Temper outbursts in Lowe syndrome: Characteristics, sequence, environmental context and comparison to Prader–Willi syndrome

Abstract Background There is limited research into the nature and aetiology of temper outbursts in people with intellectual disabilities. In this study, we describe the phenomenology and environmental context of temper outbursts in Lowe syndrome, a rare genetic syndrome in which outbursts are purportedly frequent. Method A temper outburst interview (TOI) was conducted with caregivers of seventeen individuals with Lowe syndrome to generate an account of the behavioural sequence, common antecedents and consequences of temper outbursts, and to enable comparisons with similar work on Prader–Willi syndrome. Results Outbursts in Lowe syndrome were frequently triggered by thwarted goal‐directed behaviour and were associated with high levels of physical aggression and property destruction. Conclusions Form and sequence of outbursts showed similarities to Prader–Willi syndrome and to behaviours reported in literature on typically developing children. The results highlight the importance of considering shared aetiology as well as syndrome‐specific pathways in the development of outbursts.

Definitions of temper outbursts in intellectual disabilities are varied and vague (Tunnicliffe, 2012). In the typical development literature, temper outbursts are described via constituent behaviours including crying, whining, yelling or shouting, screaming, hitting, kicking, stiffening body, pushing/pulling/grabbing, throwing objects and running away (Potegal & Davidson, 2003). Although there is little consensus on specific behaviours which constitute an outburst there is some agreement that an outburst consists of a cluster of behaviours, critically including an emotional component, and cannot be defined by one behaviour alone.
The majority of the literature on challenging behaviour in intellectual disabilities has adopted a functional analytic approach based on operant learning theory (Emerson, 1993) which asserts that behaviours are maintained by positive and negative social, and automatic reinforcement. For example, an attempt to calm or soothe a child by providing attention or distraction with tangible items may reward behaviours. The child is then more likely to repeat these behaviours when in the same situation in the future (Carr & Durand, 1985). There is a strong evidence base for the functional analytic approach (e.g., Beavers, Iwata, & Lerman, 2013) but an exclusively operant learning approach cannot adequately explain why some challenging behaviours are demonstrated more frequently by people with particular genetic syndromes (Arron et al., 2011;Waite et al., 2014). Temper outbursts are more prevalent in several genetic syndromes, including PWS, Cri-du-Chat, Smith-Magenis (Dykens, Hodapp, & Finucane, 2000) and LS (Kenworthy et al., 1993). This partial specificity (see Dykens et al., 2000) is difficult to explain either from an exclusively biological or operant conditioning perspective. A perspective that incorporates an interaction between the biological/developmental consequences of genetic difference and environmental factors is needed (Tunnicliffe & Oliver, 2011).
Peak prevalence of temper outbursts in typical development lies between 2 and 5 years (Bhatia et al., 1990;Potegal & Davidson, 2003), which coincides with developments in executive function as well as communication and social skills. Outbursts have also been noted in a mixed population of typically developing and intellectual disability older children (aged 5-12 years) referred for inpatient psychiatric treatment (Carlson, Potegal, Margulies, Gutkovich, & Basile, 2009). In this population, outbursts were referred to as "rages" or "angry-agitated outbursts" and understood as an impairment of selfregulatory executive function mechanisms.
Evidence is emerging for the potential importance of cognitive deficits as an explanation for behavioural phenotypes (Tunnicliffe & Oliver, 2011). For example, recent work by Woodcock, Humphreys, Oliver, and Hansen (2010) using functional magnetic resonance imaging (fMRI) techniques has linked temper outbursts in PWS to cognitive impairments related to task switching. If temper outbursts in LS are characterized by similar patterns of behaviour to the outbursts in PWS, there may be a common impairment of executive function which interacts with environmental contingencies. This paper provides a descriptive comparison of outburst behaviours in LS and PWS, laying the foundations of potential future research on the links between executive function and outbursts in LS.
This study seeks to increase understanding of common antecedents and behavioural sequence in temper outbursts in LS to inform future studies on the aetiology of outbursts in people with intellectual disabilities and contribute to development of effective intervention strategies. It applies methods from Tunnicliffe et al., (2014), who described temper outbursts in PWS. By adopting this "bottom-up" descriptive approach, this study will be able to identify both emotional characteristics and typically operant behaviours that comprise the temper outbursts. The replication of Tunnicliffe et al.'s interview methodology also allows for direct comparisons to be made between the two papers.

