This article has supplementary material on the web site: www.jdbp.org.
Index terms: preterm, intervention, neurodevelopment, cognition, motor abilities
Abstract
Objective:
To systematically review nonpharmaceutical interventions aiming to enhance neurodevelopment in preterm children and adolescents (aged 4–18 years).
Method:
A systematic review of the literature was conducted for all studies published up to May 1, 2022, across Medline, Web of Science, and PsycINFO databases. Studies were evaluated for inclusion by 2 independent reviewers using predetermined inclusion criteria. The Risk of Bias In Non-randomized Studies of Interventions and the Cochrane risk-of-bias tool for randomized trials (RoB 2) tools were used to assess bias in the selected studies.
Results:
Of the 1778 articles identified, 23 were included. Quality assessment revealed moderate bias in 52.2%, low bias in 21.7%, and serious bias in 26.1%. The selected studies comprised 60.9% randomized controlled trials and 21.7% pre- versus postdesigns. Interventions included Cogmed Working Memory Training® (43.5%), BrainGame Brian (13%), physiotherapy (13%), and others (30.4%). Qualitative analysis showed the limited impact of interventions on neurodevelopmental outcomes in preterm children aged 4–18 years.
Conclusion:
Despite recent efforts to use more rigorous methodologies, current research on school-age interventions for preterm neurodevelopment exhibits methodological limitations. There is a pressing need for well-designed, large-scale clinical trials to evaluate the efficacy of nonpharmaceutical interventions in this vulnerable population.
Children born preterm (born <37 weeks of gestation) are at increased risk of neurodevelopmental disabilities in the long term.1 Even in children without obvious neurological deficits, subtle challenges have been reported, including lower cognitive test scores and increased behavioral and motor problems.2 The magnitude of the effect of preterm birth on cognitive and behavioral outcomes at school age has been found to be directly proportional to their immaturity at birth, with children born extremely preterm (EPT) (born <28 weeks of gestation) having high rates of intellectual, learning, behavioral disabilities and motor impairment.3 These neurodevelopmental difficulties have an important impact on academic achievement and daily life activity with life-long consequences.4,5 In this context, the impact of nonpharmaceutical intervention at school age has become a topic of concern for preterm-born children and adolescents to decrease and prevent the long-term deleterious effect of prematurity on neurodevelopmental outcomes.
A myriad of interventions have been developed and implemented for preterm infants and their parents in the neonatal intensive care unit or in different contexts during the first 2 years of life, including therapeutic developmental interventions targeting the infant or psychosocial support and parent education.6 However, interventions specifically designed for school-aged preterm children and adolescents are rare, and their potential benefits remain poorly understood.
Given the significant neurodevelopmental challenges faced by preterm children and adolescents, there is a pressing need to describe existing school-age interventions in this population and assess their potential benefits. To address this gap, this study aimed to systematically analyze the current literature on all school-aged nonpharmaceutical interventions conducted in preterm children and adolescents aged 4–18 years and targeting neurodevelopmental outcomes. The decision to focus exclusively on nonpharmaceutical interventions arises from their potential to address a broad spectrum of neurodevelopmental difficulties, encompassing motor, cognitive, and socio-emotional domains. In addition, the growing interest in these interventions in recent years underscores the necessity for a systematic review to provide insights into this evolving area.
METHODS
This review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines.7 The review protocol was registered on PROSPERO (28/06/2022): CRD42022340509.
Eligibility Criteria
Selected studies included school-age nonpharmaceutical interventional studies conducted in preterm children and adolescents, i.e., born before 37 completed weeks of gestation, aged 4–18 years, and targeting neurodevelopmental outcomes. All neurodevelopmental outcomes were considered. Studies conducted in children younger than 4 years and in young adults older than 18 years were excluded.
Human studies, clinical trials, and randomized controlled trials (RCT) published up to May 1, 2022, in peer-reviewed journals in English and French were considered for inclusion. Feasibility, protocols, and review studies were also examined but not considered for inclusion, as they could potentially provide complementary information for the included studies.
Search Strategy and Information Source
This systematic review was performed in July 2022, targeting nonpharmaceutical interventional studies conducted in children born prematurely aged 4–18 years and aiming to improve neurodevelopmental outcomes. Searches were conducted on Web of Science, PsycINFO, and PubMed. The search strategy for Web of Science consisted of using the following terms: “preterm OR pre-term OR low birth weight (Title) and intervention OR training (Title) not early (Title) not infant (Title)”. The search strategy for PsycINFO included the keywords: preterm (any field) AND intervention (any field) NOT infant (Title); the age groups: preschool age (2–5 years), school age (6–12 years), and adolescence (13–17 years) were selected. The search strategy for PubMed consisted of the following terms: “((preterm[Title/Abstract] OR pre-term[Title/Abstract] OR low birth weight[Title/Abstract] OR extremely low birth weight (ELBW) [Title/Abstract] OR VLBW[Title/Abstract]) AND (training OR intervention)) NOT (infant[Title] OR early[Title]))”.
In addition, citation lists of eligible articles were screened for any relevant publications that could meet the inclusion criteria of the current systematic review.
Study Selection
Studies were included in the current systematic review if they met the predefined inclusion criteria. Two independent expert assessors (R.H.V.L. and V.S.) screened the titles and abstracts generated from the search strategy for eligibility after removing duplicates. Following the initial abstract review, the authors retrieved and evaluated the full-text articles that were potentially eligible for final inclusion based on the abstracts. Disagreements between individual judgments were resolved through mutual consensus. The EndNote and Excel software were used for article extraction and management.
Data Collection Process
Two authors (R.H.V.L and V.S.) extracted data from each included article independently using an Excel sheet. Disagreements between individual judgments were resolved through consensus.
The following data were extracted from the included articles: (1) type of the study (e.g., feasibility, effectiveness, efficacy, review); (2) study participant demographics and study groups (e.g., preterm, full-term; gestational age, age at assessment, gender); (3) methodology of the study (e.g., RCT, non RCT); (4) intervention (e.g., name, format, duration, aims, target); (5) comparison group (e.g., pre vs post, intervention vs placebo vs active control); and (6) primary outcome variables.
