Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Aug 27;19(10):e70350. doi: 10.1002/aur.70350

Predictors of Change in Severity of Self‐Injurious Behaviors in Preschool‐Aged Autistic Children

Joseph Boyle 1, Andrew Dakopolos 1, Olivia Surgent 1, Joshua K Lee 1, Derek S Andrews 1, Brianna Heath 1, Jason Smucny 1, Marjorie Solomon 1, Sally J Rogers 1, Christine Wu Nordahl 1,✉
PMCID: PMC13674366  PMID: 42657937

ABSTRACT

Self‐injurious behaviors (SIBs) affect up to 50% of autistic individuals and significantly impact quality of life and everyday functioning. Cross‐sectional studies link the presence of SIB to lower IQ, altered sensory responsivity, and poor emotion regulation, yet their effect on changes in SIB severity from early childhood to school age in autism remains poorly understood. We examined SIBs in 325 autistic children assessed at age 2–5 years (T1) and 161 autistic children at 4–9 years (T3). This included a subset of 122 autistic children evaluated longitudinally at both time points. SIBs were classified into subgroups based on severity of SIB (no‐SIB, mild, moderate, severe). We assessed cognitive ability, dysregulation, and sensory features at each time point. Cross‐sectional and longitudinal analyses were used to examine group differences. At T1, 74% of the sample were characterized as having mild, moderate, or severe SIB. At T3, 67% had mild, moderate, or severe SIB. There was substantial variability in SIB severity subgroup membership across time; however, only 9% did not have SIBs at either timepoint. At both timepoints, children with more severe SIBs had increased dysregulation (p < 0.001) and sensory features (p < 0.001), but differences in developmental/intellectual quotient (DQ/IQ) emerged only at T3. Longitudinally, improvement in sensory features (p = 0.026) and higher T3 IQ (p = 0.039) significantly raised odds of decreasing SIB severity. Of the baseline measures of sensory features, dysregulation, DQ, and SIB severity, higher baseline SIB severity was the only predictor of increasing SIB severity. SIBs are highly prevalent in preschool‐aged autistic children. Sensory responsivity and dysregulation predict severity of SIBs, and reductions in sensory features may lead to decreased SIB severity across time.

1. Introduction

Self‐injurious behaviors (SIBs) are characterized by repetitive, self‐directed, and often dangerous actions including head banging, self‐biting, skin‐picking, and eye poking (Bodfish et al. 2000). SIBs represent a significant challenge for many autistic individuals, as lifetime prevalence estimates approach 50% across all ages and intellectual ability levels in the autistic population (Dimian and Symons 2022; Richards et al. 2012, 2017; Steenfeldt‐Kristensen et al. 2020). While repetitive self‐directed behaviors such as head‐banging and body‐rocking do occur in many infants, they are often seen as non‐injurious and resolve within the early years of life (De Lissovoy 1961; Berkson and Tupa 2000). However, in children with developmental disabilities, these behaviors often persist, resulting in clinically significant self‐injurious behaviors (Berkson and Tupa 2000). Further, SIBs vary widely in presentation and frequency, and although not all SIBs are dangerous or of immediate clinical concern, severe SIBs can lead to serious or even permanent physical harm to the body, retinal degeneration, and in rare cases mortality (Dimian and Symons 2022; 2023; Knapp et al. 2023; Nissen and Haveman 1997). SIBs can also have a profound negative impact on social functioning and quality of life of individuals and their families, serving as a barrier to participation in social, school, and other community activities (Flowers et al. 2020; Steenfeldt‐Kristensen et al. 2020; MaClean Jr et al. 2010; Pritchard and Roy 2006). Characterizing the developmental course of SIBs in young autistic children and identifying predictors of SIB severity and change across time may help inform targets and timing of SIB interventions.

Existing longitudinal research indicates that SIBs can emerge early in development and, for a significant proportion of autistic individuals, persist throughout childhood and into adulthood. In an 18‐month prospective study of 254 infants and toddlers at risk for developmental delay, including those later diagnosed with autism, there was moderate stability in SIB across an approximately 18‐month period (Dimian et al. 2017). Extending these findings into later development, another longitudinal study followed 67 autistic adolescents and found that 44% continued to exhibit SIB a decade later in early adulthood (Laverty et al. 2020). These findings highlight the need for more longitudinal research beginning in early childhood to help clarify the emergence and persistence of SIBs in autism across development.

Intellectual functioning, broad‐based dysregulation, and sensory features have all been examined in the context of SIB in autism. Co‐occurring intellectual disability (ID) is common in autism, affecting approximately 40% of individuals (Maenner et al. 2023). Several meta‐analyses have identified significant associations between lower IQ and increased SIB among autistic children (e.g., Steenfeldt‐Kristensen et al. 2020; Weiss 2002), though these associations are generally not as strong in autistic adults (Licence et al. 2020; Rattaz et al. 2015). Cross‐sectional studies further demonstrate that lower intellectual functioning predicts presence of SIBs (Duerden et al. 2012; Rattaz et al. 2015; Richards et al. 2012, 2017). Overall, the evidence suggests a moderate association between lower cognitive ability and SIB in autism. While intellectual ability is likely relevant to SIB in early childhood, other neurodevelopmental mechanisms, including broad‐based dysregulation and sensory responsivity may also play a role in the presence or severity of SIB across development (Dempsey et al. 2016; Duerden et al. 2012; Flowers et al. 2020; Richman et al. 2013). Clarifying the unique and overlapping contributions of intellectual functioning, dysregulation, and sensory features may help advance understanding of SIB in autism during development and inform more targeted early intervention approaches.

Sensory features, including both hyper (i.e., high sensitivity to sensory stimuli) and hypo‐responsivity (i.e., low sensitivity to sensory stimuli), are distinguishing characteristics of autism compared to the general population (Dwyer et al. 2022; Kern et al. 2006) and have been linked to SIB (Duerden et al. 2012, Duerden et al. 2014; Frazier et al. 2022; Ferguson, Russell, et al. 2025, Ferguson, Spackman, et al. 2025). In autistic children, altered sensory responsivity emerged as the strongest predictor of SIB, stronger even than cognitive ability, social communication, or insistence on sameness (Duerden et al. 2012). Other studies further suggest that altered sensory responsivity co‐occurs with SIBs across the lifespan in autism, regardless of age, sex, socioeconomic status, or comorbidities (Soke et al. 2017; Carter Leno et al. 2019). Recent survey data also showed that 17%–35% of caregivers of autistic individuals rated behaviors such as nail‐biting, skin‐picking, and head‐banging as “serious” or “extreme” problems, and that sensory hyperresponsivity significantly predicted the severity of these SIB topographies (Ferguson, Russell, et al. 2025; Ferguson, Spackman, et al. 2025). In sum, qualitative, epidemiological, and clinical evidence links SIBs with altered sensory features and position sensory responsivity as a possible salient driver of SIBs in autistic individuals.

Broad‐based dysregulation, encompassing both emotional and behavioral domains, may also be a key driver of SIBs for autistic individuals, who often experience heightened levels of arousal. Emotion dysregulation has been linked to thepresence of SIBs in autism (Ferguson, Russell, et al. 2025; Ferguson, Spackman, et al. 2025; Goldfarb et al. 2021; Mazefsky et al. 2013, 2018; Samson et al. 2015; Uljarević et al. 2022). Indeed, there is evidence that self‐injury serves to modulate strong or negative emotions in autism (Ferguson, Russell, et al. 2025; Ferguson, Spackman, et al. 2025; Furniss and Biswas 2020; Moseley et al. 2019), as well as in non‐suicidal self‐injury in non‐autistic individuals (Chapman et al. 2006; Hintikka et al. 2018; Kao et al. 2024). In parallel, behavioral dysregulation, including externalizing behaviors like aggression, has been shown to co‐occur with SIBs (Northrup et al. 2022). Therefore, it is critical to better understand the relationship of broad‐based dysregulation processes and how they relate to the severity of SIB.

