Table 4.

Subgroup analysis of smoking effects on pediatric systolic blood pressure

Category	Model	Publication bias	Heterogeneitya	Beta (95%CI)	Pfor effect size
Age group (year)
0–7	FEM	p = 0.55	33.0% (p = 0.11)	0.39 (0.24 to 0.55)	< 0.001
7–13	FEM	p = 0.04	0.0% (p = 0.86)	0.14 (− 0.12 to 0.40)	0.31
12–18	FEM	p = 0.68	53.2% (p = 0.12)	0.21 (0.09 to 0.33)	0.001
Type of study
Cohort	FEM	p = 0.04	27.2% (p = 0.14)	0.25 (0.16 to 0.34)	< 0.001
Cross-sectional	REM	p = 0.68	62.5% (p = 0.03)	0.21 (−0.32 to 0.74)	0.44
Smoking definition
Not reported	FEM	p = 0.88	28.8% (p = 0.24)	0.16 (−0.38 to 0.70)	0.57
At least 1 cigarette per month	FEM	p = 0.47	29.0% (p = 0.21)	0.02 (−0.09 to 0.13)	0.71
Current daily smoking	FEM	p = 0.83	45.7% (p = 0.14)	0.25 (0.08 to 0.41)	0.003
At least 1 cigarette per month	REM	p = 0.82	55.3% (p = 0.004)	0.30 (0.10 to 0.50)	0.003
Period of exposure
Pregnancy	REM	p = 0.02	35.4% (p = 0.08)	0.26 (0.12 to 0.41)	< 0.001
Domestic (postnatal)	REM	p = 0.07	59.8% (p = 0.03)	0.28 (−0.04 to 0.59)	0.08
Parental smoking habit
Mother	REM	p = 0.04	43.2% (p = 0.04)	0.25 (0.09 to 0.41)	0.002
Father	NA	NA	NA	NA	NA
Both	REM	p = 0.05	56.8% (p = 0.03)	0.34 (0.01 to 0.67)	0.04
Sample size
< 1000 subjects	FEM	p = 0.91	27.5% (p = 0.25)	0.61 (−0.41 to 1.63)	0.24
≥ 1000 subjects	REM	p = 0.07	54.6% (p = 0.003)	0.25 (0.11 to 0.39)	< 0.001
BP measurement device
Mercury/aneroid	REM	0.576	56.5% (p = 0.011)	0.11 (0.03 to 0.20)	0.007
Automatic/semiautomatic	FEM	0.257	26.7% (p = 0.190)	0.33 (0.17 to 0.48)	< 0.001

^aHeterogeneity was reported as I-squared and corresponding p value. CI Confidence interval, FEM Fixed effect model, NA Not applicable due to lack of included studies, REM Random effect mode