Myths of Childhood

Joel Paris

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781134873258
• Edition: 1
• Topic: Psicologia
• Subtopic: Storia e teoria della psicologia

Childhood has long been considered the major factor in determining adult life. It sets us on the path toward or away from happiness, shapes our personality, and is a major cause of mental disorders. Or is it?
Myths of Childhood strongly challenges these assumptions usually taken for granted in contemporary society and the mental health community. With a healthy dose of scepticism toward clinical impressions and using empirically-based research from areas including behavioral genetics and attachment, Dr. Paris builds a convincing case against the primacy of childhood in the development of adult personality and psychopathology. In its place, he offers an alternative model for development and shows how mental health professionals can apply this model to clinical pracitce.
Myths of Childhood represents an important addition to the ongoing debate between mental health professionals regarding nature vs. nurture. For supporters of either side, this book is a valuable resource for further exploration of this controversy.

PART

I

CHILDHOOD AND ADULTHOOD: THE EVIDENCE

Part I consists of four chapters examining the empirical evidence concerning the impact of childhood experience on adult functioning.

Chapter 1 discusses the problems in establishing cause and effect relationships between life experiences and psychological symptoms. Relying on simple associations between risks and outcomes can lead to surprisingly incorrect conclusions about the causes of psychological problems. The social psychology of attributions helps to explain some of these common errors in logic.

Chapter 2 reviews the theoretical assumptions behind primacy, and shows how the scientific evidence contradicts them. The disconfirming data include discontinuities between childhood and adulthood, the reversibility of effects of early experiences through later experiences, and the fact that an unhappy childhood has very different effects on different people.

Chapter 3 reviews research on the long-term sequelae of various adversities during childhood: poor quality of parenting, marital conflict, separation or divorce, traumatic experiences, and socioeconomic deprivation. In each case, single adversities have weak relationships with outcome, while impacts are greater when adverse event are mulitple and cumulative.

Chapter 4 reviews research on resilience. Only a minority of children exposed to traumatic experiences suffer long-term sequelae. The mechanisms by which people survive a traumatic childhood include intelligence, favorable personality traits, the availability of alternate attachments, and the buffering effects of extended family and social community.

1

CHAPTER

Establishing Cause and Effect

Behavior does not follow simple laws. This is why psychology does not resemble classical physics. The simplest outcome can emerge from a complex matrix of factors. The same outcome can result from many different causes. The same causes can lead to entirely different outcomes.

All these complexities make it difficult to determine whether a risk factor, such as an adverse event during childhood, is the true cause of an outcome, such as problems during adulthood. This chapter will describe a series of problems in scientific methodology bearing on this difficulty:

1. the base rate problem,

2. the difference between risk factors and causes,

3. the difference between correlation and causation,

4. the difference between clinical and community populations,

5. distinguishing the roles of shared environment and shared genes,

6. the problem of retrospective methods,

7. the difference between statistical and clinical significance,

8. the contrast between clinical inference and research data.

When these pitfalls go unrecognized, researchers and clinicians can reach incorrect conclusions. The chapter concludes with a discussion of attribution theory, which sheds light on the reasons why we make false inferences.

The Base Rate Problem

Most people experience their share of adversity. This is true at any stage of life. A trauma-free childhood is probably exceptional. Therefore, to the extent that negative childhood experiences are frequent in the general population, many patients are bound to report them. It is important to remember that large groups of nonpatients have had parallel experiences, yet do not develop the same sequelae.

This is a classic example of a base rate problem. If a risk factor is sufficiently ubiquitous, then even when it is commonly associated with an outcome, the relationship cannot be considered as truly causal. Thus, reported childhood adversities, even when reported frequently, do not necessarily, by themselves, account for symptoms in adult patients.

To establish whether an adverse life event is a valid cause of clinically significant symptoms, we also need to decide what cut-off point to use. In other words, What is a case? This well-known problem in clinical epidemiology arises from the fact that most forms of disorder lie on a continuum with normality, with pathology emerging only on the extremes (Goldberg & Huxley, 1992).

