Skip to main content
. 2023 Mar 24;25:e43213. doi: 10.2196/43213

Table 3.

Unadjusted and adjusted structural equation modeling showing whether self-esteem and peer connectedness mediated the relationship between active social media use and mental health.

Effect Active social media use at age 12 or 13 coefficient Self-esteem coefficient Number of close friends coefficient Happiness with friends coefficient
Unadjusted model (N=2569)

Self-esteem


Direct −0.10 (−0.18 to −0.02)a N/A N/A N/A


Indirectb N/Ac N/A N/A N/A

Number of close friends


Direct −0.33 (−0.37 to 1.03) 2.48 (0.72 to 4.23)d N/A N/A


Indirect N/A N/A N/A N/A

Happiness with friends


Direct −0.27 (−0.49 to −0.05)e 0.98 (0.47 to 1.50)f N/A N/A


Indirect N/A N/A N/A N/A

Poor mental health at age 14 or 15 years


Direct 0.33 (−0.37 to 1.03) −6.80 (−8.22 to −5.36)f −0.02 (−0.08 to 0.04) −0.15 (−0.70 to 0.41)


Indirect N/A 0.70 (0.15-1.30g 0.01 (−0.03 to 0.06) 0.04 (−0.12 to 0.22)
Adjusted model (N=2316)

Self-esteem


Direct −0.05 (−0.11 to 0.02) N/A N/A N/A


Indirect N/A N/A N/A N/A

Number of close friends


Direct −1.05 (−2.50 to −0.05) 0.68 (−1.05 to 2.41) N/A N/A


Indirect N/A N/A N/A N/A

Happiness with friends


Direct −0.26 (−0.53 to 1.73) 1.11 (0.51 to 1.72)f N/A N/A


Indirect N/A N/A N/A N/A

Poor mental health at age 14 or 15 years


Direct 0.02 (−0.72 to 0.75) −5.34 (−7.31 to −3.36)f −0.05 (−0.12 to 0.02) 0.10 (−0.41 to 0.60)


Indirect N/A 0.24 (−0.12 to 0.66) 0.06 (−0.03 to 0.20) −0.03 (−0.20 to 0.12)

aP=.01.

bIndirect effects are Monte Carlo estimates. Coefficients unstandardized.

cN/A: not applicable.

dP=.006.

eP=.01.

fP<.001.

gP=.02.