| Ethical approval
Ethical approval was provided by the NHS Research Ethics Committee (Wales-REC-4), and written consent obtained from all informants. Pictorial consent forms were used as part of a wider LS study to explain the purpose of the research to people with LS and where possible, to gain direct consent from them to talk to caregivers about their daily lives.

| Recruitment
Primary caregivers of 17 people with LS were recruited via an existing study being conducted into the behavioural phenotype of LS, with support from the Lowe Syndrome Trust in the United Kingdom and the Lowe Syndrome Association in the United States. Seven participants were recruited from an existing database held by the research institution.

| Participants
Primary carers (informants) were fourteen mothers, one adoptive mother and three fathers, with one couple interviewed together. The people they cared for (participants) were male, had LS diagnosed by a paediatrician, ophthalmologist or geneticist and were aged between eight and 37 years (M = 18.29 years; n = 9, under 18 years; n = 8, 18 years or over). Eleven resided in the USA, five in the UK and one in Australia. Adaptive functioning and developmental age were measured using the Vineland Adaptive Behavior Scale-version II (VABS-II; CRESSEY Et al. Sparrow, Cicchetti, Balla, & Doll, 2005) (see Table 1). Developmental age, (calculated as an age equivalent score using the average of 11 VABS-II subscale scores) ranged from 0 years and 10 months to 10 years and 8 months (M = 4 years and 6 months; n = 11, less than five years; n = 5, 5 years or more). All were living in the family home with the informant, except for one who had died six months prior to the interview and had lived part-time with the informant. The interview schedule was adapted for this informant, for example, asking them to describe temper outbursts in the last month of their son's life rather than the last calendar month.

| Procedure
Semi-structured interviews were conducted by the same researcher, by telephone or video call. Interview duration ranged from 54 to 86 min.

| Measures
The semi-structured Temper Outburst Interview schedule (TOI; Tunnicliffe et al., 2014) comprised 32 questions intended to elicit a phenomenological account of temper outbursts from a caregiver perspective. It included open-ended questions to encourage description of idiosyncratic behaviours. Questions covered the latency and duration of outbursts; common antecedents; precursor behaviours; type and sequence of behaviours during a typical outburst; and the success or otherwise of management strategies used by caregivers to alleviate harm or reduce outbursts. Coding instructions for each question were taken from the study by Tunnicliffe et al. (2014)  The interview was previously shown to have good convergent validity with parental diary records of temper outbursts in PWS (66%-100%; Tunnicliffe et al., 2014). In order to reduce research burden on informants, a decision was taken not to include a diary study as part of the present research. A proportion of the questions were taken directly from the Challenging Behaviour Interview, which has established reliability (Oliver et al., 2003; inter-rater reliability: 0.69, test-retest reliability: Pearson's r = 0.90). Five of the interviews were coded independently by two researchers to assess inter-rater reliability. This was calculated as the percentage agreement on each question of the interview schedule. Agreement ranged between 60% and 100%, overall agreement was 85%. Fourteen of 30 questions had 100% agreement.

| Coding and data analysis
For data analysis purposes, information from the joint interview was combined and treated as for a single informant. Where differences existed in initial response a consensus was agreed between the two informants. All behaviours noticed by either informant were included in the descriptive account.
To reduce descriptors of specific behaviours to a manageable number and allow comparison, behaviours were grouped into categories (

