columns). The final factors pattern

was interpreted using factor loadings of ≥ 0.4. Extracted

factors or number of factors to be retained was based on

eigenvalue criteria ≥ 1.0. Eigenvalues indicate the

amount of variance explained by each factor. A factor

with low eigenvalue has a little contribution to explain

the variances in the variables and may be ignored. The

first and second principal components were identified

through largest and second largest amount of variance in

the data and so on. Communality is the squared multiple

correlation for the variable (as dependent) using as pre-

dictors. Hence, the communality estimates is the meas-

ure the percent of variance in a given variable explained

by all factors. A communality of 0.75 and 0.25 consid-

ered large and low respectively. Low communality indi-

cates variables are negligibly related to each other. The

probability values less than or equal to 0.05 (two-tailed)

were considered to be significant.

4. RESULTS

Table 1 presents the mean, standard deviation (SD) and

skewness of anthropometric and physiometric variables.

All right skewed distributions have converted to a nor-

mal distribution by square root transformation whereas;

reciprocal transformation is used for left skewed distri-

bution among type 2 diabetic males, pre- and post-

menopausal females. The highest mean age for onset of

T2DM was found among post-menopausal females

(52.46 ± 6.22years) and the lowest mean age for the on-

set of disease was found among pre-menopausal females

(36.97 ± 5.96 years) as compared to males (45.19 ± 7.79

years). The other highest mean values of important an-

thropometric indicator such as BMI, hip circumference,

biceps skinfold, triceps skinfold and arm circumference

were found among diabetic post-menopausal females as

compared to males and pre-menopausal females. The

diabetic males have higher mean values for WHR and

waist circumference. The physiometric variables such as

SBP, DBP, pulse rate and pulse pressure have not shown

any specific trend among three groups of diabetic sub-

jects. Bivariate correlations of the traits were examined

among type 2 diabetic males, pre- and post-menopausal

females and are presented in Tables 2 to 4. Waist cir-

cumference, hip circumference, biceps skinfold, triceps

skinfold, arm circumference and calf circumference with

Weight and BMI; hip circumference, biceps skinfold,

triceps skinfold, arm circumference and calf circumfer-

ence with waist circumference have been found signifi-

Badaruddoza et al. / Natural Science 2 (2010) 741-747

743

Table 1. Descriptive Statistics of Anthropometric and Physiometric variables among type 2 diabetic male, pre- and post-menopausal

females in the present study population (n = 349).

FEMALE (N = 192)

MALE (N = 157) PRE-MENOPAUSAL (N = 88) POST-MENOPAUSAL (N = 104)

VARIABLES

MEAN SD SKEW-

NESS MEAN SD SKEWNESSMEAN SD

SKEW-

NESS

Age (yrs) 53.994 8.093 0.466 46.89 7.85 −0.009 58.39 7.49 0.45

Onset age (yrs) 45.197 7.792 0.836 36.97 5.96 −0.28 52.46 6.22 1.02

Height (cm) 169.131 8.582 −5.437 154.95 5.599 0.19 155.80 6.95 1.57

Weight (Kg) 76.697 6.976 0.533 63.52 7.55 0.36 67.70 8.20 0.43

BMI(kg/m2) 26.555 3.870 0.370 26.25 3.34 0.21 28.00 5.60 0.62

WHR 0.941 0.0732 0.586 0.88 0.07 0.75 0.87 0.07 −0.17

Waist circumference(cm) 96.014 8.512 0.267 87.23 9.90 0.597 90.26 9.32 0.49

Hip circumference (cm) 101.94 9.056 0.242 99.18 9.59 0.68 103.07 10.97 0.52

Biceps skinfold (mm) 10.806 5.968 1.745 13.11 5.05 0.395 13.26 5.58 1.03

Triceps Skinfold (mm) 14.005 5.053 3.201 17.12 5.19 0.299 17.54 6.76 0.58

Subscapular skinfold (mm) ------- ------- -------- 24.94 4.88 −0.39 26.66 5.399 0.04

