Skip to main content

Table 1 Characteristics of studies included in the meta-analysis

From: Breastfeeding and the risk of childhood cancer: a systematic review and dose-response meta-analysis

First author, year, country Data collection years Design No. of cases/controls1 Age (years) Source of cases Source of controls Cancer type Breastfeeding categories (months) Method of assessing breastfeeding Variables matched or adjusted NOS Risk of bias
Davis, 1988, USA [33] 1976–1983 PC-CS 201/181 1.5–15 Colorado Central Cancer Registry Population Leukemia (N = 63); brain tumors (N = 38); lymphoma (N = 26); STS (N = 15); other cancers (N = 59) Never/≤ 6/> 6 (lymphoma); ≤ 6/> 6 (leukemia, brain tumors, STS) Interview of parents Matched: age (±3 years), sex, telephone exchange. Adjusted: none. 7 Moderate
Magnani, 1988, Italy [34] 1981–1984 HC-CS 182/307 NA The main pediatric hospital of Turin Hospital patients Leukemia (N = 163); NHL (N = 19) Never/1–6/> 7 Interview of parents Matched: none. Adjusted: none. 5 Serious
Van Duijn, 1988, Netherlands [35] 1973–1979 PC-CS 492/480 1–15 The national registry of the Dutch Childhood Leukemia Study Group Population ALL (N = 492) Never/≤ 6/> 6 Questionnaires mailed to parents Matched: age (±3 years), sex. Adjusted: age, sex, birth order, social class, maternal education, and maternal factors (age, smoking, alcohol use in pregnancy). 7 Moderate
Shu, 1995, China [36] 1981–1991 PC-CS 241/241 1–15 Shanghai Cancer Registry Population Lymphoma (N = 82); leukemia (N = 159) Never/1–6/> 6 Interview of parents Matched: sex, year of birth. Adjusted: maternal age at birth, birthweight, maternal working status and occupational exposure to chemicals during infancy. 8 Moderate
Shu, 1995, USA and Canada [17] 1982–1989 PC-CS 105/639 0–15 The Children’s Cancer Group (CCG), 1982–1989 Population Malignant germ cell tumors (N = 105) Never/1–6/7–12/> 12 Self-administered questionnaire to mother Matched: none. Adjusted: age, gender, gestational age, number of livebirths, maternal education, and smoking during pregnancy. 7 Moderate
Petridou, 1997, Greece [37] 1993–1994 HC-CS 153/300 0–14 A nationwide network of childhood hematologists/oncologists Hospital patients Leukemia (N = 153) Ever vs. never Interview of guardians Matched: place of residence, gender and age. Adjusted: sex, age, place of residence, sociodemographic, lifestyle, environmental and biomedical variables. 7 Serious
Schüz, 1999, Germany [38] 1992–1997 PC-CS 1001/1001 0–14 German Children’s Cancer Registry (GCCR) Population Leukemia (N = 1001) ≤ 1/2–6/> 6 Questionnaire and telephone interview Matched: gender, date of birth and district. Adjusted: socioeconomic status (SES). 7 Moderate
Shu, 1999, USA and Canada [39] 1989–1993 PC-CS 2200/2418 1–17 The Children’s Cancer Group (CCG), 1989–1993 Population Leukemia (N = 2200) Never/1–3/4–6/7–9/10–12/> 12; Ever vs. never Interview of mothers Matched: age, geographic location and ethnicity. Adjusted: maternal race, maternal education, and family annual income. 8 Moderate
Smulevich, 1999, Russia [40] 1986–1988 PC-CS 593/1181 0–14 Russian Cancer Research Center RAMS Population Lymphoma (N = 117); leukemia (N = 199); brain tumors (N = 57); neuroblastoma(N = 42); nephroblastoma (N = 48); STS (N = 53); other cancers (N = 77) < 1/1–2/3–4/5–6/7–12/> 12 (leukemia, STS, neuroblastoma, nephroblastoma); ≤ 2/3–4/5–6/7–12/> 12 (lymphoma, brain tumors) Interview of mothers Matched: age, gender and residence. Adjusted: none. 7 Moderate
Dockerty, 1999, New Zealand [41] 1991–1995 PC-CS 97/303 0–14 National databases including the New Zealand Cancer Registry Population ALL (N = 97) Never/≤6/> 6–12/> 12; Ever vs. never Interview of mothers Matched: age and sex.Adjusted: age, sex, child’s social class, mother’s education, household crowding, delay from reference date to interview. 8 Moderate