The present study evaluated the severity of SIB in young autistic children around the age of diagnosis (Time 1; 2–5 years) and then again approximately two years later (Time 3; 4–9 years). We assessed how developmental/intellectual abilities, sensory features, and generalized behavioral/emotional dysregulation were associated with SIBs both concurrently and across time. We employed a severity‐based method for characterizing SIB profiles which yielded four severity subgroups (no‐SIB, mild, moderate, and severe). Severity ratings were calculated at each study time point. The present study posed the following questions: (1) What is the composition of SIB severity groups at both time points and how stable are severity groups across time? (2) Do SIB severity groups differ in terms of developmental/intellectual ability, dysregulation, and sensory features at each time point? (3) Does developmental quotient (DQ), dysregulation, or sensory features at T1 differentiate SIB severity groups at T3? (4) Do baseline measures predict whether children will increase or decrease in SIB severity? (5) Do changes in dysregulation or sensory features differentiate autistic children who either increase or decrease in SIB severity from baseline to follow‐up? We additionally assessed whether there were sex differences in SIB severity at each timepoint.

2. Methods

2.1. Participants

The present study included 363 autistic children (131 females) enrolled in the UC Davis MIND Institute Autism Phenome Project (APP) or the Girls with Autism—Imaging of Neurodevelopment (GAIN; Nordahl et al. 2022) studies. Study protocols are identical, and the cohorts were combined. Behavioral data for the present analyses were collected at Time 1 (T1; n = 325, mean age = 3.1 years) and Time 3 (T3; n = 161, mean age =5.5 years). While the APP/GAIN study includes a Time 2 visit, behavioral data is not collected and thus Time 2 was omitted from the present study. A subset of the sample (n = 122) was assessed longitudinally at both time points; 203 participants had T1 data only and 39 participants had T3 data only. See Table 1 for demographics of the cross‐sectional samples at each time point as well as the subset with longitudinal data.

TABLE 1.

Demographic and clinical characteristics of participants at Time 1 and Time 3.

Cross‐sectional sample Longitudinal sample
Time 1 Time 3 Time 1 Time 3
n 325 (118 females) 161 (57 females) 122 (44 females)
Age, months 37.63 (6.15) 66.57 (10.16) 36.8 (5.34) 67.2 (10.7)
Mean (SD), range 24.48–63.87 49.68–112.16 24.5–50.8 49.7–112
ADOS CSS 7.60 (1.77) 7.37 (1.92) 7.38 (1.74) 7.01 (2.03)
Mean (SD), range 4–10 4–10 4–10 1–10
DQ/IQ 62.92 (20.84) 80.92 (31.63) 67.8 (22.5) 82 (32)
Mean (SD), range 20.51–132.49 17.91–133 27.7–128 19.1–133
Self‐Injurious Behavior Subscale score 3.34 (3.71) 2.72 (2.80) 3.01 (3.08) 2.56 (2.87)
Mean (SD), range 0–20 0–18 0–14 0–12
Short Sensory Profile Total score 137.0 (19.86) 134.75 (19.83) 137 (19.6) 133 (19.2)
Mean (SD), range 80–179 92–183 98–179 92–183
CBCL‐DP Dysregulation Profile 59.55 (6.93) 58.38 (6.59) 60.1 (7.04) 59 (6.62)
Mean (SD), range 50–83.66 50–78.66 50–82.7 50–78.7
Open in a new tab

Note: Values are presented as Mean (SD) with ranges. DQ/IQ measures reflect Developmental Quotient (MSEL) at Time 1 and Intelligence Quotient (DAS‐II) at Time 3. Dysregulation Profile combines CBCL Aggressive, Anxious/Depressed, and Attention Problems subscales from the Child Behavior Checklist. Self‐Injurious Behavior (SIB) subscale score is derived from the Repetitive Behavior Scale–Revised (RBS‐R). The Calibrated Severity Score (CSS) from the Autism Diagnostic Observation Schedule (ADOS) is also shown for descriptive purposes. An additional subset included in cross‐sectional analyses was missing RBS‐R data at T1 (n = 26) or were first enrolled at T3 (n = 1).

At study entry, diagnostic confirmation was conducted using the Autism Diagnostic Interview–Revised (ADI‐R; Lord et al. 1994) and the Autism Diagnostic Observation Scales (ADOS‐G or ADOS‐2) (Lord, Rutter, et al. 2012; Lord, Luyster, et al. 2012) which were administered by research‐reliable licensed clinical psychologists in accordance with the Collaborative Programs of Excellence in Autism (CPEA). Participants met ADOS‐2 cutoffs for autism or autism spectrum, exceeded the ADI‐R autism cutoff on the social or communication subscale, and were within two points of the autism cutoff on all remaining ADI‐R subscales. All participants had ADOS calibrated severity scores (ADOS‐CSS) ≥ 4 at enrollment. At T3, diagnoses were confirmed by licensed psychologists via ADOS‐2 and a DSM‐based checklist. All participants were English‐speaking, ambulatory, and free of visual, auditory, or neurologic impairments that could preclude participation. All procedures were approved by the UC Davis IRB, with informed consent provided by parents or legal guardians.

2.2. Measures

2.2.1. Self‐Injurious Behavior

At both T1 and T3, SIB was evaluated using the Repetitive Behaviors Scale–Revised (RBS‐R; Bodfish et al. 2000). Parents or primary caregivers rated eight self‐harm items: (1) hits self with body part; (2) hits self against surface or object; (3) hits self with object; (4) bites self; (5) pulls hair or skin; (6) rubs or scratches self; (7) inserts finger or object into eyes or ears; (8) skin picking, on a 4‐point Likert scale (0–3), indicating whether the behavior does not occur or is a mild, moderate, or severe problem.

2.2.2. SIB Severity Group Classification

We defined SIB severity subgroups based on the highest severity rating among any of the eight items (i.e., (0) no‐SIB, (1) mild, (2) moderate, or (3) severe). If a parent rated a child's behavior as severe (3) in any of the 8 SIB items, they were classified in the severe SIB subgroup, regardless of ratings of the other behaviors. We used this classification approach instead of the total SIB composite score (0–24 summed across all items) because using the total score may inflate the score of a child with multiple mild behaviors versus a child with fewer, but more intensely impairing self‐injurious behaviors. SIB severity grouping was determined independently for each time point. As such, T3 SIB severity groupings were based on the highest severity rating at T3, irrespective of T1 ratings (See Figure 1C).

FIGURE 1.

FIGURE 1

Open in a new tab

Panel A shows the bar plot distribution of self‐injurious behavior (SIB) severity groups, no‐SIB, mild, moderate, or severe, at Time 1. Panel B displays the corresponding distribution of SIB severity groups at Time 3. Panel C shows within‐child severity change from Time 1 to Time 3. At both timepoints, the majority of autistic participants endorsed some form of SIB (mild, moderate, severe) versus no SIB.

2.2.3. Developmental and Intellectual Abilities

At T1, DQ was derived from the Mullen Scales of Early Learning (MSEL; Mullen 1995) by averaging age‐equivalent scores across subscales, dividing by chronological age, and then multiplying by 100. At T3, intellectual functioning was assessed using the Differential Abilities Scale–Second Edition (DAS‐II; Elliott 2021). Within our sample, 18 children received the MSEL at T3 because they did not meet basal levels on the DAS‐II; thus, these individuals' DQ scores were used at T3.

