Table 2 Characteristics of included studies.

Authors	Study characteristics						Patient characteristics
	Date published	Study type	Country	Centres	Assessors blinded?	Assessment tool	Dates of birth	Birth weight	Duration of follow‐up (months)*	Follow‐up (%)	Number with NEC (n)	Total number of patients (N)
Castro17	2004	Cohort	USA	Multi	Yes	BSID II	1993–1994	<1000 g	18–22	78	72	1483
Chacko18	1999	Case–control	Australia	Single	No	GQ	1990–1993	<1000 g	12–59	100	40	60
Hintz21	2005	Cohort	USA	Multi	Yes	BSID II	1995–1998	401–1000 g	18–22	79	532	4933
Holmsgaard19	1996	Cohort	Denmark	Single	No	Clinical	1987–1990	994 g mean	42 (mean)	99	4	136
Mayr12	1994	Cohort	Austria	Single	No	Denver	1978–1991	<1501 g	12–156†	63	12	18
Simon24	1993	Cohort	USA	Single	No	BSID I	1986–1988	<1500 g	15†	100	18	18
Sonntag20	2000	Case–control	Germany	Single	No	GQ	1992–1996	<1500 g	20	91	20	60
Tobiansky10	1995	Case–control	Australia	Single	No	GQ	1986–1991	⩽1500 g	36–60	91	49	89
Walsh14	1989	Cohort	USA	Single	No	BSID I	1975–1983	<1500 g	20	90	36	802
Waugh22	1996	Cohort	Australia	Single	Yes	GQ	1977–1990	<1000 g	24	89	23	199
Yeh11	2004	Case–control	Taiwan	Single	No	BSID II	1991–2002	<1500 g	18	54	15	45

*All duration of follow‐up times refer to corrected age unless marked with †. Corrected age is gestational age at birth plus chronological age (ie, corrected for prematurity). BSID, Bayley Standardised Infant Development Scale (I and II); GQ, Griffiths Quotient; Denver, Denver developmental screening test.