At a time when there are concerns that Covid is being over-diagnosed, and its impact exaggerated, it is useful to look at another claim: that traumatic events in childhood have a major effect for the rest of a person’s life.
On 18 November 1987 a fire at King’s Cross Underground station in London killed 31 people, and the next day, with two colleagues, we established a trauma service for survivors and their families. In the next decades I worked on many disasters, and in all of these we were not only providing therapy, but were also interested in understanding whether the severity of the actual trauma or the vulnerability of the victim was the main factor in them being disabled by the event. Most of the events were horrendous, with high death rates, in one case 50% dead, so we had no doubt that they close to combat stresses in their intensity. Our impression was that both the severity of the event and the personality of the victim were equally important, but the correlations for each factor with psychological symptoms were only r=.35 so not that big.
At that time the concept of post-trauma stress was far less known and accepted, but as time went by it became applied to a very broad range of life events. In particular, one type of trauma, child sexual abuse, became far more frequently used as a diagnosis, usually on the basis of therapist-assisted recall of childhood events. This became a minefield, as more and more patients claimed their lives had been ruined by their parents.
There were certainly well documented cases of child sexual abuse, so distinguishing between real and possibly imagined cases became important, but very difficult to achieve in therapeutic practice. In a disturbing move, some therapists said it was wrong to doubt any patient’s account of having been abused as a child. Other therapists circulated credulous accounts of satanic ritual abuse which, if true, would have been far easier to track down and prosecute than the hidden abuse taking place by individuals in private houses or children’s care homes, simply because there were more witnesses, and more chance of blackmail and denunciations and a collapse of the conspirators. Also, in the accounts of child sacrifices, there were further lines of police inquiry.
There is no doubt that some adults abuse children and are rightly convicted of their crimes, and that the plea of “no harm done” is rightly rejected because children cannot give informed consent. There is no doubt that child abuse is often difficult to prosecute, because children may not be able to understand what is happening. At the same time, some assumptions may be wrong. At a big conference of therapists I was astounded to hear a lecture discussing satanic rituals as if they were both real and frequent, and even more astounded to find audience members agreeing, without supportive evidence.
Bewildered by this, I wondered why no one was following up actual documented cases of child abuse, to see what psychological effects could be traced back to actual real events. I knew some people were doing a prospective study, and also knew it would take a long time to be completed, so it was a matter of waiting. Now that study has been published, and in my view the results are explosive. The big determinant of bad effects from traumatic childhood events is whether you think they happened, not whether they actually happened.
Objective and subjective experiences of child maltreatment and their relationships with psychopathology Andrea Danese & Cathy Spatz Widom. Nature Human Behaviour volume 4, pages 811–818 (2020)
The authors say:
Does psychopathology develop as a function of the objective or subjective experience of childhood maltreatment? To address this question, we studied a unique cohort of 1,196 children with both objective, court-documented evidence of maltreatment and subjective reports of their childhood maltreatment histories made once they reached adulthood, along with extensive psychiatric assessment. We found that, even for severe cases of childhood maltreatment identified through court records, risk of psychopathology linked to objective measures was minimal in the absence of subjective reports. In contrast, risk of psychopathology linked to subjective reports of childhood maltreatment was high, whether or not the reports were consistent with objective measures. These findings have important implications for how we study the mechanisms through which child maltreatment affects mental health and how we prevent or treat maltreatment-related psychopathology. Interventions for psychopathology associated with childhood maltreatment can benefit from deeper understanding of the subjective experience.
This is a pivotal issue in psychology. Should we be helping people overcome a real-life event, or correcting a damaging mis-perception? The therapeutic approaches would probably be very different. Court cases are highly detailed specific events, which act like a radioactive tracer for later life outcomes. If childhood traumas have the status they are accorded in Freudian thinking, then children whose abuse is bad enough to have the parents taken to Court should be highly disturbed, and permanently so. The results will be worse than that caused by ordinary parents, who sometimes argue, punish unfairly and are generally tedious, or worse, embarrassing. If parents mess you up, then these Court cases will be picking up the extreme examples of parental mistreatment.
