Table 5.

Epidemiologic studies on marijuana use and childhood cancers

Study location, period, author, referenceCancer siteCharacteristics of casesCharacteristic of controls or cohortExposure assessmentExposure categoriesRR or OR (95% CI)Adjustment for potential confounders and other notes
Multicenter, US, and
  Canada, 1980–1984,
  Robison
  et al. (31)
Childhood acute
  nonlymphoblastic
  leukemia
204 cases, identified in
  registry of Children's
  Cancer Study Group,
  diagnosed at <18 years of
  age, Response rate: 77.9%
204 subjects from random
  digit dialing, individually
  matched on date of
  birth, race, telephone
  area code, and exchange
  Response rate: 78%
Telephone interviews
  of mothers and fathers
  of subjects, with structured
  questionnaire
Maternal use of
  mind-altering drugs
  during or in
  the year
  before the pregnancy

  Paternal use







11.0 (1.42–85.20)
1.47, P = 0.32
  • Adjusted for date of birth, race, residence, telephone area code

  • 9 of 11 cases had maternal marijuana only

  • The authors reported that adjustment for mother's age, education, tobacco use, alcohol did not result in reduction in risk or loss of statistical significance

Multicenter, US,
  Canada, and
  Australia, 1983–1993,
  Wen et al.
  (32)
Childhood leukemia1,805 acute
  lymphoblastic leukemia
  cases, 528 acute
  myelogenous leukemia,
  age ≤18 months,
  selected from registration
  files of the
  Children's Cancer
  Study Group
2,723 subjects from
  random digit dialing,
  individually matched on year
  of birth, telephone
  area code,
  and exchange number
Telephone interviews
  of mothers and
  fathers of subjects,
  with structured
  questionnaire
Ever marijuana use
  by father
1.5 (P < 0.05)
  • Adjusted for year of birth, telephone area code, exchange number

  • Dose–response assessment not available

Pennsylvania, New Jersey,
  Delaware, US, 1980–1986,
  Kujiten et al.
  (33)
Childhood astrocytoma163 cases, identified
  through 8 hospital
  tumor registries, diagnosed
  at <15 years,
  Response rate: 80%
163 subjects from random digit dialing, individually matched on birth date, race, telephone exchangeTelephone interviews
  of mothers and
  fathers of subjects,
  with structured
  questionnaire
Gestational
  marijuana
  exposure
2.8 (0.9–9.9)
  • Adjusted for age, race, residence, dose–response relations not assessed

  • Data on paternal use not presented

Multicenter, US, 1982–1988, Grufferman et al. (34)Childhood rhabdomyosarcoma322 cases, identified
  in the registry of
  the Children's
  Cancer Study Group,
  diagnosed at 0–20 years
  of age
Response rate: 79.8%
322 subjects from random
  digit dialing, individually
  matched on age,
  sex, race
Telephone interviews of
  mothers and fathers of
  subjects, with structured
  questionnaire
Maternal use of
  marijuana
  Paternal use
3.0 (1.4–6.5)





2.0 (1.3–3.3)
  • Adjusted for age, sex, race, birthmarks on child, bleeding/cramping during pregnancy, prematurity of child

  • Factors associated with rhabdomyosarcoma in data were adjusted for

Multicenter, US,
  and Canada, 1989–1993,
  Trivers et al.
  (35)
Childhood acute
  myeloid leukemia
638 cases age
  <18 identified from
  the Children's
  Cancer Group. Response
  rate: 83%
610 matched controls.
  Matched to the
  cases on age and
  residential location.
  Using a random
  digit dialing procedure.
Response rate: 79%
Telephone interviews of
  mothers and fathers
  of subjects,
  with structured
  questionnaire
Maternal
Never
Ever

Paternal
Never
Ever

1.00 0.89 (0.66–1.19)


1.00 1.37 (1.02–1.83)
  • Adjusted for age at the index child's birth, parental education, household income, alcohol consumption and cigarette smoking before, during, and after pregnancy, maternal history of fetal loss before the index pregnancy, and birth order

  • Maternal and paternal marijuana use in specific time periods did not support an association

  • Frequency of marijuana use was not associated (data not presented in publication)

Multicenter, North
  America, 1992–1994,
  Bluhm et al.
  (36)
Childhood
  neuroblastoma
Mothers of 538 children with
  neuroblastomas identified
  through the Children's
  Cancer Group and
  Pediatric Oncology
  Group. Age: birth to
  19 years old.
Response rate: 73%
504 age-matched
  control identified
  by random digit dialing.
Response rate: 72%
Maternal interview by
  trained interviewers using
  a standardized
  questionnaire
Maternal
None
Any


Frequency in
first trimester
<1 pipeful per day
1 or more pipefuls per day

1
1.37 (0.77–2.49)



4.16 (1.52–14.61)

4.42 (1.09–29.58)
  • Adjusted on household income, age at diagnosis, other drugs used

  • Maternal marijuana use in specific time periods did not support an association, except for in the first trimester (OR, 4.75; 95% CI, 1.55–16.48)