Supra-iliac skinfold (mm) ------- -------- -------- 21.37 3.82 −0.14 24.37 4.43 0.10

Arm Circumference (cm) 27.838 3.192 0.282 27.27 3.39 0.18 28.20 3.90 0.87

Calf circumference (cm) 33.548 3.868 0.502 31.97 3.62 0.24 32.34 3.67 0.04

Systolic blood pressure

(mmHg) 124.656 9.266 1.802 121.72 8.82 0.86 124.95 9.49 0.798

Diastolic blood pressure

(mmHg) 80.083 10.475 1.090 79.55 9.93 −0.12 79.70 9.81 1.04

Mean Blood Pressure

(mmHg) 95.924 9.384 2.081 93.56 9.67 0.57 96.14 8.27 2.695

Pulse Rate 83.229 8.471 0.251 85.10 8.85 0.12 83.98 8.24 0.15

Pulse Pressure 44.790 7.232 1.660 42.27 7.58 0.77 46.41 7.08 1.296

Significant at least at p ≤ 0.05

SD = Standard Deviation

cantly associated at least 5% level (p < 0.05) among all

diabetic males, pre- and post-menopausal females. SBP

and DBP were found to be significantly associated with

weight, BMI, waist circumference, hip circumference,

biceps skinfold and triceps skinfold at least 5% level (p

< 0.05) among diabetic pre- and post-menopausal fe-

males. Whereas, WHR was found significantly associ-

ated (p < 0.05) with other anthropometric variables

among only type 2 diabetic males.

The comparison of factor loading pattern of six factors

(components) is presented in Table 5 among diabetic

males, pre- and post-menopausal females. Only variables

with factor loading greater than or equal to 0.4 were

considered for present interpretation among three groups.

After Varimax rotation, weight, BMI, waist circumfer-

ence, hip circumference, arm and calf circumferences

are relatively large and positively loaded (> 0.7) on fac-

tor 1 among males, pre- and post-menopausal females.

However, on factor 1 highest loading was found in

weight (0.944) for males, hip circumference (0.883) for

pre-menopausal females and BMI (0.940) for post-

menopausal females. The physiometric variables such as

SBP, DBP and pulse pressure are grouped together and

loaded positively on factor 2 among three groups.

Maximum loading has found for SBP (> 0.90) on factor

2 for all three groups, whereas loading of DBP for this

factor is just above the cut-off value (0.4) for post-

menopausal females. Both mean age (actual age and

onset age of type 2 diabetes) have found maximum

loading (0.952 and 0.940) on factor 3 among males,

whereas, skinfold thickness (biceps, triceps, sub-scapular

and supra-iliac) grouped together and loaded signifi-

cantly among pre- and post-menopausal females. Only

triceps skinfold and biceps skinfold among males have

positive loading on factor 4, whereas, actual mean age

and the mean age of onset of the disease have grouped

for higher positive loading among pre- and post-meno-

pausal females. Only WHR has positive loading on fac-

tor 5 among males, whereas, WHR and waist circum-

ference for pre-menopausal females and WHR, height

and waist circumference for post-menopausal females

have positive loading on this factor. However WHR has

maximum positive loading (≥ 0.80) among pre- and

post-menopausal females but for males it is just above

cut-off value (0.55). Only Height and pulse rate have

positive loading on factor 6 among pre- and post-meno-

pausal female whereas, all variables are extracted on this

factor among males. The five factors explained 79% of

the total variance among males in which the first two

factors cumulatively explained 54% of the total variance.

Whereas, the six factors explained 78% of the total

variance among pre- and post-menopausal females in

which first two factors cumulatively explained 48% and

49% of the total variance respectively. The eigenvalue of

the first two factors have also been seen maximum

among males, pre- and post-menopausal females. The

Badaruddoza et al. / Natural Science 2 (2010) 741-747

744

Table 2. Inter-correlation matrix of selected anthropometric variables among males of Type 2 Diabetes Mellitus (T2DM).

