Levothyroxine use and longitudinal changes in thigh muscles in at-risk participants for knee osteoarthritis: preliminary analysis from Osteoarthritis Initiative cohort

Table 4 Mediatory role of thigh muscle MRI markers in the effect of levothyroxine use on KOA incidence

	Mediatory variables^a	Estimate (95%CI), P
	Mediatory variables^a	Total association of levothyroxine use with KOA incidence (through Eq. 4 in Fig. 2)	Direct association of levothyroxine use with KOA incidence^a (through Eq. 3 in Fig. 2)	Mediatory role of thigh muscle biomarkers in the association of levothyroxine use and KOA incidence^a (through Eqs. 1 and 2 in Fig. 2)
KOA incidence	4-year changes in CSA (mm²)
Radiographic	Quadriceps	5.519 (2.751–8.082), P < 0.001	4.939 (2.215–7.524), P < 0.001	0.580 (0.209–1.047), P < 0.001
Radiographic	Total thigh muscles	5.456 (3.136–7.837), P < 0.001	5.108 (2.909–7.466), P < 0.001	0.348 (0.144–0.596), P < 0.001
Symptomatic	Quadriceps	7.842 (3.469–11.937), P < 0.001	7.116 (2.915–11.306), P < 0.001	0.726 (0.346–1.223), P < 0.001
Symptomatic	Total thigh muscles	7.445 (4.463–11.313), P < 0.001	7.121 (4.119–10.977), P < 0.001	0.324 (0.100–0.626), P < 0.001

CI confidence interval, CSA cross-sectional area, intra-MAT intra-muscular adipose tissue, JSN joint space narrowing, KOA knee osteoarthritis
^aSince mixed models could not be used in mediation analysis of the R mediate package, here, longitudinal changes in the muscle biomarkers were assessed as relative change index between baseline and 4th-year visits, i.e., (baseline − 4th-year)/baseline

ISSN: 1478-6362