Study Risk of Bias Assessment
To evaluate the quality of included studies and risk of bias, we assessed studies using the Risk Of Bias In Nonrandomized Studies of Interventions (ROBIN-I) tool for all studies included in the systematic review.8,9 The risk of bias judgments in ROBINS-I assesses bias across 7 domains: (1) bias due to confounding: evaluates the presence of confounding that systematically differ between groups; (2) bias in selection of participants: evaluates systematic errors arising from the process of participant recruitment, introducing differences in characteristics between groups; (3) bias in classification of interventions: evaluates the presence of misclassified intervention status; (4) bias due to deviations from intended interventions: evaluates systematic differences between the care provided to experimental intervention and comparator groups, beyond the assigned interventions; (5) bias due to missing data: evaluates systematic loss in the availability of outcome data, potentially resulting in an incomplete or skewed representation of study results; (6) bias in measurement of outcomes: evaluates inaccuracies in the assessment or measurement of study outcomes; and (7) bias in selection of the reported result: evaluates the systematic choice of specific outcomes for reporting based on their statistical significance or other characteristics. Each domain is described thoroughly, and guidelines are provided by Sterne et al.8 Each domain was evaluated for low, moderate, serious, or critical risk of bias.
In addition, version 2 of the Cochrane risk-of-bias tool for randomized trials (RoB 2) was used to assess the risk of bias in randomized trials included in the current systematic review.10 RoB 2 assesses bias across 5 components: (1) bias arising from the randomization process: evaluates whether the randomization process was conducted appropriately to minimize selection bias, (2) bias due to deviations from the intended interventions: evaluates whether there were any deviations from the intended interventions and whether these deviations could lead to bias, (3) bias due to missing outcome data: evaluates systematic loss in the availability of outcome data, potentially resulting in an incomplete or skewed representation of study results, (4) bias in measurement of the outcome: evaluates inaccuracies in the assessment or measurement of study outcomes, and (5) bias in selection of the reported result: evaluates the systematic choice of specific outcomes for reporting based on their statistical significance or other characteristics. Each domain was evaluated for low, some concerns, or high risk of bias.
Of note, for both ROBIN-I and RoB 2 and following the guidance of RoB 2, the assessment of bias due to missing outcome data was estimated as moderate/some concerns if the outcome data were missing for more than 20% of the participants.10
Two authors (R.H.V.L. and V.S.) independently extracted data using the ROBIN-I and RoB 2 tools. For all studies, an overall risk-of-bias judgment was determined using ROBIN-I, and for randomized trials, an additional risk-of-bias judgment was determined using Rob 2. Any disagreements between the assessors were resolved through consensus.
Evidence Synthesis
We conducted a narrative synthesis summarizing the key findings of nonpharmaceutical interventions on primary outcomes in preterm children and adolescents aged 4–18 years. To enhance readability, we structured this comprehensive evidence synthesis by categorizing studies based on the type of intervention, followed by organizing studies from the same research group, and finally, arranging them by their respective publication years.
RESULTS
Study Selection
From database search, 1778 articles were extracted using Web of Science, PubMed, and PsycINFO (Figure 1). After removing duplicates, 1711 articles were screened for inclusion criteria. Using database search, 22 studies met the eligibility criteria. Using citation search, 2 studies were screened for inclusion criteria. After screening titles, abstracts, and full texts, both reviewers agreed to include 23 articles in the final review.
Figure 1.
PRISMA flow diagram.
Quality Assessment
The quality assessment of all studies included in the systematic review was completed using ROBIN-I (Table 1). Among the 23 studies included in the systematic review, the majority (56.5%) were classified with a moderate risk of bias (n = 13), while 17.4% had an overall low risk of bias (n = 4) and 26.1% had a serious risk of bias (n = 6). Among the 6 studies flagged for serious risk of bias, notable bias was identified in the category (1) bias due to confounding. This was attributed to the omission or absence of reporting of critical confounders, specifically neonatal risk factors and/or socioeconomic status, which were predetermined during the preliminary consideration of confounders in the ROBIN-I assessment process.
Table 1.
Quality Assessment of all Studies Included in the Systematic Review Using the Risk of Bias In Nonrandomized Studies of Interventions (ROBIN-I) Tool
| (1) Confounding |
(2) Selection of Participants |
(3) Classification of Interventions |
(4) Deviations from Intended Interventions |
(5) Missing Data |
(6) Measurement of Outcomes |
(7) Selection of Reported Results |
Overall Judgment for Risk of Bias | |
| Aarnoudse-Moens et al., 201811 | Serious | Low | Low | Moderate | Low | Moderate | Low | Serious |
| Anderson et al., 201812 | Low | Low | Low | Low | Low | Low | Low | Low |
| Brown et al., 201713 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Brown et al., 201814 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Brown et al., 202015 | Low | Low | Low | Low | Low | Low | Low | Low |
| Di Lieto et al., 202016 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Everts et al., 201717 | Low | Low | Low | Low | Moderate | Moderate | Low | Moderate |
| Garci-Bermudez et al., 201918 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Grunewaldt et al., 201319 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Grunewaldt et al., 201620 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Huotilainen et al., 201121 | Serious | Low | Low | Low | Serious | Low | Low | Serious |
| Jaekel et al., 202122 | Low | Low | Low | Low | Low | Low | Low | Low |
| Kelly et al., 202023 | Low | Low | Low | Low | Moderate | Low | Low | Moderate |
| Kelly et al., 202124 | Low | Low | Low | Low | Moderate | Low | Low | Moderate |
| Lee et al., 201725 | Serious | Low | Low | Low | Low | Moderate | Low | Serious |
| Lohaugen et al., 201126 | Serious | Low | Low | Moderate | Low | Moderate | Low | Serious |
| Pascoe et al., 201927 | Low | Low | Low | Low | Low | Low | Low | Low |
| Perez-Fernandez et al., 201728 | Serious | Low | Low | Moderate | Low | Moderate | Low | Serious |
| Perricone et al., 201229 | Serious | Low | Low | Moderate | Low | Moderate | Low | Serious |
| Siffredi et al., 202130 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Tseng et al., 201931 | Low | Low | Low | Low | Moderate | Low | Low | Moderate |
| Van Houdt et al., 201932 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
| Van Houdt et al., 202133 | Low | Low | Low | Low | Low | Moderate | Low | Moderate |
For each study, domains of ROBIN-I were evaluated as having a low, moderate, or serious risk of bias.