To know what a case is, we must address the difference between distress and disorder. Everyone is affected by adverse life experiences, but few become clinically ill as a result. For example, although most of us have times when we are unusually sad, the prevalence of clinical depression in the population is only about 5% at any given time, and no more than 20% over the course of a lifetime (Kessler et al., 1994; Robins & Regier, 1991). Thus, while distress is almost universal, disorder is less common.

This problem with determining “caseness” has important implications for measuring resilience. When we are conservative about defining pathology, most people appear highly resilient. But when we are liberal about defining pathology (as the psychodynamic model tends to be), resilience can appear exceptional.

Some might argue that we miss important effects of adversity by counting only those who have clinical levels of psychopathology. In this view, we should give equal weight to those affected in any perceptible way, even if they have no diagnosable disorder. The main reason for rejecting this line of argument derives from the base rate problem.

Epidemiological surveys (Kessler et al., 1994; Robins & Regier, 1991) show that one third of the population have a measurable psychological disorder at any given time. But if we limit consideration to those mental disorders with a severe effect on functioning (i.e., schizophrenia, major mood disorders, severe substance abuse, and severe personality disorders), only about 10% to 15% of the population are affected.

These figures do not take distress into account. Some symptoms, that is, complaints that do not meet criteria for a diagnosable disorder, can be found in about another third of the population (Srole, 1980). At some point, being mildly troubled becomes perfectly normal! This issue is particularly relevant to issues raised in the next few chapters of this book. Unfortunately, many of the studies of the impact of adversity use distress rather than disorder as an outcome (e.g., studying the effects of early experience on self-esteem).

In summary, we need to define caseness in such a way as to avoid making the distinction between normality and pathology meaningless. Life involves many travails, and no one can expect to be happy most of the time. Models that pathologize all human suffering run the risk of applying a standard of mental health that makes almost everyone into a patient. Using distress as a criterion also trivializes the question of whether unhappiness in childhood causes unhappiness in adulthood.

This is not to deny that people can be deeply affected by their childhood. Intuitively, most people are aware of carrying some scars from our early lives. But when one measures outcomes too broadly, one ends up concluding that there is no such thing as resilience.

In summary, psychopathology is not just a matter of distress. Clinicians should not diagnose illness without a clearcut impairment in functioning. In that context, when we ask whether childhood experiences cause mental disorders, the answer is: usually not.

Risk Factors and Causes

We can open the newspaper almost any day and read about the results of the latest research on risk factors for disease. Over morning coffee, we learn that researchers are arguing about whether too much caffeine increases the risk for cancer. Reading on, we discover that it is a bad idea to drink too much alcohol, that it is dangerous to be too fat or too thin, and that although we need to exercise, we should be careful not to overdo it. One can only conclude that life is a fatal disease!

But we need not become unnecessarily concerned. All of these purported dangers are risks, but none of them are causes. Risk factors consist of anything that makes disease more likely. But we need to know how likely that outcome is. If the base rate in the population is 1 %, and a risk factor doubles it to 2%, such information provides very little reason to change one’s lifestyle.

When are risk factors truly etiological? Only when: (a) they precede the development of pathology; (b) they are consistently, strongly, and specifically associated with the disorder; and (c) there is a plausible mechanism linking them with outcome (Regier & Burke, 1989).

Let us consider each one of these criteria in turn. First, a risk factor must precede the development of pathology. If not, both risk and outcome may well be due to a third factor, termed a latent variable. Moreover, it is difficult to determine whether risk or outcome takes precedence when adversities are assessed retrospectively.

Second, the relationship must be consistent, strong, and specific—if not, a risk is unlikely to play a major role in outcome. Actually, the consistency and strength of risks in childhood are not as great as one might expect. As Chapter 3 will show, few adversities during childhood lead to predictable long-term effects. Even in the short term, most risks, by themselves, will affect only a minority of those exposed. Even the most severe traumas are associated with adult pathology in about 25% of cases.

In many ways, 25% looks like a high rate. If, for example, a couple were receiving genetic counseling, that level of risk would be high enough to raise serious questions about conceiving. But we are not comparing 25% to 0%. These numbers have to be seen in the context of relatively high base rates of mental disorder in the population.

