Review procedures were based on recommendations from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Guidelines.10 The study question was based on the PECO strategy: Participants: Adolescents and adults with CP; Exposure: exposure to any factors shown to influence motor outcomes; Comparator: non-applicable; Outcome: any motor-related outcomes. The research question was defined as: Which factors are relevant for the motor outcomes of adolescents and adults with cerebral palsy?

This systematic literature review was registered in the International Prospective Registries of Systematic Reviews (PROSPERO): registration number CRD42022311783.

The checklist to assess study bias was scored based on the clear description of several relevant items, assigning 1 point for studies reporting the item and 0 for those that did not clearly report this item. The maximum score for each study is 16 points. Studies with scores between 12 and 16 were considered to have few methodological limitations, being classified as having good quality; scores between 7 and 11 points indicate moderate methodological limitations, with the study being classified as fair; a score below 7 points represents significant methodological limitations, and thus the study quality was classified as poor.13,14

The heterogeneity of the outcomes did not allow meta analyses, therefore the results were analyzed descriptively. The results were summarized based on the factors related to each category of motor-related outcomes.

Sixteen studies identified factors relevant that resulted in gross motor function decline. The decline in gross motor function in individuals with CP has been consistently described through GMFM, both in overall and in standing scores over development.26,27 This may be a relevant tool to follow-up youth with CP across transition, however validity for adults needs to be demonstrated. A prominent relevant factor that may contribute to the decrease in gross motor scores is the presence of pain.15,28,29 Pain is therefore a factor that should not be neglected by healthcare professionals working with individuals with CP as it may precede relevant functional losses.

As expected, results highlighted age as a frequent factor relevant for the decline in gross motor function in adolescents and adults with CP, along with GMFCS levels.29–36 The combination of age and GMFCS predicted Pediatric Evaluation of Disability Inventory (PEDI) mobility scores, with higher development and longer period in development for individuals with mild to moderate GMFCS levels (I-III), compared to those at more severe GMFCS levels.35 Similarly, these factors predicted earlier peak GMFM scores and significant declines in adolescence and adulthood in those between levels III to V.31,36 Several other studies supported the influence of the GMFCS level as a factor relevant for the decline in gross motor function in individuals with CP in the first decade of life and in adults, with results similar to those previously discussed.15,30,31,33,34,36–39 Collectively, these findings underscore the vulnerability of non-ambulant individuals to early motor decline.

Musculoskeletal conditions were also pointed out as relevant factors for gross motor change. These included muscle strength,30 ROM in the upper limbs and in hip, knee, and ankle joints, spinal alignment,28 and presence of stiffness and deformities.29 Such findings point to the relevance of monitoring and intervening early in life on these factors because interventions addressing these issues may be strategies to prevent motor decline. Among the interventions targeting musculoskeletal aspects, single-event multilevel surgery (SEMLS) has been suggested to favor stability of GMFCS levels.38

In summary, the findings are quite consistent in demonstrating that during transition to adulthood there is a decline in gross motor function and mobility, with age, GMFCS level, and body structure and function aspects such as pain and neuromusculoskeletal aspects being the primary relevant factors. Thus, monitoring these factors throughout the development of children, adolescents, and young adults with CP is important so timely interventions to prevent decline and/or maintain the functional level are offered.

Gait deterioration was reported as an indicator of motor decline in 11 studies. GMFCS level was pointed out as a relevant factor in several studies,15,16,19,26,31,39,40 often interacting with age to determine changes in gait.15,31,39,41,42 Other aspects of gross motor function such as the performance of the Timed Up and Go (TUG) test, the ability to walk and climb stairs, and the need for assistive devices were also relevant predictors of gait stability and decline.19 It is therefore relevant to consider the implementation of validated tests in transition programs to help predict the risk of gait deterioration.

The presence of pain was another relevant factor, predicting gait endurance,19 and gait in general,15 with reports of loss of the ability to walk in individuals who had pain.16 However, the impact of the association of pain and gait decline on functioning was not explored in the included studies, and should be addressed in the future. Another limitation was that the assessment of pain did not include detailed information such as type and location of the pain, which warrants further investigation.

The presence of musculoskeletal deformities was also reported as an important contributor to gait loss,40 and to the performance in gait tests.19,41 However, there is conflicting evidence on the impact of interventions targeting such deformities on gait. Bell et al.41 showed that gait function declined over 4 years in both functional and less functional walkers without surgery, illustrated by decline walking velocity and other kinematic parameters. In contrast, a group that underwent orthopedic surgery had a more positive prognosis, suggesting that this intervention may minimize the decline over time. Possible positive effects of surgical interventions were also reported by Gannotti et al.16 and Veerbeek et al.17 These findings are however in contrast with the decline after multilevel surgery over an 11-year period reported in another study.42 Also, Church et al.26 have suggested a mild decline (increased popliteal angle and decreased gait speed) in a 12-year follow-up after multilevel surgery. Given these findings, particularities such as type of surgery, participants' age, and other characteristics should be explored in future studies to elucidate the impact of surgery on motor prognosis.