2.2.4. Dysregulation

Broad‐based emotional and behavioral dysregulation was measured using the Child Behavior Checklist (CBCL; Achenbach and Ruffle 2000, Althoff et al. 2010). The CBCL is a caregiver report that assesses current behavioral, social, and emotional problems. Depending on each child's age, parents completed either the preschool (1.5–5 years) or school‐age (6–18 years) version. The Dysregulation Profile (DP) score is derived by averaging t‐scores from three syndrome subscales: Anxious/Depressed, Aggressive Behavior, and Attention Problems (Althoff et al. 2010; Keefer et al. 2020; Miller et al. 2019).

2.2.5. Sensory Features

Sensory features were assessed using the Short Sensory Profile (SSP; Dunn and Westman 1997), a 38‐item standardized caregiver‐report measure of children's responses to sensory experiences across domains (e.g., tactile sensitivity and auditory filtering). Lower total scores indicate more pronounced sensory features. The SSP‐2 (released 2014) replaced the SSP mid‐study; because the two versions are not compatible, only SSP data were used, restricted to participants with data at both T1 and T3 (n = 70). Table S1 depicts data availability for each measure at each time point.

2.3. Analytic Plan

First, we examined whether developmental/intellectual ability (DQ/IQ), dysregulation (CBCL‐DP), and sensory features (SSP total score) differed by SIB severity groups (no‐SIB, mild, moderate, severe) separately at each time point (T1 and T3) by conducting ANCOVAs, controlling for sex and age at each time point. Because dysregulation and sensory features are related constructs, we examined Pearson correlations between CBCL‐DP and SSP total scores. We subsequently conducted MANCOVAs (separately for T1 and T3) to jointly test across both outcomes. All models included age, DQ/IQ, ADOS‐CSS, and sex assigned at birth. In addition, we conducted follow‐up ANCOVAS predicting SSP and CBCL‐DP while controlling for the other (e.g., CBCL‐DP was a covariate in the SSP models and SSP was a covariate for CBCL‐DP) to partition the unique variance that overlap SSP and CBCL‐DP.

Next, to determine whether developmental ability, dysregulation, or sensory features at T1 differed by T3 SIB severity groups, we conducted separate ANCOVAs (with outcomes of DQ, CBCL‐DP, and SSP total score) adjusting for T1 age and sex.

We then examined change in SIB severity by characterizing individual change in SIB severity groupings from T1 to T3 for the 122 participants with RBS data at both time points. We indexed change in SIB severity as “increased” (moving from a less severe to a more severe grouping), “decreased” (moving from a more severe to a less severe grouping) or “stayed the same” (remained in the same severity group). A fourth group had no SIBs endorsed at either time point. Notably, for the ‘increased’ and ‘decreased’ groups, change between severity groupings was weighted equally—for example, participants who went from no SIB to mild SIBs and participants who went from mild SIBs to severe SIBs were included in the ‘increased’ group. The primary difference between the increase and decrease groupings was the direction of SIB severity change.

To examine whether T1 measures of DQ, sensory features, dysregulation, and baseline SIB severity predicted changes in SIB severity, we conducted a logistic regression with increased vs. decreased SIB severity as the outcome. Predictors included T1 scores for SSP total, CBCL‐DP, DQ, and T1 SIB severity group along with T1 age, interval between visits, and sex.

Finally, to examine whether changes in dysregulation and sensory features from T1 to T3 predicted changes in SIB severity, we focused on the subset of participants who had both CBCL‐DP and SSP data at T1 and T3 who either increased or decreased in severity grouping from T1 to T3 (n = 40). To examine how change in CBCL‐DP and SSP scores from T1 to T3 predicted change in SIB‐trajectory groups, we performed a logistic regression (outcome: increased vs. decreased SIB severity) with predictors including relative change in CBCL‐DP score [(T3 − T1)/T1 × 100], relative change in SSP total score [(T3 − T1)/T1 × 100], T1 age, interval between visits, T3 IQ, and sex.

3. Results

Table 1 depicts demographic and clinical characteristics for participants at each time point. Information pertaining to missing data and attrition can be found in Figure S1. Importantly, participants in the longitudinal sample did not differ from participants who only had T1 data on autism severity, dysregulation, sensory features, or total SIB score (Table S2). A significant difference in DQ was observed between participants with only T1 data and those with longitudinal data (Table S2). However, the mean difference was relatively small (longitudinal sample M = 68, T1 only M = 62). Notably, the longitudinal sample represents a comprehensive range of DQs (19–133) and was inclusive of females (36%). SSP subscale ANCOVAS were examined at T1 (Figure S2) and at T3 (Figure S3). Figure 1A,B illustrate the composition of SIB severity subgroups at each of the time points. Overall, 74% of autistic children at T1 and 67% at T3 had either mild, moderate, or severe SIBs. Sex differences were not observed in any of the severity groups at T1 (χ2 (3) = 2.13, p = 0.546) nor T3 (χ2 (3) = 0.85, p = 0.837). Table 2 depicts clinical characteristics for each SIB severity group across both time points. Figure 1C depicts change in severity groups from T1 to T3, highlighting substantial individual variability in change across severity groups.

TABLE 2.

Cross‐sectional: Clinical characteristics of SIB severity subgroups at Time 1 and Time 3.

SIB severity groupings Time 1 Time 3
None (n = 87) Mild (n = 113) Moderate (n = 83) Severe (n = 42) None (n = 53) Mild (n = 55) Moderate (n = 32) Severe (n = 20)
Sex (n) (m/f) 56/31 73/40 55/28 32/10 33/20 38/17 20/12 12/8
Age (months) (M, SD, range) 37.44 (5.8) 24.87–53.49 38.05 (5.7) 25.26–57.72 38.20 (5.9) 25.2–54.41 35.52 (7.8) 24.48–63.87 66.29 (10.10) 51.09–101.72 65.47 (9.23 m) 49.68–99.45 70.75 (12.24) 50.46–112.16 63.85 (7.75) 53.88–89.30
DQ/IQ (M, SD, range) 65.94 (21.12) 28.95–131.62 63.81 (21.22) 20.51–132.49 66.48 (23.01) 27.67–118.73 56.87 (18.01) 29.43–91.01 92.08 (28.65) 31.71–133 83.72 (29.83) 19.15–126 68.42 (28.98) 20.74–13 63.70 (31.79) 23.09–133
CBCL‐DP (M, SD, range) 56.09 (5.45) 50–76 58.79 (5.84) 50.33–74.66 61.19 (6.84) 50–82.66 66.35 (8.17) 52.33–84.0 55.94 (5.10) 50–69.99 57.58 (6.04) 50.33–78.33 61.92 (7.21) 50.33–78.67 62.09 (6.9) 54–78
SSP total score 147.76 (18.78) 94–179 135.85 (17.9) 98–171 133.09 (18.53) 99–176 123.96 (20.33) 80–157 142.61 (22.54) 100–183 135.26 (18.03) 92–167 126.58 (17.36) 98–176 123.18 (12.63) 109–143
RBS SIB total 0 (N/A) 2.05 (1.24) 1–8 5.18 (2.96) 2–16 8.81 (4.54) 3–20 0 (N/A) 2.65 (2.55) 1–8 5.34 (2.59) 2–10 6.22 (3.37) 3–14
ADOS CSS 7.57 (1.77) 4–10 7.61 (1.74) 4–10 6.89 (1.77) 4–10 7.5 (1.72) 5–10 6.35 (2.02) 1–10 7.54 (1.82) 2–10 7.1 (1,89) 4–10 7.12 (2.47) 4–10
Open in a new tab

3.1. Differences in Sensory Features and Dysregulation, but Not DQ Predict SIB Severity Groups at T1

We first examined cross‐sectional SIB severity group (no SIB, mild, moderate, severe) differences in developmental ability (DQ), dysregulation (CBCL‐DP), and sensory features (SSP total score) at T1. SIB severity groups did not differ by DQ (F(3, 309) = 1.87, p = 0.134; Figure 2A) but significantly differed by SSP total score (F (3,215) = 11.88, p < 0.001; Figure 2B) and CBCL‐DP score, (F (3,309) = 25.83, p < 0.001, Figure 2c). Post hoc pairwise comparisons with Tukey HSD multiple comparison correction showed that the no‐SIB group had a higher SSP total score (i.e., fewer sensory features) than the mild (p = 0.0014), moderate (p = 0.0003), and severe (p < 0.0001) groups, but the SIB severity groups did not differ from each other. Post hoc pairwise comparisons for CBCL‐DP scores revealed that the no‐SIB group had lower scores (i.e., better regulation) than the mild (p = 0.0148), moderate (p < 0.0001), and severe (p < 0.0001) groups. Compared to the severe group, children in both the mild group (p < 0.0001) and moderate group (p = 0.004) had lower CBCL‐DP scores. For illustrative purposes, Figure S2 depicts SIB severity group differences in each SSP subscale domain at T1 and T3.

