حساسیت پردازش حسی پاسخ درمانی به یک برنامه پیشگیری از افسردگی مبتنی بر مدرسه را پیش بینی می کند: مدارک و شواهد از برتری حساسیت

Objective Treatment effects of preventative mental health interventions for adolescents tend to be relatively small. One reason for the small effects may be individual differences in the response to psychological treatment as a function of inherent characteristics, a notion proposed in the concept of Vantage Sensitivity. The current study investigated whether the personality trait Sensory-Processing Sensitivity moderated the efficacy of a new school-based intervention aimed at the prevention of depression. Method Using a two-cohort treatment/control design with one cohort serving as the control group (N = 197) and a subsequent cohort as the treatment group (N = 166) it was tested whether Sensory-Processing Sensitivity predicted depression trajectories from pre-treatment up to a 12 months follow-up assessment in 11-year-old girls from an at-risk population in England. Results Sensory-Processing Sensitivity emerged as a significant predictor of treatment response. The prevention program successfully reduced depression scores in girls scoring high on Sensory-Processing Sensitivity but was not effective at all in girls scoring low on the same measure. Conclusions This study provides first empirical evidence for Vantage Sensitivity as a function of the personality trait Sensory-Processing Sensitivity regarding treatment response to a school-based depression prevention intervention.

Rising rates of depressive disorders during childhood and adolescence pose a major public health concern in most Western societies (e.g., Collishaw, Maughan, Goodman, & Pickles, 2004). Not only are depressive symptoms in adolescence often associated with social, academic, and physical health difficulties, but they also tend to predict subsequent major depression in adulthood (Aalto-Setala, Marttunen, Tuulio-Henriksson, Poikolainen, & Lonnqvist, 2002). Children growing up in economically deprived neighborhoods (Yoshikawa, Aber, & Beardslee, 2012) and girls (Hyde, Mezulis, & Abramson, 2008) are at a particularly high risk for the development of depressive disorders. According to a recent study in England the percentage of youth reporting frequent feelings of depression and anxiety doubled over the last two decades, with girls being almost three times more likely to suffer from depression/anxiety than boys (Collishaw, Maughan, Natarajan, & Pickles, 2010). Given the detrimental effects of depression and the recent increase of depressive disorders in adolescent populations, substantial efforts have been directed towards the prevention of depression in childhood—usually through school-based promotion of adaptive coping skills and related competencies (Sutton, 2007). According to several meta-analyses such preventative interventions have generally been found effective regarding the reduction of depression symptoms (Brunwasser et al., 2009, Horowitz and Garber, 2006 and Stice et al., 2009). However, the average treatment effects tend to be modest at best (r = .11–.24) and treatment efficacy appears to vary as a function of intervention delivery and sample demographics ( Brunwasser et al., 2009, Durlak et al., 2011, Horowitz and Garber, 2006 and Stice et al., 2009). What has been neglected in existing work, until very recently ( Eley et al., 2012), is the notion that intervention effects may differ as a function of inherent child characteristics (e.g., personality traits, genetics). It is widely accepted that some individuals are more vulnerable to the negative effects of adversity as a function of individual traits, be they of psychological ( Kochanska & Kim, 2012), physiological ( Cummings, El-Sheikh, Kouros, & Keller, 2007), or genetic ( Caspi et al., 2002) nature. Extending this Diathesis-Stress perspective ( Zuckerman, 1999), the Differential Susceptibility framework ( Belsky & Pluess, 2009) suggests that such inherent traits may not just increase vulnerability to adversity but rather sensitivity to a variety of environmental influences, with more susceptible individuals being more affected by both negative as well as positive experiences ( Pluess, in press). In other words, the same characteristics that make children more vulnerable to adverse experiences may also make them more responsive to beneficial exposures ( Belsky & Pluess, 2009). The proposition—derived from Differential Susceptibility reasoning—that individuals may differ generally in their response to positive experiences as a function of inherent characteristics has recently been articulated in more detail in the concept of Vantage Sensitivity ( Pluess & Belsky, 2013). According to this framework some people are more likely to benefit from positive exposures while others appear to be less responsive or even resistant to the positive effects of the same supportive experience. The suggested reason for such differences in response to positive experiences is that people differ fundamentally in their environmental sensitivity with some being more and some less sensitive ( Pluess, in press). Although a fairly new concept, a growing body of empirical evidence reports individual differences in Vantage Sensitivity as a function of different psychological, physiological, and genetic characteristics in response to a wide range of positive exposures—including psychological intervention (for an overview, see Belsky & Pluess, 2013). For example, in their pioneering experimental study evaluating genetic moderation of a psychological intervention, Bakermans-Kranenburg, van IJzendoorn, Pijlman, Mesman and Juffer (2008) investigated whether a genetic polymorphism in the dopamine receptor D4 (DRD4) gene moderated the positive effects of a video-feedback parenting intervention on children’s externalizing behaviour in a randomised controlled trial. Providing evidence for Vantage Sensitivity as a function of genetic differences of the child, the intervention proved effective in decreasing externalizing behaviour—but only for children carrying the DRD4 7-repeat gene variant. Children without this gene variant did not benefit from the intervention at all. In the current study we sought to investigate Vantage Sensitivity as a function of Sensory-Processing Sensitivity (SPS)—a personality trait measured with the Highly Sensitive Person (HSP) Scale ( Aron & Aron, 1997)—in response to a new universal school-based preventative depression intervention, the SPARK Resilience program ( Boniwell & Ryan, 2009). About 20% of the general population is estimated to score particularly high on SPS, characterized by increased awareness and deeper processing of environmental subtleties as well as the tendency to be more easily overwhelmed when in very stimulating situations. SPS has been hypothesized to be the manifestation of a highly sensitive central nervous system, on which environmental influences register more easily and more deeply (2012). In a first experimental study 160 undergraduate students were randomly allocated to solve either very easy or very difficult math problems ( Aron, Aron, & Davies, 2005). Students scoring high on SPS reported the highest negative affect when assigned to the “difficult” math problems condition but also the lowest negative affect when allocated to the “easy” condition, compared to students low on SPS in either experimental condition, providing the first empirical evidence that SPS may increase sensitivity to both negative and positive experiences. The current study involved a sample of 363 11-year-old girls at a state school in one of the most deprived neighborhoods of England, representing the population most at risk for depressive disorders in the United Kingdom. Applying a nonrandomized two-cohort treatment/control design, the intervention was conducted in the treatment cohort only, which included all children in the same year at the same school, while the complete year-ahead cohort served as a control group. Based on the Vantage Sensitivity framework ( Pluess & Belsky, 2013), it was hypothesized that girls scoring high on SPS would show a greater positive response (i.e., steeper decline of depression symptoms over time) to the preventative intervention than girls scoring low on SPS.