Infante-Rivard, 2000, Canada [42] 1980–1993 PC-CS 491/491 0–9 Tertiary care centers Population ALL (N = 491) Never/≤ 3/> 3 Interview of mothers Matched: age, sex, and region of residence at the time of diagnosis.Adjusted: none. 8 Moderate
UKCCS Investigators, 2001, UK [43] 1991–1996 PC-CS 3500/6964 1–14 UKCCS Population Lymphoma (N = 342); leukemia (N = 1637); other cancers (N = 1521) Never/< 1/1–6/≥ 7; Ever vs. never Interview of mothers Matched: month and year of birth, sex and region of residence at diagnosis.Adjusted: age at diagnosis, sex, region, birth order and deprivation index. 7 Moderate
Hardell, 2001, Sweden [14] 1993–1996 PC-CS 835/860 0–14 Swedish Cancer Register Population Leukemia (N = 235); lymphoma (N = 99); brain tumors (N = 264); neuroblastoma (N = 34); retinoblastoma (N = 22); germ cell tumors (N = 21); STS(N = 37); other cancers (N = 123) 0– < 1/1– < 6/≥ 6;≥ 1 vs. 0- < 1 Routine child health records Matched: age and sex.Adjusted: none. 8 Moderate
Schüz, 2001, Germany [44] 1988–1994 PC-CS 465/2442 0–14 German Children’s Cancer Registry (GCCR) Population Brain tumors (N = 281); other cancers (N = 184) ≤ 1/2–6/> 6 Questionnaire and telephone interview Matched: gender and date of birth within 1 year.Adjusted: degree of urbanization and socioeconomic status. 7 Moderate
Daniels, 2002, USA and Canada [45] 1992–1994 PC-CS 393/376 0.5–19 The Children’s Oncology Group in the United States and Canada, 1992–1994 Population Neuroblastoma (N = 393) Never/0–3/4–6/7–9/9–12/≥ 13; Ever vs. never Interview of mothers Matched: date of birth.Adjusted: income and age. 6 Moderate
Murray, 2002, UK [46] 1975–1986 Historical cohort 178/409969 0–16 The Northern Ireland Cancer Registry Population ALL (N = 178) Ever vs. never Child health records Matched: none.Adjusted: none. 6 Serious
Perrillat, 2002, France [47] 1995–1999 HC-CS 247/237 2–15 The hospitals of Lille, Lyon, Nancy and Paris Hospital patients Leukemia (N = 247) Never/< 3/3–5/6–11/≥ 12; Ever vs. never Interview of mothers Matched: age, gender, hospital, hospital catchment area and ethnic origin.Adjusted: age, gender, hospital, ethnic origin, maternal educational level and parental socio professional category, birth weight, length of pregnancy and number of previous pregnancies 8 Moderate
Lancashire, 2003, UK [13] 1972–1981 PC-CS 3376/3376 1–14 The Oxford Survey of Childhood Cancers (OSCC) Population Leukemia (N = 1342); other cancers (N = 2034) Never/< 1/1–6/≥ 7; Ever vs. never Interview of parents Matched: sex and date of birth.Adjusted: sex, age at death, father’s occupational social class, sibship position, maternal age at child’s birth. 7 Moderate
Jourdan-Da, 2004, France [48] 1995–1998 PC-CS 452/530 1–15 The National Registry of Childhood Leukemia and Lymphoma Population Leukemia (N = 452) Never/< 3/3–6/> 6; Ever vs. never Self-administered questionnaire to mother Matched: age, sex and region.Adjusted: gender, age at diagnosis, region of residence at diagnosis. 7 Moderate
Altinkaynak, 2006, Turkey [49] 1990–2000 PC-CS 137/146 1–16 The Ataturk University Medicine Faculty Population Leukemia (N = 87); lymphoma(N = 50) 0–6 vs. > 6 Interview of mothers Matched: sex and age.Adjusted: none. 7 Moderate
Saddlemire, 2006, USA and Canada [50] 1999–2002 PC-CS 461/443 0.5–16 The Children’s Oncology Group in the United States and Canada, 1999–2002 Population Wilms tumor (N = 461) Never/0–3/4–6/7–9/10–12/≥ 13; Ever vs. never Telephone interviews of mothers Matched: age strata and geographic region of residence.Adjusted: child’s age at reference date, geographic region of residence, household income and mother’s education. 7 Moderate
Shaw, 2006, Canada [51] 1980–1993;1995–1999 PC-CS 272/272 0–14 Tertiary care centers in the province of Quebec Population Brain tumors (N = 272) Never/≤ 8 weeks/9-24 weeks/> 24 weeks; Ever vs. never Interview of parents Matched: sex and age at diagnosis.Adjusted: mother’s education level. 6 Moderate