FIGURE 2.

FIGURE 2

Open in a new tab

Cross‐sectional differences in developmental ability, sensory features, and dysregulation by SIB severity group at T1 (top row A‐C) and T3 (bottom row D–F).

3.2. Differences in Cognitive Ability, Sensory Features, and Dysregulation Predict SIB Severity Groups at T3

At T3, SIB severity groups differed significantly by IQ (F(3, 137) = 5.87, p = 0.001), SSP total score (F(3, 116) = 5.00, p = 0.003), and CBCL‐DP (F(3, 148) = 8.96, p < 0.001; Figure 2D–F). Post hoc pairwise comparisons for IQ indicated that the moderate (p = 0.0114) and severe (p = 0.0039) groups had significantly lower IQs than the no‐SIB group. The mild SIB group did not differ from the no‐SIB group (p = 0.496). Post hoc pairwise comparisons for SSP total scores revealed group differences between the no‐SIB group and both the moderate (p = 0.02) and the severe (p = 0.008) groups. Post hoc pairwise comparisons for CBCL‐DP scores revealed that the moderate (p = 0.0009) and severe (p = 0.0002) groups had higher scores (i.e., worse regulation) than the no‐SIB group. The no‐SIB and mild groups did not differ (p = 0.411).

3.3. Dysregulation and Sensory Are Both Jointly and Uniquely Associated With SIB Severity Group Differences

Because CBCL‐DP and SSP total scores were negatively correlated at both time‐points (T1: r = −0.58, 95% CI [−0.66, −0.49], p < 0.001; T3: r = −0.53, 95% CI [−0.65, −0.39], p < 0.001), MANCOVAs were conducted at each time point to test the multivariate effect of severity group on the two outcomes jointly, controlling for age, sex, DQ/IQ, and overall autism characteristics (ADOS‐CSS). At T1, a significant multivariate (joint) effect of SIB severity was observed (Pillai's Trace = 0.201, F (6,364) = 6.77, p < 0.001). No other predictor reached significance (ps ≥ 0.18). Post hoc ANCOVAs (Table 3) that accounted for the partial variance explained by the other dependent variable revealed significant SIB group differences in CBCL‐DP (F(3, 181) = 3.27, p = 0.023) and SSP total score (F(3, 181) = 4.32, p = 0.006). ADOS‐CSS was not a significant predictor of dysregulation or SSP total score at T1 (ps ≥ 0.48).

TABLE 3.

Residualized ANCOVA models partitioning unique variance between dysregulation and sensory responsivity.

Univariate outcome Predictor Df F p
Time 1
CBCL‐DP (controlling SSP Total) SIB severity 3, 181 3.27 0.023
Age 1, 181 0.35 0.553
ADOS severity 1, 181 0.36 0.548
IQ 1, 181 0.38 0.538
Sex 1, 181 2.60 0.108
SSP Total 1, 181 56.35 < 0.0001
SSP Total (controlling for CBCL‐DP) SIB severity 3, 181 4.32 0.006
Age 1, 181 0.54 0.463
ADOS severity 1, 181 0.49 0.483
IQ 1, 181 2.52 0.114
Sex 1, 181 0.00 0.988
Dysregulation 1, 181 56.35 < 0.0001
Time 3
CBCL‐DP (controlling SSP Total) SIB severity 3, 93 2.46 0.068
Age 1, 93 13.05 < 0.001
ADOS severity 1, 93 0.01 0.914
IQ 1, 93 0.08 0.783
Sex 1, 93 0.44 0.509
SSP Total 1, 93 27.47 < 0.0001
SSP Total (controlling for CBCL‐DP) SIB severity 3, 93 0.94 0.426
Age 1, 93 0.01 0.931
ADOS severity 1, 93 0.46 0.500
IQ 1, 93 0.34 0.562
Sex 1, 93 0.04 0.849
Dysregulation 1, 93 27.47 < 0.0001
Open in a new tab

At T3 the joint CBCL‐DP and SSP total multivariate outcome was significantly associated with SIB group (Pillai's Trace = 0.214, F(6, 188) = 3.75, p = 0.002) and age in months (Pillai's Trace = 0.156, F(2, 93) = 10.5, p < 0.001). All other predictors failed to reach significance (Pillai's ≤ 0.011, ps ≥ 0.57). In post hoc T3 ANCOVAs, SIB did not reach significance for CBCL‐DP (F(3, 93) = 2.46, p = 0.067) or SSP total score (F(3, 93) = 0.93, p = 0.43) (Table 3). ADOS‐CSS was not a significant predictor of CBCL‐DP nor SSP total score at T3 (ps ≥ 0.50).

Comparison of effect‐sizes shows that the joint effect of SIB severity group was comparable at T1 (partial η2 = 0.10) and T3 (partial η2 = 0.11). However, while SIB severity group accounted for similar variance in SSP (η2p = 0.07) and CBCL‐DP (η2p = 0.05) at T1, at T3 the CBCL‐DP effect was retained (η2p = 0.07) while the SSP effect was numerically smaller (η2p = 0.03). The significant multivariate age effect at T3 but not T1 likely reflects increased age variance at T3, with the univariate effect driven by CBCL‐DP (η2p = 0.12) rather than SSP (η2p < 0.001), but may also partially reflect an emerging difference in the developmental stability of SSP (r = 0.54) and CBCL‐DP (r = 0.35) across time (z = 1.60, p = 0.10).

3.4. Developmental Ability, Dysregulation, and Sensory Features at T1 Do Not Differentiate SIB Severity Groupings at T3

Time 1 DQ (F (3,131) = 0.23, p = 0.8750), SSP (F(3,86) = 2.03, p = 0.1163), and CBCL‐DP (F(3,118) = 1.90, p = 0.089) scores did not significantly differ across the T3 SIB severity groups.

3.5. Categorizing Increasing vs. Decreasing SIB Severity From T1 to T3

Based on our longitudinal sample seen at both T1 and T3 (n = 122), we first categorized participants based on change in SIB severity from T1 to T3 into four SIB‐trajectory groups: increased SIB (i.e., moved from a less to more severe group), decreased SIB (i.e., moved from a more to less severe group), no‐SIB (did not endorse SIB at either time point), and stable SIB (i.e., did not change SIB group from T1 to T3) (Figure 3A). Across the two time points, 31.9% of the sample decreased in SIB severity, 27.5% increased in SIB severity, 31.9% had stable SIB, and 8.7% had no‐SIBs at either timepoint. Figure 3B,C depict distributions of relative‐change in SSP total and CBCL‐DP scores across each SIB‐trajectory group at T1 and T3. The stable SIB groups (stable with‐SIB in pink and stable with no‐SIB in green) demonstrated very little change in CBCL‐DP and SSP scores across time and were not included in further analyses.

FIGURE 3.