| Latency and duration of temper outbursts
Data on the latency and duration of outbursts are shown in Table 3.
Temper outbursts were a frequent occurrence for all 17 participants.
Latency data indicated that 14 out of 17 participants experienced outbursts at least once a day with 8 informants reporting expected outbursts within the next hour. Typically outbursts lasted less than 15 min for 12 out of 17 participants. Of the three informants reporting longer latency periods two reported weekly occurrence of outbursts lasting between 5 and 15 min, and one would only expect to see the next outburst within a month and typically lasting less than 5 min. Of those reporting durations of between 15 and 60 min, one reported a latency of fifteen minutes, one expected an outburst within the next hour, and one reported outbursts of between 30 and 60 min as a daily occurrence. One informant experienced daily outbursts of more than one hour, with a maximum duration of 3 hr.
Informants were asked to identify factors likely to lead to prolonged outbursts. Saying "no," "not getting his own way," or "being forced to do something he did not want to do" were cited by eight informants as the main reason for extended outbursts.
"Frustration" was identified by three informants, anxiety by two and ignoring or not paying sufficient attention by a further four.
Obsessive behaviours and an inability to "let go of an issue" were mentioned by two informants. Table 4 provides a list of the setting events identified. Physiological or internal setting events, included tiredness (n = 7), hunger (n = 4), anxiety/fear (n = 5) and physical pain or discomfort (n = 5). Low mood (n = 1) and thirst (n = 1) were also mentioned. Environmental factors included time pressure (n = 2), or generalized change to routine such as being on holiday (n = 5) or being in unfamiliar surroundings (n = 3).

| Setting events
Many informants noted that high ambient or unexpected noise levels (n = 9) or crowded situations (n = 4) increased the likelihood that an outburst would be triggered. Table 5 provides information about the principal antecedent identified by informants. Two informants said that they could not identify a trigger for the specific episode described but were able to report the most common trigger for outbursts in general. Nine out of 17 informants indicated that some form of thwarted desire was the most prevalent trigger for an outburst. This included frustrated goals (n = 1), delayed gratification (n = 1), "not getting what he wants" or "not getting his own way" (n = 6), "not being able to do something he wants to do" (n = 1). Two other informants stated that "being asked to do something he does not want to do" leads to most outbursts.

| Antecedents
Change to routine or uncertainty about expectations provoked regular outbursts for three participants. Two informants noted that unexpected change in auditory stimulation such as a car engine stopping, or the TV or radio switching to advertising, triggered outbursts.
One informant identified boredom or frustration as the main trigger.
There was no robust evidence of an association between the principal antecedent and individual characteristics such as chronological or developmental age or additional diagnoses.
All informants reported multiple potential triggers for outbursts, ranging from five to 18 out of 21 possible antecedents suggested. The results are presented in Figure 1 together with the results from Tunnicliffe et al. (2014) for parents/carers of people with PWS. The graph shows that all LS informants (n = 17; 100% 1 ) reported witnessing a temper outburst triggered by the participant being asked to do something they did not want to do. The next most commonly reported antecedents in LS were change in expectation (n = 16; 94%), change in own routine (n = 14; 82%), not getting something they want (n = 14; 82%) and interruption to preferred activities (n = 14; 82%). Denial of food and disagreements were both reported in 76% (n = 13) of LS participants, and imperfections and concerns that belongings had been stolen were reported in 59% (n = 10) of participants. These results are consistent with the individual antecedents reported above.

| Sequence of behaviours during an outburst
Individual behavioural sequences, using coded topographies from  Latency to the next outburst: Within the next 15 min 2 Within the next hour 6 By this time tomorrow 6 By this time next week 2 By this time next month 1 Duration of longest outburst in the last month: Less than a minute 0 Less than 5 min 4 Less than 15 min 4 Less than an hour 5 More than an hour 4 Duration of typical outburst: Less than a minute 2 Less than 5 min 5 Less than 15 min 5 Less than an hour 4 More than an hour 1 Length of the longest outburst over one hour

Setting event Nª
Physiological (any of the below list) 12 Tiredness 7 Hunger 4 Thirst 1 Low mood 1 Anxiety/fear 5 Physical pain or discomfort 5 Environmental (any of the below list) 17 Time pressure 2 General change to routine (e.g., holidays) 5 Coming home 1 Unfamiliar setting 3 Crowds 4 Noise levels high or unexpected 9 Social (any of the below list) 7 When with certain person 5 Relationship difficulties 5 Embarrassment 1 ª Some informants reported more than one setting event within each category.

