1. Introduction

Cocaine use remains a major and growing public-health concern worldwide [1–4]. The UN Office on Drugs and Crime’s World Drug Report (2025) notes that the number of cocaine users continued to increase, being cocaine the fastest-growing illicit drug market globally [5]. The European Drug Report (2025) similarly highlights that cocaine, after cannabis, is the second most commonly used illicit substance, with rising availability across countries and signals of increasing health and social harms. The report documents a growing burden on treatment services and emergency departments (cocaine is now frequently reported among first-time treatment entrants and is prominent in acute drug-toxicity presentations), as well as upward trends in wastewater indicators in many cities, together suggesting wider geographical and social distribution of cocaine use [6].

Among the many public-health problems associated with cocaine use [7–10], children’s exposure to this substance has received growing attention [11,12]. This exposure may occur through different routes, including intrauterine, breastfeeding, accidental intake, passive inhalation, and intentional administration.

Cocaine use during pregnancy represents a clinically and public-health relevant concern, although the actual prevalence of PCE remains difficult to estimate because of limited national data, variability in ascertainment methods, and underreporting by mothers [13]. In a US national survey from 1992, 1.1% of women under 44 years reported cocaine use during pregnancy; another report found that, among pregnant women reporting any illicit drug use (2.8%), 10% reported cocaine use [13].

In the 1990s, commentary in the lay press and some scientific discussions framed prenatal cocaine exposure (PCE) as producing dramatic and unavoidable developmental sequelae for exposed children [14]; however, assessing long-term consequences of PCE is challenging, particularly with respect to the specificity and generalisability of cocaine-related effects across studies. The limited availability of human cohorts followed into adulthood and the pervasive influence of environmental and co-exposure factors indeed complicated efforts to isolate cocaine-specific effects [15]. Likewise, differences between study populations (such as variations in prenatal care, concurrent substance use, and living environment) have been identified as key reasons why conclusions about neurobehavioral effects remain difficult to generalise [16].

Given these uncertainties, and in light of the increasing global availability of cocaine together with the limited number of cohorts followed into adulthood, a contemporary, systematic synthesis of the human literature on long-term outcomes following children’s exposure to cocaine is timely. The present review therefore aims to provide an updated, developmentally informed overview of human studies that have assessed outcomes from infancy through adolescence and early adulthood, to summarise the domains in which effects have been reported, and to highlight methodological limitations and implications for prevention and intervention.

2. Materials and methods

This systematic review was conducted and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [17]. A protocol was developed prior to study selection to define the research question, eligibility criteria, and data extraction framework (S1 Text); the review was not registered in PROSPERO.

Studies evaluating and reporting long-term effects of children’s exposure to cocaine were searched in the bibliographic databases PubMed (including PubMed Central and Medline), and Scopus, from inception until August 2025. The search terms were related to exposure to cocaine in the paediatric population and its long-term effects. The following search string was used: (children AND exposure AND cocaine AND long-term AND effects). The PubMed search was run using Title/Abstract field tags, while the Scopus search was run using TITLE-ABS-KEY (title, abstract and keywords). All types of studies were searched. Narrative or systematic reviews and meta-analyses, as well as book chapters, editorials, and conference abstracts, were excluded but screened to identify other potential studies to be included. Cross-sectional studies, case series, and case reports were included. After removing duplicates from the different databases, one legal medical doctor (CC) and one PhD student (RM) screened the titles and abstracts of records identified to remove articles that were clearly irrelevant. The full texts of the selected articles were then reviewed by the two authors to define whether they met the inclusion criteria, consisting of the evaluation of long-term effects of children’s exposure to cocaine, whether occurring in utero or postnatally during childhood. For the purposes of eligibility, “long-term effects” were defined as outcomes assessed beyond the immediate neonatal or acute post-exposure period, rather than by applying a predefined minimum age at follow-up. This definition included developmental, cognitive, behavioural, neurobiological, or physical outcomes detectable after the acute phase, including manifestations that persist, emerge, or remain clinically relevant across subsequent developmental stages. Accordingly, studies assessing outcomes in infancy or toddlerhood were considered eligible when they addressed effects not limited to transient neonatal manifestations. In cases of disagreement between the two reviewers at either the title/abstract screening or full-text eligibility stage, discrepancies were resolved through discussion. When consensus could not be reached, a third senior author (CT) acted as an independent adjudicator to resolve conflicts.

