The Role of Adverse Childhood Experiences on People in Opiate Agonist Treatment: The Importance of Feeling Unloved

Introduction: Adults in opiate agonist treatment (OAT) often have a background of adverse childhood experiences (ACEs) and are more likely to be exposed to a variety of risks that may trigger post-traumatic stress disorder (PTSD). Summative ACE scores are often used to identify individuals at risk of PTSD and continued substance use. What has not been addressed is whether specific ACE factors are exerting a greater influence on the individual. This study investigated whether specific ACEs predicted PTSD, and current continued substance use among adults in long-term OAT. Methods: An analysis of data that were collected at the follow-up stage of a study among 131 adults who attended OAT was conducted. Participants attended one of six OAT settings, covering 45% (n = 890) of clients in a defined area of Dublin, Ireland in 2017. Interviews were conducted with 104 participants, 66 males (63%) and 38 females (37%), with an average age of 43 years (SD = 7.4). The Adverse Childhood Questionnaire (ACQ); PTSD checklist (PCL-5); heroin; tranquilliser; cannabis; alcohol; and cocaine used in the previous 28 days were measured using the quantity used score within the Opiate Treatment Index. Socio‐demographics and age of first use of these four substances were also collected. The analysis has focussed on relating ACEs to PTSD, age of first drugs use, and current drug use of the participants. Results: Bivariate analysis showed that the summative ACQ score was significantly correlated with age of first opiate use (p = 0.004). Multiple regression analysis showed that the summative ACQ score and tranquilliser use predicted higher levels of PTSD (R2 = 0.50). Four specific ACEs predicted 54% of the variance in PTSD, these were feeling unloved (β = 0.328) living with a household member who had a problem with alcohol or used illicit street drugs (β = 0.280); verbal abuse (β = 0.219); and living with a person who had a mental illness (β = 0.197). Conclusions: While a summation of all ten ACEs predicted higher levels of PTSD, the factor “feeling unloved” as a child provided the single strongest predictor and may represent an overarching risk of PTSD and continued substance use in later life among adults in treatment for an opiate use disorder.