In addition, the quality assessment of randomized trials included in the current systematic review was completed using RoB 2 (Table 2). Of the 23 studies included in the current systematic review, 14 were RCT and assessed using RoB 2. Most (64.3%) were classified with some concerns (n = 9), while 28.6% had an overall low risk of bias (n = 4) and 7.1% had a high risk of bias (n = 1).
Table 2.
Quality Assessment of Randomized Trials Only Included in the Current Systematic Review Using the Version 2 of the Cochrane Risk-Of-Bias Tool for Randomized Trials (RoB 2)
| (1) Randomization |
(2) Deviations from Intended Interventions |
(3) Missing Outcome Data |
(4) Measurement of Outcomes |
(5) Selection of Reported Results |
Overall Judgment for Risk of Bias | |
| Anderson et al., 201812 | Low | Low | Low | Low | Low | Low |
| Brown et al., 201814 | Low | Low | Low | Some concerns | Low | Some concerns |
| Brown et al., 202015 | Low | Low | Low | Low | Low | Low |
| Di Lieto et al., 202016 | Low | Low | Low | Some concerns | Low | Some concerns |
| Everts et al., 201717 | Low | Low | Some concerns | Some concerns | Low | High |
| Grunewaldt et al., 201319 | Low | Low | Low | Some concerns | Low | Some concerns |
| Jaekel et al., 202122 | Low | Low | Low | Low | Low | Low |
| Kelly et al., 202023 | Low | Low | Some concerns | Low | Low | Some concerns |
| Kelly et al., 202124 | Low | Low | Some concerns | Low | Low | Some concerns |
| Pascoe et al., 201927 | Low | Low | Low | Low | Low | Low |
| Siffredi et al., 202130 | Low | Low | Low | Some concerns | Low | Some concerns |
| Tseng et al., 201931 | Low | Low | Some concerns | Low | Low | Some concerns |
| Van Houdt et al., 201932 | Low | Low | Low | Some concerns | Low | Some concerns |
| Van Houdt et al., 202133 | Low | Low | Low | Some concerns | Low | Some concerns |
For each study, domains of RoB2 were evaluated as having a low, some concerns, or high risk of bias.
In the study by Evert et al., 17 the discrepancy between ROBIN-I (rated as moderate overall) and ROB 2 (rated as high overall) results from differences in the criteria used for the overall risk of bias assessment between the 2 tools.
Study Characteristics
Detailed characteristics for each study, including characteristics of the participants, study design, description of the intervention, and main outcomes, are summarized in Table 3.
Table 3.
Characteristics of Included Studies
| Participants | Intervention | Primary/Secondary Outcome and Assessment Modalities | Timing of Assessment after Intervention | |||||
| Preterm Intervention Group | Preterm Control Group | Full-Term Group | Study Design | Intervention Program | Control | |||
| Aarnoudse-Moens et al., 201811 | n = 12; mean GA = 28.7; mean age = 10.2; age range = 9.5–11.0 | NA | NA | Pilot feasibility intervention study: pre- vs postdesign | BrainGame Brian [computerized training program targeting EF, 25 sessions (45 min each) for 6 wk] | NA | Neurocognitive measures: Visual WM, impulse control, cognitive flexibility, speed variability. Parent questionnaire: Attention | Immediately postintervention |
| Anderson et al., 201812 | n = 45; mean GA = 27.3; mean corrected age = 7.6; corrected age range = 7–7.9 | n = 46; mean GA = 26.9; mean corrected age = 7.6; corrected age range = 7–7.9 | NA | Double-blind placebo RCT | Cogmed RM [computerized training program targeting WM, 20–25 sessions (35–50 min each) for 5–7 wk] | Placebo Cogmed [placebo computerized program] | Neurocognitive measures: Academic achievement (primary outcome); WM, attention (secondary outcomes). Parent questionnaires: EF (secondary outcomes) | 2-wk 12-mo, and 24-mo postintervention |
| Brown et al., 201713 | n = 24; median GA = 26; mean corrected age = 49.6 mo; corrected age range = 4–4.5 | NA | NA | Analyses pre vs post conducted in the context of an RCT | Physiotherapy intervention program [group-based weekly session targeting gross motor function and postural stability, 6 sessions (60 min) and 30-min parent discussion and home program] | NA | Neuromotor measures: Goal attainment (primary outcome), motor coordination, postural stability, lower limb strength, and control (secondary outcome) | Immediately postintervention |
| Brown et al., 201814 | n = 24; median GA = 26; mean corrected age = 49.6 mo; Corrected age range = 4–4.5 |
n = 24; median GA = 26; mean corrected age 50.1 mo; corrected age range = 4–5 | NA | RCT | Physiotherapy intervention program [group-based weekly session targeting gross motor function and postural stability, 6 sessions (60 min) and 30-min parent discussion and home program] | Standard care | Parent questionnaire: Behavior | Immediately and 1-yr postintervention |
| Brown et al., 202015 | n = 24; median GA = 26; corrected age mean = 49.6 mo; Corrected age range = 4–4.5 |
n = 24; median GA = 26; corrected mean age 50.1 mo; corrected age range = 4–5 | NA | RCT | Physiotherapy [group-based weekly session targeting gross motor function and postural stability, 6 sessions (60 min) and 30-min parent discussion and home program] | Standard care | Neuromotor measures: Motor performance, postural control and balance, lower limb strength and control | Immediately and 1-yr postintervention |
| Di Lieto et al., 202116 | n = 19; mean GA = 31; mean age = 7.3; age range: 4.1–13.1 |
NA | NA | Stepped wedge randomized trial | Cogmed JM and RM depending on the age of the participant [computerized training program targeting WM, 25 sessions (35–50 min each) for 5 wk] | NA | Neurocognitive measures: Cogmed improvement index (primary outcome), neuropsychological assessment battery (secondary outcome) | Immediately postintervention |