On the other hand, Court cases are not sensitive, in the sense of picking up all cases of abuse, but they will provide a solid benchmark for evaluating outcomes in those actually abused.
We have studied one such unique sample. Maltreated participants (n = 908) were identified as victims of child abuse or neglect based on official records from juvenile (family) and adult criminal courts in a metropolitan area in the Midwest United States during 1967–1971. A comparison group was painstakingly drawn of children without official records of abuse or neglect matched on the basis of age, sex, race/ethnicity and approximate family social class at the time of the child maltreatment (n = 667). During a follow-up assessment between 1989 and 1995 (mean age 28.7 years), 1,196 study participants underwent a 2-hr in-person interview, which included assessment of retrospective reports of childhood physical abuse, sexual abuse and neglect, as well as assessment of current and lifetime psychopathology.
Objective and subjective measures of child maltreatment identified largely distinct groups of participants (Cohen’s κ = 0.25) with poor agreement across all maltreatment types (child physical abuse κ = 0.09; child sexual abuse κ = 0.17; child neglect κ = 0.32, consistent with meta-analytical findings.
It was possible to separate the relative contribution of objective and subjective measures of child maltreatment to psychopathology studying three target groups: (1) adult participants who were identified as victims of child maltreatment by virtue of official records but did not retrospectively recall the experience (objective measure); (2) adult participants who were identified as victims of child maltreatment by virtue of official records and also retrospectively recalled the experience (objective and subjective measures); and (3) adult participants who retrospectively recalled being maltreated in childhood but were not identified as victims of child maltreatment by virtue of official records (subjective measure).
We found that participants identified as victims of child maltreatment only by virtue of official records did not have different risk of any lifetime psychopathology compared to those with neither objective nor subjective measure of childhood maltreatment (objective measure: risk ratio, RR = 0.92). In contrast, participants with both objective and subjective measures of child maltreatment showed greater risk of any lifetime psychopathology compared to those with neither measure (objective and subjective measures: RR = 1.35. Finally, participants identified as victims of child maltreatment only through retrospective recall also showed elevated risk of any lifetime psychopathology compared to those with neither measure (subjective measure: RR = 1.29).
So, if you had actual maltreatment as shown by a Court case, but you did not recall it, you were left no more disturbed than the general public. Those who had been maltreated and recalled it were more likely to be disturbed, and those who reported abuse even without objective evidence were also more disturbed.
These are interesting charts. First, if you look at the Venn diagrams it is clear that the most common form of child maltreatment to come before the Courts is neglect. Physical and sexual abuse are far rarer (or far more difficult to detect). This has been known for a long time, but media tend to publicize sexual abuse. Second, if you look at the histograms, the column for the control group has been shown in such light grey that it is invisible. You can only detect it by looking at the error bars. These show that children who have gone through bad events are no more affected than children who haven’t gone through those events, so long as they don’t recall them. The devil lies in the recall. A corollary is that if your recall is faulty, you may suffer from your mis-perceptions.
In their discussion, the authors say:
The risk of psychopathology linked to objective experiences of childhood maltreatment, even for severe cases of maltreatment identified through official court records, is minimal in the absence of a subjective appraisal. In contrast, the risk of psychopathology linked to subjective experiences of childhood maltreatment is high, whether or not subjective appraisal is consistent with objective measures. The findings were remarkably invariant across different types of maltreatment and psychopathology and across genders and races, expanding initial observations made with regard to drug abuse in this cohort. These results suggest that psychopathology emerges as a function of subjective rather than objective experience of childhood maltreatment.
As discussed before, Courts won’t pick up all the cases of child mistreatment, so there will be many false negatives. Those with a “subjective” appraisal may have been maltreated, and never properly investigated. Nonetheless, the apparent preponderance of the “subjective” group leads us to doubt the directly causal nature of childhood events.