VARI-

ABLES

Age

(yrs)

(cm)

(Kg)

BMI

(kg/m2) WHR WC

(cm)

BSkn

(mm)

T Skn

(mm)

(cm)

SBP

(mmHg)

DBP

(mmHg)

MBP

(mmHg) PR PP

Age (yrs) .81* −.05 −.06 −.007 .03 .05.03−.003 −.08 −.04 −.09 −.05 −.05 −.10 −.12 .02

Onset age

(yrs) −.03 .09 .16 .09 .17.16.09.01.10.04−.08 −.005 −0.12 −.13 −.06

Height (cm) .09 −.15 −.12 −.06 −.007 −.17 −.27 −.02 −.03 −.007 −.07 −0.04 −.04 .03

Weight (Kg) .91* .48* .87*.84*.58*.40*.73*.73*.08 .18 0.11 −.11 .05

BMI(kg/m2) .50* .85*.80*.57*.46*.73*.73*.07 .18 0.11 −.12 .03

WHR .68*.20*.33*.40*.42*.34*.03 .12 0.04 −.19 −.001

Waist cir-

cumference

(cm)

.82*.62*.45*.75*.70*.01 .13 0.05 −.20* −.004

Hip cir-

cumference

(cm)

.60*.32*.70*.70*−.006.09 0.05 −.09 −.009

Biceps skin-

fold (mm) .72*.64*.50*.003 .07 −0.02 −.13 .011

Triceps

Skinfold

(mm)

.46*.32*−.04 .04 −0.04 −.13−.03

Arm Cir-

cumference

(cm)

.68*.03 .13 0.06 −.13 −.005

Calf cir-

cumference

(cm)

.001 .04 0.01 −.10 −.001

Systolic

blood pres-

sure

(mmHg)

.73 0.86* .08.90*

Diastolic

blood pres-

sure

(mmHg)

0.80* −.004 .40*

Mean Blood

Pressure

(mmHg)

.053.67*

Pulse Rate .09*

Pulse Pres-

sure

Significant at least at p ≤ 0.05; OA = Onset age, yrs = years, Ht = Height, Wt = Weight, BMI = Body Mass Index, WHR = Waist Hip Ratio, WC =

Waist Circumference, HC = Hip Circumference, BS = Biceps skinfold, TS = Triceps skinfold, SS = Sub-scapular skinfold, SiS = Supra-iliac skinfold,

AC = Arm Circumference, CC = Calf Circumference, SBP = Systolic Blood Pressure, DBP = Diastollic Blood Pressure, MBP = Mean Blood Pressure,

PR = Pulse Rate, PP = Pulse Pressure.

common greater communality estimates (> 0.70) have

found on age, onset age of disease, weight, BMI, waist

circumference, hip circumference, triceps skinfold, SBP

among three groups. WHR has maximum communality

estimates among pre- and post-menopausal females.

5. DISCUSSION

The present quantitative analysis have shown that which

of the anthropometric and physiometric traits (BMI,

WHR, weight, waist circumference, hip circumference,

skinfolds, SBP, DBP, and pulse pressure) are more

closely associated and act as a good predictors for fur-

ther risk among three groups of T2DM individuals such

as males, pre- and post-menopausal females in North

Indian Punjabi population. The present study also pro-

vides through PCFA among three groups that which of

the traits would require more attention to clinicians for

raised risk of T2DM.

The many previous studies suggested that obesity, over-

weight, glucose intolerance, hypertension and elevated

blood pressures are closely associated with T2DM [12-

18].The present analysis showed a common association of

BMI, WHR, waist circumference, hip circumference and

subcutaneous fat with T2DM incidence among males,

pre-and post-menopausal females.

Badaruddoza et al. / Natural Science 2 (2010) 741-747

745

Table 3. Inter-correlation matrix of selected variables among pre-menopausal females of Type 2 Diabetes Mellitus (T2DM).