| Everts et al., 201717 | Memo training—n = 23; mean GA = 30; mean age = 10.2; age range = 7–12 | BrainTwister—n = 23; mean GA = 29.9; mean age = 10.7; age range = 7–12/waiting list—n = 23; mean GA = 28.5; mean age = 9.8; age range = 7–12 | NA | RCT | Memo training [computerized training program targeting memory strategies, 4 sessions (60 min)] | Brain twister [computerized training program targeting intensive WM practice, 20 sessions (10 min)]/waitlist | Neuroimaging measure: Visual WM functional MRI paradigm. Neurocognitive measures: Verbal learning and recall, verbal and visual WM complex reasoning, mental arithmetic, divided attention, inhibition | Immediately postintervention |
| Garci-Bermudez et al. 201918 | n = 36; mean GA = 35.18; mean age = 5; age range = 4–5 | n = 32; mean GA = 34.73; mean age = 4.9; age range = 4–5 | NA | Intervention study: 2 groups (intervention and control) using pre- vs postdesign | PEFEN program [stimulation program with activities targeting EF, 2h 30 per wk] | Routine curricular intervention | Neurocognitive measures: Neuropsychological assessment battery; child questionnaire: maturity and development; parent questionnaire: EF | Immediately postintervention |
| Grunewaldt et al., 201319 | n = 20; GA≤32; mean GA = 28.8; age range = 5–6 | NA | NA | Stepped wedge randomized trial | Cogmed JM [computerized training program targeting WM, 5 sessions (10–15 min) a wk for 5 wk] | Treatment as usual | Neurocognitive measures: Cogmed training/improvement index, verbal and visual WM, attention, EF language, memory, and learning; parent questionnaire: hyperactivity/inattention, anxiety | 4-wk post intervention |
| Grunewaldt et al., 201620 | n = 20; mean GA = 28.8; mean age = 5.8; age range = 5–6 | n = 17; mean GA = 29.6; mean age = 5.4; age range = 5–6 | NA | Intervention study: 2 groups (intervention and control) 7-mo postintervention | Cogmed JM [computerized training program targeting WM, 5 sessions (10–15 min) a wk for 5 wk] | Treatment as usual | Neurocognitive measures: Visual and verbal WM, attention, EF, language, memory, and learning; parent questionnaire: hyperactivity/inattention, communication, daily living skills, socialization, and problem behavior | 4-wk and 7-mo post intervention |
| Huotilainen et al., 201121 | n = 11; mean GA = 27; age = 6 | n = 11; mean GA = 27; age = 6 | NA | Intervention study: 2 groups (intervention and control) using pre- vs postdesign | Audilex [computerized dyslexia remediation program, for 5 wk, minimum 130 min] | Entertaining computer games [easy and entertaining computer games of nonverbal visual tasks for 5 wk, minimum 130 min] | Neurocognitive measures: Mismatch negativity component of auditory event-related brain potentials to sound changes (primary outcome), Audilex ability test, reading and writing (secondary outcomes) | Immediately and 2-yr postintervention |
| Jaekel et al., 202122 | n = 33; mean GA = 32.73; mean age = 6.99 | n = 32; mean GA = 32.81; mean age = 6.91 | NA | RCT multicenter |
XtraMath [computerized training program targeting working mathematical skills, 5 session per wk minimum (10–15 min) for 5 wk] |
Cogmed JM [computerized training program targeting WM, 5 session per wk minimum (10–15 min) for 5 wk] |
Teacher questionnaire: Academic attainment (primary outcome); neurocognitive measures: Mathematical skills (secondary outcome) | Immediately and 1-yr postintervention |
| Kelly et al., 202023 | n = 23; mean GA = 27.0; mean corrected age = 7.9; corrected age range = 7–7.9 | n = 25; mean GA = 26.5; mean corrected age = 7.8; corrected age range = 7–7.9 | NA | Double-blind placebo RCT (participants included in neuroimaging analysis) | Cogmed RM [computerized training program targeting WM, 20–25 sessions (35–50 min each) for 5–7 wk] | Placebo Cogmed [placebo computerized program] | Neuroimaging measures: Cortical morphometry, white matter microstructure, n-back WM fMRI paradigm | 2-wk postintervention |
| Kelly et al., 202124 | n = 28; mean GA = 27.0; mean age = 7.8; corrected age range = 7–7.9 | n = 27; mean GA = 26.6; mean age = 7.8; corrected age range = 7–7.9 | NA | Double-blind placebo RCT (participants included in neuroimaging analysis) | Cogmed RM [computerized training program targeting WM, 20–25 sessions (35–50 min each) for 5–7 wk] | Placebo Cogmed [equivalent placebo computerized program] | Neuroimaging measure: Graph theoretical measures extracted from white-matter structural connectivity networks | 2-wk post intervention |
| Lee et al., 201725 | n = 12; mean GA = 28.3; mean age = 5.6; age range = 4–6 | NA | n = 10; mean age = 5.5; age range = 4–6 | Intervention study: 2 groups (intervention and full-term control) using pre- vs post- vs 5-wk follow-up design | Cogmed JM [computerized training program targeting WM, 5 sessions a wk (15 min) for 5 wk] | Cogmed JM [computerized training program targeting WM, 5 sessions a wk (15 min) for 5 wk] in full-term participants | Neurocognitive measures: Verbal and visual WM, visual attention; parent questionnaire: EF | Immediately and 5-wk postintervention |
| Løhaugen et al., 201126 | n = 16; mean GA = 25.8; mean age = 14.1; age range = 14–15 | NA | (i) Cogmed RM—n = 19; (ii) treatment as usual - n = 11; mean age = 14.3; age range = 14–15 | Intervention study: 3 groups (intervention and 2 control groups) using pre- vs post- vs 6-mo follow-up design | Cogmed RM [computerized training program targeting WM, 5 sessions a wk (30–40 min) for 5 wk] | Full-term participants: (i) cogmed RM [computerized training program targeting WM, 5 sessions a wk (30–40 min) for 5 wk]; (ii) treatment as usual | Neurocognitive measures: Cogmed training/improvement index, verbal and visual WM, verbal learning and memory; parent questionnaire: hyperactivity/inattention | Immediately and 6-mo postintervention |