Petridou, 2006, Greece [52] 1996–2002 HC-CS 71/71 0–14 A nationwide network of childhood hematologists/oncologists Hospital patients HD (N = 71) Ever vs. never Interview of guardians Matched: gender and age (±6 months).Adjusted: demographic, socioeconomic, anthropometric and perinatal variables. 7 Serious
Bener, 2008, United Arab Emirates [53] 1983–2004 HC-CS 169/169 0–15 Tawam Hospital, Al Ain Hospital (healthy subjects) ALL (N = 103); HD (N = 32); NHL (N = 34) 0–6 vs. > 6 Telephone interviews with mothers Matched: age and sex.Adjusted: none. 6 Moderate
MacArthur, 2008, Canada [54] 1990–1994 PC-CS 399/399 0–14 Pediatric oncology treatment centers and cancer registries Population Leukemia (N = 399) Never/0–3/4–6/7–12/> 12 Interview of parents or guardians Matched: age, gender, and area.Adjusted: maternal age at birth, maternal education, annual household income, ethnicity, and number of residences since birth. 6 Moderate
Flores-Lujano, 2008, Mexico [55] 1998–2006 HC-CS 57/218 0–19 The institutions that treat children with cancer in Mexico Children with Down syndrome Leukemia (N = 57) Never/≤ 6/≥ 7; Ever vs. never Interview of parents Matched: none.Adjusted: sex, weight at birth < 2500 g, age of child, firstborn child, low standard of living and cardiovascular diseases. 6 Moderate
Rudant, 2010, France [56] 2003–2004 PC-CS 720/1494 1–14 ESCALE, 2003–2004 Population ALL (N = 634); AML (N = 86) Never/≤ 2/3–5/6–11/≥ 12; Ever vs. never Telephone interview of mothers Matched: age and gender.Adjusted: age, gender, parental professional category, degree of urbanization, and maternal age at child’s birth. 8 Moderate
Rudant, 2011, France [57] 2003–2004 PC-CS 291/2149 2–14 ESCALE, 2003–2004 Population HD (N = 127); NHL (N = 164) Never/< 6/≥ 6 Telephone interview of mothers Matched: age and gender.Adjusted: age, gender, parental professional category, degree of urbanization, maternal age at child birth and housing, for contact with animals. 8 Moderate
Waly, 2011, Sultanate of Oman [58] 2008–2010 HC-CS 70/70 NA2 The Sultan Qaboos University Hospital Relatives of the cases ALL (N = 70) < 6/6–12/12–24/ > 24 Interview of mothers Matched: none.Adjusted: none. 6 Serious
Urayama, 2012, USA [59] 1995–2008 PC-CS 507/762 1–14 The Northern California Childhood Leukemia Study Population ALL (N = 507) Ever vs. never Questionnaire to parents or guardians Matched: date of birth, sex, Hispanic status, and maternal race.Adjusted: age, sex, maternal age, maternal education, annual household income and race/ethnicity. 8 Moderate
Lupo, 2014, USA [60] 1982–1988 PC-CS 319/317 0–20 Intergroup Rhabdomyosarcoma Study Group (IRSG) Population Rhabdomyosarcoma (N = 319) Never/< 6/6–12/≥ 12; Ever vs. never Telephone interview to mothers Matched: race, sex, and age.Adjusted: age, race, sex, household income, maternal, and paternal education. 7 Moderate
Mattioli, 2014, Italy [61] 1998–2003 PC-CS 82/1040 0–10 Childhood acute leukemia in 14 Italian Regions Population AnLL (N = 82) Never/1–3/4–6/> 6 Interview of parents Matched: date of birth, sex and residence area.Adjusted: none. 6 Serious
Schraw, 2014, USA [62] 1997–2011 PC-CS 142/284 0–14 The Texas Children’s Cancer Center Healthy subjects ALL (N = 142) Ever vs. never Questionnaire to mothers Matched: age, sex, race, and ethnicity.Adjusted: none. 7 Moderate
Ajrouche, 2015, France [63] 2010–2011 PC-CS 617/1225 1–14 ESTELLE, 2010–2011 Population ALL (N = 617) Never/< 6/≥ 6; Ever vs. never Telephone interview of mothers Matched: age and gender.Adjusted: age, gender, parental professional category, maternal age, and degree of urbanization. 8 Moderate
Greenop, 2015, Australia [64] 2003–2006;2005–2010 PC-CS 592/1389 0–14 Aus-ALL and Aus-CBT Population ALL (N = 314); brain tumors (N = 278) Never/< 3/3- < 6/≥ 6; Ever vs. never Questionnaire to mothers Matched: age, sex, and state of residence.Adjusted: Matched variables, maternal education, maternal age at birth, birth order, proportion of optimum birth weight. 7 Moderate