Introduction
Globally opiate agonist treatment (OAT) is recommended as a harm reduction approach for people who use heroin or other opiates.Adults in receipt of OAT often have a background of adverse childhood experiences (ACEs) such as verbal, physical, or sexual abuse.As demonstrated in a recent systematic review of the literature [1], these experiences are recognised to negatively impact children and may contribute to subsequent drug use and a variety of risk factors and incidents that may trigger post-traumatic stress disorder (PTSD), continued drug use while in treatment, and retention in long-term treatment [1][2][3].
Developmental psychological theory generally contends that children grow up most effectively in a safe environment with high levels of positive stimulation, attachment to caregivers, and positive caregiver involvement, unconditional positive regard, reinforcement, and adaptive modelling [4,5].Conversely, a family environment with abuse, violence, and parental psychopathology is strongly associated with the onset of mental health disorders and substance misuse in the children brought up within it [6][7][8][9][10].A recent systematic review and meta-analysis have demonstrated that there is a strong relationship between childhood maltreatment and opiate use disorder [1].Violence, abuse, and other ACEs unsettle the processes whereby children attach to caregivers and experience a sense of safety and predictability.In this environment, they may also not receive consistent reinforcement for the behaviours essential to become successful adults and also, they may see maladaptive behaviours and coping modelled that are more likely to get them in trouble [11].Khantzian [12] suggests that human behaviour is governed by coping skills developed and acquired within a care taking environment.Specifically, if a parent using substances is also the child's primary caregiver the risk of the child engaging in addictive behaviours are markedly increased [13].
Underlying processes in the link between childhood maltreatment and substance use include impairments in relationship formation and stress regulation [14].Research has shown consistently that adverse experiences in early life can be chronic stressors that increase the risk of developing serious health and behavioural problems [15,16], "including internalising (e.g., anxiety, depression) and externalising (e.g., aggression, acting out) behaviour" ( [17] p211).This can have a harmful impact on the developing person and continue to have negative implications throughout the lifespan [18].Prolonged activation of an infant's stress response systems can disrupt the architecture of the developing brain thereby increasing the "lifelong risk for physical and mental disorders" [16 p1].According to McLaughlin, Koenen [19] childhood adversities are known to specifically predict an increase in the risk of PTSD [9,[20][21][22].An ACEs study, conducted in California between 1995 and 1996, retrospectively measured the long-term impact of household dysfunction among a sample of almost 10,000 adults [6].In this study, a strong correlation emerged between the aggregate number of ACEs and the risk of disease and death [6,8].Researchers have also reported a graded relationship between the number of ACEs and the risk of poor health outcomes, including depression, PTSD, and substance misuse [15,22,23].
This "summative approach" suggests that the number of ACEs a child has experienced increases proportionally the risk of serious illness and mental health problems.This perspective has gained significant support over recent years [6,10,24,25].Harris [15] suggests that people who report having experienced four or more ACEs are twice as likely to smoke, more than five times more likely to be dependent on alcohol, and ten times more likely to intravenously use substances than people with no ACEs [15].There are, however, signs that some ACEs, in particular those occurring earlier in infancy, may be more important than others in predicting a long-term impact [26].There are signs that substance dependency may have its origin in earlier rather than later occurring ACEs [13,27].For instance, the self-medication hypothesis [28] of substance use is rooted in neural pathways potentially established in response to early childhood distress [25].
While the current study takes a mostly exploratory perspective, these divergent perspectives have been investigated and related to PTSD and continued substance use while in OAT.Implications for treatment services [29,30] and the support they provide will be discussed.OAT has been shown to be effective in maintaining people in treatment, reduce mortality rates, and the harms caused by opiate use [31].However, the fact that people remain in treatment for a long time [32], remains problematic.Since the experience of ACEs can have lifelong consequences on how people respond to stressful situations, identification of specific ACEs is important.Also, how they affect those receiving OAT at present may have implications for how they can be supported on their recovery journey [29,30].This study was aimed at providing an overview of the background, continued substance use, and PTSD among a group in long-term OAT.Two hypotheses and two exploratory research questions around the relationship between ACEs, age of first opiate use, current substance use, and PTSD will be examined.

Participants and Design
The study was conducted with adults who attended substance use treatment services in 2017 in Dublin, Ireland.Participants came mainly from a lower socio-economic background identified as unemployment black spots by the Irish population census data [33].The participants (n = 112) were purposefully followed up from April to October 2019 among 131 people recruited from approximately 890 service users for the study to develop the Healthy Addiction Treatment Recovery Model (HAT) [34,35], within six substance treatments centres.Of the 112 people successfully located, 104 agreed to participate in the follow-up.Individual appointments were made with each participant.Data collection took place using several validated measures to which the participants responded in a structured interview.In addition to the measures utilised for the HAT study [34], two new additional instruments were included to measure PTSD and the occurrence of ACEs among this cohort.These are the focal point of this paper.

Measures
The research instrument contained an extensive demographic questionnaire developed for the HAT study.However, only relevant demographics are reported in this paper.
Substance use aspects were measured using the Opiate Treatment Index (OTI).The OTI assesses substance use (heroin; nonprescribed methadone; prescribed and nonprescribed tranquillizers substances; cocaine; cannabis; inhalants; barbiturates; alcohol, and tobacco) [36].The substance use data was collected on and consolidated to provide a total score for polysubstance use within the OTI.
ACEs data were collected using the ACE questionnaire (ACQ).The ACQ consists of 10 items listed in Table 1.Each item is designed to measure a different aspect of childhood neglect, childhood abuse, and household dysfunction.The items are scored, 1 = yes and zero = no.All 10 items are summed to give a total score ranging from zero to 10, representing the number of ACEs experienced by the participant [8].Reliability tests for the ten items among this sample returned a Cronbach score, α= 0.81.Previous research posits that individuals who have experienced four or more ACEs are at greater risk of serious health problems, psychopathology, and substance use in adulthood [6,8,22].Therefore a cut-off score of less than 4 ACEs and ≥4 ACEs has been chosen for the dichotomised ACE variable.
PTSD was measured using the PCL-5, a 20 item PTSD checklist that measures items corresponding to the 20 PTSD symptoms contained in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [27].Participants were asked to indicate how often they have been bothered by a symptom in the last month for each of the 20 items.Each response ranges from zero = not at all to 4 for extremely and all 20 items are summed up to provide a total score for the instrument with a range of zero-80.Reliability analysis for the PCL-5 returned a Cronbach score, α= 0.94.Bovin, Marx [27] reported the PCL-5 returned good internal consistency (α= 0.96) and suggest that a cut-off range of 31 to ≥33 is appropriate for a preliminary diagnosis of PTSD.However, Walker and Newman [37] found a cut-off score of 30 was optimal for a diagnosis of PTSD among a large sample of hospitalised women.The present study has taken a conservative approach by adopting the upper value of ≥33 in this study.Therefore, the continuous PTSD variable was dichotomised; PTSD scores below 33 = no; PTSD scores ≥33 = yes.

