Prevalence of obesity and associated risk factors in chinese pre school children aged 6 to 72 months old in singapore 3

Figure 10 shows the distributions of skinfold Triceps measurements according to different age groups for all children n=923.. Similarly, Figure 12 shows distributions of skinfold measure

58

CHAPTER 3

3 RESULTS

3.1 Characteristics of the study population

Table 3 shows the anthropometric and demographic characteristics of the

participants Boys were heavier, taller and had higher BMI than girls Fathers who

had university (and above) education constituted 33.0% and the percentage of

mothers with university education (and above) was 28.0% About half of the

families (45.2%) had total combined monthly income of ≥S$ 5,000

Table 4 shows birth weight and behavioral characteristics of the

participants The mean time spent in total on sedentary activities was about five

times greater than that of total outdoors activities [3.4±2.5(SD) vs 0.7±0.9(SD)

hours per day, p=<0.001] Among sedentary activities, watching television time

was longer than time spent on computer/ handheld video games [1.9±1.5(SD) vs

0.3±0.8(SD) hours per day, p=<0.001] and reading, drawing, coloring time

[1.9±1.5(SD) vs 1.2±1.3(SD) hours per day, p=< 0.001]

Majority of mothers (77.8%) breastfed their children and the most

common type of breastfeeding was partial breastfeeding (77.6% with 1 to 6

months breastfeeding and 60.2% with 6 to 12 months breastfeeding)

59

Table 1 Anthropometric and demographic characteristics of the participants

Secondary/ ‘O’/ ‘N’ level 856(29.5) 437(28.9) 419(30.3)

‘A’ level/ Polytechnic/ Diploma/ ITE/

Certificate

794(27.4) 432(28.6) 362(26.1) University (degree and above) 957(33.0) 488(32.3) 469(33.9)

None/Primary 199(6.8) 106(7.0) 93(6.7)

Secondary/ ‘O’/ ‘N’ level 1032(35.4) 543(35.6) 489(35.1)

‘A’ level/ Polytechnic/ Diploma/ ITE/

Certificate

869(29.8) 456(29.9) 413(29.6) University (degree and above) 818(28.0) 419(27.5) 399(28.6)

60

Table 2 Birth weight and behavioral characteristics of the participants

n (%) Mean (SD) n (%) Mean (SD) n (%) Mean (SD)

(458.8)

3137.8 (466.0)

3036.7 (444.9)

Outdoor leisure activities

Playing computer/ television/

hand held video games

(hour/day)

0.3(0.8) 0.4(1.0) 0.2(0.5)

Reading, drawing and

coloring activities (hour/day)

61

3.2 Participants and Non-participants

Table 5 shows the comparison between participants and non-participants

The participants were mostly from the Jurong West area (42.5%) and the

participation of boys (52.2%) was slightly more than that of girls (47.8%)

However, the numbers of participants from different age groups were similar

except for the 6–11.9 months age group

Table 6 highlights age differences between participants and

non-participants The mean age of participants and non-participants were similar

(40.48 vs 39.78 months) with standard deviations of 18.57 and 18.48

respectively There was no significant difference in ages between participants and

non-participants

62

Table 3 Comparison between participants and non-participants

Non-Participants (n=1,155) N Percent

Participants (n=3,009) p

Table 4 Age difference between participants and non-participants

Participation N Mean S.D Median Range Skewness Kurtosis P

No 1,043 39.78 18.48 40.00 66.00 -0.05 -1.19

Yes 3,009 40.48 18.57 41.20 65.92 -0.07 -1.14

63

3.3 Distributions of Height, Weight, Body Mass Index and

Skinfold Measurements

3.3.1 Distributions of Height (centimeters)

Among the 3,009 participated children of aged 6–72 months old, 2,964 children (98.5%) were involved in the analysis of distributions of height, weight

and body mass index

Figure 1 shows the distributions of height according to different age

groups It shows that height was normally distributed among different age groups

(kurtosis=0.25 to 3.84 and skewness=-0.49 to 0.29) It also highlighted that there

were significant differences in height (mean) among different age groups

(p=<0.001)

