A meta-analysis of the effects of early developmental prevention programs in at-risk populations on non-health outcomes in adolescence

Abstract

We present the results of a meta-analytic review of early developmental prevention programs (children aged 0–5: structured preschool programs, center-based developmental day care, home visitation, family support services and parental education) delivered to at-risk populations on non-health outcomes during adolescence (educational success, cognitive development, social–emotional development, deviance, social participation, involvement in criminal justice, and family well-being). This review improves on previous meta-analyses because it includes a more comprehensive set of adolescent outcomes, it focuses on measures that are psychometrically valid, and it includes a more detailed analysis of program moderator effects. Seventeen studies, based on eleven interventions (all US-based) met the ten criteria for inclusion into the analysis. The mean effect size across all programs and outcomes was 0.313, equivalent to a 62% higher mean score for an intervention group than for a control group. The largest effect was for educational success during adolescence (effect size 0.53) followed by social deviance (0.48), social participation (0.37), cognitive development (0.34), involvement in criminal justice (0.24), family well-being (0.18), and social–emotional development (0.16). Programs that lasted longer than three years were associated with larger sample means than programs that were longer than one year but shorter than three years. More intense programs (those with more than 500 sessions per participant) also had larger means than less intense programs. There was a marginally significant trend for programs with a follow-through component into the early primary school years (e.g. preschool to Grade 3) to have more positive effects than programs without a follow-through. We conclude that the impact of well-conducted early development programs on quality of life in adolescence can be substantial for social policy purposes.

Introduction

Longitudinal research has confirmed the benefits of well-designed and carefully implemented early developmental prevention (EDP) programs (Homel, 2005). Evidence from a series of interventions demonstrates that intervening with the child and the family early in the developmental pathway can minimize future health, educational, behavioral and crime-related problems (Farrington & Welsh, 2002). This is important given evidence demonstrating deterioration in health (e.g. asthma, diabetes, obesity, intellectual disability, eating disorders, depression, attention deficit disorder, hyperactivity) and non-health outcomes (e.g. crime and educational success) over time for children, particularly from at-risk populations (Stanley, Richardson, & Prior, 2005).

This paper presents results of a meta-analytic review of the effectiveness in at-risk populations of EDPs involving children up to 5 years of age who do not have mental health or severe developmental problems. Outcomes encompass indicators of individual and family quality of life during adolescence, defined as ages 10–19 years (World Health Organisation, 1986). In the longitudinal literature we reviewed, at-risk populations were mainly socio-economically disadvantaged, operationalized as people with poor levels of education, living in areas of high unemployment, living in poverty according to local income standards, and perhaps isolated as a result of ethnicity and language.

Earlier systematic reviews and meta-analyses have focused on crime, educational, and family-related outcomes in the childhood and adolescent life phases. However, these studies are incomplete, as they do not include a full review of the salient outcomes in adolescence. Additionally, the methodology employed in earlier meta-analyses for defining quality of life outcome domains and their respective indicators in adolescence is undeveloped. To fill this gap, we carried out a meta-analysis that incorporated a structured and scientific method of identifying and analyzing important outcome domains and indicators. We focused on adolescence because of its importance both in terms of individual development and social investment. It is a phase of life when young people in developed countries are presented with many new opportunities for positive development, but also face “historically unprecedented levels of physical, behavioral, and social health risks” (Lerner & Castellino, 2002, p. 122). In recognition of these risks and opportunities, much emphasis is placed by government on reducing negative outcomes during adolescence (e.g. school drop-out, formal contact with the criminal justice system). Our results are consistent with the argument that governments should invest more in the development and implementation of good EDPs to balance the more usual strategy of relying on costly remedial interventions (Manning, 2008).

The meta-analysis was carried out as the first part of a two stage process in a study conducted by Manning (2008). The second stage by Manning adapts the Analytical Hierarchy Process (Saaty, 1980) to develop a method for making complex multi-criteria decisions with respect to policy options for early childhood interventions. Nagin (2001) argued that a method was required that would permit analysts to identify common metric outcomes across competing and often disparate programs, such as home visitation to pregnant teenage women and center-based developmental day care, with the goal of eliciting preferences and relative utility values. The second stage also provided an outline of how relative utility values, derived using the Analytical Hierarchy Process approach, may be used to identify the economic benefits of developmental prevention programs on non-health-related quality of life outcomes in adolescence, as proposed by Nagin.