Table 5. The association between Modic changes and activity limitation.

MCs type(s)	Study	Outcome measure	OR (95% CI)	Mean difference •
Clinical studies
MCs, any type	Kleinstuck 2006 [42]	RMQ (0–24)		1.1 (-1.75–3.95) p = 0.45
	Annen 2016 [51]	ODI (0–100)		3.32 (-0.73–7.37) p = 0.11
MCs1	Schistad 2014 [49]	ODI (0–100)		2.8 (-4.65–10.25) p = 0.46
MCs2 or 3	Schistad 2014 [49]	ODI (0–100)		5.4 (-0.33–11.13) p = 0.07
Non-clinical studies
MCs, any type	Koyama 2013 [63]	Osaka City University Questionnaire (OCU-test)	0.20 (0.01–4.27)£
	Mok 2016 [64]	ODI (0–100)	NS*
	Mok 2016 [64]	RMQ (0–24)	NS*
	Määttä 2016 [67]	ODI (0–100)	1.47 (1.04–2.10)
MCs1	Määttä 2016 [67]	ODI (0–100)	1.23 (0.67–2.24)
MCs2	Määttä 2016 [67]	ODI (0–100)	1.56 (1.06–2.31)
Case-control studies
MCs1	Rannou 2007 [68]	Quebeck Disability Score (0–100)		1 (-10.65–12.65) p = 0.87
MCs2	Rannou 2007 [68]	Quebeck Disability Score (0–100)		4 (-6.68–14.68) p = 0.48

•: (MCs+)-(MCs-) calculated from raw data where available, using t-test.

*: No raw data supplied.

£: Added 0.5 to all cells in 2x2 table.