VARIABLES Age

(yrs)

(cm)

(Kg)

BMI

(kg/m2) WHR WC

(cm)

BSkn

(mm)

T Skn

(mm)

SiS

(mm)

(cm)

SBP

(mmHg)

DBP

(mmHg)

MBP

(mmHg) PR PP

Age (yrs) .54* −.08 −.05 −.14 .10 .08.03.05.07.06.04−.09 −.14.13 −.04 .02 −.06.22

OA (yrs) −.03 −.03 −.06 −.05 −.03.02 −.02.08 −.006 .19−.04.06.17 .10 .10 .05 .19

Ht (cm) .30−.08 .10 .06.0097.013.04.16.06.15.09.02 .03 .03 .22 .006

Wt (Kg) .83* .15 .80*.82*.53*.42.36*.50*.70*.51*.17 .11 .19 −.06.13

BMI(kg/m2) .08 .68*.73*.58*.45.34*.43*.68*.61*.16 .09 .18 −.13.15

WHR .51*−.16 .009−.12.10.14 −.03 −.02−.05 −.02 .007 .04 −.08

WC(cm) .75*.48*.30*.22*.42*.58*.46*.18 .20* .24* −.08.08

HC (cm) .56*.45*.21*.40*.69*.53*.24*.24* .28* −.14.14

BS (mm) .80*.45*.46*.66*.49*.15 .03 .099 −.30*.19

T S (mm) .38*.36*.54*.38*.13 −.07 .04 −.25*.25

SS (mm) .66*.24*.16−.05 −.05 −.06 −.19 −.03

SiS (mm) .35*.31*.12 .09 .10 −.05.10

AC (cm) .66*.008−.03 .03 −.13.02

CC (cm) .008.04 .05 −.14 −.008

SBP (mmHg) .71* .91* −.09.87*

DBP

(mmHg) .85* .09 .29*

MBP

(mmHg) −.009.65*

P R −.17

P P

Significant at least at p ≤ 0.05; OA = Onset age, yrs = years, Ht = Height, Wt = Weight, BMI = Body Mass Index, WHR = Waist Hip Ratio, WC =

Waist Circumference, HC = Hip Circumference, BS = Biceps skinfold, TS = Triceps skinfold, SS = Sub-scapular skinfold, SiS = Supra-iliac skinfold,

AC = Arm Circumference, CC = Calf Circumference, SBP = Systolic Blood Pressure, DBP = Diastolic Blood Pressure, MBP = Mean Blood Pressure,

PR = Pulse Rate, PP = Pulse Pressure.

Table 4. Inter-correlation matrix of variables among post-menopausal females of Type 2 Diabetes Mellitus (T2DM).

VARIABLES Age

(yrs)

(cm)

(Kg)

BMI

(kg/m2) WHR WC

(cm)

BSkn

(mm)

T Skn

(mm)

SiS

(mm)

(cm)

SBP

(mmHg)

DBP

(mmHg)

MBP

(mmHg) PR PP

Age (yrs) .73* −.12 .007 .06 −.01 .17.15−.08 −.02 −.20* −.07 −.04.03.06 −.07 .14 .15 .09

OA (yrs) −.08 −.03 .006 .04 .07.03−.04 .04 −.09 −.05 −.08 −.02.06 .099 .16 .15 .005

