Table 2. Assessment of study quality.

Quality Criteria	Quality Score
Representativeness of the population	Population-based representative = 1
	Not representative, selected group, volunteers, or no description = 0
Ascertainment of exposure to child abuse and neglect	Data on child maltreatment collected prospectively = 1
	Data on child maltreatment collected retrospectively = 0
Selection of the non-exposed cohort/controls	Drawn from the same population = 1
	Drawn from a different source or no description = 0
Assessment of child abuse and neglect	Secure official record (court-substantiated abuse) = 1
	Self-reported or structured interview or self-administered questionnaire or no description = 0
Case definition for child abuse and neglect	Uses WHO definitions of child maltreatment or court-substantiated abuse or Barnett-Cicchetti Maltreatment Classification System = 1
	Marks and bruises (physical abuse), questions from scales (e.g., Childhood Trauma Questionnaire), published surveys, or own system = 0
Assessment of outcome	Use of structured clinical interview for DSM-III/IV (DIS, DISC, CIDI) (mental health); direct physical measurements or blood tests (physical health) = 1
	Questions from published health surveys/screening instruments, own system, symptoms described, no system, not specified, or self-reported = 0
Adequacy of follow-up of cohorts (where relevant) or response rate	Completeness good (≥80%), with description of those lost to follow-up = 1
	Completeness poor (<80%) or no statement = 0
Appropriate statistical analysis	Yes = 1
	No = 0
Appropriate methods to control confounding	Yes = 1 (multivariable adjusted OR including SES, education, or family dysfunction in models)
	No = 0 (univariate analysis or controls for age/sex only)
Source of funding declared	Yes (financial disclosure, funding/support/grant declared) = 1
	No = 0

CIDI, Composite International Diagnostic Interview; DIS, Diagnostic Interview Schedule; DISC, Diagnostic Interview Schedule for Children; SES, socioeconomic status.