Because of the low agreement between objective and subjective measures of childhood maltreatment, aetiological studies based on subjective measures of maltreatment are unlikely to identify damages or abnormalities linked to actual exposure to maltreatment; rather, they are likely to identify correlates of unhelpful cognitions/memories about the self and the environment, which appear crucial to understanding risk of psychopathology.
Unhelpful cognitions are of course a big part of all emotional distress, regardless of life events. The authors continue:
Studies with sampling based on objective measures will be needed to understand why some maltreated children develop subjective appraisal of their ordeal while others do not, for example investigating the severity of the actual maltreatment experience, the intensity of subjective distress reported by maltreated children, the age at which the abuse took place, the role of social care involvement in buffering or accentuating the distress and the experience of later adversity. Finally, studies with sampling based on objective measures will also be needed to understand why some adults develop subjective appraisal of childhood maltreatment in the absence of objective experience, including the role of residual memory biases linked to previous psychopathology, personality, suggestibility and source-monitoring errors.
Even in the absence of an actual experience of child maltreatment, some individuals may endorse cognitions/memories about themselves and their childhood environment that amount to a subjective experience of maltreatment, which is also associated with elevated risk of psychopathology. Faulty memories, or apparent memories elicited by faulty procedures may be the causes of psychological distress.
In fact, the most common form of difficulty in this group was alcohol misuse, then drug misuse; and a step down after that PTSD, depression and anti-social personality. There was little generalized anxiety, which is a surprising finding.
When drawing up the control group, it was found from Court records that 11 children had been maltreated, so they were dropped from the group. This gives a very rough estimate of the rate of abuse (mostly neglect) in this sort of child population, which is 11/667 or 1.7%. This is a useful statistic, and serves as a benchmark for the prevalence of child maltreatment as detected by legal procedures in this US Mid-West population in the 1970s.
Talking about those abused, the authors say: “Approximately half the sample are female (48.7%) and about two-thirds are white (62.9%)”. Not mentioned is that in 1970, the US Mid-West population of 56.5 million was 91.4% white, 8.1% black.
https://www.census.gov/content/dam/Census/library/working-papers/2002/demo/POP-twps0056.pdf The relevant table is on page 21.
The composition of the abused sample (for lifetime pathology) is n=735 participants of White ethnicity (65.4%) and n=389 participants of Black ethnicity (34.61%). If the court cases are representative of the extreme end of child maltreatment in the community, it would appear that black children were abused at 4.3 times their expected rate. US nationwide assessments made in 1996 give the non-white percentage of abused children at 23% to 26% (pages 187 and 189).
In summary, this good paper by Danese and Widom has taken a forensic look at child maltreatment, and has shown that there is no direct link between the event and subsequent emotional responses. Responses depend on personal interpretation and recall. Even more, it has shown that simply thinking you have been maltreated, even when there is no objective proof of it, is sufficient for emotional distress to occur. The authors are too cautious to spell it out, but some persons alleging that they were maltreated in childhood may be mistaken. Worse, they may be making false allegations against their parents.
The paper does not report assessing the parents in any way as regards their own mental health. Perhaps there are data in the Court records somewhere. If so, it would be essential to include them, because as it currently stands there is a profound genetic confound here. Disturbed parents are more likely to have disturbed children, and incidentally are also more likely to not care for them properly, because these parents are subject to emotional and educational difficulties of their own which diminish their ability to care for their children. On this important matter, the paper is silent. In other work the lead author has looked at the interplay between trauma and the polygenic risk scores for major depression, finding that they interact in a complex manner, so there is more that can be said about this.
But, to sum it all up, as Marcus Aurelius did in his Meditations:
If you are distressed by anything external, the pain is not due to the thing itself, but to your estimate of it; and this you have the power to revoke at any moment.”