| Pascoe et al., 201927 | n = 45; mean GA = 27.3; mean corrected age = 7.6; corrected age range = 7–7.9 | NA | NA | Double-blind, placebo-controlled RCT (participants included in the Cogmed arm) | Cogmed RM [computerized training program targeting WM, 20–25 sessions (35–50 min each) for 5–7 wk] | NA | Neurocognitive measures: Verbal and visual WM, intrinsic motivation for school learning, training-related intrinsic motivation | 2-wk, 12-mo, and 24-mo postintervention |
| Perez-Fernandez et al., 201728 | n = 1; BW = 1400g; age = 8 | NA | NA | Case study of intervention: pre- vs postdesign | Go/no go task [computerized training program targeting EF, 7 consecutive d (4 min per day)] | NA | Neurocognitive measures: Go/NoGo performance (omission, commission, total accuracy, reaction time) | Immediately postintervention |
| Perricone et al., 201229 | Moderately preterm—n = 35; mean GA = 34.6; mean age = 62 mo; age range = 57–67 mo/severely preterm—n = 20, mean GA = 29; mean age = 64 mo; age range = 61–66 mo | NA | NA | Intervention study: pre- vs postdesign | Rehabilitative training of 6 mo for children, parent, teacher, pediatrician/neonatologist targeting ADHD risk | NA | Parent questionnaire: ADHD detection; teacher questionnaire: ADHD detection. Observation during training program: selective attention, self-regulation, and problem-solving | Immediately postintervention |
| Siffredi et al., 202130 | n = 63; mean GA = 29.2; mean age = 12.24; age range = 10–13.9 | NA | NA | Stepped wedge randomized trial | Mindfulness-based intervention [group-based training program consisting in the ongoing monitoring of present-moment experience while attending to it with openness, nonjudgment, and acceptance, 8 sessions (l90 min) for 8 wk and invitation to practice daily at home] | Treatment as usual | Neurocognitive measures: Executive, socio-emotional abilities; self-reported questionnaire: Well-being, self-compassion, social goals; parent questionnaire: EF, behavior and socio-emotional abilities | Immediately and 3-mo postintervention |
| Tseng et al., 201931 | n = 12; mean GA = 27.4; mean age = 7.7; age range = 7–7.9 | n = 9; mean GA = 26.4; mean age = 7.8; age range = 7–7.9 | NA | Double-blind placebo RCT (participants included in neuroimaging analysis) | Cogmed RM [computerized training program targeting WM, 20–25 sessions (35–50 min each) for 5–7 wk] | Placebo Cogmed [placebo computerized program] | Neuroimaging measures: Resting state functional connectivity changes in intra– and inter–resting-state functional connectivity networks | 2-wk post intervention |
| Van Houdt et al., 201932 | n = 29; mean GA = 28.2; mean age = 10.2; age range = 8–12 | (i) Placebo—n = 26; mean GA = 28; mean age = 10.2; age range = 8–12; (ii) waitlist - n = 30; mean GA = 27.8; mean age = 10.3; age range = 8–12 | NA | Multicenter, double-blind, waiting list and placebo RCT | BrainGame Brian [computerized training program targeting EF, 25 sessions (30–45 min) for 6 wk] | (i) Placebo BrainGame Brian [equivalent placebo computerized program]; (ii) waiting | Neurocognitive measures: BrainGame Brian training performance, attention; child-questionnaire: Self-perception profile; parent questionnaire: behavior and socio-emotional abilities | Immediately and 5-mo postintervention |
| Van Houdt et al., 202133 | n = 29; mean GA = 28.2; mean age = 10.2; age range = 8–12 | (i) Placebo—n = 26; mean GA = 28; mean age = 10.2; age range = 8–12; (ii) waitlist—n = 30; mean GA = 27.8; mean age = 10.3; age range = 8–12 | NA | Multicenter, double-blind, waiting list, and placebo RCT | BrainGame Brian [computerized training program targeting EF, 25 sessions (30–45 min) for 6 wk] | (i) Placebo BrainGame Brian [equivalent placebo computerized program]; (ii) waiting | Parent questionnaire: inattention/hyperactivity (primary outcome), EF (secondary outcome); teacher questionnaire: inattention/hyperactivity (primary outcome); neurocognitive measures: verbal and visual WM, inhibition, cognitive flexibility, arithmetic, reading (secondary outcomes) | Immediately and 5-mo postintervention |
GA, gestational age; mean age relates to the mean age at baseline, and it refers to the mean age in years unless otherwise specified; if the age range is not provided, it is that this information was not available in the given publication; RCT, randomized controlled trial; EF, executive functions; WM, working memory; attention deficit hyperactivity deficit ADHD; Cogmed JM for preschoolers; Cogmed RM for school age.
Of the 23 studies included, 60.9% were RCT (n = 14), 21.7% used a pre- vs postdesign and included an intervention group and at least 1 control group (n = 5), and finally, 17.4% used a pre- vs postdesign with an intervention group only (n = 4). Different types of intervention were used: 43.5% of the studies used Cogmed Working Memory Training® (n = 10); 13% of the studies used BrainGame Brian (n = 3); 13% of the studies used physiotherapy intervention program (n = 3); and the remaining 30.4% of studies reported other interventions, including mindfulness-based intervention (MBI), Memotraining/Brain Twinster, PEFEN, Audilex, XtraMath, Go/NOGO, and rehabilitative training (n = 7).
Evidence Synthesis of the Impact of Nonpharmaceutical Interventions on Neurodevelopmental Outcomes in Preterm Children and Adolescents Aged 4–18 Years
This section provides a summary of the Comprehensive Evidence Synthesis available in the Supplementary Methods (Supplemental Digital Content 1, http://links.lww.com/JDBP/A468).