Statistical Analysis
The following demographic data were used in the analysis: age, gender, educational attainment, employment, and time spent in prison.Specific drug use-related questions queried and included in the analysis were current substance use (heroin, other opiates, cocaine, tranquilizers, alcohol) and age of first heroin use.Descriptive statistics for substance use ACEs and PTSD were computed using means, standard deviations, proportions, and frequency tables.
Further inferential analysis included Spearman rho correlations, Pearson's χ 2 , and a linear multiple regression procedure.Tabachnick and Fidell [38] suggest that multiple regression using a sample size of 104 is restricted to the analysis of six independent factors.Therefore, no more than six factors were entered in the models.The process leads to the multiple regressions in three steps.
In the first step, an additional dummy variable (0 = no PTSD; 1 = PTSD) was created for the PTSD variable.This was done by dividing the continuous PTSD scores into two groups; PTSD scores below the cut-off of 33 and scores ≥33.Thus, PTSD could be included both as a categorical variable (based on the diagnostic cut-off point) in χ 2 analysis and as a continuous variable for the regression analysis.For all other variables, the raw scores were maintained.In the second step, Bivariate Spearman's rho correlations were computed to test Hypotheses 1 and 2 which addressed the relationships between the ACQ score, PTSD and four of the substance classes measured within the OTI.
In the third and final step, the multiple regression procedure was performed.Corrections for multiple comparisons are included as standard in the multiple regression procedure.Exploratory analysis 1, whether the number of ACEs and current substance predict PTSD, was tested in regression model 1.The dependent variable was the continuous PTSD factor, and the independent variables were, the summative ACQ score and four classes of substances used in the last 28 days: alcohol; cocaine; tranquillizers, and cannabis (Q scores) [36].Heroin was not included, as too few participants had reported its use (n = 31).Exploratory analysis 2, whether any of the ACE factors predict PTSD, was tested in regression model 2. The dependent variable was the continuous PTSD factor, and the six categorical ACE items were the independent variables.All the variables were entered into the regression models simultaneously with the least significant factors removed sequentially through backward elimination until factors not contributing to model had been removed.Regression results are presented as goodness of fit R 2 and β values with 95% confidence intervals (Table 2).Pearson χ 2 analysis was used to substantiate associations between PTSD and the individual ACE factors.

