| Name of cohort study, country*, first year of recruitment | Location† | Aim and selection criteria of primary study, age at recruitment to primary study | Number recruited/number eligible (%)‡ | First author, year, (quality rating) | Aim of paper, selection criteria (number of children studied) and duration of follow-up from recruitment (% of those recruited followed-up) |
| Cohort from Baise, China (M), 2002 | Baise City, Guangxi Zhuang Autonomous Region (U) | Aim: to describe patterns of non-fatal unintentional injuries. Selection: adolescents from 36 randomly selected classes in 9 randomly selected schools. Age at recruitment: 11–18 years | 1840/1855 (99) | Chen, 200513 (A) | Aim: as primary study. Selection: children aged 11–18 years (n = 1840). Follow-up: 1 year (99.2%) |
| Chen, 200514 (A) | Aim: as primary study plus investigation of the association of psychological symptoms with injury. Selection: children aged 13–18 years (n = 1474). Follow-up: 1 year (95.2%) | ||||
| Cohort from Maanshan, China (M), 2001 | Maanshan city (U) | Aim: to study the incidence of injuries and the relationship with behaviour problems. Selection: cluster sampling from years 1–5 in 3 primary schools. Age at recruitment: 6–12 years | 2005/NS (nk) | Peng, 200312 (B) | Aim: as primary study. Selection: as primary study (n = 1983). Follow-up: 1 year (98.9%) |
| Cohort from Kaohsiung, Taiwan (M), 1995 | Kaohsiung city (U) | Aim: to study the incidence of non-fatal school-related injuries over one academic year. Selection: adolescents aged 13–15 years (grades 7, 8 and 9) attending 6 randomly selected schools. Age at recruitment: 13–15 years | 13335/NS (nk) | Yang, 199811 (B) | Aim: as primary study. Selection: as primary study (n = 13335). Follow-up: 1 academic year (nk) |
| West of Scotland 11–16 Study, UK (H), 1994 | Central Clydeside (U) | Aim: to study teenage health and the factors that influence it. Selection: pupils entering 43 randomly selected post-primary schools from randomly selected classes in 135 primary schools. Age at recruitment: 11 years | 2586/2793 (93) | West, 200430 (A) | Aim: to test the hypothesis of equalisation in health between childhood and adolescence. Selection: participants who were surveyed at recruitment (age 11), 13 and 15 years (n = 2196). Follow-up: 5 years (11 years, 93%; 13 years, 84.9%; 15 years, 78.6%) |
| National Longitudinal Survey of Children & Youth, Canada (H), 1994 | Nationwide (M) | Aim: to follow the development and well-being of children from birth to early adulthood. Selection: a random probability sample of residential households with children aged 0–11 years. Age at recruitment: 0–11 years | 22831/NS (nk) | Soubhi, 200434 (B) | Aim: to study the relationship between child, family and neighbourhood characteristics on medically attended injuries. Selection: children aged 4–11 years living in 10261 households responding to cycle 2 of the study (n = 5357). Follow-up: 1 year (63.3%) |
| Soubhi, 200458 (B) | Aim: to study the relationships between injury, behaviour, parenting, family functioning and neighbourhood characteristics. Selection: children aged 4–11 years living in 10261 households responding to cycle 2 of the study (n = 5357). Follow-up: 1 year (63.3%) | ||||
| Add Health Study, USA (H), 1994. | Nationwide (M) | Aim: to study a nationally representative sample of public and private school students. Selection: clustered sampling of 145 middle junior and high schools. Age at recruitment: 11/12–17/18 years | 90118/e118576 (76) | Hammig, 200136 (B) | Aim: to identify behaviours associated with injuries among boys who fight. Selection: boys involved in fights in past 12 months (n = 1314) from a random sample of cohort participants. Follow-up: 1 academic year (100%) |
| Cohort from Kamphaeng Phet Province Vaccination Study, Thailand (M), 1991 | Kamphaeng Phet Province (R) | Aim: to study the efficacy of an inactivated hepatitis A vaccine. Selection: children attending 148 largest community primary schools in the study province. Age at recruitment: school entry to 16 years | 40119/130000 (31) | Kozik, 199910 (B) | Aim: to describe mortality and self-reported injury morbidity in a cohort of schoolchildren. Selection: a randomly selected subset of 20% of the cohort, chosen for sequential serological tests as part of the vaccine trial (n = 6378). Follow-up: 2 years (81.0%) |
