Table 1.

Methodology summary of the eligible articles.

Author, year Study	Sample sizes (at follow-up/recruited) Age at follow-up	Analyses comparison groups by marijuana exposure	Prenatal exposure to substances other than marijuana and other covariates that analyses controlled for.	Study biases and limitations (reported by the authors and additionally identified by the reviewers)
Hayes, 1991 Jamaica study	54/59 4 and 5 years	Non-users Light users (<10 marijuana cigars or ‘spliffs’ per week) Moderate (11–20) Heavy (21–70)	Quality of Housing Index	(a) Results may not be generalizable as sample consisted of lower income rural women. (b) Impossible to distinguish between exposure effects of marijuana and child’s environment. (c) Authors do not provide results of the analyses for 5-year-olds.
Fried, 1988 OPPS	153/217 1 year and 2 years	Joints per week (range 0–153, mean 15 in 12-month sample, and 18 in 24-month sample) Heavy use (>5 joints/week).	Cigarettes, alcohol Family income, maternal age and education, maternal caffeine, protein, and caloric intake during pregnancy, difficulties during pregnancy, maternal and paternal health history, exposure to X-rays or rubella, gestation, birth weight, parity, method of feeding, and the HOME scale.	(a) Prenatal marijuana use was confounded by nicotine and alcohol use. (b) Study results may not be generalizable.
Fried, 1990 OPPS	133/190 3 and 4 years	Infrequent/ no use Moderate (> 1 - < 6 joints/week) Heavy (≥ 6 joints/week)	Cigarettes, alcohol Family income, mother’s weight and pregnancy weight gain, age, education, nutrition, and the two drugs not of primary interest. Perinatal controls were sex, parity, gestation, birth weight, and HOME scale	(a) Volunteer subjects are a low risk sample which may represent a conservative estimate of drug effects. (b) Variance explained by maternal drug use was relatively small compared with the HOME test.
Fried, 1992 OPPS	139/190 5 and 6 years	Infrequent/no use (≤1 joint/week) Moderate (>1 - <6 joints/week) Heavy (≥6 joints/week)	Cigarettes, alcohol Family income, mother’s pregnancy weight and pregnancy weight gain, mother’s age at delivery, average level of parental education, parental relationship, predominant language spoken by the child, child’s gender, HOME scale, and the two drugs not of primary interest	(a) Instruments that provide a general description of cognitive abilities may not be capable of identifying nuances in neuro-behavior that may discriminate between marijuana and non-marijuana exposed children. (b) Very low risk sample which may represent a conservative estimate of drug effects. (c) Potency of marijuana preparations has increased several fold since the entrance of pregnant women in the study.
Fried, 1992 OPPS	127/190 6 years	Infrequent/no use Moderate (>1 - <6 joints/week) Heavy (≥6 joints/week)	Cigarettes, alcohol Family income, mother’s pre-pregnancy weight, mother’s caffeine intake and nutrition during pregnancy, pregnancy difficulties, mother’s age at delivery, average level of parental education, parity, child’s sex, predominant language spoken by child, parental relationship, the two drugs not of primary interest and the HOME scale	(a) Fetal drug exposure measurements do not distinguish timing of exposure or account for sporadic heavy use during pregnancy. (b) Measures of home environment were not statistically associated with attention-related outcomes, other postnatal factors not assessed may influence child’s performance on these tasks. (c) Interaction between drug exposure and parenting and/or personality is currently being investigated in this sample.
O’Connell, 1991 OPPS	56/56 6–9 years	Non-users Users (>1 joint/week) Reported range 1.5–50 joints/week (mean=14.4, standard deviation=15)	Cigarettes, alcohol Mother’s age at delivery, mother’s education level, father’s education level, highest occupational status of parents, family income, number of parents in the home, the number of parents working outside the home, the number of children in the family, the birth order of the subject child, the principal language of the home, the principal language of instruction in school, presence of problems in school, history of eye and/or ear infections, the need for visual correction, the presence of special conditions at the time of testing, and HOME scale	(a) Home environment measures are viewed as legitimate outcomes, rather than potential confounders. (b) Low risk sample which may represent a conservative estimate of drug effects. (c) Potency of marijuana preparations has increased since the entrance of pregnant women in the study.
Fried, 1997 OPPS	146/190 9–12 years	Infrequent/no use (≤1 joints/week) Moderate (>1 - <6 joints/week) Heavy (≥6 joints/week)	Cigarettes, alcohol Family income, mother’s age at delivery, mother’s weight before pregnancy, mother’s total pregnancy weight gain, average level of parental education, other maternal drug use, and prenatal passive smoke exposure. Postnatal variables: sex of the child, the home environment, the mother’s personality, child’s level of depression and anxiety, secondhand smoke exposure of child, and current maternal sociodemographic characteristics and marijuana use at the time of child’s testing.	(a) Small number of subjects in the group of children exposed to moderate marijuana use limits confidence in the results.