Demographics
All 104 participants were ethnic European, with an average age of 42.7 years (SD = 7.4).Approximately, two-thirds of the sample (63%, n = 66) were male with a The Importance of Feeling Unloved among People in OAT  median age of 44 years (SD = 7.9) and 40.5 years (SD = 5.0) for female.The mean numbers of years spent in formal education was 9.9 years (male, mean = 9.6, female, mean = 10.5), with 14% of people working full time and 7% in part-time employment.The average length of time in their current treatment was 11.2 years (SD = 6.9) with a maximum length of 27 years, while for 59% (n = 61), this was not their first time in OAT.The median age for opioid use initiation was 18 (IQR of 16-22 years), with 44% of first-time use occurring before the age of 18 years.
Descriptives: All Substance Use, ACEs, and PTSD Over two-thirds of the participants (70%, n = 73) had abstained from heroin use in the previous 28 days, however, 18% still consumed heroin more than once a week (Table 3).Tranquillisers both licit and illicit were used more than weekly (63%), while cannabis (38%), alcohol (31%), and cocaine (19%) were regularly consumed by participants.
The average the number of ACEs was 4 (SD = 2.9) and more than four ACEs were reported by 55% of participants (Table 1).The current study does not address gender differences, however, a greater proportion of female participants, 62% reported four or more ACEs when compared to 32% of male participants.PTSD scores of ≥33 were reported by 40% of participants, with a higher proportion of females than males indicating a diagnosis for PTSD may be appropriate (Table 3).

Hypothesis 1: Number of ACEs and Age of First Opiate Use
Bivariate analysis showed a low but negative relationship between ACEs and the age of first opiate use, supporting the hypothesis that participants with higher ACEs scores started using opiates at a younger age than people with lower ACE scores (Table 4).

Hypothesis 2: Number of ACEs and Current Substance Use
The correlations between current substance use and number of ACEs showed only a significant relationship for cocaine use (Table 4).It would seem that overall, this hypothesis was not supported, and current substance use is not strongly associated with ACEs.The strongest relationship was shown between ACEs and PTSD.This is the focus of the further analysis using multiple regression.
Exploratory Analysis 1: ACEs, Current Substance Use, and PTSD The high correlation between ACEs and PTSD presented in Table 4 is confirmed in the results of the multiple regression analysis presented in Table 2. Model 1 showed that summative ACEs and tranquilliser use explained 50% of the variance in PTSD scores (R 2 = 0.50, F [2, 100] = 51.203,p < 0.001) with summative ACEs emerging as the strongest predictor (β = 0.674) and with a 95% confidence interval between 3.7 and 5.7 (Table 2).These results suggest that for each unit increase in the ACE score, PTSD scores would increase by 0.674 units, demonstrating a strong predictive relationship between summative ACEs on levels of PTSD.Furthermore, the emergence of tranquilliser use as a significant predictor suggests that participants may be using tranquillizers to medicate their PTSD symptomology.

Exploratory Analysis 2: The Different ACEs Factors as Predictors of PTSD
The six ACE factors with the strongest association with PTSD were chosen for entry in model 2 (Table 1) as predictors of continuous PTSD.Four ACE factors explained 54% of the variance in PTSD scores (R 2 = 0.54, F [4, 98] = 30.285,p < 0.001).It is particularly notable that the feeling that they were unloved as a child was found to be the strongest predictor of PTSD (β = 0.328), followed by growing up in a household with a person who was a problem drinker or used any mood-altering substances (β = 0.280).
To substantiate the findings of the multiple regression, a Pearson's χ 2 contingency table was computed for the 10 ACE factors and the categorical PTSD variable discussed above.The findings confirmed the outcomes of the multiple regression.Again, the strongest significant association, those with the highest χ 2 values was shown for ACE 4, "no one in your family loved you or thought you were important or special" (χ 2 = 29.122)(Table 4).Furthermore, bivariate categorical analysis was conducted between PTSD and subjects reporting 4 or more aces and those reporting less than four ACES.The findings showed a significant difference between the two groups, with a mean PTSD score of 41 among those reporting ≥4 ACEs (Table 1).