The findings of this review highlight that gait decline is prevalent and relevant in the motor decline context. When untreated, it can result in loss of mobility and quality of life.43 Thus, the identification of relevant factors for changes in gait can prevent a cascade of other potentially aggravating events for health in general. Monitoring these factors should be a priority in transition planning.

The decline in neuromusculoskeletal functions (muscle strength, ROM, worsening of hip condition, among others) was related to factors such as GMFCS level, altered gait parameters, and age26,37,38,41,44 Interestingly, worsened conditions in the hips, knees, and ankles may already appear during childhood, as noted by Bell et al.41 in children followed over 4 years starting at ages as young as 5 years. Less functional walkers presented particular characteristics, such as more changes in the foot and ankle, which may relate to the increased risk of mobility decline in this subgroup and points to the importance of prevention by means of appropriate tone management and physical activity early on.41 Over a 12-year follow-up, however, changes in knee angles did not result in significant decline in the gait of young adults,16 which suggests that outcomes related to ROM must be accompanied by measures that inform about the capacity and performance of individuals, so that the proposed interventions can be functionally relevant.

One study focused more directly on hip status of non-ambulant children with CP,44 pointing that a postoperative hyperlordosis greater than 60° was related to the increased risk of worsening the hip condition at 5 years after surgical correction of scoliosis. The implementation of adequate postural management through positioning and assistive devices is therefore a critical strategy to prevent such complications. Although not reported in the included studies, other studies indicate that in addition to spinal alignment, other factors such as age, GMFCS level, and type of gait as determinants for hip dislocation.26,41 Considering the incidence of hip dislocation of up to 35 % (increasing with GMFCS) in CP, which can result in pain, reduced function, and reduced quality of life,44 the factors relevant for this condition should be closely monitored and addressed as early as possible. The implementation of hip surveillance, which is the strategy with the highest level of evidence to minimize complications related to hip dislocation, is necessary from the first year of life of children with CP and in some cases needs to continue until skeletal maturity.45,46

Given these findings, it is clear that musculoskeletal comorbidities and multimorbidities arise early and deteriorate with advancing age in adults with CP and may be more critical in the non-ambulant individuals. Thus, access to preventive services and therapeutic interventions at the right time can have a positive impact on the functioning and well-being of individuals with CP, especially those at higher risk for musculoskeletal complications, thus providing more favorable health conditions as they transition to adulthood.

In general, the reviewed studies showed a low risk of bias, which gives more credibility and support to the conclusions made in this review. However, heterogeneity in the use of assessment instruments was found, for example, the GMFCS in some studies was used as a criterion to indicate gait ability, while in others as an indicator of general gross motor function. There was also a lack of validated assessments for older age groups of people with CP. Although the GMFM is the gold standard in CP assessment, it is not validated for adults yet. This aspect also deserves attention in future studies.

Limitations of the study include the fact that the factors were selected based on their statistical significance and the lack of measures of the strength of the recommendations. We suggest that future reviews also take into account the clinical significance of the findings and use tools to provide accurate recommendations. Interestingly, indicators of functional decline that were not directly related to motor outcomes were assessed in some studies, including decline in self-care skills, health-related quality of life, and performance and satisfaction with lifestyle habits. These indicators should not be disregarded as they illustrate areas of functioning that are essential for wellbeing. Other factors identified as relevant for motor outcomes but not classified as significant as they did not reach the criteria established in this review included depression, emotional support,15 mental health,16 housing,41 balance,18 fatigue,15 and CP type.19 Their contribution should be investigated in future, high quality studies. Still, although the distribution of motor impairment and the presence of comorbidities have been previously pointed out as relevant factors in the literature,18 they were not identified in the results of this review. Such variability of outcomes and results is probably due to the multifactorial nature of motor decline, and indicates that individualized biopsychosocial care must be assured when planning transition services to adult life.

Regarding implications for physical therapy, the results of this review support the continuity of follow-up of individuals with CP during the transition to adulthood, especially the ones at higher risk for motor decline, while discharge due to lack of motor progress is not recommended as it may contribute to further deterioration in health status. Future studies should summarize evidence on recommended physical therapy interventions for this population. Studies that longitudinally follow adolescents and young adults with CP are also scarce, especially in low and middle-income countries, and the determinants for the motor outcomes in these countries need to be investigated in future studies.