Fried, 1998 OPPS	146/190 9–12 years	No use Infrequent/moderate (>0 - <6 joints/week) Heavy (≥6 joints/week)	Cigarettes, alcohol Family income, mother’s age at delivery, mother’s weight before pregnancy, average level of parental education, other maternal drug use, and prenatal passive smoke exposure. Postnatal variables: sex of the child, home environment, mother’s personality, child’s level of depression and anxiety, secondhand smoke exposure of the child, current maternal sociodemographic characteristics and marijuana use at the time of child’s testing.	(a) Some mothers continued to use marijuana after the pregnancy (b) Data had extreme univariate outliers (z score >4): 2 marijuana and alcohol, one nicotine value.
Fried, 2000 OPPS	146/190 9–12 years	No use Infrequent/moderate (>0 - <6 joints/week) Heavy (≥6 joints/week)	Cigarettes, alcohol Average level of parental education, other maternal drug use, prenatal passive smoke exposure, and sex of the baby. Postnatal variables: home environment, current socioeconomic status, child’s gender, and the environmental tobacco smoke exposure of the child.	(a) Could not categorize marijuana use into three levels due to inadequate cell size (delineated into heavy use and infrequent or moderate or no use). (b) Unclear whether deficits observed in visuoperceptual tasks are due to the perceptual demands of these tests or due to one or more non-perceptual requirements that are necessary for their successful performance.
Day, 1994 MHPCD	672/763 3 years	Average daily number of joints (ranges: 0–8.8 in first trimester, 0–6.5 in second trimester, and 0–9.4 in third trimester)	Alcohol, tobacco, amphetamines, tranquilizers, heroin, cocaine Maternal education, current work status, family income, home environment, number and distance in age between siblings, maternal levels of: depression, anxiety, hostility, self-esteem, mother’s perception of how difficult the child was	(a) Only 55% of the children completed the quantitative reasoning subscale. (b) Significant differences between non-completion and age at assessment were found, but not by prenatal marijuana exposure. (c) The effects reported are not clinically significant for an individual.
Leech, 1999 MHPCD	608/763 6 years	Abstainers >0 to <0.4 joint/day 0.4 to <1 joint/day ≥1 joint/day	Alcohol, tobacco, amphetamines, tranquilizers, heroin, cocaine Child Characteristics: age at assessment, gender, number of hospitalizations, number of illnesses, race Environmental Characteristics: Home Screening Questionnaire, male in household, maternal work/school status Maternal Characteristics: Hostility, life events, maternal age	(a) The Continuous Performance Test varies across studies in terms of modality (visual, auditory), type of stimulus (color, letter, number, animal), and difficulty of task. It may not have been difficult enough, did not allow comparison of different types of commission errors, and did not include a measure of reaction time. (b) All subjects were assessed by Stanford-Binet test but results were not reported by marijuana exposure status.
Goldschmidt, 2008 MHPCD	648/763 6 years	Abstainers/light/ moderate (≥0 and <1 joint/day) Heavy (≥1 joint/day)	Alcohol, tobacco, amphetamines, tranquilizers, heroin, cocaine. Maternal variables: cognitive ability, age at delivery, ethnicity, current level of education, income, work status, marital status, depression, hostility, social support, number of life events Environmental variables: total number of people in the household, drug and alcohol problems of the man in the household, current home environment Child variables: sex, nutrition, number of siblings, poor speech/vision/hearing, number of injuries, hospitalizations, and illnesses	(a) The sample was predominantly of lower socio-economic status.
Goldschmidt, 2000 MHPCD	636/763 10 years	First trimester users: Abstainers Light/moderate (0–0.89 joints/day) Heavy (>0.89 joints/day) Second/third trimester users: No use Light use (0–0.4 joints/day) Moderate/heavy (>0.4 joints/day)	Alcohol, tobacco, amphetamines, tranquilizers, heroin, cocaine Maternal variables: Number of years of education, working/studying outside the home, monthly family income, race/ethnicity, presence of husband or boyfriend in the household, depression, hostility, and number of reported life events Child’s environment variables: cognitive stimulation and emotional support provided by the child’s family, level of overt aggression among family members, number of siblings, child in maternal custody, gender, age, number of illnesses, number of injuries	(a) Mothers reported 21 children (3.3%) taking medication for attention attention-deficit/hyperactivity disorder.
