Table 5. The selection of Newcastle-Ottawa Scale.

Author Publication y	Representativeness of the exposed cohort (a or b = 1, c or d = 0)	Selection of the not exposed cohort (a = 1)	Ascertainment of exposure (a or b = 1)	Demonstration that outcome of interest was not present at start of study (a = 1, b = 0)
Harousseau 1997[55]	b	a	a	a
Slovak 2000[6]	b	a	a	a
Suciu 2003[5]	b	a	a	a
Tsimberidou 2003[56]	a	a	a	a
Brunett 2006[12,64]	a	a	a	a
Cornelissen 2007[11]	a	a	a	a
Pfirrmann 2012[57]	b	a	a	a
Zhu 2013[58]	b	a	a	a
Stelljes 2014[59]	a	a	a	a

y indicates year.