FIGURE 3

Open in a new tab

Distribution of participants across SIB‐severity change groups (Panel A), relative change in SSP total score from T1 to T3 (Panel B), and relative change in CBCL‐DP score from T1 to T3 (Panel C) by severity group.

3.6. Developmental Ability, Dysregulation, and Sensory Features at T1 Do Not Predict Change in Severity Groupings From T1 to T3

The logistic regression model assessing whether DQ, SSP total score, CBCL‐DP, and SIB severity group at T1 predict SIB change group status (increase or decrease) was significant χ2 (6) = 13.72, p = 0.008, Nagelkerke R2 = 0.26. However, consistent with findings above that T1 scores did not predict T3 severity groups, T1 SSP total score (OR = 1.01, 95% CI [0.96, 1.06], p = 0.718), T1 CBCL‐DP (OR = 1.02, 95% CI [0.91, 1.15], p = 0.688), and T1 DQ (OR = 0.98, 95% CI [0.94, 1.01], p = 0.174) did not predict increasing versus decreasing SIB severity change. Age at T1 (OR = 0.96, 95% CI [0.83, 1.12], p = 0.608), interval between visits (OR = 1.01, 95% CI [0.93, 1.10], p = 0.799), and sex (OR = 0.51, 95% CI [0.09, 2.59], p = 0.417) were also non‐significant predictors. The only significant predictor of change in SIB was T1 SIB severity group (OR = 0.14, p = 0.002), indicating that children with higher initial SIB severity were more likely to be in the decreasing SIB severity group likely due to ceiling effects for participants who started out with a higher baseline SIB severity score.

3.7. Improvement in Sensory Features Increases Likelihood of Decreasing SIB From T1 to T3

The logistic regression model assessing whether relative change in SSP and CBCL‐DP from T1 to T3 predicted membership in the increasing or decreasing SIB trajectory groups was significant, χ2 (6) = 17.45, p = 0.008 (Nagelkerke R2 = 0.31). As seen in Figure 4, positive relative change in SSP total score (i.e., decreasing sensory features) significantly raised the likelihood of being in the decreasing‐SIB trajectory group (OR = 0.92; p = 0.026). Higher IQ at T3 also raised the likelihood of being in the decreasing‐SIB trajectory subgroup (OR = 0.97; p = 0.039); no other predictors reached significance.

FIGURE 4.

FIGURE 4

Open in a new tab

Forest plot of logistic regression predicting SIB‐trajectory groups that increased or decreased in SIB severity. Sex not depicted.

4. Discussion

The present study sought to characterize SIB severity profiles in autistic children and to identify predictors of SIB severity, both concurrently and across time. Using a novel SIB severity grouping approach, we found that 74% of our cohort at T1 (mean age 3 years) exhibited some form of SIB, with 38% having moderate or severe SIBs. Approximately 2–3 years later at T3 (mean age 5.5 years), 67% of children in our sample had some form of SIB, and 33% had moderate or severe SIBs. Although the proportions of individuals comprising severity groups at T1 and T3 are similar, considerable individual variability in severity group change was noted across time points. Out of all participants with longitudinal data, only 9% of autistic children had no SIBs at either time point, suggesting that between T1 and T3, the majority of autistic children experience SIB to some extent.

We found that at both T1 and T3, autistic children with greater SIB severity had more prominent sensory features and more broad‐based emotional and behavioral dysregulation. However, dysregulation, sensory features, and DQ at T1 did not predict SIB severity at T3. Among autistic children whose SIB severity changed over time, those with relative improvement in sensory features or higher IQ at T3 were more likely to be in the decreasing SIB severity group compared to the increasing group. Interestingly, at T1, DQ was not associated with SIB severity groups, but by T3, the children in the moderate and severe groups had, on average, lower IQs than the mild and no‐SIB groups. Collectively, these findings represent the first longitudinal assessment of SIB severity change in autistic children and highlight that both dysregulation and sensory features independently differentiate SIB severity.

4.1. Characterization of SIBs

We utilized a novel severity subgrouping approach to overcome a limitation of the standard SIB‐total score approach (generated from the sum of the Likert ratings across all 8 RBS‐R SIB items). While the traditional total score approach can indicate higher levels of SIB in a child, it may also limit the measure's ability to differentiate between a child with several mild versus fewer more severe behaviors. By classifying each child according to the highest endorsed SIB score (no‐SIB, mild, moderate, severe), we sought to preserve clinical relevance and emphasize patterns that a total score may mask. Finally, categorizing SIB severity allows for tracking whether a child is improving, remaining stable, or increasing in SIB severity, patterns that a continuous score may be less sensitive to if symptom profiles change (e.g., moving from a single severe SIB to three mild SIBs would yield the same total score). Using this framework, we showed that SIBs are highly dynamic on an individual level from preschool to early school age.

4.2. Dysregulation and Sensory Features Are Associated With SIB Severity at T1 and T3

At both T1 and T3, we found that children in the severe SIB group had heightened dysregulation and sensory features. These findings are consistent with prior cross‐sectional studies linking sensory responsivity (Duerden et al. 2012, 2014; Fetta et al. 2021) and dysregulation (Antezana et al. 2025; Ferguson, Russell, et al. 2025; Ferguson, Spackman, et al. 2025; Mazefsky et al. 2013; Northrup et al. 2022) to SIB. In a community cohort of 241 autistic 2‐ to 19‐year‐olds, Duerden et al. (2012) showed that elevated sensory features predicted SIB above and beyond any other clinical measure. Similar patterns have been reported in surveillance data (Soke et al. 2017) and prospective infant‐sibling work (Dimian et al. 2023), suggesting sensory responsivity strongly predicts SIB. Prior work has also implicated emotion and behavioral dysregulation in SIB (Mazefsky et al. 2018; Richards et al. 2017), as well as in adolescent non‐suicidal self‐injury in non‐autistic participants (Fong et al. 2022). However, little work has examined these constructs together across time to determine whether they each have unique contributions toward SIBs—or rather—are overlapping mechanisms and similar constructs. Indeed, sensory features and dysregulation are highly correlated, yet findings suggest that broad‐based dysregulation and sensory features may differently map onto to the severity of SIBs in autistic children.

These results could be interpreted to support a longitudinal framework in which sensory responsivity and dysregulation reflect partially distinct but interacting pathways that contribute to SIB. Here sensory responsivity could be conceptualized as a predominantly bottom‐up process, reflecting the child's threshold for and reactivity to environmental input, and dysregulation as a predominantly top‐down process, reflecting effortful control, inhibitory capacity, and modulation of emotional and behavioral arousal. Although these constructs are moderately correlated, they are not redundant and may have distinct mechanistic drivers. Sensory responsivity may increase the likelihood that environmental stimuli exceed tolerability thresholds, whereas dysregulation may increase the likelihood that, once arousal is elevated, the child cannot effectively modulate or inhibit escalating responses. At T1, SIB severity differentiated both dysregulation and sensory features after accounting for their shared variance, suggesting that each domain contributed unique information related to SIB severity at T1. In contrast, at T3, SIB severity no longer significantly predicted dysregulation or sensory features after controlling for the alternate variable. These findings may suggest that the unique associations between SIB severity and these constructs are more pronounced earlier in childhood, whereas by T3 the relationships between SIB severity, dysregulation, and sensory responsivity may be less distinct or more difficult to disentangle statistically.