| Adolescent Injury Control Study, USA (H), 1990 | Allegheny County, Pittsburgh, Pennsylvania (U) | Aim: to investigate the incidence and risk factors for adolescent injuries. Selection: 7th–9th grade students in one school district. Age at recruitment: 12–16 years | 1245/1400 (89) | Anderson, 199437 (A) | Aim: to examine the patterns of socioeconomic status and injury morbidity. Selection: as primary study (n = 1245). Follow-up: 2 years (89.0%) |
| Cohort from Eastern Shore, Maryland, USA (H), 1986 | 3 counties on Eastern Shore, Maryland, Baltimore (M) | Aim: to investigate factors associated with use of tobacco, drugs and alcohol, and early unprotected sexual intercourse among rural youth. Selection: all 8th grade students in three counties. Age at recruitment: 12–14 years | 758/1930 (39.3) | Alexander, 199235 (B) | Aim: to study behavioural risk factors for medically attended injuries. Selection: as primary study + having completed data from both parent and child (n = 632). Follow-up: 2 years (72.0%) |
| Carolina Longitudinal Study, USA (H), 1981 | Carolina (M) | Aim: to study social development. Selection: students in two bi-racial school districts in either the 4th grade (9–10 years) or 7th grade (12–13 years). Age at recruitment: 9/10 and 12/13 years | 695/e993 (70) | Cobb, 199526 (B) | Aim: to study the relationships between child factors and socioeconomic status and injury/“close calls” (near accidents). Selection: sub-sample of students responding to questions on injury and “close calls” during interviews (n = 271). Follow-up: to 12th grade (∼2–7 years) (39.0%) |
| Christchurch Child Development Study, New Zealand (H), 1977 | Christchurch (U) | Aim: to examine the social, environmental and other risk factors related to child morbidity and explore factors related to health service use, family functioning and well-being. Selection: all hospital births in the urban region of Christchurch, New Zealand between period 15 April 1977 and 5 August 1977. Age at recruitment: birth | 1265/1310 (96.4) | Horwood, 198931 (B) | Aim: to describe participant’s medical history at 5–10 years of age. Selection: all children traceable, with data up to 10 years (n = 1079). Follow-up: 10 years (84.3%) |
| Fergusson, 199546 (B) | Aim: to study relationship between antisocial behaviour in adolescence and injury. Selection: respondents from cohort with complete data (n = 954). Follow-up: 16 years (75.4%) | ||||
| McKinley, 200259 (B) | Aim: to study the effect of mild head injury before age 10 on children in mid to late childhood. Selection: respondents from cohort with data available. (n = 939). Follow-up: 13 years (74.2%) | ||||
| Cohort from Seattle, USA (H), 1975 | Seattle, Washington, (nk) | Aim: to study the relationship between injury and risk-taking behaviour or stressful life events. Selection: all 7th grade boys enrolled in physical education at one middle school. Age at recruitment: 12–13 years | 138/150 (92) | Padilla, 197616 (C) | Aim: as primary study. Selection: as primary study. Follow-up: 5 months (68.7%) |
| Dunedin Multidisciplinary Child Development Study, New Zealand (H), 1975 | Dunedin metropolitan area, Otago (U) | Aim: to study the health and development of children and adolescents, the influences and events contributing to morbidity and health behaviour. Selection: all surviving infants born at Dunedin’s maternity hospital between 1 April 1972 and 31 March 1973, whose mothers resided in the metropolitan area during pregnancy and were still living in the province of Otago when children were age 3. Age at recruitment: 3 years | 1037/1139 (91) | Langley, 198117 (B) | Aim: to describe injuries experienced by children aged 6–7 years. Selection: all traceable 7-year-olds from the original cohort plus those eligible and added to the cohort (n = 1072). Follow-up: 4 years (92.4%) |
| Langley, 198518 (B) | Aim: to describe injuries experienced by children aged 8–9 years. Selection: all 9-year-olds from the original cohort assessed at the research centre. (n = 818). Follow-up: 6 years (78.9%) | ||||