Figure 2 shows the distributions of height according to different age

groups for boys The results were normally distributed except for the 36.0–47.9 months age group (kurtosis=1.50, skewness=-0.46 to 0.66) Significant

differences in height (mean) were also seen between the different age groups

(p=<0.001)

Figure 3 highlights the distributions of height according to different age

groups for girls Although they had kurtosis of 4.36, 1.24, 3.54 and 6.00 for age

groups of 6-–11.9, 12–23.9, 24–35.9 and 60–72 months, respectively, the distributions were normal

64

There was a significant difference in height (mean) between boys and girls

(p=0.01) and as the age increased, the mean value of height also increased, as

seen in all these figures

65 Figure 1 Distribution of Height (centimeters) by age groups

66 Figure 2 Distribution of Height (centimeters) of boys by age groups

67 Figure 3 Distribution of Height (centimeters) of girls by age groups

68

3.3.2 Distributions of Weight (kilograms)

Figure 4 demonstrates the distributions of weight according to different

age groups The weight of age groups 6–11.9 and 12–23.9 months were normally distributed (p=0.76 and p=0.41, respectively), but the distributions of the other

age groups were skewed towards higher values (skewness ranged from 0.66 to

1.97) and they had kurtosis which ranged from 1.70 to 9.39

The distributions of weight according to different age groups for boys are

shown in Figure 5 Among six age groups, three age groups (6.0–11.9, 12.0–23.9 and 24.0–35.9 months) showed a normal distribution (p=0.43, p=0.90 and p=0.41, respectively) However, the distributions of the other three age groups (36–47.9,

48–59.9, 60–72 months) were skewed towards higher values of 1.71, 1.63 and 1.21 and kurtosis of 5.19, 6.15 and 2.09 respectively

As shown in Figure 6, similarly skewed distributions towards higher

values were also seen in girls of the same age groups (36–47.9, 48–59.9, 60–72 months) with the skewness of 2.29, 1.62 and 2.35 and kurtosis of 15.53, 5.27 and

13.72, respectively On the other hand, the age groups of 6–11.9, 12–23.9 and 24–

35.9 months showed normal distribution (p=0.54, p=0.60 and p=0.12,

respectively)

Similarly, significant differences of weight (mean) were found between

boys and girls with p-value of <0.001 and as in height, the mean value of weight

increased as the age increased

69 Figure 4 Distribution of Weight (kilograms) by age groups

70 Figure 5 Distribution of Weight (kilograms) of boys by age groups

71 Figure 6 Distribution of Weight (kilograms) of girls by age groups

72

3.3.3 Distributions of Body Mass Index (BMI)

The distributions of BMI according to different age groups for both

genders are shown in Figure 7 Only the 12.0–23.9 months age group demonstrated that BMI was normally distributed (p=0.20) The other age groups

showed distributions with skewness towards higher values (skewness ranged from

0.53 to 3.35 and kurtosis ranged from 4.28 to 28.35)

In addition, as seen in Figure 8, the distributions skewed towards higher

values for age groups 6.0–11.9, 36.0–47.9, 48.0–59.9 and 60.0–72 months in boys, with skewness ranging from 1.24 to 2.5 and kurtosis from 2.50 to 7.77

Nevertheless, two age groups (12.0–23.9 and 24.0–35.9 months) described the normal distribution pattern with p-value 0.73 and 0.15, respectively

According to Figure 9, girls of all age groups in our study showed

distributions of skewness towards higher values with skewness ranging from 1.14

to 4.61 and kurtosis from 2.41 to 46.33

Overall, the majority of the distributions of BMI skewed towards higher

values for both genders except from only two age groups (12.0–23.9 and 24.0–

35.9 months of boys) Moreover, significant difference in distributions of BMI

(mean) between boys and girls (p=<0.001) was also found

73 Figure 7 Distributions of BMI by age groups

74 Figure 8 Distribution of BMI of boys by age groups

75 Figure 9 Distribution of BMI of girls by age groups

76

3.3.4 Distributions of Skinfold (Triceps) Measurements

1,176 children were 48 months or older However, only 923 of these

children (78.5%) were involved in the analysis of distributions of skinfold

(Triceps) measurements due to missing values

Figure 10 shows the distributions of skinfold (Triceps) measurements

according to different age groups for all children (n=923) Both the distributions

of the 48.0–59.9 and 60.0–72.0 months age groups were skewed towards higher values (skewness=0.95, 1.25 and kurtosis=1.31, 1.86, respectively) In addition,

there was no significant difference of mean skinfold measurements between these

two age groups (p=0.11)

