Traumatic childhood events of parents enrolled in the Avon Longitudinal Study of Parents and Children (ALSPAC)

Background: Early life experiences can have a significant impact on an individual’s later behaviour, the way they view the world, their beliefs and their success at forming strong interpersonal relationships. These factors may subsequently influence the way that the individual may parent their children, which in turn may have an effect on their child’s behaviour, mental health and world view. Research has linked early traumatic life experiences in the parent’s childhood to disorganised attachment to their own child. In this paper we describe the data collected from parents enrolled in the Avon Longitudinal Study of Parents and Children (ALSPAC) on traumatic events experienced during their childhood, so that it can act as a resource for researchers in the future when considering outcomes on the adult, their children and grandchildren. Methods: Data were collected via multiple questionnaires completed by parents enrolled into the ALSPAC study. During pregnancy and post-delivery, questionnaires were administered between 1990 and 1992 via post to the study mothers and their partners. Data were collected on life events including bereavement, sexual abuse, physical abuse, abandonment, neglect, memories of childhood and accidents. Other reports of traumatic events in childhood were reported by parents using free text. This can be made available to researchers for coding on request.


Introduction
In creating the ALSPAC resource, the research team included questions designed to examine childhood experiences of trauma for the study mothers, their partners and the offspring. This paper focuses on the data collected on the childhoods of the mother and her partner.
Psychological trauma is a complex psychological state resulting from life events that are physically or emotionally harmful and have lasting adverse effects on wellbeing. It is believed that these effects can be passed down intergenerationally to the children of trauma sufferers in the form of relational trauma 1 .
Traumatic events can include (but are not limited to) events such as war/conflict, childhood physical and sexual abuse, natural disasters, traumatic accidents, illness or witnessing events that result in death, threaten death, serious injury or a threat to the physical integrity of self or others 2 . The symptoms resulting long after the trauma has occurred can include depression, anger, anxiety, dissociative disorders and posttraumatic stress disorder (PTSD) 3 .
Whilst we were able to find a large amount of research into the effects of trauma during childhood, and into the effects on the children of parents who experience trauma in adulthood, there is much less research on the effects on the children of parents who have suffered trauma in childhood.
The research that has been carried out to date regarding the relationship between childhood trauma and the impact it has on the offspring of those who experience childhood trauma, is predominantly centred around PTSD caused by parental neglect, domestic or sexual abuse 4,5 and war 6,7 .
However, there appear to be fewer research papers available regarding parental childhood trauma from other sources such as natural disasters 8 , illness 9 , extreme poverty 10 , accidental injury 11 , loss of a parent 12 or sibling, or being a witness to a distressing event 13 . Fisher et al. (2013) 14 , studied data from ALSPAC and found an association between adverse early life experiences, harsh parenting and bullying, and psychotic symptoms in the study children. The researchers used variables relating to the child's lifestyle and life events as reported by the study mother, and evaluated the child's early environment and bullying data, self-reported by the child to test for adversity. They then evaluated the child's neurocognitive and psychological markers using these measures: Locus of Control gathered using a 12 item version of the Nowicki-Strickland Internal-External scale (NSIE) 15 , self-esteem using a shortened form of Harter's Self Perception Profile for Children 16 , affective symptoms using the Development and Wellbeing Assessment (DAWBA) 17 and the Short Moods and Feelings Questionnaire (SMFQ) 18 both completed by the child's parents, and finally, 12 core questions from the semi-structured psychosis interview (PLIKSi) 19 . Fisher et al. concluded that harsh parenting was not a factor for psychotic symptoms in the child but was indicated in depressive symptoms, anxiety, external locus of control and low self-esteem. It would be interesting to see whether the parent exhibiting the harsh behaviour also suffered from these symptoms as a result of similar or other traumatic experiences during their own childhoods and whether there is a link to intergenerational trauma here. Plant et al. (2018) 20 analysed 9,397 ALSPAC mother child dyads with the aim of examining whether children whose mothers had a history of childhood trauma, were at increased risk of exhibiting psychopathology. The researchers tested three main hypotheses, that maternal child maltreatment (defined as whether the mother had been maltreated as a child) would predict child internalising and externalising difficulties, that maternal antenatal depression, maternal postnatal depression, maternal maladaptive parenting and child maltreatment would operate as independent mediators and finally, that there would be an indirect effect of maternal postnatal depression through maladaptive parenting and child maltreatment demonstrating separate cumulative effects. The researchers measured whether the mother had experienced physical abuse, sexual abuse, emotional abuse or neglect during their childhood (defined as <18 years). If the mother answered 'yes' to any of these exposures, they were considered to have experienced childhood maltreatment. They also measured whether the mothers showed signs of depression (antenatally and postnatally) using the Edinburgh Postnatal Depression Scale (EPDS) 21 ; and tested for symptoms of maladaptive parenting such as shouting, slapping and feelings of hostility towards their child. Child maltreatment (physical, sexual or emotional abuse) was elicited through maternal self-report. At 8 years of age the children were asked to self-report whether they had been bullied by their peers. The child was considered to have experienced maltreatment if a positive response had been given to any of these questions. Child depressive symptoms were measured at ages 10 and 13 by asking the mothers to complete the Development and Wellbeing Assessment (DAWBA) 17 and at 11 years, the Strengths and Difficulties (SDQ) 22 questionnaire. The researchers then looked at symptoms of attention-deficit hyperactivity disorder (ADHD), conduct disorder and oppositional defiant disorder (ODD), and found that maternal childhood maltreatment directly predicted the study child being exposed to maltreatment and subsequently developing psychopathology. This effect was greater when tested alongside maternal depression. Children of maltreated mothers had significantly greater emotional and behavioural difficulties at the ages of 10, 11 and 13, as well as greater peer conduct difficulties and hyperactivity problems. Maternal antenatal depression, postnatal depression and child maltreatment had an effect of independent mediation of the association between maternal child maltreatment and both internalising and externalising behaviour difficulties. Maladaptive parenting did not show this result.
In this paper we describe the data collected from parents enrolled in the Avon Longitudinal Study of Parents and Children (ALSPAC) on traumatic events experienced during their childhood, so that it can act as a resource for researchers in the future when considering outcomes on the adult, their children and grandchildren.