| 1221
Published for the British Institute of Learning Disabilities CRESSEY Et al.
during outbursts was shown by 9/17 participants. The small sample size did not allow for identification of statistically significant associations.
All informants were able to identify precursor behaviours which alerted them to a potential outburst but preceded the point at which an outburst could no longer be avoided. Seven informants reported perseverative requests, and four mentioned emotional vocalizations as a warning sign (e.g., shouting or yelling), differentiated from the  and emotional state as if nothing had happened, but in four of those cases this only occurred after reassurance had been provided by caregivers. Attempts to self-soothe were also present, with self-talk reported by three informants.
Specific behaviours coded as "antisocial" were identified for five participants. These behaviours included spitting (n = 2) and deliberate urination, defaecation and smearing (n = 4). All five of these individuals also demonstrated physical aggression towards others, and either aggression towards property (n = 4) or verbal aggression (n = 1). Three of these participants also exhibited self-injurious behaviours. There was insufficient evidence to suggest an association between this behaviour and participant characteristics such as age or developmental abilities.
The most frequently reported perceived emotions during an outburst were frustration (n = 12) and anger (n = 8). These sometimes occurred together.

| Management strategies used by caregivers
At the precursor stage, seven informants reported that the most successful strategy was distraction or redirection to an alternative activity. Other strategies included calm reasoning (n = 2), removal of choice (n = 1), providing attention (n = 1), offering help (n = 1), reiterating clear routine (n = 1), removing other children from the room (n = 1) or giving in (n = 1). Informants estimated that success rates for avoidance of an outburst were between 40% and 90% at this stage.
Only one informant felt that there was nothing that could be done even at the precursor stage. Table 6 gives a list of principal strategies and the success rate for each.
During an outburst, the chances of successful intervention reduced and the main aim of intervention at this stage appeared to be harm reduction, either to the person with LS, others at risk of aggression, or to avoid damage to property. Removal of the participant to a quiet location or withdrawal of the caregiver avoided further escalation but did not immediately stop an outburst.
Redirection, humour or distraction was reported to be successful in 60%-90% of outbursts if the intervention was made early enough.
The most common reason for variation in intervention strategies was location (n = 10). Concern for the judgement of others and risk to others' safety when in public were given as reasons for variation.
Informants also reported that they would be more likely to intervene directly rather than ignore behaviour, or might withdraw for their own safety, when the participant became aggressive.

| Comparison with Prader-Willi syndrome
Data were extracted from the Tunnicliffe et al. paper, and the two samples compared to check for differences in mean age or adaptive abilities (Tunnicliffe et al., 2014). No significant differences were found in VABS-II adaptive behaviour composite scores (Mann-Whitney U, p = 0.377), but a difference was found in the chronological age profiles of the two samples (t (29) = −1.44; p = 0.018), with a higher mean age in years reported for PWS. When a comparison was made based on age group (<18 years; ≥18 years) no significant difference was found between the two groups (p > 0.05). An important difference between the two samples is that participants with PWS were selected on the basis that routine change was a trigger for outbursts. This was not a requirement for the LS participants due to the rarity of the disorder but should be taken into consideration when interpreting the data.
The following differences were noted between the two samples. change as an antecedent enables other comparisons to be made despite this difference in initial selection criteria. Differences in the pattern of antecedents reported during the last twelve months were significant at p < 0.01 for "asked to do something they don't want to do" (ᵡ 2 = 7.24; p = 0.007) and for "interruption of preferred activity" (ᵡ 2 = 7.04; p = 0.008). Both these factors were reported more frequently in LS than in PWS. The sudden resumption of activities as if nothing had happened was not reported at all in PWS but was spontaneously mentioned by eight informants in the LS study (Fisher's exact, p = 0.003). No other significant differences were found.