Cogmed
Cogmed Working Memory Training® is a computerized training program targeting working memory.34 In a pre-post design focusing on ELBW adolescents (n = 16), improved trained visuospatial and verbal working memory were found, as well as verbal tasks after training and at 6-months follow-up.26 Another study in preterm (n = 12) and full-term (n = 10) preschoolers showed enhanced nontrained working memory after Cogmed and at 5-week follow-up.25 Using a stepped wedge RCT, Cogmed in 20 VPT preschoolers led to improved trained and nontrained working memory, auditory attention, and memory skills.19 A follow-up after 7 months showed significant gains in narrative memory, memory for faces, and spatial span backward compared with controls.20 Another study in 19 preterm children with spastic diplegia demonstrated enhancements in trained working memory and visuospatial processing, inhibition, and phonological skills after Cogmed.16 In the IMPRINT study, a double-blind RCT compared Cogmed with a placebo training in 91 extremely preterm children; findings showed no significant improvement in working memory, attention, or academic skills over 2 weeks, 12 months, and 24 months post-training.12,35 Intrinsic motivation positively affected working memory, and vulnerable families showed greater improvement.27 Investigation of structural and functional neuroplasticity revealed no specific effects of Cogmed.23,24,31
Physiotherapy Intervention Program
Brown et al.15 conducted an RCT testing a group-based physiotherapy program in 50 extremely preterm preschoolers (intervention, n = 24; control, n = 26). Improved motor skills as well as behavioral and emotional scores were observed in the intervention and control groups, immediately postintervention and at 1-year follow-up.14 Individual goal achievement also increased after physiotherapy in the intervention group.13
BrainGame Brian
BrainGame Brian is a computerized program designed to enhance executive functions. A feasibility study with 12 VPT children showed improved visual working memory and speed variability. An RCT with 8-year-old VPT children and attentional difficulties (n = 85) found training benefits in trained executive functions only, with no overall improvement in other untrained attentional, executive, behavioral, and academic performances.32,33
Other Interventions
In 6-year-old ELBW children, a dyslexia remediation program (Audilex, n = 11) compared with an active control (n = 11) showed increased auditory processing immediately after the intervention but no improvement in reading skills.21
Perricone et al.29 implemented a 6-month rehabilitative training targeting ADHD risk in 55 preterm children (pre-post design). Results showed significant improvements in attention and self-regulation abilities.
In an 8-year-old preterm child using visual Go/NoGo training (pre-post design), sustained attention, cognitive flexibility, and attentional control improved after training.28
In an RCT conducted in 69 VPT children aged 7–12 years, Everts et al.17 investigated the effect of 2 different memory trainings. In both memory training groups, different trained and nontrained memory and cognitive performances improved immediately after training but not in the waiting group. Results also suggested that both memory trainings decreased brain activation during a visual working memory task, but no correlation was found between changes in behavioral scores and neural activation scores.
Garcia-Bermudes et al. 18 introduced the PEFEN program, incorporating mindfulness-based approaches, showing improvements in executive functions and memory compared with a control group (n = 68 preterm children aged 4–6 years).
This aligns with the RCT study of Siffredi et al.36 on MBI conducted in 56 VPT young adolescents aged 10 to 14 years. Findings showed enhancements in executive, behavioral, and socio-emotional competencies immediately after the intervention.
In an RCT, Jaekel et al.22 compared XtraMath (n = 33), an adaptive mathematics training, with Cogmed (n = 33) in 6-year-old preterm children. Results showed a gain in total academic performance immediately after XtraMath, although not sustained at the 12-month follow-up.
DISCUSSION
This study systematically reviewed all nonpharmaceutical interventions conducted to improve neurodevelopmental outcomes in preterm children and adolescents aged 4–18 years.
Despite extensive evidence of neurodevelopmental difficulties associated with preterm birth during childhood and adolescence,5,37 we only identified a total of 23 nonpharmaceutical intervention studies published. Most of the studies were RCT (60.9%), but we also included studies using a pre- vs postdesign with at least 1 control group (21.7%), and studies using a pre- vs postdesign without control group (17.4%).
We identified 10 types of interventions aimed at improving neurodevelopmental outcomes in preterm children and adolescents aged 4–18 years, including Cogmed, BrainGame Brian, physiotherapy intervention program, MBI, Memotraining/Brain Twinster, PEFEN, Audilex, XtraMath, Go/NOGO, and rehabilitative training. The most frequently used intervention was Cogmed, which was used in 43.5% of the selected studies. Preliminary studies with designs such as pre-post intervention or stepped wedge randomized trials reported significant cognitive gains up to 7 months postintervention. However, the IMPRINT study, conducted with a large sample size (n = 91) and showing low risk of bias, found no significant evidence of Cogmed on working memory, attention, academic skills, or brain structure and function.12,23,24,27,31 BrainGame Brian was used in 13% of the selected studies. After encouraging results of a feasibility pilot study with serious risk of bias,11 an RCT showed no evidence of a beneficial effect of BrainGame Brian in any outcomes, including academic, executive, attentional performances and behavioral functions.32,33 A research team implemented an RCT of a physiotherapy intervention program, which was documented in 3 separate studies, constituting 13% of the selected studies. In this RCT, there was no specific effect of the intervention on 4-year-old EPT children for gross motor (low to moderate risk of bias13,15) and for behavioral outcomes (moderate risk of bias14) immediately after the intervention, but also at 1-year follow-up. The remaining 30.4% of studies each focused on a distinct type of intervention. Notably, Jaekel et al.22 conducted a low-bias study on adaptive mathematics training (XtraMath), which showed short-term academic performance improvements compared with an active control group (Cogmed), although these gains were not maintained at 12 months. Three studies with moderate bias risk reported positive outcomes in preterm children and adolescents from different interventions: Everts et al.17 found significant memory performance and brain activity improvements from 2 memory training programs; Garcia-Bermudes et al.18 reported cognitive improvements from the PEFEN program; and Siffredi et al.30 observed improvements in executive function, behavior, and socio-emotional competencies immediately after an MBI. Finally, 3 studies showed high risk of bias when exploring the impact of a dyslexia remediation program Audilex on reading skills,21 of a rehabilitative training on attention and executive functions,29 and of a Go/No Go training on inhibitory control.28
Throughout the completion of this systematic review, we identified numerous difficulties and aspects to consider when conducting interventional studies in preterm children and adolescents.
First, nonpharmaceutical interventional studies targeting neurodevelopment in school-age children and adolescents are highly demanding for participants and their families. In the case of RCT, in addition to the intervention itself, it requires families to complete several assessments that can also be highly demanding. Altogether, it usually requires substantial resources from the family to take part in this kind of research. In many studies, we observe low participation and high dropout rates. As an example, in Lee et al.25, among the initial 120 VPT preschool children invited to complete a Cogmed intervention, 15 agreed to participate and 12 completed all training sessions. These challenges likely play a substantial role in the small sample sizes observed in some of the studies included. Notably, 39% of the studies had a sample size smaller than 30 participants.