Methods
Participants 14,541 pregnant mothers' resident in the former Avon county of South West England were recruited into the ALSPAC study. These mothers all had an expected delivery date of 1 st April 1991 to 31 st December 1992. From these pregnancies, there were a total of 14,676 fetuses and 14,062 live births. Of these children, 13,988 were still alive at 1 year of age. Mothers were considered enrolled if they had returned at least one questionnaire or attended a "Children in Focus" clinic by 19 th July 1999. At the age of 7, the study team reached out to mothers who had previously not been included in the study and recruited additional families in order to boost the number of participants. As such, from the age of 7 the total sample number is 15,454 live births, resulting in 15,589 foetuses, of which 14,901 alive at 1 year of age 23,24 . However, these additional parents did not have data collected on their own childhood exposures.
In order to protect the confidentiality of the sample, data from triplet and quad multiple births have been removed as these children were considered to be at risk of identification. This leaves 14,691 eligible participants remaining. ALSPAC is continuing to monitor these families and are recruiting the Children of the Children of the 90's 25 . The ALSPAC team continue to gather data concerning the parents and grandparents of the study children, enabling further intergenerational research.
Following the advice of the ALSPAC Ethics and Law Committee, partners were recruited into the study only if the mothers wished them to be included. Questionnaires were sent to the mother who then passed the questionnaire on to the partner with a separate pre-paid return envelope. This method meant that ALSPAC were unable to follow up or communicate directly with the partners 24,26 . Therefore, the numbers of partners' questionnaires returned were less than those received for the mother's questionnaires. Around 75% of the partners participated in the study.

