Skip to main content
. 2023 Jun 2;10:1094280. doi: 10.3389/fmed.2023.1094280

Table 5.

Mediation analysis: effects of multimorbidity on the MCS-12 subscale scores.

Natural direct effect Natural indirect effect Marginal total effect
β (95% CIϯ) β (95% CIϯ) β (95% CIϯ)
Vascular-Inflammatory class
Mediator: recent service use a −1.44* (−2.85, −0.11) −0.22* (−0.42, −0.08) −1.66* (−3.01, −0.37)
Mediator: private insurance b −1.44* (−3.00, −0.19) 0.09 (0.02, 0.25) −1.35* (−2.90, −0.09)
Mediator: public insurance c −1.44* (−2.82, −0.11) 0.01 (−0.10, 0.15) −1.42* (−2.75, −0.13)
Respiratory class
Mediator: recent service use a −2.24 (−5.48, 0.55) −0.25* (−5.39, −0.06) −2.49* (−5.57, 0.33)
Mediator: private insurance b −2.24 (−5.44, 0.51) −0.02 (−0.12, 0.06) −2.25 (−5.48, 0.43)
Mediator: public insurance c −2.10 (−6.24, 1.48) −0.38 (−2.65, 1.45) −2.48 (−5.52, 0.63)

MCS-12, mental component summary score. ϯBias-corrected confidence interval. ¥Exposure–mediator interactions were not significant at the p < 0.05 level and not included in the models.

a

Adjusted for age, sex, education, employment, wealth, current alcohol use, past or present smoking status, physical activity level, public insurance coverage, and private insurance coverage.

b

Adjusted for age, sex, education, employment, wealth, current alcohol use, past or present smoking status, physical activity level, public insurance coverage, and recent service use.

c

Adjusted for age, sex, education, employment, wealth, current alcohol use, past or present smoking status, physical activity level, private insurance coverage, and recent service use.