These findings may be consistent with this dual‐pathway model wherein SIB emerges through several partially distinct but potentially overlapping mechanisms, including intrinsic reinforcement processes in which SIB modifies sensory input, and positive and negative reinforcement processes in which SIB reduces escalating arousal when regulatory capacity is exceeded. Importantly, although neither sensory features nor dysregulation at T1 predicted later SIB severity, improvements in sensory features and increased cognitive ability at T3 were associated with decreasing SIB severity over time, suggesting that changes in these behaviors and cognitive systems and processes may likely alter the likelihood or persistence of SIBs. While we cannot infer the specific function of SIB from these data, our findings generate testable longitudinal hypotheses: namely, that sensory responsivity may influence whether SIB is reinforced by changes in environmental input, whereas dysregulation may influence when SIB occurs in the context of heightened arousal and limited inhibitory control. Future work leveraging more dense longitudinal sampling and formal tests of interaction between these domains will be critical for disentangling how these pathways jointly shape the developmental course of SIB.

4.3. Associations Between SIB Severity and Intellectual Ability Are Present at T3, but Not at T1

Building on several decades of work showing that children and adults with intellectual disability are at elevated risk for self‐injury (Bartak and Rutter 1976; McClintock et al. 2003; Cooper et al. 2009; Duerden et al. 2012), we found that while developmental abilities at T1 did not predict SIB severity, by T3 children with moderate and severe SIBs had lower IQ scores than children with mild or no SIBs. This likely reflects reduced precision in developmental assessments at younger ages, but by T3, cognitive assessments may be more consistent and more reliably associated with SIBs. Higher IQs at age T3 were also associated with decreasing SIB severity from T1 to T3.

4.4. Strengths and Limitations

This study should be considered in the context of several strengths and limitations. A major strength is the deeply phenotyped, longitudinal cohort of autistic children beginning in toddlerhood that is inclusive DQ/IQs in the range of intellectual disability. This cohort allows for simultaneous examination of developmental ability, general sensory features, and broad‐based emotional and behavioral dysregulation as predictors of SIB across early childhood development.

However, the natural history design limits our ability to make causal inferences regarding the mechanisms underlying changes in SIB, sensory features, or dysregulation. Our SIB severity classification approach, while clinically interpretable, introduces potential ceiling effects. Because severity groups were defined based on the highest‐rated SIB item, individuals in the severe group at baseline could not worsen further within this framework, potentially constraining variability in longitudinal change. This approach also does not differentiate between specific SIB topographies, thus changes in severity may reflect shifts across behavior types rather than within the same behavior. Accordingly, it is not clear whether differentiating a child with more than one moderate SIB from a child with only one severe SIB is clinically meaningful. Second, our sensory analyses relied on a global sensory composite score rather than more granular modality‐specific scores, which may obscure associations between particular sensory patterns and SIB. Similarly, while we modeled dysregulation as a composite construct, this approach may mask differential contributions of its component domains (e.g., anxiety, attention, aggression). Lastly, we acknowledge that calculating percent change scores for SSP and CBCL‐DP may be susceptible to regression‐to‐the‐mean effects. Future work using more advanced longitudinal approaches (e.g., mixed‐effects or latent change models with larger samples) will be important to further refine these findings.

4.5. Conclusions and Future Directions

In conclusion, our findings reinforce a multilevel theoretical model in which early intellectual ability, sensory responsivity, and broad‐based behavioral and emotion regulation capacity interact at behavioral and contextual levels to contribute to the severity and trajectory of SIBs in autism. This work provides a foundational step toward understanding developmental predictors of SIB in early childhood. Ongoing data collection within this cohort, including additional behavioral timepoints and neuroimaging, will enable future studies to identify more precise mechanisms and potential intervention targets. Such interventions that simultaneously address sensory features and strengthen emotion‐regulation strategies while also targeting overall developmental delays and supporting communication may yield the greatest reduction in SIB across early development.

Funding

This work was supported by Autism Science Foundation (PRO‐25‐002), National Institutes of Health (R01MH127046, R01MH128814, R01MH103284), MIND Institute Intellectual and Developmental Disabilities Research Center (P50 HD103526).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Availability of cognitive and behavioral data among participants with SIB severity scores at T1 and T3.

Table S2: Mean differences of longitudinal versus T1 only group.

Figure S1: Flowchart depicting longitudinal participants and reasons for missing T3 data.

Figure S2: Violin plots depicting SSP subscale ANCOVAS at T1.

Figure S3: Violin plots depicting SSP subscale ANCOVAS at T3.

AUR-19-0-s001.docx (589.4KB, docx)

Acknowledgments

The authors would like to thank the families and children who participated in the Autism Phenome Project and Girls with Autism Imaging of Neurodevelopment studies.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