Data collection
Since its inception, the ALSPAC study has collated thousands of variables from the study children and their families using questionnaires and clinics to collect the data. This paper focuses on the variables we identified as being indicative of maternal childhood trauma experiences. There were 186 maternal variables which were collected from the A, B, C and D questionnaires, and 160 variables from the Partner's collected via the PA, PB and PF questionnaires (Copies of these questionnaires can be viewed on the ALSPAC website). The variable naming convention for these questionnaires usually follows the format of the questionnaire's assigned alphabetical letter followed by the number of the variable allocated in a sequential pattern in the data set.
The first four questionnaires were sent out at specific time points based on the mother's gestation at enrolment. Partners were also surveyed at the mother's discretion. The questionnaires were sent to the mother and the mother decided whether or not to pass that questionnaire to the partner.
Provided the mother enrolled before 14 weeks gestation, questionnaire A "Your Environment" and PA "You and Your Environment" were sent out on enrolment. If the mother enrolled after 14 weeks, the questionnaire was not sent until 22 weeks gestation. This was to prevent the mother feeling overwhelmed with questionnaires and to avoid clashes with the Having a Baby and Your Pregnancy questionnaires which were confined to certain gestation periods. In total, the Your Environment questionnaire was sent to 45% of participants before 15 weeks gestation, 32% between 22-25 weeks gestation and 23% were sent out later than 25 weeks.
Questionnaire B "Having a Baby" was sent out from 18 weeks gestation until 23 weeks gestation. If the mother enrolled before 24 weeks gestation, she also received the PB "Partners Questionnaire" alongside the B questionnaire. If the mother enrolled after 24 weeks, the PB "Partners Questionnaire" was sent out with the mothers "Your Home and Lifestyle" questionnaire. The "Your Home and Lifestyle" questionnaire is a version of the A "Your Environment" questionnaire which has been adapted for use by mothers who enrolled later into the study and therefore this data is coded alongside the A questionnaire data.
Questionnaire C "Your Pregnancy" was sent out from 32 weeks gestation until 40 weeks gestation. For questionnaire D "About Yourself" the timing of when this questionnaire was considered to be less important and therefore the bulk of these were sent out from 14 weeks gestation until 37 weeks gestation. However, in a small number of cases, questionnaire D was sent after the birth of the study child.
The H "Your Health Events and feelings" questionnaire was sent to mothers at 33 months postpartum and was accompanied by the PF "Partners Health Events and Feelings" questionnaire.
Please note that the study website contains details of all the data that is available through a fully searchable data dictionary and variable search tool.
The ALSPAC team have also provided a Questionnaire Topic guide which summarises the topics in each questionnaire.
The majority of the other questions available were drawn up by either the ALSPAC team or the European Longitudinal Study of Pregnancy and Childhood (ELSPAC) team. However, where suitable established measures existed, these measures were used within the questionnaire, occasionally with minor modifications. Table 1a and Table 1b show the demographics of the parents who completed the questionnaires that contained the childhood

Scale of life events
The life events questions to the mother (Table 2a) comprised a set of 31 specific questions based on the previous work of Coddington 27 , regarding events that may or may not have happened to the respondent before the age of 17. These questions were included in the questionnaire C sent to the mother at around 32 weeks gestation. 30 similar questions were also sent to the mother's partners in their own questionnaire administered during pregnancy, the question to the mothers concerning whether they had acquired a stepsibling was inadvertently omitted from the partner's questionnaire (Table 2b). For each specified item the respondent was asked to choose between five possible answers: 'yes, affected me a lot'; 'yes, moderately affected'; 'yes, mildly affected'; 'yes, but did not affect me'; 'no, did not happen'. For each parent a question concerning anything else that occurred was asked, with a description written as text; these responses are not included here.
From the life events scale, two variables were created. The first, the weighted life events scores (c432 and pb481) gives an indication of the perceived degree of effect of each life event the parent experienced. Thus, c432 was created by selecting variables c400-c430, reversing the coding for these variables and setting 'no' to 0 (unless all were missing in which case the variable was left as missing). Therefore, greater perceived effects will have a higher score; addition of all life events variables together provides the overall score. [Variable c431 was not included in the score as analysis showed that when another event was noted, it closely matched another variable between c400-c430 and had been coded into that variable]. The same coding system was used for the partners' variables, pb450 to pb479 (Figures 1a, 1b). The range for mothers and partners were 0-107 and 0-78, with medians of 6 and 7 respectively.
The second derived variables, the Life Events Scores (c433 and pb482), gives the number of life events experienced by each parent and was created by recoding the variables to make all 'yes' responses as 1 and the 'no' responses as 0, and then adding these variables together. For the mother's data, the variables used to create the score were c400-c430. The variables used to create the partners' score were pb450 to pb479. The ranges were 0-27 and 0-22 and medians 3 and 3.5 for mothers and partners respectively (Figures 2a, 2b).

Other childhood circumstances
Questions relating to other circumstances during the parents' childhood can be found in Table 3a and Table 3b. Variables indicating further questions where the respondent was asked to indicate their actual age at which circumstances occurred are shown in Table 4a and Table 4b. Although not listed in this table, ages are also available for when specific accidents and injuries had occurred, thus enabling childhood events to be computed (see variables d080 to d130 for the mothers and pa110 to pa163 for their partners). Subsequent to pregnancy, at 33 months after delivery, further questions were asked about abuse during childhood as well as of violent events in the home. The reason for the delay in asking these questions concerned the realisation that they should have been asked earlier but would still be valid to ask at this stage in the parents' life course (Table 5a and Table 5b).