Richardson, 2001 MHPCD	636/763 10 years	No use Light (0–0.4 joints/day) Moderate (>0.4–0.89 joints/day) Heavy (>0.89 joints/day)	Alcohol, tobacco, amphetamines, tranquilizers, heroin, cocaine Maternal variables: education, monthly family income, race Child characteristics: Age, anxiety, gender, cognitive stimulation and emotional support provided by the child’s family, uncorrected vision problems	(a) Magnitude of marijuana effects was small and limited to only a few aspects of functioning. (b) Difficult to compare Continuous Performance Test measure of inattention to parental reports of inattention. (c) Possible that marijuana effects on these and additional domains may be found when the children reach 14 years of age.
Goldschmidt, 2004 MHPCD	636/763 10 years	Light/moderate (<1 joints/day) Heavy (≥1 joints/day).	Alcohol, tobacco, amphetamines, tranquilizers, heroin, cocaine Maternal variables: age, education, family income, presence of an adult male in the household, ethnicity, working status, depression, hostility, number of life events, and support from friends and relatives Child characteristics: home environment, child in maternal custody, number of siblings, age between oldest and youngest child, child’s gender	(a) Variables such as motivation, social skills, and parent involvement in child’s education were not taken into account. (b) Generalizability is somewhat limited as cohort is low-income and only women who had received prenatal care.
Faden, 2000 NMIHS	8285/9953 3 years	No use <1/month 1/month 2–3/month 1–2/week >3/week	Alcohol, tobacco (cocaine use collected but too rare to be analyzed) Birthweight, child’s exact age in months, child’s sex, mother’s race, mother’s level of education, and mother’s Hispanic status	(a) Parental report and self-report of marijuana use may cause reporting bias. (b) Biased estimates of effects from exposure – important covariates left out or incorrectly modeled in the regression analysis.
Noland, 2003 Prenatal cocaine exposure study	316/415 4 years	Exposed Unexposed No/light use Heavier use (>5 joints per week)	Cocaine, alcohol, tobacco Race, gender, birth mother characteristics (age, education, verbal ability, block design and picture completion, SES, psychiatric symptoms, and marital status), and current caregiver characteristics (education, verbal ability score, block design and picture completion, SES, psychiatric symptoms, marital status, and HOME interview).	(a) Prenatal marijuana exposure effect on speeded and organized responding may not be apparent until subsystem develops more fully. (b) Atypical levels of gestational stress associated with sample may limit generalizability.
Noland, 2005 Prenatal cocaine exposure study	330/415 4 years	Exposed Unexposed No/light use Heavier use (>5 joints per week)	Cocaine, alcohol, tobacco Gender, African-American ethnicity of birth mother, maternal age at birth, parity, prenatal care visit(s), maternal years of education, marital status, low socioeconomic status, biological and current caregiver mental functioning variables	(a) Prenatal substance exposure accounts for a very small percent of the variance in performance.
Bennett, 2008 Developmental effects of prenatal substance exposure study	185/231 4, 6 and 9 years	0 joints/day 0.01–0.5/day 0.51–1/day >1/day	Cocaine, alcohol, cigarettes, opiates, phencyclidine, tranquilizers, amphetamines, barbiturates Environmental risk, maternal verbal intelligence, gender, and neonatal health problems	(a) Main focus of study was cocaine exposure. (b) Maternal marijuana use was associated with cocaine, alcohol and tobacco use. (c) Results are not generalizable as study enrolled a convenience sample of urban, predominantly African American and low socio-economic status population. (d) Other environmental factors may have affected children’s IQ but were not controlled for in this study.
Carmody, 2011 Developmental effects of prenatal substance exposure study	210/321 6, 9 and 11 years	Joints/day (range 0.022–0.497)	Cocaine, alcohol, cigarettes, opiates, phencyclidine, tranquilizers, amphetamines, barbiturates Environmental risk, medical complications, and gender	(a) Main focus of study was cocaine exposure. (b) Maternal marijuana use was associated with cocaine, alcohol and tobacco use. (c) Results are not generalizable as study enrolled a convenience sample of urban, predominantly African American and low socio-economic status population. (d) Other environmental factors may have affected children’s IQ but were not controlled for in this study.
Chakraborty, 2015 IDEAL	165/170 4.5 years	Frequency of use (days per week): <1 1–4 5–7 Amount of drug (joint per occasion): Light (<1) Moderate (1–2) Heavy (>2)	Methamphetamine, nicotine, alcohol Sex, ethnicity, stereoacuity, visual acuity, and verbal IQ	(a) Results cannot be extrapolated beyond global motion perception or interpreted as marijuana having beneficial effects on fetal development. (b) Average motion coherence thresholds reported for non-drug exposed children are slightly elevated (worse) compared to previous studies of global motion perception in preschool children. (c) Study has a small sample size in which the majority of participants were poly drug users.

OPPS – Ottawa Prenatal Prospective Study, Canada

MHPCD – Maternal Health Practices and Child Development Project, USA, Pennsylvania

NMIHS – National Maternal and Infant Health Survey, USA

IDEAL – Infant Development, Environment, and Lifestyle study, New Zealand

IQ – Intelligence quotient