| D ISCUSS I ON
Temper outbursts have previously been shown to be more prevalent in LS than in other people with intellectual disabilities or in typical development (Dolinsky et al., 2008;Kenworthy & Charnas, 1995).
The primary aim of this investigation was to generate a description of temper outbursts in LS based on informant accounts. Seventeen interviews with eighteen caregivers provided detailed accounts of the antecedents, behavioural and emotional sequence, and the consequences of temper outbursts in 17 people with LS.
It is notable that all participants with LS were eight years or older, putting them above the expected chronological age of five years for reduction or cessation of temper outbursts in typically developing children (Potegal & Davidson, 2003). Developmental age, however, as measured using age equivalent scores from the VABS-II, showed that more than half the participants had a developmental age of below five years. The topographies of behaviour during outbursts in LS bear marked similarities to those described for temper outbursts in typically developing children aged 2-5 years (Österman & Björkqvist, 2010;Potegal & Davidson, 2003), and "angry-agitated outbursts" in paediatric inpatients Potegal, Carlson, Margulies, Gutkovitch, & Wall, 2009).
In pre-school children, Wakschlag et al., (2007) suggested that both quality of behaviours (severity) and pervasiveness (frequency and duration) should be considered when determining the degree of pathological emotional dysregulation. In the current study of people with LS, most informants reported outbursts as a daily occurrence and nearly half reported a latency of an hour or less. Durations varied between less than five minutes and over an hour, compared with 0.5 to 40 min (M = 3 min) previously reported in typically developing children (Potegal, Kosorok, & Davidson, 1996). The prevalence of physical and verbal aggression towards others and towards property is indicative of a high level of severity, with implications for the wellbeing of both carers and people living with LS.
The reason for occurrence of distressing behaviours such as smearing, deliberate defaecation or urination and spitting is unclear.
One hypothesis proposed by several informants in this study was the difficulty carers had in disregarding such behaviours. They reported feeling obliged to respond, particularly when the behaviours impacted on others. This also applied to extreme aggression towards carers, attacks on siblings or strangers, or dangerous behaviours such as kicking windows. This hypothesis would be consistent with a functional behavioural analysis of challenging behaviours being inadvertently reinforced by attention, escape from demands or distraction with tangible reward (Iwata et al., 1994;Warren & Mondy, 1971) although further research is needed to understand the functions of these behaviours.
When exploring the aetiology of temper outbursts in genetic syndromes it is important to consider the role of physical differences. LS is characterized by significant physical impairment (Lewis et al., 2012) with associated limitations to independent access to tangible items, and the possibility of physical pain and discomfort. Frustration was cited as both a triggering event and a reason for prolonged outbursts. Informants sometimes perceived this as resulting from inability to perform a physical task to a desired standard, or without physical assistance from a caregiver (e.g., toileting) or lack of independent choice over timing or content of activities.
Physiological setting events were commonly identified as increasing the likelihood of an outburst, including hunger, thirst and tiredness. It is also interesting to note the environmental factors which impact on outbursts. Change in ambient noise or sudden changes in auditory stimuli were reported by more than half the respondents as increasing the likelihood of an outburst. Increased sensitivity to noise (hyperacusis) has been noted as a feature of other genetic disorders such as Cri-du-chat, and Williams syndromes but was not previously found to be associated with LS (Cornish & Pigram, 1996). Increased physiological arousal or anxiety caused by unusual sensitivity to sensory stimuli have been noted as a potential contributory factor in challenging behaviour in other disorders such as autistic spectrum disorders (ASD; Grapel, Cicchetti, & Volkmar, 2015) and Williams syndrome (John & Mervis, 2010). One of the participants for whom change in ambient noise was the principal antecedent for outbursts had a comorbid diagnosis of ASD but the other did not have this diagnosis. Another interesting aspect of environmental setting is the reported absence of temper outbursts outside the home, and a difference in behaviours dependent on who the carer is (e.g., mother or father). The "context-specificity" of outbursts may offer scope for environmental interventions to reduce the frequency or intensity of outbursts but further research would be needed to understand why self-regulation is possible in some circumstances but not in others.
This study has highlighted the potential importance of frustration intolerance as a factor in temper outbursts in LS. The current research did not use functional analysis methodology but the fact that all informants reported outbursts in response to unwanted demands suggests that escape may be a prominent driver of outbursts for this population. Further functional analytical research would be required and comparison with other intellectual disability populations to establish whether this is true.
The absence of difference between LS and PWS populations in reports of routine change as an antecedent is interesting given that the PWS group were selected on this basis, and the LS group were not. Change to routine has been noted as a potential trigger for temper outbursts in a number of genetic syndromes including PWS, LS, fragile X and Smith Magenis syndromes (Bull, Oliver, & Woodcock, 2017). A link has also been made between intolerance of change and repetitive behaviour as a precursor to outbursts (Moss, Oliver, Arron, Burbridge, & Berg, 2009). In the current study, perseverative requests were frequently reported as a precursor and change to routine or expectation was reported as antecedent to temper outbursts.
More than half of respondents spontaneously identified some form of thwarted goal-directed behaviour as the principal antecedent for outbursts. Thwarted goal-directed behaviour differs conceptually from traditional functional behavioural analysis. It incorporates elements of all three operant conditions including demands generated by the individual (i.e., access to attention or tangible reward) and desire to escape from an unwanted demand or unpleasant stimulus (such as high ambient noise). This concept also appears to be supported by findings from a recently published study by Rice, Woodcock, and Einfeld (2018)  The inability to tolerate frustration in young typically developing children and in older paediatric psychiatric inpatients is thought to be due to underdevelopment of cognitive mechanisms which control and regulate emotions and behaviour, known as executive functions (Hunter & Sparrow, 2012). Executive functions cover a range of cognitive abilities including judgement, planning, impulsivity, behavioural inhibition and task switching. It is of interest that some participants reported that their children showed remorse or relationship repair following an outburst. This may suggest that, at least for a subset of individuals with LS, temper outbursts were experienced as a loss of control over emotions, which later gave rise to the motivation to apologize or express remorse.
Executive function deficits may be implicated in temper outbursts in a range of genetic syndromes. For example, in PWS, a strong association has been found between task switching deficits, change to routine and temper outbursts (Woodcock et al., 2010). The cognitive challenge of moving from a well-rehearsed sequence of behaviour to adapt to a new task is thought to be aversive for individuals with PWS, which then overwhelms emotional coping skills. The discovery of task switching difficulties in PWS has led to the development of promising interventions to support transitions between activities and reduce the incidence of outbursts (Bull et al., 2017