Studies conducted on animal models and studies published before 2000 were excluded. The exclusion of studies published before 2000 was intended to focus on more contemporary cohorts with improved exposure assessment methods.

Data from each article included in the systematic review were extracted by the two reviewers. For each article, data on the following items were retrieved: author(s), year of publication, number of children enrolled, age of children, timing of exposure and method of its assessment, exposure to other substances, context, long-term effects at follow-up.

A narrative synthesis approach was adopted, structured by developmental domain (cognitive and neurodevelopmental, behavioural and emotional, neurobiological and functional, and physical growth and health outcomes) and age at assessment.

Risk of bias was assessed using a modified Newcastle–Ottawa Scale, with selected items adapted to better reflect the characteristics of the included studies; the modified tool is provided in S2 Text.

3. Results

The PRISMA flowchart summarizing the study selection process is shown in Fig 1 [17]. The search strategy allowed the identification of a total of 201 records. After removing duplicates, 153 articles remained for title and abstract screening. Of these, 118 articles were excluded at title-abstract screening, consisting in animal studies (n = 13), review articles (n = 14), published before 2000 (n = 36), or not relevant to the topic (n = 55). Therefore, 35 full-text articles were sought for retrieval; however, 2 reports could not be retrieved, and 33 full-text articles were assessed for eligibility. Of these, further 14 were excluded at full-text reading because they were animal studies (n = 1), review articles (n = 3), or not relevant (n = 10). By screening the reference lists of the 19 studies that met the eligibility criteria, 4 additional eligible studies were identified. Three further studies were included through additional targeted searching, as they referred to longitudinal cohorts already represented or cited in the included articles. Consequently, a total of 26 studies were included in the final review. The characteristics of the included studies are summarized in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Methodological approaches and findings of the reviewed studies.

https://doi.org/10.1371/journal.pone.0352587.t001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flow diagram.

https://doi.org/10.1371/journal.pone.0352587.g001

All included articles dealt with cocaine exposure that occurred in utero. The results are presented according to developmental domains, with attention to the age at which outcomes were assessed.

4. Discussion

This review aimed to explore evidence on long-term effects of children’s exposure to cocaine. By drawing exclusively on human cohorts followed from infancy into childhood, adolescence and early adulthood, the review aims to provide a focused, developmentally framed picture of the range of cognitive, behavioural, neurobiological and health outcomes reported in the literature. Interestingly, all studies meeting selection criteria and included in this review refer to exposures that occurred in utero.

Unlike previous reviews examining a broad range of substances (e.g., alcohol, nicotine) [42,43] or focusing on prenatal cocaine effects on a specific age (e.g., childhood) [44,45], this review examines the long-term effects of PCE across a wide developmental span, from infancy to young adulthood. This lifespan perspective allows for a more comprehensive understanding of the risks associated with in utero cocaine exposure. Moreover, whereas some prior syntheses have relied on animal models [46,47], this review draws exclusively on human research to characterise the nuanced developmental sequelae of PCE. By adopting this human-focused lens, the analysis provides insights directly applicable to real-world populations.

The reviewed studies collectively suggest that prenatal exposure to cocaine is associated with a broad spectrum of effects across multiple developmental domains. These sequelae are generally described as modest and domain specific. Early neurodevelopmental impairments, along with growth deficits, have been documented in infancy and toddlerhood [21,32], while later cognitive and behavioural sequelae tend to be focal (perceptual reasoning, specific language and phonological skills, executive/attention networks, and elevated externalizing behaviours). Specific language and phonological deficits have been reported from childhood to adolescence [22,24–28], while impairments in perceptual reasoning have been observed in middle childhood into young adulthood [16,30,31]. Behavioural elevations, particularly in externalizing problems, were described in several cohorts [14,35,36]. Neurobiological studies (although based on relatively small samples) point to focal alterations in white-matter microstructure and functional recruitment during tasks of inhibition and attention [23,37], along with atypical ERP responses during language processing [22]. Overall, these converging findings support the possibility of persistent adverse effects of prenatal cocaine exposure on brain development [21].