Discussion
In summary, the findings show that participants were in OAT for a prolonged period of time (on average 11 years).While not particularly unusual [32], this remains a problematic and striking finding.The The Importance of Feeling Unloved among People in OAT effectiveness of OAT treatment was confirmed by the reduced heroin use as reported by participants.However, this did not necessarily extend to other drugs.Tranquillizers and cannabis were reportedly in daily use, while crack cocaine was also frequently used.
In terms of their childhood experiences and how these may relate to their current situation, the participant group reported a high number of ACEs and elevated levels of PTSD.In regard to the relationship between the number of ACEs and the age of first opiate use (Hypothesis 1), findings revealed a small correlation and demonstrated that a higher number of ACEs was associated with an earlier start of opiate use, supporting the hypothesis.Nonetheless, the number of ACEs experienced was found to be mostly unrelated to current substance use (Hy-pothesis 2).Only cocaine use could be predicted somewhat from the number of ACEs, leaving this hypothesis mostly unsupported.What was also quite noteworthy was that the number of ACEs and current tranquilliser use explained 50% of the variance in PTSD scores (Exploratory Analysis 1).Another significant finding is that four of the ACEs predicted 54% of the variance in PTSD scores (Exploratory Analysis 2).Of which, feeling unloved turned out to be the most significant predictor.Feeling unloved may be of particular importance and provides a context in which the lasting impact of traumatic childhood experiences may be exacerbated.This may in fact be at the heart of how substance use started, persists during OAT, and may even be related to why OAT treatment seems to turn into a continuous affair.
As was indicated in the introduction, substance misuse may have its origins in neglect in infancy.While we should avoid jumping to conclusions, an indicator for this is that the ACE least specifically related to events, "feeling unloved" emerged as the most important factor.This unsettling sentiment in early childhood suggests the possibility that normal attachment with caregivers may not have occurred in the participants in the study who reported this ACE.The importance of attachment or lack thereof in studies of substance misuse has been documented [29].Attachment theory argues that the cognitive processes of the child construct internal working models which can influence how they respond to their social environment [39,40].Rewarding relationships with caregivers provide security and a source of meaning and happiness, while separation and loneliness constitute potential risk factors for psychopathology through social isolation [41].Secure attachment relationships in childhood provide a mechanism for the development of emotional regulation, therefore influencing social relationships in later life [26,39,42,43].Insecure or avoidant attachment may reduce psychological wellbeing, and lead to attempts to soothe past traumas with polysubstance use [29,44].
While the connection between the three variables, ACEs, substance use, and PTSD, is not well researched, the separate relationships are well-established.A meta-analysis by Breslau [45] identified childhood trauma as one of three risk factors for the development of PTSD.Furthermore, substance dependency may develop within individuals, as part of efforts to regulate a sense of self and relationships with others [46].Also, the link between PTSD and substance use is widely accepted by multiple researchers.
Our study suggests that PTSD potentially originating in ACEs may be implicated in maintaining people in long-term OAT and current substance use.In addition to the psychological risk of lack of attachment for the development of substance use, a physiological pathway has also been established.Substance use may be a response to early childhood distress because it increases the level of dopamine in the brain, which may help the individual to feel better about themselves, therefore increasing the individuals desire to repeatedly consume substances [12,13,44].This element of functionality is part of the addiction.The importance of dopamine in attachment and bonding between infants and mothers has been known for many years [13,47].An individual experiencing deficits in dopamine levels may turn to substance use, in the absence of protective factors such as social relationships [48].Based on this principle, it would be expected that ACEs occurring in early infancy disrupting attachment, might be of particular importance in precipitating substance misuse and "feeling unloved" may well be a core indicator of this risk factor.The frequent use of tranquilizers by the cohort in the study fits in with this pattern of unease about childhood memories and a felt need to reduce symptoms of PTSD.