Ht (cm) .19 −.23* .17 .06−.05.04 .10 .03−.08 −.09 −.02 −.15 −.07 −.11 −.03 −.19

Wt (Kg) .90* .18 .80*.83*.53*.40*.54*.52*.76*.74*.13 .18 .10 .05 −.005

BMI(kg/m2) .07 .77*.85*.50*.35*.52*.55*.80*.74*.20 .20* .16 .08 .09

WHR .47*−.08.10 .10 .10.03.01−.02.002−.09 .10 −.03.05

WC(cm) .796*.53*.42*.495*.46*.65*.62*.19 .083 .20 .10 .15

HC (cm) .50*1.0.55*.53*.77*.75*.16 .12 .14 .05 .09

BS (mm) .78*.50*.47*.44*.44*.06 .05 .009 .013.06

T S (mm) .44*.37*.33*.33*.03 −.05 .01 .013.10

SS (mm) .60*.49*.43*−.09 −.02 −.15 −.002 −.12

SiS (mm) .48*.49*.18 .12 .03 −.03.12

AC (cm) .69*.12 .21* .12 .05 .007

CC (cm) .20*.16 .12 .06 .14

SBP (mmHg) .59* .72* .13 .86*

DBP (mmHg) .56* .13 .16

MBP (mmHg) .003.53*

PR .06

Significant at least at p ≤0.05; OA = Onset age, yrs = years, Ht = Height, Wt = Weight, BMI = Body Mass Index, WHR = Waist Hip Ratio, WC =

Waist Circumference, HC = Hip Circumference, BS = Biceps skinfold, TS = Triceps skinfold, SS = Sub-scapular skinfold, SiS = Supra-iliac skinfold,

AC = Arm Circumference, CC = Calf Circumference, SBP = Systolic Blood Pressure, DBP = Diastolic Blood Pressure, MBP = Mean Blood Pressure,

PR = Pulse Rate, PP = Pulse Pressure.

Badaruddoza et al. / Natural Science 2 (2010) 741-747

746

Table 5. Comparison of factor loadings by principal component analysis with Varimax rotation and communalities of the risk factors

among type 2 diabetic male, pre-menopausal and post-menopausal females (n = 349).

FACTOR 1 FACTOR 2 FACTOR 3 FACTOR 4 FACTOR 5 FACTOR 6

COMMU-

NALITY

ESTIMATES

VARIABLES

M PRF POF M PRF POFMPRF POFMPRF POFMPRFPOF M PRF POF MPR

Age (yrs) −.046 −.083.042 −.025 .038 .041.952.057 −.072.008 .860 .930 .053 .136−.021 ---- −.134 .010 .912.787.874

Onset age (yrs) .114 .013−.035 −.054 .105 .048.940.021.021.015.868.873.054 −.097.006 ---- .111 .168 .902.787.794

Height (cm) .056 .041−.099 −.035 .030 −.217 −.080 .303.104−.827 −.084 −.196.118.007.652 ---- .726 .404 .709 .629 .695

Weight (Kg) .944 .833.903 .091 .117 .017−.147 .260.160−.043 −.042 −.029.102.255.241 ---- .173 .129 .912.872.917

BMI(kg/m2) .905 .841 .940 .088 .101 .112 .054.171 .111.154−.108.054 .083 .154−.057 ---- −.118 −.026 .861.796 .916

WHR .432 −.029.072 .061 −.060 .092.007.107.034.278 .025 .080 .550 .913.831 ---- .006 −.296 .569 .849.799

Waist circumference(cm) .900 .735 .796 .033 .133 .139.078 .087.215.117 .039 .182 .253.620.412 ---- −.030 −.072 .894 .953.908

Hip circumference (cm) .911 .883 .903 .002 .202 .048 .088.060 .184−.031 .039.103 −.088.019 −.064 ---- −.037 .027 .846.827.867

Biceps skinfold (mm) .663 .633.419 −.020 .067 .025 −.002 .527.794.473.043−.046 .106−.125 .064 ---- −.271 .045 .675.775.814

Triceps Skinfold (mm) .429 .511 .245 −.051 .033 .021 −.110.540 .880.680.114.045 .245−.267 .103 ---- −.257 .036 .722.704.848

Subscapular skinfold (mm) -- .137.571 -- −.071 −.197-- .871.469---−.030 −.179-- .103.005 ---- .032 −.037 ---- .795.619

Supra-iliac skinfold (mm) -- .346.577 -- .079 .091 -- .684.413 ---.125−.104-- .202−.143 ---- .105 −.107 ----- .661 .554