  1. Achenbach, T. M. , and Ruffle T. M.. 2000. “The Child Behavior Checklist and Related Forms for Assessing Behavioral/Emotional Problems and Competencies.” Pediatrics in Review 21, no. 8: 265–271. 10.1542/pir.21.8.265. [DOI] [PubMed] [Google Scholar]
  2. Althoff, R. R. , Verhulst F. C., Rettew D. C., Hudziak J. J., and van der Ende J.. 2010. “Adult Outcomes of Childhood Dysregulation: A 14‐Year Follow‐Up Study.” Journal of the American Academy of Child and Adolescent Psychiatry 49, no. 11: 1105–1116. 10.1016/j.jaac.2010.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Antezana, L. , Conner C. M., Eldeeb S., et al. 2025. “Using Machine Learning to Identify Features Associated With Different Types of Self‐Injurious Behaviors in Autistic Youth.” Psychological Medicine 55: e98. 10.1017/S0033291725000637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bartak, L. , and Rutter M.. 1976. “Differences Between Mentally Retarded and Normal Intelligence Autistic Children.” Journal of Autism and Childhood Schizophrenia 6, no. 2: 109–120. 10.1007/BF01538054. [DOI] [PubMed] [Google Scholar]
  5. Berkson, G. , and Tupa M.. 2000. “Early Development of Stereotyped and Self‐Injurious Behaviors.” Journal of Early Intervention 23, no. 1: 1–19. [Google Scholar]
  6. Bodfish, J. W. , Symons F. J., Parker D. E., and Lewis M. H.. 2000. “Varieties of Repetitive Behavior in Autism: Comparisons to Mental Retardation.” Journal of Autism and Developmental Disorders 30, no. 3: 237–243. 10.1023/a:1005596502855. [DOI] [PubMed] [Google Scholar]
  7. Carter Leno, V. , Vitoratou S., Kent R., et al. 2019. “Exploring the Neurocognitive Correlates of Challenging Behaviours in Young People With Autism Spectrum Disorder.” Autism 23, no. 5: 1152–1164. 10.1177/1362361318769176. [DOI] [PubMed] [Google Scholar]
  8. Chapman, A. L. , Gratz K. L., and Brown M. Z.. 2006. “Solving the Puzzle of Deliberate Self‐Harm: The Experiential Avoidance Model.” Behaviour Research and Therapy 44, no. 3: 371–394. 10.1016/j.brat.2005.03.005. [DOI] [PubMed] [Google Scholar]
  9. Cooper, S. A. , Smiley E., Jackson A., et al. 2009. “Adults With Intellectual Disabilities: Prevalence, Incidence and Remission of Aggressive Behaviour and Related Factors.” Journal of Intellectual Disability Research 53, no. 3: 217–232. 10.1111/j.1365-2788.2008.01127.x. [DOI] [PubMed] [Google Scholar]
  10. De Lissovoy, V. 1961. “Head Banging in Early Childhood. A Study of Incidence.” Journal of Pediatrics 58: 803–805. 10.1016/s0022-3476(61)80135-2. [DOI] [PubMed] [Google Scholar]
  11. Dempsey, J. , Dempsey A. G., Guffey D., et al. 2016. “Brief Report: Further Examination of Self‐Injurious Behaviors in Children and Adolescents With Autism Spectrum Disorders.” Journal of Autism and Developmental Disorders 46: 1872–1879. 10.1007/s10803-016-2704-x. [DOI] [PubMed] [Google Scholar]
  12. Dimian, A. F. , Botteron K. N., Dager S. R., et al. 2017. “Potential Risk Factors for the Development of Self‐Injurious Behavior Among Infants at Risk for Autism Spectrum Disorder.” Journal of Autism and Developmental Disorders 47, no. 5: 1403–1415. 10.1007/s10803-017-3057-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dimian, A. F. , Estes A. M., Dager S., Piven J., Wolff J. J., and IBIS Network . 2023. “Predicting Self‐Injurious Behavior at Age Three Among Infant Siblings of Children With Autism.” Autism Research: Official Journal of the International Society for Autism Research 16, no. 9: 1670–1680. 10.1002/aur.2981. [DOI] [PubMed] [Google Scholar]
  14. Dimian, A. F. , and Symons F. J.. 2022. “A Systematic Review of Risk for the Development and Persistence of Self‐Injurious Behavior in Intellectual and Developmental Disabilities.” Clinical Psychology Review 94: 102158. 10.1016/j.cpr.2022.102158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Duerden, E. G. , Card D., Roberts S. W., et al. 2014. “Self‐Injurious Behaviours Are Associated With Alterations in the Somatosensory System in Children With Autism Spectrum Disorder.” Brain Structure and Function 219, no. 4: 1251–1261. 10.1007/s00429-013-0562-2. [DOI] [PubMed] [Google Scholar]
  16. Duerden, E. G. , Oatley H. K., Mak‐Fan K. M., et al. 2012. “Risk Factors Associated With Self‐Injurious Behaviors in Children and Adolescents With Autism Spectrum Disorders.” Journal of Autism and Developmental Disorders 42, no. 11: 2460–2470. 10.1007/s10803-012-1497-9. [DOI] [PubMed] [Google Scholar]
  17. Dunn, W. , and Westman K.. 1997. “The Sensory Profile: The Performance of a National Sample of Children Without Disabilities.” American Journal of Occupational Therapy 51, no. 1: 25–34. [DOI] [PubMed] [Google Scholar]
  18. Dwyer, P. , Ferrer E., Saron C. D., and Rivera S. M.. 2022. “Exploring Sensory Subgroups in Typical Development and Autism Spectrum Development Using Factor Mixture Modelling.” Journal of Autism and Developmental Disorders 52, no. 9: 3840–3860. 10.1007/s10803-021-05256-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Elliott, C. D. 2021. Differential Ability Scales—Second Edition [Dataset]. American Psychological Association (APA). 10.1037/t15074-000. [DOI] [Google Scholar]
  20. Ferguson, E. F. , Russell K. M., Licona S. J., et al. 2025. “Toward Improved Understanding and Treatment of Self‐Injurious Behaviors in Autistic Individuals With Profound Intellectual Disability.” Autism Research 18, no. 2: 261–272. 10.1002/aur.3289. [DOI] [PubMed] [Google Scholar]
  21. Ferguson, E. F. , Spackman E., Cai R. Y., Hardan A. Y., and Uljarević M.. 2025. “Exploring the Heterogeneity of Self‐Injurious Behaviors in Autistic Youth: Patterns, Predictors, and Implications for Intervention.” Autism Research 18, no. 1: 133–151. 10.1002/aur.3269. [DOI] [PubMed] [Google Scholar]
  22. Fetta, A. , Carati E., Moneti L., et al. 2021. “Relationship Between Sensory Alterations and Repetitive Behaviours in Children With Autism Spectrum Disorders: A Parents' Questionnaire Based Study.” Brain Sciences 11, no. 4: 484. 10.3390/brainsci11040484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Flowers, J. , Lantz J., Hamlin T., and Simeonsson R. J.. 2020. “Associated Factors of Self‐Injury Among Adolescents With Autism Spectrum Disorder in a Community and Residential Treatment Setting.” Journal of Autism and Developmental Disorders 50, no. 8: 2987–3004. 10.1007/s10803-020-04389-4. [DOI] [PubMed] [Google Scholar]
  24. Fong, Z. H. , Loh W. N. C., Fong Y. J., Neo H. L. M., and Chee T. T.. 2022. “Parenting Behaviors, Parenting Styles, and Non‐Suicidal Self‐Injury in Young People: A Systematic Review.” Clinical Child Psychology and Psychiatry 27, no. 1: 61–81. 10.1177/13591045211055071. [DOI] [PubMed] [Google Scholar]
  25. Frazier, T. W. , Crowley E., Shih A., et al. 2022. “Associations Between Executive Functioning, Challenging Behavior, and Quality of Life in Children and Adolescents With and Without Neurodevelopmental Conditions.” Frontiers in Psychology 13: 1022700. 10.3389/fpsyg.2022.1022700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Furniss, F. , and Biswas A. B.. 2020. Self‐Injurious Behavior in Individuals With Neurodevelopmental Conditions. Springer Nature Switzerland AG. 10.1007/978-3-030-36016-0. [DOI] [Google Scholar]
  27. Goldfarb, Y. , Zafrani O., Hedley D., Yaari M., and Gal E.. 2021. “Autistic Adults' Subjective Experiences of Hoarding and Self‐Injurious Behaviors.” Autism 25, no. 5: 1457–1468. 10.1177/1362361321992640. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hintikka, J. , Tolmunen T., Rissanen M.‐L., Honkalampi K., Kylmä J. M., and Franklin J. C.. 2018. “Why Do People Hurt Themselves? A New Conceptual Model of Nonsuicidal Self‐Injury.” Clinical Psychological Science 6, no. 3: 428–451. 10.1177/2167702617745641. [DOI] [Google Scholar]
  29. Kao, H.‐T. , Mürner‐Lavanchy I., Lerch S., et al. 2024. “Longitudinal Associations Between Beta‐Endorphin, Nonsuicidal Self‐Injury and Comorbid Psychopathology.” Psychiatry Research 340: 116142. 10.1016/j.psychres.2024.116142. [DOI] [PubMed] [Google Scholar]
  30. Keefer, A. , Singh V., Kalb L. G., Mazefsky C. A., and Vasa R. A.. 2020. “Investigating the Factor Structure of the Child Behavior Checklist Dysregulation Profile in Children and Adolescents With Autism Spectrum Disorder.” Autism Research: Official Journal of the International Society for Autism Research 13, no. 3: 436–443. 10.1002/aur.2233. [DOI] [PubMed] [Google Scholar]