Implications for Treatment
The view of addiction as partly an attachment disorder suggests the need for a different approach in substance treatment.This might be aimed at breaking the attachment bond an individual has formed with their substance of choice [12,29,30].For people who are insecurely attached a drug substance may be the only attachment object they have in their lives [29], therefore, an attachment-focused therapy which centres on human relationships may open a path to recovery and abstinence [29].Traditionally addiction treatment has focussed on treating substance use without addressing the underlying co-occurring PTSD, which can have negative implications for treatment outcomes [49].Attachment-focused therapy aims to introduce a more person-centred approach to addiction treatment by reacquainting a person with their emotional self, therefore assisting them in building resilience through fostering healthy interpersonal relationships [29].Bettmann and Jasperson [50] found that attachment-oriented family treatment can be effective in developing positive feelings between parent and adult child enhancing positive effect.Therefore, working to mitigate the emotional repression associated with substance use can become a substitute for and barrier to the formation of long-term human relationships [29,30,44].There are good reasons to consider more alternative approaches for interventions for people in OAT.The length of time a person remains in OAT is costly for both the health care system and the person [51].Also, longitudinal research on people in treatment suggests many current treatment methods are not cost-effective and providers need to look beyond short-term stabilisation towards long-term treatment recovery models [29].While higher ACE scores are directly related to substance misuse, as people get older individual resilience may enable them to recover.In contrast, chronic forms of ACEs may endure across the lifespan which may require a different more attachment informed approach to treatment [29,30,52].A new avenue in treatment should consider the relationship found in the present study between feeling unloved and PTSD in which substance use may well be part of a core effort to manage the associated distress.Breaking through this impasse may require addressing the fundamental impact of feeling unloved.Understanding these particular childhood factors influencing the treatment seeking individual may help service providers to deliver a more person centric and cost-effective journey for each individual on their road to long-term recovery.

Strengths and Limitations
This study collected primary data from a cohort of people attending OAT using internationally validated survey instruments.The data were collected using selfreporting and may contain elements of bias.Participants were asked to recall their early childhood experiences many years after the actual events, therefore recall bias of events may also be a factor [8, Furthermore, the personal nature of the questions may have affected the participants desire to provide socially desirable responses, including underreporting their current substance use.However, Dube and Felitti [8] suggest underreporting of ACEs is more likely therefore misclassifying people as underexposed.A further strength of the study is that the results provided confirmatory evidence of the prevalence of emotional neglect among people with an OUD [3].Although the sample size was appropriate for modelling six independent factors using multivariate regression analysis [38], robust in-depth analysis of co-variates and interaction effects was not possible.A further limitation is the ACE instrument does not capture all potential childhood adversities such as social or racial discrimination, food insecurity, and issues around housing which may have implications when attempting to generalise the findings of this study to racial or ethnic minorities.While there were some gender differences in the findings, overall these were not robust enough to warrant reporting them.

Further Research
The findings from the current study suggest that the lifelong impact on an individual's psychological health from feeling unloved as a child could be a chronic form of ACE and is not explained by a purely summative approach.Further research is needed to investigate the effects of ACEs on people in long-term OAT, treatment, particularly for those who grew up in households without emotional support, parental mental illness, and parental substance misuse.

Statement of Ethics
Ethical approval was granted by the Faculty of Health Sciences Ethics Committee in Trinity College Dublin in December 2018.The committee informed us that the project has ethical approval to proceed subject to compliance with all relevant regulations, including DATA PROTECTION and HEALTH AND SAFETY (Ref.18120, March 3, 2019).The explicit written informed consent was obtained from all the participants in this study.
More than once a week 12

Table 1 .
χ 2 associations between ACEs and PTSD Did a parent or other adult in the household often or very often. . .Swear at you, insult you, put you down, or humiliate you? or Act in a way that made you afraid that you might be physically hurt?Did you often or very often feel that . . .No one in your family loved you or thought you were important or special?or Your family didn't look out for each other, feel close to each other, or support each other?Did you often or very often feel that . . .You didn't have enough to eat, had to wear dirty clothes, and had no one to protect you? or Your parents were too drunk or high to take care of you or take you to the doctor if you # 1. # 4. # 7. Was your mother or stepmother: Often or very often pushed, grabbed, slapped, or had something thrown at her? or Sometimes, often, or very often kicked, bitten, hit with a fist, or hit with something hard? or Ever repeatedly hit over at least a few minutes or threatened with a gun or knife?37 (36) 12.874*** 41 24 (65) # 8. Did you live with anyone who was a problem drinker or alcoholic, or ns, not significant.Significance levels: *p < 0.05; **p < 0.01; ***p < 0.001.# Selected for multivariate regression modelling.

Table 3 .
Demographics and substance use characteristics, ACEs, and PTSD by gender

Table 4 .
Bivariate relationships between ACEs, PTSD, and substance use