Arm Circumference (cm) .838 .863.871 .030 −.071 .037−.006 .175.077.115−.041 −.078 .113−.096 −.053 ---- .014 .070 .729.792.780

Calf circumference (cm) .837 .763.811 −.025 −.046 .103−.062 .048.141−.011−.021 −.006 .002−.135 −.053 ---- .031 .057 .705.606.694

Systolic blood pressure

(mmHg) .006 .056.102 .970 .975 .957−.028 .050.031−.013 .103 −.002.036−.034 −.057 ---- −.066 .101 .944.972.941

Diastolic blood pressure

(mmHg) .097 .076.190 .823 .817 .579−.009 −.117 −.163.061 −.063 −.135.086.082 −.084 ---- .178 .539 .698.729.714

Mean Blood Pressure

(mmHg) .035 .104.102 .934 .965 .833−.082 −.043 −.115 −.007 −.014.103 −.003.046.112 ---- .046 .108 .880.948.752

Pulse Rate −.049 −.084.028 .060 −.032 .035 −.097 −.254 .062.088.065.235−.862 −.005 −.059 ---- .752 .642 .769.642.477

Pulse Pressure .007 .015−.032 .814 .767 .846.020.165.178−.029 .207.066−.084 −040 −.060 ---- −.209 −.215 .671 .721.802

Eigenvalue 6.016 5.917 6.408 3.190 3.159 2.876 1.8371.762 1.8041.318 1.574 1.444 1.034 1.2281.201 --- 1.204 1.034 ---------

Variance Explained(%) 35.389 31.140 33.726 18.767 16.626 15.135 10.8069.2749.4967.7568.2867.601 6.0846.4646.322 --- 6.336 5.440 -----------

Cumulative Variance(%) 35.389 31.140 33.726 54.157 47.767 48.860 64.96257.04

1 58.35672.71865.32765.95778.80171.79172.28 --- 78.127 77.719 ---------

Factor Loadings ≥ 0.4; M = Males, PMF = Pre-menopausal females, POF = Post-menopausal females.

PCFA is applied to identify the significant association

with T2DM among three groups. As far as concern in the

North Indian Punjabi population, very little information

[15,19-21] to identify the underlying factors/components

of the T2DM are available. In this consideration the

present work has been undertaken among the males and

females (pre- and post-menopausal). PCFA have identi-

fied five factors with 79% explained variance among

male diabetic subjects and six factors with 78% ex-

plained variance among pre- and post-menopausal dia-

betic female subjects.

It is important to note that neither of the anthropomet-

ric and physiometric variables equally loaded on all five

or six components. Factor 1 is the most diverse among

three groups. It could be identified as weight, BMI, waist

circumference for males; hip circumference, BMI,

weight for pre-menopausal females and BMI, weight,

hip circumference for post-menopausal females. How-

ever weight for males, hip circumference for pre-meno-

pausal and BMI for post- menopausal females are heav-

ily loaded. The second factor could be identified as SBP

and DBP for males and pre-menopausal females and

DBP for post-menopausal females. This component is

most clearly and heavily loaded. Therefore, among dia-

betic individuals, males and pre-menopausal females

were very closely associated with SBP and DBP whereas,

post-menopausal diabetic females were more concerned

with DBP only. Among male diabetic subjects factor

three was grouped with actual age and age of onset of

the disease, whereas, subcutaneous fat was identified as

factor three. Factor four could be identified as subcuta-

neous fat for males, whereas actual age and age of onset

of the disease were grouped together for factor four.

Factor five could be identified as WHR for the three

groups. However, WHR is heavily loaded for pre- and

post-menopausal females. Pulse rate could be identified

as factor six for both pre- and post-menopausal females,

whereas, no sixth factor is identified for males.

