The mechanisms through which cannabis affects the brain of the human fetus and causes neurochemical and neuroanatomical changes are not well known. Cannabinoid receptors are present in the placenta and they appear in the fetal brain at 14 weeks after conception and increase in density throughout the third trimester. Given its lipophilic nature, one third of the THC in the maternal plasma crosses the placenta and it can be excreted into breastmilk. Cannabis could alter the function of fetal cannabinoid receptors [14, 26] and it may lead to changes in the dopaminergic and opioid system [15, 16].

There is a high density of cannabinoid receptors in the frontal lobe and the cerebellum, and these structures show a parallel and late maturation when compared with other structures of the brain [25]. Studies on the exposure of human fetuses to cannabis have reported alterations and a different impact on the mesocorticolimbic system, which is in charge of the regulation of emotions [15, 16]. The CB1 receptor has been associated with the dopaminergic and opioid systems (neuropeptidase) [15], and it has been hypothesized that an increase of its expression in the hippocampus and amygdala of the fetus may suggest that these structures are more vulnerable to prenatal exposure to cannabis [15].

Intrauterine exposure to cannabis may causes different pattern of the anatomical organization of the CB1 mRNA expression in the mid-gestation fetal and adult human brain; has been found high CB1 mRNA expression in the fetal hippocampus and amygdala [15], a decrease in the expression of proenkephalin mRNA in the fetal striatum with a dose-dependent effect [20, 47], and an up-regulation of proenkephalin mRNA in the mesolimbic area in adult life [20, 47]. There is also an increased expression of u-opioids in the amygdala and a decrease of k-opioid receptor mRNA in the mid-dorsal thalamus [17–20, 47]. This all suggests that in utero exposure to cannabis fundamentally changes the systems that regulate the emotional life, such as the mesocorticolimbic pathway, and these changes may even be persistent in this individuals [17, 20, 47].

There is one study that assess the different consequences of the maternal use of cannabis in the newborn [9]. From a somatic perspective, the use of cannabis has been associated, both in the early and the late stages of pregnancy, with a higher risk of restricted fetal growth [9], low birth weight, a shorter birth length and a low head circumference [18, 27, 58, 59], hypertelorism, and epicanthus [9, 56]. From a cardiovascular perspective, an association has been observed with ventricular septal defect [28]. The in utero exposure to cannabinoids may result in a dysfunction of the T lymphocytes and a decreased immune response to viral antigens. These effects may be mediated by epigenetic mechanisms such as alterations of micro RNA, DNA methylation and modification of histone profiles. Therefore, prenatal exposure to cannabis may cause epigenetic changes that could have consequences on the later development or even long-term transgenerational effects [29]. In children who were exposed to cannabis in the prenatal stage, a thicker prefrontal cortex was observed when compared with non-exposed children [43] (Table 1).

From a psychiatric perspective, cannabis withdrawal syndrome has been described in newborns [9], although other authors have been unable to prove the existence of negative perinatal effects in children whose mothers used cannabis [58].

There are few review studies suggesting prenatal exposure to cannabis may be associated with mood and behavioral alterations, that could be related with affective mental disorders, and depressive symptoms, as well as ADHD [19, 31–34]. There are no studies that establish a connection with the presence of psychotic disorders [35].

There are longitudinal studies which describe behavioral and cognitive disorders associated with uterine exposure to cannabis, such as the Ottawa Prenatal Prospective Study [60] and the Maternal Health Practices and Child Development Project [36]. These studies did not find significant behavioral alterations in the newborns whose mothers used cannabis, but they observed that they sleep fewer hours and show habituation deficit to visual stimuli [36, 37, 61]. At age two, no cognitive alterations have been found on those children [38]. At age three, an alteration of short-term memory has been observed on verbal and abstract reasoning and on verbal skills, without any effect on intelligence [38].

The study carried out by Day et al. (2011) describes the association between prenatal exposure to cannabis and a low IQ during the school age [39]. An alteration of the executive functions regulated by the prefrontal cortex has also been described [32], including visual-spatial reasoning, response inhibition and working memory, which may last until ages 13–16 [32]. At these ages, the children may present higher impulsiveness and hyperactivity, lower attention capacity and a higher prevalence of delinquent behavior, which could be partially related to the alterations mentioned above [39]. Have been observed in two studies that the executive dysfunction could be present even into early adulthood [37, 40]. On the other hand, an increase in the use of cannabis and nicotine at ages 14–21 has been observed among individuals who had been exposed during the prenatal stage, particularly in men [37, 40]. Another longitudinal study, The Generation R Study [62], still ongoing, has observed an association between the maternal use of cannabis and an increase in aggressive behavior and attention disorders in girls at 18 months of age, and this association stops being statistically significant at 36 months [42, 62]. At 30 months, no differences were found in the nonverbal cognitive scales or in language development, regardless of sex [62]. Finally, a recent study establishes an association between the maternal use of alcohol and cannabis and Tourette syndrome [44].

A maternal questionnaire could not be considered an efficient screening tool for the detection of the maternal use of cannabis one of the reasons is the underreporting among pregnant women [11, 12, 46, 66], which justifies the need to use biomarkers [8]. Both the use of maternal hair and the meconium of the newborn have shown a prevalence of substance use that is higher than what was detected through clinical interviews [8].

Maternal hair provides direct information on THC use over the last months, or even years [8]. It offers a direct estimation of maternal exposure to drugs, but only an indirect estimation of the substances that reach the fetus through the placenta.

Meconium is the most widely used fetal matrix to reveal prenatal exposure to drugs of abuse [6]. It can be easily obtained, but it has a collection window of 72 h, it could be less sensitive for detection of exposure during the first trimester and testing is more expensive and less available [7, 8]. Screening of blood and urine samples during pregnancy is simpler and more readily available, but it only provides information about substance use over the last 24–48 h [10, 46].