  31. Kern, J. K. , Trivedi M. H., Garver C. R., et al. 2006. “The Pattern of Sensory Processing Abnormalities in Autism.” Autism 10, no. 5: 480–494. 10.1177/1362361306066564. [DOI] [PubMed] [Google Scholar]
  32. Knapp, M.‐C. , Saunders N. R., Huang A., et al. 2023. “Self‐Harm Events and Suicide Deaths Among Autistic Individuals in Ontario, Canada.” JAMA Network Open 6, no. 8: e2327415. 10.1001/jamanetworkopen.2023.27415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Laverty, C. , Oliver C., Moss J., Nelson L., and Richards C.. 2020. “Persistence and Predictors of Self‐Injurious Behaviour in Autism: A Ten‐Year Prospective Cohort Study.” Molecular Autism 11, no. 1: 8. 10.1186/s13229-019-0307-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Licence, L. , Oliver C., Moss J., and Richards C.. 2020. “Prevalence and Risk‐Markers of Self‐Harm in Autistic Children and Adults.” Journal of Autism and Developmental Disorders 50, no. 10: 3561–3574. 10.1007/s10803-019-04260-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Lord, C. , Luyster R. J., Gotham K., and Guthrie W.. 2012. Autism Diagnostic Observation Schedule, Second Edition (ADOS‐2) Manual (Part II): Toddler Module. Western Psychological Services. [Google Scholar]
  36. Lord, C. , Rutter M., DiLavore P. C., Risi S., Gotham K., and Bishop S.. 2012. Autism Diagnostic Observation Schedule, Second Edition. Western Psychological Services. [Google Scholar]
  37. Lord, C. , Rutter M., and Le Couteur A.. 1994. “Autism Diagnostic Interview‐Revised: A Revised Version of a Diagnostic Interview for Caregivers of Individuals With Possible Pervasive Developmental Disorders.” Journal of Autism and Developmental Disorders 24, no. 5: 659–685. 10.1007/BF02172145. [DOI] [PubMed] [Google Scholar]
  38. MaClean, W. E., Jr. , Tervo R. C., Hoch J., Tervo M., and Symons F. J.. 2010. “Self‐Injury Among a Community Cohort of Young Children at Risk for Intellectual and Developmental Disabilities.” Journal of Pediatrics 157, no. 6: 979–983. 10.1016/j.jpeds.2010.05.052. [DOI] [PubMed] [Google Scholar]
  39. Maenner, M. J. , Warren Z., Williams A. R., et al. 2023. “Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8 Years ‐ Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2020.” Morbidity and Mortality Weekly Report. Surveillance Summaries (Washington, D.C.: 2002) 72, no. 2: 1–14. 10.15585/mmwr.ss7202a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Mazefsky, C. A. , Herrington J., Siegel M., et al. 2013. “The Role of Emotion Regulation in Autism Spectrum Disorder.” Journal of the American Academy of Child and Adolescent Psychiatry 52, no. 7: 679–688. 10.1016/j.jaac.2013.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Mazefsky, C. A. , Yu L., White S. W., Siegel M., and Pilkonis P. A.. 2018. “The Emotion Dysregulation Inventory: Psychometric Properties and Item Response Theory Calibration in an Autism Spectrum Disorder Sample.” Autism Research: Official Journal of the International Society for Autism Research 11, no. 6: 928–941. 10.1002/aur.1947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. McClintock, K. , Hall S., and Oliver C.. 2003. “Risk Markers Associated With Challenging Behaviours in People With Intellectual Disabilities: A Meta‐Analytic Study.” Journal of Intellectual Disability Research 47, no. 6: 405–416. 10.1046/j.1365-2788.2003.00517.x. [DOI] [PubMed] [Google Scholar]
  43. Miller, M. , Iosif A.‐M., Young G. S., et al. 2019. “The Dysregulation Profile in Preschoolers With and Without a Family History of Autism Spectrum Disorder.” Journal of Child Psychology and Psychiatry, and Allied Disciplines 60, no. 5: 516–523. 10.1111/jcpp.13003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Moseley, R. L. , Gregory N. J., Smith P., Allison C., and Baron‐Cohen S. J. M. A.. 2019. “A ‘Choice’, an ‘Addiction’, a Way ‘Out of the Lost’: Exploring Self‐Injury in Autistic People Without Intellectual Disability.” Molecular Autism 10, no. 1: 18. 10.1186/s13229-019-0267-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Mullen, E. M. 1995. Mullen Scales of Early Learning, 58–64. AGS. 10.1016/j.biopsych.2010.04.015. [DOI] [Google Scholar]
  46. Nissen, J. M. , and Haveman M. J.. 1997. “Mortality and Avoidable Death in People With Severe Self‐Injurious Behaviour: Results of a Dutch Study.” Journal of Intellectual Disability Research: JIDR 41, no. Pt 3: 252–257. [PubMed] [Google Scholar]
  47. Nordahl, C. W. , Andrews D. S., Dwyer P., et al. 2022. “The Autism Phenome Project: Toward Identifying Clinically Meaningful Subgroups of Autism.” Frontiers in Neuroscience 15: 786220. 10.3389/fnins.2021.786220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Northrup, J. , Simon T. J. B., Goodwin M. S., et al. 2022. “Mapping the Time Course of Overt Emotion Dysregulation, Self‐Injurious Behavior, and Aggression in Psychiatrically Hospitalized Autistic Youth: A Naturalistic Study.” Autism Research 15, no. 10: 1855–1867. 10.1002/aur.2773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Pritchard, A. , and Roy A.. 2006. “Reversing the Export of People With Learning Disabilities and Complex Health Needs.” British Journal of Learning Disabilities 34, no. 2: 88–93. [Google Scholar]
  50. Rattaz, C. , Michelon C., and Baghdadli A.. 2015. “Symptom Severity as a Risk Factor for Self‐Injurious Behaviours in Adolescents With Autism Spectrum Disorders.” Journal of Intellectual Disability Research: JIDR 59, no. 8: 730–740. 10.1111/jir.12177. [DOI] [PubMed] [Google Scholar]
  51. Richards, C. , Davies L., and Oliver C.. 2017. “Predictors of Self‐Injurious Behavior and Self‐Restraint in Autism Spectrum Disorder: Towards a Hypothesis of Impaired Behavioral Control.” Journal of Autism and Developmental Disorders 47, no. 3: 701–713. 10.1007/s10803-016-3000-5. [DOI] [PubMed] [Google Scholar]
  52. Richards, C. , Oliver C., Nelson L., and Moss J.. 2012. “Self‐Injurious Behaviour in Individuals With Autism Spectrum Disorder and Intellectual Disability.” Journal of Intellectual Disability Research: JIDR 56, no. 5: 476–489. 10.1111/j.1365-2788.2012.01537.x. [DOI] [PubMed] [Google Scholar]
  53. Richman, D. M. , Barnard‐Brak L., Bosch A., Thompson S., Grubb L., and Abby L.. 2013. “Predictors of Self‐Injurious Behaviour Exhibited by Individuals With Autism Spectrum Disorder.” Journal of Intellectual Disability Research 57, no. 5: 429–439. 10.1111/j.1365-2788.2012.01628.x. [DOI] [PubMed] [Google Scholar]
  54. Samson, A. C. , Hardan A. Y., Podell R. W., Phillips J. M., and Gross J. J.. 2015. “Emotion Regulation in Children and Adolescents With Autism Spectrum Disorder.” Autism Research: Official Journal of the International Society for Autism Research 8, no. 1: 9–18. 10.1002/aur.1387. [DOI] [PubMed] [Google Scholar]
  55. Soke, G. N. , Rosenberg S. A., Hamman R. F., et al. 2017. “Factors Associated With Self‐Injurious Behaviors in Children With Autism Spectrum Disorder: Findings From Two Large National Samples.” Journal of Autism and Developmental Disorders 47, no. 2: 285–296. 10.1007/s10803-016-2951-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Steenfeldt‐Kristensen, C. , Jones C. A., and Richards C.. 2020. “The Prevalence of Self‐Injurious Behaviour in Autism: A Meta‐Analytic Study.” Journal of Autism and Developmental Disorders 50, no. 11: 3857–3873. 10.1007/s10803-020-04443-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Uljarević, M. , Frazier T. W., Jo B., et al. 2022. “Big Data Approach to Characterize Restricted and Repetitive Behaviors in Autism.” Journal of the American Academy of Child and Adolescent Psychiatry 61, no. 3: 446–457. 10.1016/j.jaac.2021.08.006. [DOI] [PubMed] [Google Scholar]
  58. Weiss, J. 2002. “Self‐Injurious Behaviors in Autism: A Literature Review.” Journal on Developmental Disabilities 9: 129–143. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1: Availability of cognitive and behavioral data among participants with SIB severity scores at T1 and T3.

Table S2: Mean differences of longitudinal versus T1 only group.

Figure S1: Flowchart depicting longitudinal participants and reasons for missing T3 data.

Figure S2: Violin plots depicting SSP subscale ANCOVAS at T1.

Figure S3: Violin plots depicting SSP subscale ANCOVAS at T3.

AUR-19-0-s001.docx (589.4KB, docx)

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Articles from Autism Research are provided here courtesy of Wiley

RESOURCES