The present factor analysis confirmed that cluster of at

least three variables such as, weight, BMI, waist cir-

cumference which have identified as factor one ex-

plained 35%, 31% and 34% of the total variance among

diabetic males, pre- and post-menopausal females re-

spectively. Therefore, the cluster of weight, BMI and

waist circumference could be classified as central obe-

sity and this cluster is equally associated with diabetic

males and post-menopausal females. Furthermore, in the

present study, the second factor, that is blood pressures

explained 19%, 17% and 15% of the total variance

among diabetic males, pre- and post-menopausal fe-

males. The blood pressures (SBP and DBP) were posi-

tively and significantly associated with diabetic males.

Therefore the above two types factors such as, central

obesity and blood pressures are more predispose among

diabetic males as compared to females. PCFA also con-

firmed that WHR and pulse pressure are significantly

Badaruddoza et al. / Natural Science 2 (2010) 741-747

747

associated with diabetic pre- and post-menopausal fe-

males as compared to diabetic males. Therefore, it is

very difficult to single out of the particular variable

which is more associated with male or pre- and post-

menopausal females due to the fact that many overlap-

ping variables have found as more than one factor

among all the three groups. Further, research with PCFA

is required on other Indian ethnic groups to compare the

present trend of the study.

6. ACKNOWLEDGEMENTS

The authors are greatful to Dr Rohit Kapoor; Dr. A. P. Singh and Dr.

Puneet Arora for their co-operation during the data collection. This

work is financially supported by University Grants Commission, New

Delhi [DRS I (UGC-SAP)].

REFERENCES

[1] Meigs, J.B. (2000) Invited commentary: Insulin resistance

syndrome? Syndrome X? Multiple metabolic syndrome?

A syndrome at all? Factor analysis reveals patterns in the

fabric of correlated metabolic risk factors. American

Journal of Epidemiology, 152(10), 908-911.

[2] Hanley, A.J., Karter, A.J. and Festa, A. (2002) Factor

analysis of metabolic syndrome using directly measured

insulin sensitivity: The insulin resistance atherosclerosis

study. Diabetes, 51(8), 2642-2647.

[3] Ghose, A., Bose, K. and Das Chaudhari, A.B. (2003)

Association of food pattern, central obesity measures and

metabolic risk factors for coronary heart disease (CHD)

in middle aged Bengalee Hindu men, Calcutta, India.

Asia Pacific Journal of Clinical Nutrition, 12(2), 166-

171.

[4] Ghose, A. (2005) Factor analysis of metabolic syndrome

among the middle-aged Bengalee Hindu men of Calcutta,

India. Diabetes/Metabolism Research and Reviews, 21(1),

58-64.

[5] Singh, I.P. and Bhasin, M.K. (1968) Anthropometry,

Kamla Raj Enterprises, Delhi.

[6] Weiner, J.S. and Lourie, J.A. (1981) Practical Human

Biology, Academic Press, London.

[7] Lemieux, S., Prud’homme, D., Bouchard, C., Tremblay,

A. and Despres, J.P. (1996) A single threshold value of

waist girth identifies normal weight and overweight sub-

jects with excess visceral adipose tissue. American Jour-

nal of Clinical Nutrition, 64(5), 685-693.

[8] Badaruddoza and Afzal, M. (1999) Age specific differ-

ences in blood pressure among inbred and non-inbred

North Indian children. Journal of Biosciences, 24(2),

177-184.

[9] Perusse, L., Rice, T., Bouchard, C., Vogler, G.P. and Rao,

D.C. (1989) Cardiovascular risk factors in the French

Canadian population: Resolution of genetic and familial

environmental effects on blood pressure by using exten-

sive information on environmental correlates. American

Journal of Human Genetics, 45(2), 240-251.

[10] Truxillo, C. (2003) Multivariate statistical methods: Prac-

tical research applications course notes. SAS Institute,

USA.

[11] Wu, C.Z., Lin, J.D., Li, J.C., Hsiao, F.C., Hsich, C.H.,

Kuo, S.W., Hung, Y.J., Lu, C.H., He, C.T. and Pei, D.