According to univariate analyses of variance (ANOVA), depression and SPS did not differ as a function of child ethnicity or paternal education in either cohort. Similarly, bivariate correlations yielded no significant association between family size and depression or SPS. Consequently, ethnicity, family size and paternal education were not included as covariates in the analyses. Descriptive statistics and bivariate correlations of depression and SPS are reported in Table 1. Importantly, SPS was not associated with depression scores at pre and post assessment, suggesting that SPS measures assessed at post-treatment were not influenced by treatment effects. In a hierarchical linear model that included both linear and quadratic slopes across the four depression assessments in the treatment cohort, SPS significantly predicted the depression intercept at the 12-months follow-up assessment (B = −.18, p = .03) as well as the linear change in depression scores over time (B = −.08, p < .01). However, there was no significant differences in depression scores between the treatment and control cohort at the 12-month follow-up assessment (t(361) = −1.64, p = .10, d = −.17), a finding consistent with the original evaluation of the study ( Pluess et al., submitted). In order to investigate the significant effects of SPS on the intercept centered at 12 months and the slope of depression, extreme groups (bottom and top 25% of the treatment cohort based on the original SPS scores) were created and model-predicted depression scores for both extreme groups plotted across the four measuring points (see Fig. 2). The top SPS group (M = 67.90, SD = 6.16) had significantly higher SPS scores (t(80) = 18.62, p < .01) than the bottom SPS group (M = 40.80, SD = 6.99). According to repeated t-tests within each extreme group between the pre-assessment and the 12-months follow-up assessment, the change within the low SPS group was not significant (t(40) = 1.45, p = .16, d = .19) whereas it was highly significant in the high SPS group (t(40) = −2.95, p < .01, d = −.40). According to t-tests between the two groups, the high SPS group did not differ from the low SPS group at pre and post assessment, but had significantly lower depression at the 6-months (t(80) = −2.04, p < .05) and the 12-months assessment (t(80) = −2.18, p = .03) 1. Comparing the 12-months assessment depression scores of both high and low SPS treatment groups with the complete control cohort revealed that the low SPS group did not differ from the control cohort (t(236) = .41, p = .68, d = .07), whereas the high SPS group had significantly lower depression scores (t(236) = −2.08, p = .04, d = −.39). These findings are illustrated in Fig. 3. Full-size image (19 K) Fig. 2. Growth curve model-predicted depression scores of the treatment cohort for Sensory-Processing Sensitivity extreme groups (top and bottom 25%, n = 41 for each) across the four measuring points in order to illustrate growth curve model findings that emerged using the whole treatment cohort (N = 166). Figure options Table 2. Descriptive statistics and unadjusted associations for outcome variables of the control (N = 197) and treatment cohort (N = 113–166). Variables Mean value Standard deviation Sample size 1 2 3 4 Depression 12M (CC) 18.40 9.27 197 1 Depression Pre (TC) 17.06 7.77 141 – 2 Depression Post (TC) 16.44 9.50 166 .53∗∗ – 3 Depression 6M (TC) 15.49 9.46 144 .36∗∗ .69∗∗ – 4 Depression 12M (TC) 16.90 10.46 113 .24∗ .44∗∗ .61∗∗ – 5 Sensory-Processing Sensitivity (TC) 54.25 11.00 166 .13 −.04 −.12 −.13 Note. CC = control cohort, TC = treatment cohort, statistically significant correlations are marked bold. ⁎ p < .05. ⁎⁎ p < .01. Table options Full-size image (23 K) Fig. 3. Depression mean scores of the complete control cohort (N = 197) and growth curve model-predicted depression scores for both Sensory-Processing Sensitivity extreme groups (top and bottom 25%, n = 41 for each) at the 12-months follow-up assessment in order to illustrate growth curve model findings that emerged across the whole sample (N = 363).