In Figure 11, the distributions of skinfold measurements for boys, also

shows that all distributions were not normal The distributions of boys, skewness

ranged from 1.03 to 1.21 and kurtosis ranged from 1.03 to 1.52 Moreover, there

was no significant differences of mean skinfold measurements between different

age groups (p=0.90)

Similarly, Figure 12 shows distributions of skinfold measurements of girls

by different age groups which were skewed towards higher values, ranging from

0.96 to 1.37 and had kurtosis ranged from 1.51 to 2.73 Mean skinfold

measurements were not significantly varied between different age groups

(p=0.72)

77

As found in height, weight and BMI distributions, there was a significant

difference in skinfold measurements between boys and girls (p=<0.001)

Figure 10 Distribution of Skinfold (Triceps) measurements by age groups

78 Figure 11 Distribution of Skinfold (Triceps) measurements of boys by age groups

Figure 12 Distribution of Skinfold (Triceps) measurements of girls by age groups

79

3.4 Prevalence of overweight and obesity according to different

definitions

Only children aged 2–5 years (1,159 boys and 1,097 girls) were involved

in the analysis of prevalence of overweight and obesity This is because the “CDC

and IOTF BMI for age” definitions define the cutoffs from 2 to 20 years

However, using the “Singapore BMI for age” reference which defines cutoffs of

overweight and obesity in children starting from birth to 72 months, 1,548 boys

and 1,416 girls were involved in the analysis

Table 7 shows the prevalence of overweight and obesity according to age

groups by the three references: the “CDC BMI for age”, “IOTF BMI for age” and

“Singapore BMI for age” references Prevalence of obesity increased with age

(p-trend=<0.001) by the “CDC BMI for age” and the “IOTF BMI for age”

references However, the age effect on obesity by the “Singapore BMI for age”

reference was not seen

Table 8 demonstrates the prevalence of overweight and obesity by gender

and age groups according to the “CDC, IOTF and Singapore BMI for age”

references For boys, age did not effect on the prevalence of overweight by all

three references while the prevalence of overweight increased with age for girls

(p-trend=0.01 for both “CDC and IOTF BMI for age” references) Similarly, the

increasing trend in the prevalence of obesity with age was seen in both genders by

the “CDC and IOTF BMI for age” references (p-trend=<0.001 [boys] and 0.002

[girls] for the “CDC BMI for age” reference, p-trend=0.002 [boys] and 0.01

80

[girls] for the “IOTF BMI for age” reference), but the trend was not found by the

“Singapore BMI for age” reference

3.4.1 Prevalence according to “CDC BMI for age” reference

As shown in Table 7, the overall prevalence of overweight was 8.1% and

that of obesity was 7.1% The percentage of boys who were obese was a little bit

higher than that of boys who were overweight (8.3% vs 7.3%) with p-value of

0.02 (Table 8) In contrast, the percentage of girls who were overweight was

about two times higher than that of girls who were obese (9.0% vs 5.7%, p=0.03)

(Table 8) The result also showed that boys were more likely to be obese than

girls (OR=1.42, 95% CI: 1.02,1.97, p=0.03) whereas there was no statistically

significant gender difference for the prevalence of overweight (p=0.10)

Furthermore, for both genders, the prevalence of obesity increased with age

(p=<0.001 for boys and p=0.002 for girls) However, the increasing prevalence of

overweight with age was seen in girls only (p=0.01) and not in boys

3.4.2 Prevalence according to “IOTF BMI for age” reference

According to the “IOTF BMI for age” cutoffs (Table 7), the overall

prevalence of overweight was 7.6% and that of obesity was 3.9% The prevalence

of overweight (boys=6.8%, girls=8.4%) was higher than that of obesity

(boys=4.6%, girls=3.2%) for both genders although it was not statistically

81

significant (p=0.09 for boys and p=0.11 for girls) (Table 8) There were also no

statistically significant difference in the prevalence of overweight (p=0.10) and

obesity (p=0.13) between genders (Table 8) The increasing trend of prevalence of

obesity with age was seen in both genders (p=0.002 for boys and p=0.01 for girls)

although the age effect on overweight was only seen for girls (p=0.01)