(2008) Factor analysis of metabolic syndrome using di-

rect measurement of insulin resistance in Chinese with

different degrees of glucose tolerance. Indian Journal of

Medical Research, 127(1), 336-343.

[12] Singh, R., Shaw, J. and Zimmet, P. (2004) Epidemiology

of childhood type 2 diabetes in the developing world.

Pediatric Diabetes, 5(3), 154-168.

[13] Li, J.K., Ng, M.C., So, W.Y., Chiu, C.K., Ozaki, R., Tong,

P.C., Cockram, C.S. and Chan, J.C. (2006) Phenotypic and

genetic clustering of diabetes and metabolic syndrome in

Chinese families with type 2 diabetes mellitus. Diabe-

tes/Metabolism Research and Reviews, 22(6), 46-52.

[14] Comuzzie, A.G., Tejero, M.E., Funahash, T., Martin, L.J.,

Kisseban, A., Takahashi, S., Tanaka, S., Rainwater, D.L.,

Matsuzawa, Y., MacCluer, J.W. and Blangero, J. (2007)

The genes influencing adiponectin levels also influence

risk factors for metabolic syndrome and type 2 diabetes.

Human Biology, 79(2), 191-200.

[15] Rai, E., Sharma, S., Koul, A., Bhat, A.K., Bhanwer, A.J.

and Bamezai, R.N. (2007) Interaction between the

UCP2-866G/A, mtDNA 10398G/A and PGC1alpha

p.Thr394Thr and p.Gly482Ser polymorphisms in type 2

diabetes in North Indian population. Human Genetics,

122(5), 535-540.

[16] Franceschini, N., Almasy, L., MacCluer, J.W., Goring,

H.H., Cole, S.A., Diego, V.P., Laston, S., Howard, B.V.,

Lee, E.T., Best, L.G., Fabsitz, R.R. and North, K.E. (2008)

Diabetes-specific genetic effects on obesity traits in

American Indian populations: The strong heart family

study. BMC Medical Genetics, 9(1), 90-96.

[17] van’t Riet, E., Alsseme, M., Nijpels, G. and Dekker, J.M.

(2008) Estimating the individual risk of diabetes: Not on

the grounds of overweight only. Nederlands Tijdschr

Geneeskd, 152(44), 2385-2388.

[18] Dunkley, A.J., Taub, N.A., Davis, M.J., Stone, M.A.

and Khunti, K.C. (2009) Is having a family history of

type 2 diabetes or cardiovascular disease a predictive

factor for metabolic syndrome? Primary Care Diabetes,

3(1), 49-56.

[19] Matharoo, K., Kumar, A., Randhawa, N.K., Arora, P. and

Bhanwer, A.J.S. (2006) Angiotensin converting enzyme

gene insertion/deletion polymorphism in type 2 diabetes

(T2DM) patients from Punjab. Journal of Punjab Acad-

emy of Sciences, 3(3), 7-12.

[20] Sanghera, D.K., Bhatti, J.S., Bhatti, G.K., Ralhan, S.K.,

Wander, G.S., Singh, J.R., Bunker, C.H., Weeks, D.E.,

Kamboh, M.I. and Ferrell, R.E. (2006) The Khatri Sikh

Diabetes Study (SDS): Study design, methodology, sam-

ple collection, and initial results. Human Biology, 78(1),

43-63.

[21] Sanghera, D.K., Ortega, L., Han, S., Singh, J., Ralhan,

S.K., Wander, G.S., Mehra, N.K., Mulvihill, J.J., Ferrell

R.E., Nath, S.K. and Kamboh, M.I. (2008) Impact of nine

common type 2 diabetes risk polymorphisms in Asian

Indian Sikhs: PPARG2 (Pro12Ala), IGF2BP2, TCF7L2

and FTO variants confer a significant risk. BMC Medical

Genetics, 9(1), 59-67.