Table 7.

The relationship between physical activity and cardiometabolic health

# of studies	Design	Quality assessment					# of participants	Absolute effect	Quality
		Risk of bias	Inconsistency	Indirectness	Imprecision	Other
Mean baseline age ranged from 3-4.9 years. One study reported only that the children were preschool age. Data were collected by non-randomized intervention, longitudinal with up to 2 years follow-up, and cross-sectional study designs. Cardiometabolic health was assessed by mean arterial pressure, DBP, SBP, total cholesterol, total serum cholesterol, HDL, triglycerides, HDL2, LDL, LDL/HDL, total serum cholesterol/HDL, HDL/total triglycerides, and clustered cardiovascular risk score (SBP, triglycerides, total cholesterol/HDL, HOMA-IR, sum of two skinfolds). All outcomes were objectively measured.
1	Non-randomized interventiona	Serious risk of biasb	No serious inconsistency	No serious indirectness	No serious imprecision	None	264	BP The PA intervention (structured/organized PA) was favourably associated with DBP during rest and activity [125].	VERY LOWc
2	Longitudinald	Serious risk of biase	No serious inconsistency	No serious indirectness	No serious imprecision	None	291	BP Aerobic PA was favourably associated with BP (SBP but not DBP, boys only, 2-year follow-up but not 1-year follow-up) in 1 study [126]. Leisure PA was unfavourably associated with BP (DBP but not SBP, boys only, 1-year follow-up but not 2-year follow-up) in 1 study [126]. Structured PA was not associated with BP (SBP or DBP) in 1 study [126]. Cholesterol TPA was not associated with cholesterol (total serum cholesterol, HDL, HDL 2 , LDL, LDL/HDL, or total serum cholesterol/HDL) in 1 study [43]. Triglycerides TPA was not associated with triglycerides in 1 study [43].	VERY LOWf
6	Cross-sectionalg	Serious risk of biash	Serious inconsistencyi	No serious indirectness	No serious imprecision	Exposure/outcome gradientj	1882	Clustered risk score TPA was favourably associated with clustered risk score (boys only, Quartile 1 vs. Quartile 5 only) in 1 study [127]. MPA was not associated with clustered risk score in 1 study [127]. MVPA was not associated with clustered risk score in 1 study [127]. VPA was favourably associated with clustered risk score (boys only, Quartile 2 vs. Quartile 5 only) in 1 study [127]. BP TPA was unfavourably associated with BP (SBP and DBP) in 1 study [117] and not associated with BP (SBP, DBP, or mean arterial pressure) in 3 studies [66, 72, 81]. Aerobic PA was not associated with BP (SBP or DBP) in 1 study [126]. Indoor PA was not associated with BP (SBP or DBP) in 1 study [81]. Leisure PA was not associated with BP (SBP or DBP) in 1 study [126]. Outdoor PA was not associated with BP (SBP or DBP) in 1 study [81]. Structured PA was not associated with BP (SBP or DBP) in 1 study [126]. Cholesterol TPA was favourably associated with cholesterol (total cholesterol but not HDL) in 1 study [81] and not associated with cholesterol (total cholesterol, HDL, or HDL/total cholesterol) in 1 study [72]. Indoor PA was not associated with cholesterol (total cholesterol or HDL) in 1 study [81]. Outdoor PA was unfavourably associated with cholesterol (HDL but not total cholesterol) in 1 study [81]. Triglycerides TPA was not associated with cholesterol (total cholesterol, HDL, or HDL/total cholesterol) in 1 study [72].	VERY LOWk

BP: blood pressure; DBP: diastolic blood pressure; HDL: high-density lipoprotein cholesterol; HOMA-IR: homeostatic model assessment – insulin resistance; LDL: low-density lipoprotein cholesterol; MPA: moderate-intensity physical activity; MVPA: moderate- to vigorous-intensity physical activity; PA: physical activity; SBP: systolic blood pressure; TPA: total physical activity; VPA: vigorous intensity physical activity

^aIncludes 1 non-randomized intervention [125]

^bNo intention-to-treat analysis; results are based on children who were measured at all 3 time points. Physical activity was not measured, so it is unknown if the intervention significantly changed physical activity

^cQuality of evidence was downgraded from “low” to “very low” because of serious risk of bias

^dIncludes 2 longitudinal studies [43, 126]

^ePotential confounders were not adjusted for in 1 study [43]. No psychometric properties were reported for the subjective physical activity measure in 1 study [126]

^fQuality of evidence was downgraded from “low” to “very low” because of serious risk of bias

^gIncludes 6 cross-sectional studies [66, 72, 81, 117, 126, 127]

^hNo potential confounders were adjusted for in 5 studies [66, 72, 81, 117, 127]. Convenience sample in 1 study [117]. No psychometric properties were reported for the subjective physical activity measure in 1 study [126]

ⁱFavourable and unfavourable associations between physical activity and cardiometabolic health observed across studies

^jA gradient for higher TPA with worse total cholesterol was observed in 1 study [81]

^kQuality of evidence was downgraded from “low” to “very low” because of serious risk of bias and serious inconsistency; because of this limitation, was not upgraded for an exposure/outcome gradient