Second, a challenge faced by researchers is the use of an equivalent and methodologically valid active control intervention. Active control groups should be engaged in activities that are cognitively demanding and trigger positive expectations about their effectiveness in the participants. Therefore, control activities should differ from the cognitive-training program regarding only the key element that is hypothesized to enhance the target cognitive skill.38 In the case of nonpharmaceutical interventional studies targeting neurodevelopment, especially for noncomputerized intervention, this is a challenging task.
Third, several studies included in this systematic review failed to blind participants, families, and/or assessors that represent an important risk of bias in the measurement of outcomes (e.g., Studies 19,30,32). The inclusion of an active control group, as discussed above, could facilitate blinding of participants and their families. Blinding of outcome assessors aims to prevent systematic differences in measurements between intervention groups. It requires the research team to have sufficient personnel to effectively organize the blinding process.
Fourth, interindividual variability in response to interventions is an important factor to consider, especially in a population such as preterm children and adolescents, who are highly heterogeneous. This point has been raised in several studies included in this systematic review.12,30 Moreover, the examination of factors that have been associated with variation in improvement in outcomes, such as intrinsic motivation and family environment, is also important to consider.27 This is again a highly challenging task as it requires large sample size to capture interindividual variability and examine influential factors that may help identify children most likely to benefit from such training.
Finally, it is noteworthy that studies included in our review included participants up to 15 years old, with the eldest cohort observed in the study conducted by Løhaugen et al.26 While many studies recognized cognitive and socio-emotional difficulties in preterm adolescents and young adults,39 the current systematic review highlights a notable gap in nonpharmaceutical interventional studies focusing on preterm adolescents beyond the age of 15 years. Considering the importance of the adolescence and young adulthood period, especially for long-term health trajectory and well-being, addressing this gap represents a crucial avenue for future studies.
When discussing the findings of the current systematic review, it is important to acknowledge both its strengths and limitations. This review stands out for its strong and reliable methodology, which follows the PRISMA guidelines. The methodology includes the use of multiple databases, extensive inclusion criteria, and the input of 2 independent reviewers.
A limitation of this study is that a meta-analysis might have provided additional insights into our research question. Nevertheless, the included studies exhibit considerable heterogeneity and we felt that the intervention studies reviewed here were too varied in their design, participants, focus, and outcomes measured to warrant such analysis. Another limitation is the exclusive focus on nonpharmaceutical interventions. The inclusion of studies implementing pharmaceutical interventions could have expanded the range of options for addressing neurodevelopmental outcomes, particularly in targeting attention preterm children and adolescents.40
To conclude, the finding underlines the scarcity of research exploring the effect of nonpharmaceutical interventions conducted to improve neurodevelopmental outcomes in preterm children and adolescents. Upon qualitative analysis, nonpharmaceutical intervention studies involving preterm children and adolescents aged 4–18 years displayed varying outcomes. Notably, interventions like Xtra-maths, certain memory training programs, and MBI seem to show promising results even when using robust methodologies and acceptable sample sizes. However, concerns persist regarding the transferability of effects and the maintenance of these effects over the long term. Hence, it remains an urgent need for well-designed large-scale clinical trials to explore the effectiveness of nonpharmaceutical interventions in this population.
Data statement: All data collected for this article, including data extraction tables and the statistical analysis, will be available from the publication date. Requests to access these data should be made to the corresponding author.
Supplementary Material
Footnotes
The authors declare no conflict of interest.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.jdbp.org).
REFERENCES
- 1.Allotey J, Zamora J, Cheong‐See F, et al. Cognitive, motor, behavioural and academic performances of children born preterm: a meta‐analysis and systematic review involving 64,061 children. BJOG. 2018;125:16–25. [DOI] [PubMed] [Google Scholar]
- 2.Bhutta AT, Cleves MA, Casey PH, et al. Cognitive and behavioral outcomes of school-aged children who were born preterm: a meta-analysis. JAMA. 2002;288:728–737. [DOI] [PubMed] [Google Scholar]
- 3.Johnson S, Strauss V, Gilmore C, et al. Learning disabilities among extremely preterm children without neurosensory impairment: comorbidity, neuropsychological profiles and scholastic outcomes. Early Hum Dev. 2016;103:69–75. [DOI] [PubMed] [Google Scholar]
- 4.Hee Chung E, Chou J, Brown KA. Neurodevelopmental outcomes of preterm infants: a recent literature review. Transl Pediatr. 2020;9:S3–S8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pascoe L, Anderson PJ. Very preterm children and the impact on neurodevelopmental outcomes. Diagnosis, Management and Modeling of Neurodevelopmental Disorders; 2021:265–274. [Google Scholar]
- 6.Beltrán MI, Dudink J, de Jong TM, et al. Sensory-based interventions in the NICU: systematic review of effects on preterm brain development. Pediatr Res. 2022;92:47–60. [DOI] [PubMed] [Google Scholar]
- 7.Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg. 2021;88:105906. [DOI] [PubMed] [Google Scholar]
- 8.Sterne JAC, Hernán MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ Open. 2016;355:i4919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sterne JAC, Higgins JPT, Elbers RG, et al. Risk of bias in non-randomized studies of interventions (ROBINS-I): detailed guidance, updated 12 October 2016. Available at: http://www.riskofbias.info
- 10.Sterne JAC, Savović J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. [DOI] [PubMed] [Google Scholar]
- 11.Aarnoudse-Moens CSH Twilhaar ES Oosterlaan J, et al. Executive function computerized training in very preterm-born children: a pilot study. Games Health J. 2018;7:175–181. [DOI] [PubMed] [Google Scholar]