| Langley, 198719 (B) | Aim: to describe injuries experienced by children aged 10–11 years. Selection: all traceable 11-year-olds from the original cohort who agreed to take part (n = 925). Follow-up: 8 years (89.2%) | ||||
| Langley, 198743 (A) | Aim: to study the relationship between child and family variables to childhood injuries. Selection: all traceable children with data for the period 7–11 years (n = 781). Follow-up: 8 years (75.3%) | ||||
| Chalmers, 198920 (B) | Aim: to describe injuries experienced by children aged 12–13 years. Selection: all traceable children completing questionnaires at the research centre (n = 738). Follow-up: 10 years (71.2%) | ||||
| Lodge, 199021 (B) | Aim: to describe injuries experienced by children aged 14–15 years. Selection: all traceable children completing questionnaires at the research centre (n = 849). Follow-up: 12 years (81.9%) | ||||
| Begg, 199060 (B) | Aim: to describe the road crash experiences of children aged 14–15 years. Selection: all traceable children completing questionnaires at the research centre (n = 848). Follow-up: 12 years (81.8%) | ||||
| Begg, 199161 (B) | Aim: to study injuries sustained in bicycle crashes in children aged 14–15 years. Selection: all traceable children completing questionnaires at the research centre (n = 848). Follow-up: 12 years (81.8%) | ||||
| Begg, 199233 (B) | Aim: to study injuries sustained in motor vehicle crashes in children aged 14–15 years. Selection: all traceable children completing questionnaires at the research centre (n = 848). Follow-up: 12 years (81.8%) | ||||
| Jones, 200248 (B) | Aim: to describe the proportion of children remaining fracture-free up to the age of 18 years. Selection: all children providing injury information at each stage of follow-up (n = variable, 739 to 984). Follow-up: 15 years (71.3–84.5%) | ||||
| Jones, 200439 (B) | Aim: to study child risk factors for fractures in cohort members. Selection: poorly specified. (n = 675–853). Follow-up: 15 years (65.1–82.3%) | ||||
| Cohort from South Wales, UK (H), 1972 | Two “industrial towns” in South Wales, (U) | Aim: to study the effects of milk supplementation on child growth to 5 years. Selection: consecutive births in two community hospitals. Twins, premature infants, and those receiving supplements excluded. Age at recruitment: birth | 1163/1288 (90.3) | Davidson, 198745 (B) | Aim: to study the relationship between maternal personality and injury in children. Selection: participants of original cohort with complete data (n = 831). Follow-up: 8 years (71.5%) |
| Davidson, 198822 (B for injury reporting, C for analysis) | Aim: to study the relationship between child behaviour and injury. Selection: all children remaining in the study at 5 years of age (n = 951). Follow-up: 8 years (81.8%) | ||||
| Child Health & Education Study (CHES), UK (H), 1970 | Nationwide (England, Scotland, Wales and Northern Ireland) (M) | Aim: to study the circumstances, health, education and social development of children through to adulthood. Selection all children born between 5–11 April 1970, alive and living in England, Wales or Scotland in 1975. Age at recruitment: birth | CHES: 16004/NS (nk). BCS70: 17196/NS (∼95) | Bijur, 198824 (A) | Aim: to study the relationship between behaviour and injury. Selection: children with data at both 5 and 10 years, who were singleton births, had an English speaking mother of British ancestry, no suspicion of child abuse as the cause of the injuries (or in care), and mother present at the 5-year-old interview (n = 10394). Follow-up: 10 years (64.9%) |
| Bijur, 198838 (B) | Aim: to study the relationship between preschool injuries and injuries in the school-aged period. Selection: as Bijur24 (n = 10394). Follow-up: 10 years (64.9%) | ||||
| Bijur, 198847 (B) | Aim: to study the relationship between family and child factors and injury. Selection: as Bijur24 (n = 10394). Follow-up: 10 years (64.9%) | ||||
| Bijur, 199062 (B) | Aim: to study the sequelae of mild head injury in children. Selection: as Bijur24 (n = 10394). Follow-up: 10 years (nk) | ||||