Heck, 2015, USA and Canada [65] 2006–2011 HC-CS 243/134 0–15 Wills Eye Hospital in Philadelphia, or the Children’s Oncology Group Friends or unrelated children of case family Retinoblastoma (N = 243) < 1/1–6/7–11/≥ 12; Ever vs. never Telephone interview of parents Matched: age.Adjusted: child’s age at interview, mother’s race/ethnicity, mother’s educational attainment, and household income. 7 Moderate
Amitay, 2016, Israeli [66] 2005–2013 PC-CS 178/357 1–18 The hematologic departments of five large Israeli hospitals Population Leukemia (N = 117); lymphoma (N = 61) ≤ 6 vs. > 6 Interview of mothers Matched: age, gender, and spoken language.Adjusted: none. 7 Moderate
Rios, 2016, France [67] 2003–2004;2010–2011 PC-CS 269/1589 0.5–6 ESCALE, 2003–2004; ESTELLE, 2010–2011 Population Neuroblastoma (N = 269) Never/< 3/3–5/≥ 6; Ever vs. never Telephone interview of mothers Matched: age and gender.Adjusted: age and sex, birth order, maternal age, urban status of the area of residence and study. 8 Moderate
Bailey, 2017, France [68] 2003–2004;2010–2011 PC-CS 469/2719 1–14 ESCALE, 2003–2004; ESTELLE, 2010–2011 Population Brain tumors (N = 469) Never/< 3/3- < 6/≥ 6; Ever vs. never Telephone interview of mothers Matched: age and gender.Adjusted: age and gender. 8 Moderate
Schraw, 2017, USA [69] 1997–2012 PC-CS 171/342 0–14 Texas Children’s Cancer Center in Houston Population ALL (N = 171) Ever vs. never Questionnaire to parents Matched: sex, ethnicity, and age.Adjusted: none. 7 Moderate
Gao, 2018, China [70] 2008–2017 HC-CS 958/785 0–14 The Children’s Hospital of Zhejiang University Hospital patients Leukemia (N = 958) Never/1–3/4–6/7–9/10–12/≥ 13 Medical records Matched: none.Adjusted: family history of cancer, the family history of neoplasm of the lymphatic/hematopoietic systems, history of bedroom decoration, smoking during pregnancy, the history of using birth control pills before pregnancy, abortion history, Down’s syndrome and parents use of hair dye. 8 Moderate
Jiménez-Hernández, 2018, Mexico [18] 2010–2015 HC-CS 1455/1455 0–17.6 Private and public hospitals in Mexico City Hospital patients Leukemia (N = 1455) Never/0–3.9/4–6.9/7–12.9/> 13 Interview of mothers Matched: none.Adjusted: none. 5 Serious
Lingappa, 2018, India [71] 2015 HC-CS 120/120 1–15 Kidwai Institute of Oncology the elder sibling of the cases Leukemia (N = 120) Never/< 4/4–6/6–9/ 9–12/13–18/ 19–24/> 24;Ever vs. never Interview of mothers Matched: parental education, socioeconomic status, and the region of residence.Adjusted: none. 5 Moderate
Rafieemehr, 2019, Iran [72] 2015–2018 HC-CS 125/130 1–12 Be’sat Hospital of Hamadan University of Medical Sciences Hospital (healthy subjects) ALL (N = 125) Ever vs. never; 1–3/4–6/7–9/ 10–12/≥ 24 Interview of parents Matched: age, sex, gender, and residence location.Adjusted: none. 6 Serious
Bauer, 2020, France [73] 2015–2018 PC-CS 112/996 0.5–11 ESTELLE, 2010–2011 Population Wilms tumor (N = 112) Never/ < 3/3–5/≥ 6; Ever vs. never Telephone interview of mothers Matched: age and gender. Adjusted: age, gender, birth order, and maternal age. 7 Moderate
  1. Abbreviations: PC-CS population-based case-control study, HC-CS hospital-based case-control study, CNS central nervous system, ALL acute lymphoblastic leukemia, AML acute myeloid leukemia, AnLL acute nonlymphoblastic leukemia, HD Hodgkin’s disease, NHL non-Hodgkin lymphoma, NA not applicable, UKCCS United Kingdom Childhood Cancer Study, Aus-ALL The Australian Study of the Causes of Acute Lymphoblastic Leukemia, Aus-CBT Australian Study of Childhood Brain Tumors, STS Soft-tissue sarcoma, ESCALE and ESTELLE The ESCALE and ESTELLE studies were two nationwide population-based studies in France, designed to investigate the role of environmental, infectious, and genetic factors of childhood cancers
  2. 1No. of cases/controls in breastfeeding analysis; 2Mean ± SD of age in cases and controls were 12.1 ± 2.72 and 13.5 ± 2.85, respectively, while age range had not been reported