Childhood happiness and unhappiness
Each parent was asked to rate their memories of happiness/ unhappiness at each of three ages: under 6, 6-11 and 12-15 years. As a validation exercise, the mother was asked to do this on two separate occasions, and the data are found in the C files and the D files. Table 6a shows these data, and Table 6b lists the partners' memories of happiness in their own childhoods as self-reported on one occasion during pregnancy.

Maternal care score
The Parental Bonding Instrument 28 is a scale that seeks to define the principal dimensions of child/parent bonding and to examine the importance of these factors in determining the strength of a child's relationship with their parent. In 1987, Gamsa 29 re-wrote several of the questions in the Parental Bonding Instrument as in the original test, five of the questions contained double negatives which were considered to be confusing to participants.
The ALSPAC team also changed the responses for the questions as piloting revealed that the participants did not like the original options given in the test. Participants were asked to look at statements and rate how like their relationship with their mother the statements were. The answers in the original scale are "Very like", "Moderately like", "Moderately unlike" and "Very unlike". These answers were changed to "Never", "Sometimes" and "Usually". In addition, three questions were removed from the test due to their similarity and the perceived repetition causing annoyance to the participants. The responses to each question used are shown in Table 7a and  Table 7b for each parent.
From the questions two separate scales were derived as specified by Gamsa indicating (a) maternal care and (b) maternal over-protection. These can be found in Table 8. To create these variables, all coding was reversed for variables d702, d703, d708, d709 and d711 to make the negative response have the higher value. The binary variables d713, d714, d715, d716, d717, d718, d719, d720 and d721 were also recoded so that they faced the same direction.
The Maternal Care Score variable comprises a continuous measure of how much the study parent thought their own mother cared for them on a scale of 0-24, with 24 being the most caring mothers. The Overprotection Scale measured how overprotective the study parent thought their own mother was on a scale of 0-20, with 20 being the most overprotective mothers. The numbers of cases in each scale are shown firstly for the scales based on complete data, secondly for scales where the missing response to the individual data have been replaced by the mode for the item, and thirdly the number of such missing items are given.

Strengths and limitations of the data
The strengths of these data include the large sample size, with over 20,000 participants with data available 23 . The only inclusion requirements at enrolment for this study were the geographical location the participating mother resided in and the expected date of delivery. The participants recruited to the study were broadly representative of the general population of new parents' resident in the area at the time in terms of sex, ethnicity and socio-economic status 24 .
All parent participants received the same questions and one of the major strengths of this study is the vast array of other       data that is available with several decades of follow up and the opportunity to examine effects across generations. Available data includes information about the study parents' relationships with the study child, biological markers from parents and children, data regarding the grandparent's health, life experiences and demographics and data gathered from the grandchildren. This makes the data very flexible and relatable to intergenerational aspects of the family's life.  A key limitation of the study is the lack of ethnic diversity. At the time of enrolment, the county of Avon was mainly Caucasian, therefore there were too few Black, Asian and Minority Ethnic (BAME) participants (<6% in total) to allow for detailed analysis by ethnic background.

Data availability
Underlying data ALSPAC data access is through a system of managed open access. The steps below highlight how to apply for access to the data included in this paper and all other ALSPAC data. 1. Please read the ALSPAC access policy (http://www. bristol.ac.uk/media-library/sites/alspac/documents/ researchers/data-access/ALSPAC_Access_Policy.pdf) which describes the process of accessing the data and biological samples in detail, and outlines the costs associated with doing so.
2. You may also find it useful to browse our fully searchable research proposals database (https://proposals. epi.bristol.ac.uk/), which lists all research projects that have been approved since April 2011.
3. Please submit your research proposal (https://proposals. epi.bristol.ac.uk/) for consideration by the ALSPAC Executive Committee using the online process. You will receive a response within 10 working days to advise you whether your proposal has been approved.
If you have any questions about accessing data, please email: alspac-data@bristol.ac.uk (data) or bbl-info@bristol.ac.uk (samples).

Ethical approval and consent
Prior to commencement of the study, approval was sought from the ALSPAC Ethics and Law Committee and the Local Research Ethics Committees 2 . Informed consent for the use of data collected via questionnaires and clinics was obtained from participants following the recommendations of the ALSPAC Ethics and Law Committee at the time. Questionnaires were completed in the participants own home and return of the questionnaires was taken as continued consent for their data to be included in the study. Full details of the approvals obtained are available from the study website (http://www.bristol.ac.uk/ alspac/researchers/research-ethics/). Study members have the right to withdraw their consent for elements of the study or from the study entirely at any time.
transmission. As investigators take advantage of the roadmap that Ellis et al provide, however, it would be well for them to have the potential for such variations in mind.