| Beattie, 199923 (B) | Aim: to describe injuries requiring medical attention in Scottish teenagers. Selection: all teenagers in the cohort traced as resident in Scotland in 1986/7 (n = 958). Follow-up: 16 years (68.0%) | ||||
| Cambridge Study of Delinquent Development, England, UK (H), 1961 | London (U) | Aim: to study offending and antisocial behaviour in London males. Selection: all boys aged 8–9 years (born 1951–4), on the registers of 6 state primary schools and one special school, within a 1-mile radius of the research office in a working class area of South London. Age at recruitment: 8–9 years | 411/411 (100) | West, 197729 (B) | Aim: to describe the lifestyles of youths at age 18. Selection: all traceable from primary study (n = 389). Follow-up: 10 years (94.6%) |
| Shepherd, 200228 (A) | Aim: to study the relationship between offending, health and injury. Selection: all traceable from primary study (n = 387). Follow-up: 10 years (94.2%) | ||||
| Shepherd, 200427 (B) | Aim: to study the relationship between childhood characteristics, teenage delinquency, injury and illness at 16–18. Selection: all traceable from primary study (n = 378). Follow-up: 26 years (94.6%) | ||||
| National Child Development Study (NCDS), UK (H), 1958 | Nationwide (England, Scotland and Wales), (M) | Aim: to monitor the social, economic, educational and health circumstances of all children in England, Scotland and Wales. Selection: all children living in England, Scotland and Wales who were born in the week 3–9 March 1958. Age at recruitment: birth | 17418/e17957 (97) | Peckham, 197325 (B) | Aim: to describe the preliminary findings at age 11 years. Selection: all children from the cohort alive and living in England, Scotland or Wales (n = “more than 15 000”). Follow-up: 11 years (nk) |
| Peckham, 197632 (B) | Aim: to describe development, illnesses, school absence, social conditions and educational progress at age 16 years. Selection: all children from the cohort traced through educational authority school registers in 1974, and agreeing to participate (n = 15245). Follow-up: 16 years (87.5%) | ||||
| Pless, 198940 (B) | Aim: to study factors that may affect the risk of having a traffic injury. Selection: all children from the cohort alive and living in England, Scotland or Wales (n = 13 653 at 11 years, n = 11 507 at 16 years). Follow-up: 16 years (78.4% at 11 years, 66.1% at 16 years) | ||||
| Bijur, 199144 (B) | Aim: to study the relationship between parent–adolescent conflict and injury. Selection: children from the cohort whose British, English-speaking mother responded to questionnaire, with >50% data complete and had data on injury episode at age 16 and 23 years (n = 8231). Follow-up: 23 years (47.3%) | ||||
| Cumberland, 200441 (B) | Aim: to study the relationship between colour vision deficiency, education and injury. Selection: unclear. Assumed to be participants from the original birth cohort that had colour vision assessed at age 11 years using the Ishihara test (n = 12534). Follow-up: 33 years (72.0%) | ||||
| Rahi, 200642 (B) | Aim: to study the relationship between amblyopia and educational, health and social outcomes. Selection: participants from the cohort at age 16, excluding those with bilateral visual loss, unilateral visual loss inconsistent with amblyopia or known eye disease (n = 8861). Follow-up: 33 years (50.9%) | ||||
| Newcastle Thousand Families Study, UK [H), 1947 | Newcastle upon Tyne, England, [U) | Aim: to describe disease and disablement in a representative sample of the city’s children. Selection: infants born to mothers resident in Newcastle from 1 May to 30 June 1947. Age at recruitment: birth | 1142/NS (nk) | Miller, 197415 (B) | Aim: to describe growth, injury, disease, social adaptation and educational attainment of children in relationship to family life and the environment of Newcastle families. Selection: all members of the cohort still enrolled at the ages of 5–15 years (n = 763). Follow-up: 15 years (66.8%) |
*High (H), middle (M) or low (L) income country.
†Urban (U), rural (R) or mixed (M).
‡e, estimated; NS, not specified; nk, not known.
¶A, B or C (see Methods).