3.4.3 Prevalence according to “Singapore BMI for age” reference

Table 7 also shows the prevalence of overweight and obesity according to

the “Singapore BMI for age” cutoffs The overall prevalence of overweight was

7.0% and for obesity was 5.3% As shown in Table 8, the prevalence of obesity

for boys and girls were similar (5.4% and 5.3%, respectively) There were no

statistically significant differences in the prevalence of overweight (p=0.39) and

obesity (p=0.94) between genders However, for both genders, overweight

prevalence (boys=7.4%, girls=6.6%) was higher than that of obesity (boys=5.4%,

girls=5.3%) and that was statistically significant with p-value for boys was

0.01and that for girls was 0.02 Trend between prevalence and age was not found

in our study, using the “Singapore BMI for age” reference

3.4.4 Comparison of prevalence according to different definitions

Figure 13 shows the comparison of overall overweight and obesity

prevalence for boys and girls with the three different references (the “CDC, IOTF

82

and Singapore BMI for age” references) To allow for a comparison of the three

different references, the prevalence, using the “Singapore BMI for age” reference,

only involved children who were 24 months and older

As shown in Figure 13, there were more obese boys than girls However,

overweight was common in girls than in boys, based on the “CDC BMI for age”

and the “IOTF BMI for age” references Using the “Singapore BMI for age”

reference, overweight was common in boys than in girls, as in the prevalence of

obesity

The prevalence of overweight and obesity for both genders were lower by

using the “IOTF BMI for age” cutoffs than using the “CDC BMI for age” and the

“Singapore BMI for age” references This can be clearly seen in the prevalence of

obesity for both genders Overall, the “IOTF BMI for age” reference

underestimated the prevalence of overweight and obesity for both genders when

compared to the “CDC BMI for age” and “Singapore BMI for age” references

84

Table 6 Frequency distribution and prevalence of overweight and obesity in Singapore Chinese Preschool children aged 6 to 72 months by gender and age groups (according to

"CDC, IOTF and Sinagpore BMI for age" references)

N

†CDC (85th to < 95thpercentile)

85

Figure 13 Prevalence of overweight and obesity in Singapore Chinese Preschool children aged 24 to 72 months by "CDC, IOTF and Singapore

BMI for age" references

From the 2,964 children with available BMI data, birth weight data were

available for 2,183 children (73.7%) (boys=1,125, girls=1,058) who were 24 to 72

months old

Table 9 shows the association between birth weight of child and the

overweight, obesity and overweight/obesity statuses The results showed a

significant relationship between birth weight and excess weight statuses:

overweight (p=<0.001with quartiles and p=0.004 with median cutoffs), obesity

(p=<0.001 with quartiles and median cutoffs) and overweight/obesity (p=<0.001

with both cutoffs for obesity) Children of the highest birth weight quartile (3,390

grams and above) had higher odds ratio (OR) of 2.62 times (95% CI: 1.66,4.14)

and 4.19 times (95% CI: 2.48,7.09) of being overweight and obese, respectively,

compared with children of the lowest birth weight quartile (less than 2,825

grams) Moreover, comparing those with birth weight less than 3,105 grams and

3,105 grams and above, the likelihood for being overweight and obesity for

children with birth weight 3,105 grams and above were 1.57 times (95% CI:

1.15,2.15) and 2.29 times (95% CI: 1.61,3.26), respectively

87

Regarding the association between birth weight and overweight/obesity,

OR for birth weight 3,390 grams and above (highest quartile) was 3.24 (95% CI:

2.27,4.62) when compared to a birth weight below 2,825 grams (lowest quartile)

Furthermore, by median cutoffs, children of birth weight 3,105 grams and above

had 1.86 times (95% CI: 1.46,2.37) higher odds of being overweight or obese

compared to those with birth weight less than 3,105 grams

The results highlighted that the higher the birth weight, the higher the risk

of being overweight and/or obesity in children aged 2–5 years