- 12.Anderson PJ, Lee KJ, Roberts G, et al. Long-term academic functioning following cogmed working memory training for children born extremely preterm: a randomized controlled trial. J Pediatr. 2018;202:92–97.e4. [DOI] [PubMed] [Google Scholar]
- 13.Brown L Burns YR Watter P, et al. Goal attainment scaling to evaluate intervention on individual gains for children born extremely preterm. Pediatr Phys Ther. 2017;29:215–221. [DOI] [PubMed] [Google Scholar]
- 14.Brown L, Burns YR, Watter P, et al. Behaviour of 4 to 5-year-old nondisabled ELBW children: outcomes following group-based physiotherapy intervention. Child Care, Health Development. 2018;44:227–233. [DOI] [PubMed] [Google Scholar]
- 15.Brown L, Burns YR, Watter P, et al. Randomised clinical trial of group-based physiotherapy in extremely low birthweight children with minimal/mild motor impairment: a preliminary study. J Paediatrics Child Health. 2020;56:727–734. [DOI] [PubMed] [Google Scholar]
- 16.Di Lieto MC, Pecini C, Brovedani P, et al. Adaptive working memory training can improve executive functioning and visuo-spatial skills in children with pre-term spastic diplegia. Front Neurol. 2020;11:601148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Everts R Murner-Lavanchy I Schroth G, et al. Neural change following different memory training approaches in very preterm born children - a pilot study. Dev Neurorehabil. 2017;20:14–24. [DOI] [PubMed] [Google Scholar]
- 18.Garcia-Bermudez O, Cruz-Quintana F, Perez-Garcia M, et al. Improvement of executive functions after the application of a neuropsychological intervention program (PEFEN) in pre-term children. Child Youth Serv Rev. 2019;98:328–336. [Google Scholar]
- 19.Grunewaldt KH, Løhaugen GC, Austeng D, et al. Working memory training improves cognitive function in VLBW preschoolers. Pediatrics. 2013;131:e747–e754. [DOI] [PubMed] [Google Scholar]
- 20.Grunewaldt KH, Skranes J, Brubakk AM, et al. Computerized working memory training has positive long‐term effect in very low birthweight preschool children. Develop Med Child Neurol. 2016;58:195–201. [DOI] [PubMed] [Google Scholar]
- 21.Huotilainen M, Lovio R, Kujala T, et al. Could audiovisual training be used to improve cognition in extremely low birth weight children? Article. Acta Paediatr. 2011;100:1489–1494. [DOI] [PubMed] [Google Scholar]
- 22.Jaekel J, Heuser KM, Zapf A, et al. Preterm children's long-term academic performance after adaptive computerized training: an efficacy and process analysis of a randomized controlled trial. Pediatr Res. 2021;89:1492–1499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kelly CE, Thompson DK, Chen J, et al. Working memory trainPng and brain structure and function in extremely preterm or extremely low birth weight children. Hum Brain Mapp. 2020;41:684–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kelly CE Harding R Lee KJ, et al. Investigating the brain structural connectome following working memory training in children born extremely preterm or extremely low birth weight. J Neurosci Res. 2021;99:2340–2350. [DOI] [PubMed] [Google Scholar]
- 25.Lee CSC, Pei J, Andrew G, et al. Effects of working memory training on children born preterm. Appl Neuropsycholchild. 2017;6:281–296. [DOI] [PubMed] [Google Scholar]
- 26.Løhaugen GC Antonsen I Håberg A, et al. Computerized working memory training improves function in adolescents born at extremely low birth weight. J Pediatr. 2011;158:555–561.e4. [DOI] [PubMed] [Google Scholar]
- 27.Pascoe L, Spencer-Smith M, Wiley J, et al. Child motivation and family environment influence outcomes of working memory training in extremely preterm children. J Cogn Enhancement. 2019;3:396–404. [Google Scholar]
- 28.Perez-Fernandez C, Canovas R, Moreno M, et al. Go/NoGo training improves executive functions in an 8-year-old child born preterm article. Rev Psicol Clin Ninos Adolesc. 2017;4:60–66. [Google Scholar]
- 29.Perricone G, Morales MR, Polizzi C, et al. Rehabilitative training of preterm children's attention: a study on sustainability. J Pediatr Neonatal Individual Med. 2012;1:87–96. [Google Scholar]
- 30.Siffredi V, Liverani MC, Hüppi PS, et al. The effect of a mindfulness-based intervention on executive, behavioural and socio-emotional competencies in very preterm young adolescents. Scientific Rep. 2021;11:19876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Tseng CEJ, Pascoe L, Roberts G, et al. Working memory training is associated with changes in resting state functional connectivity in children who were born extremely preterm: a randomized controlled trial. J Cogn Enhancement. 2019;3:376–387. [Google Scholar]
- 32.van Houdt CA, Aarnoudse-Moens CSH, van Wassenaer-Leemhuis AG, et al. Effects of executive function training on attentional, behavioral and emotional functioning and self-perceived competence in very preterm children: a randomized controlled trial. Front Psychol. 2019;10:2100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.van Houdt CA, Van Wassenaer-Leemhuis AG, Oosterlaan J, et al. Executive function training in very preterm children: a randomized controlled trial. Article. Eur Child Adolesc Psych. 2021;30:785–797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Klingberg T, Fernell E, Olesen PJ, et al. Computerized training of working memory in children with ADHD-a randomized, controlled trial. J Am Acad Child Adolesc Psychiatry. 2005;44:177–186. [DOI] [PubMed] [Google Scholar]
- 35.Pascoe L, Roberts G, Doyle LW, et al. Preventing academic difficulties in preterm children: a randomised controlled trial of an adaptive working memory training intervention—IMPRINT study. Article. BMC Pediatr. 2013;13:12–144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Siffredi V Liverani MC Smith MM, et al. Improving executive, behavioural and socio-emotional competences in very preterm young adolescents through a mindfulness-based intervention: study protocol and feasibility. Early Hum Dev. 2021;161:105435. [DOI] [PubMed] [Google Scholar]
- 37.Larroque B, Ancel PY, Marret S, et al. Neurodevelopmental disabilities and special care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): a longitudinal cohort study. Lancet. 2008;371:813–820. [DOI] [PubMed] [Google Scholar]
- 38.Gobet F, Sala G. Cognitive training: a field in search of a phenomenon. Perspect Psychol SciJ Assoc Psychol Sci. 2023;18:125–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Crump C. An overview of adult health outcomes after preterm birth. Early Hum Dev. 2020;150:105187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Catalá-López F, Hutton B, Núñez-Beltrán A, et al. The pharmacological and non-pharmacological treatment of attention deficit hyperactivity disorder in children and adolescents: a systematic review with network meta-analyses of randomised trials. PloS One. 2017;12:e0180355. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.