| Limitations of the study
The TOI semi-structured interview schedule included open-ended questions to allow for the emergence of a detailed descriptive account of temper outbursts in LS. It provided structure for the collection of frequency data for comparison with behaviours in other populations. The schedule had been written however specifically for research on PWS and may therefore have overemphasized factors important to that population. A large-sample questionnaire study by Rice et al. (2018), published since the current study, reports that over 90% of respondents with PWS have had outbursts within the last year provoked by "being told no" or "being asked to do something they do not want to do." This is higher than the percentages reported in Tunnicliffe at al. (2014) used for comparison in the current study.
This suggests that the differences between PWS and LS reported within the current study may have arisen because of specific selection criteria in the Tunnicliffe et al. paper, the use of interview schedules as opposed to questionnaires, or because of small sample sizes.
Importantly, cross-syndrome comparisons of LS and PWS using consistent methodology are needed in the future to further delineate the similarities and differences between these groups.
Reliance on informant report gives limited insight into the perspective of the person with LS, which is a major flaw in the methodology. In future, it would be important to devise methods for enabling those with limited communication skills to contribute to understanding of outburst behaviours and provide insight into internal emotional state before, during and after an episode.
Informant report is also potentially problematic as caregivers will be influenced by their own attributions about the cause of behaviours that challenge, available social discourse on temper outbursts in typical development and their own levels of stress tolerance.

CRESSEY Et al.
Future research could include objective observation of temper outburst behaviours in an experimental or naturalistic setting to provide additional scientific rigour and more accurate measurement and description of the frequency, duration, severity and sequence of component behaviours to complement descriptive accounts. Statistical comparisons have been made with findings from Tunnicliffe et al., 2014 in PWS but should be treated with caution as the number of participants in each of these studies is small.

| CON CLUS IONS
This is the first paper to provide a robust descriptive analysis of  (Bull et al., 2017) and in the use of effective parenting techniques to teach emotional recognition and control to preverbal typically developing children (e.g., Douglas, 2007). With better understanding of the gene-environment-behaviour pathway (Tunnicliffe & Oliver, 2011), it is possible that these techniques could be adapted for children with LS and other syndromes in which temper outbursts are frequent. This paper has therefore made an important contribution to the literature on temper outbursts in intellectual disability populations and may have moved us closer to understanding the complex interplay between emotion, behaviour regulation and neurological difference in genetic syndromes.