A recurring and central theme across the literature is that PCE rarely occurs in isolation. Concurrent prenatal exposure to alcohol, tobacco and marijuana was common across cohorts, and most samples were drawn from low-income, inner-city populations. For example, Arendt et al. [18] observed that both exposed and unexposed children from disadvantaged backgrounds often perform below expectations, and poverty-related risks can produce progressive developmental lags that may rival or exceed the impact of prenatal drug exposure. Socioeconomic adversity and polysubstance exposure therefore appear important contextual determinants that complicate causal attribution and magnify risk.

The pattern of results (consistent perinatal effects, selective cognitive deficits, focal neurobiological changes, and influence of postnatal context) suggests a multifactorial model. Biological effects of PCE (e.g., on fetal growth and brain development) are supported by mediational findings in some cohorts, but post-natal environmental risks may attenuate or explain observed associations. Thus, causal interpretation should be cautious: some impairments may be directly related to prenatal cocaine’s teratogenic potential, while others reflect the combined influence of prenatal polysubstance exposure and adverse postnatal environments.

Importantly, multiple cohorts highlighted that postnatal environment matters for later outcomes. According to Singer et al. [16], while some cognitive differences may have a biological basis, functional outcomes are modifiable through environmental interventions. Similarly, as Arendt [18] reported, although PCE may confer vulnerability in specific domains (for example, visual–motor skills), inadequate rearing environments are often stronger predictors of children’s developmental trajectories than prenatal exposure alone. Across studies, these observations are supported by evidence that foster/adoptive care, higher HOME scores, and higher caregiver vocabulary or IQ were associated with better cognitive or language outcomes among children with PCE, whereas violence exposure, lead exposure, and other environmental risks were associated with poorer outcomes [16,18,24–27,29,40]. These findings suggest an important, positive implication: targeted postnatal supports and interventions may reduce functional disadvantage and promote resilience. The reviewed evidence thus supports dual-action strategies that combine prenatal substance-exposure prevention with postnatal family- and environment-focused interventions.

5. Limitations and future perspectives

Several limitations of the primary literature constrain definitive conclusions. Heterogeneity across cohorts (exposure assessment methods, covariate control, outcome measures, ages at follow-up) is substantial; some reviewed studies relied primarily on maternal self-report or on maternal/infant urine toxicology to identify exposed children, whereas other studies were based on meconium testing, and only one employed hair analysis. In this regard, it should be noted that hair analysis is a very effective method for drug-use investigation and offers a long detection window, but the finding of cocaine in the hair of infants should be interpreted cautiously: infant hair is thinner and more porous and therefore more prone to external contamination, and in the first months of life it can be difficult to disentangle in utero incorporation from postnatal exposures [11,48]. This further highlights the challenges in accurately characterising exposure in the reviewed cohorts. Sample sizes for neuroimaging studies are small; moreover, polysubstance use and socioeconomic confounding remain difficult to fully disentangle. Given this substantial heterogeneity across studies, a narrative synthesis without quantitative meta-analysis was performed; consequently, pooled effect estimates could not be generated, limiting the ability to precisely quantify effect sizes. Finally, publication bias and the focus on low-SES cohorts may reduce generalisability to other populations.

Future research should prioritise large, well-characterised longitudinal cohorts with rigorous, multi-method exposure assessment, better control for co-exposures and socioeconomic confounders, and sufficiently powered neuroimaging and intervention studies to test causality and plasticity. Evaluations of postnatal interventions would be particularly valuable to determine which strategies most effectively affect outcomes.

6. Conclusions

This review suggests, on the one hand, that prenatal cocaine exposure is associated with cognitive, behavioural and neurobiological effects, and, on the other, that supportive postnatal environments may mitigate many functional outcomes. In particular, higher-quality home environments and higher caregiver IQ or vocabulary were associated with better cognitive or language outcomes among children with PCE, whereas violence exposure and other environmental risks were associated with poorer outcomes. Considering the observed importance of the postnatal environment, the reviewed evidence supports the implementation of dual-action strategies combining prevention of prenatal substance use with postnatal supports, including coordinated medical monitoring of infant health and growth, maternal substance-use treatment when needed, caregiver support, and early developmental services [13]. Thus, while some biological vulnerabilities may persist, clinical and public-health responses should focus on reducing modifiable environmental risks and promoting stable, enriched developmental contexts for children with PCE.

Supporting information