Moreover, the risk of sequelae for any adversity is rarely as high as 25%. The vast majority of exposed children do not develop any serious pathology. This is why clinicians can conclude very little, simply from observing symptoms, about the probability of finding any specific risk factor. Moreover, most risks in child development are not very specific. Chapter 3 will provide examples of how similar outcomes can arise from many different causes, while causes do not lead to predictable outcomes. This is another reason why the presence of childhood adversity cannot, by itself, explain the development of adult symptoms. Some other factor is required; Chapters 9 and 10 will suggest what that is likely to be. Finally, we cannot accurately describe the mechanisms by which childhood adversities cause adult pathology. We have many theories, but few certainties.

There are several good examples in medicine of risk factors that meet most criteria for being true risk factors. A good example is the relationship of cigarette smoking to lung cancer and cardiovascular disease. We know that smoking frequently and consistently leads to serious disease, and we also have a very good idea of the mechanism by which tobacco causes pathology. Therefore, there can be little doubt that this relationship is etiological.

In contrast, the association between childhood trauma and mental illness is neither consistent, strong, nor specific. Moreover, it lacks a plausible pathogenic mechanism. Even if the criteria defining a valid risk factor listed above are met (and they rarely are), most people exposed to risk factors never become ill, and if they do, they rarely become ill in the specific way suggested by reported associations with disorders. As will be discussed later in this chapter, when relationships are purely statistical, they need not, and usually do not, apply to any particular individual.

Thus, from the retrospective point of view embodied in the clinical encounter, we cannot readily understand why anyone develops pathology. Simply knowing that someone has been exposed to specific risk factors is not enough. Even from the prospective point of view of the researcher, we cannot predict which individuals exposed to any risk factor will develop mental disorders.

In summary, risk factors often contribute to a pathological outcome but should not generally be thought of as either necessary or sufficient conditions for their development. The causes of psychological symptoms are almost never single but depend on complex interactions between many risk factors.

Correlation and Causation

Correlation does not prove causation. This saying is so well known that it has become a cliché. Yet the principle is often ignored. The difference between correlation and causation is crucial to understanding the relationships between childhood experiences and adult disorders. When researchers find such associations, they need not be etiological in any way. As discussed in the previous section, we need to know whether risk factors are consistently, strongly, and specifically associated with outcome and to establish a plausible mechanism before we can feel confident that any such relationship is truly causal.

Moreover, associations between risk and outcome may result from latent variables not measured in the study. For example, people who drink too much are more likely to develop lung cancer. But alcohol does not cause lung cancer. The explanation is that people who drink more also smoke more, so that tobacco is the latent variable leading to disease.

To consider an example closer to the experience of psychotherapists, children who are sexually abused are more likely to develop a wide range of psychological problems as adults (Browne &Finkelhor, 1986). Yet abuse, by itself, need not be the primary or only cause of these outcomes. Children who are traumatized by abuse also tend to come from dysfunctional families. This latent variable is a much more powerful cause of later difficulties than child abuse itself (Beitchman et al., 1992; Nash et al., 1993; Rind & Tromofitch, 1997; Rind, Tromofitch, & Bauserman, 1998).

In summary, childhood adversities are multiple and intercorrelated. As a result, researchers are likely to draw faulty conclusions unless they carry out multivariate studies, in which they measure more than one risk. These methods provide a way of assessing a wide range of potential factors, rather than being confined to one or two. Yet, no matter how many additional variables are added, important ones can still be missed. This is why scientific papers so often end with a call for further research!

Clinical and Community Populations

Therapists do not see a random sample of the population. Therefore, observations on patients cannot be generalized to nonclinical groups. Clinical impressions often lead to incorrect conclusions about cause and effect. Practitioners may find associations between adversities and disorders in the people they see, but these relationships may not apply to community populations, since patients are, almost by definition, more vulnerable to adversity.

A project conducted at my own univ...

Table of contents

• Cover Page
• Half-Title Page
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Introduction
• Part I Childhood and Adulthood: The Evidence
• Part II Childhood and Adulthood: Myths
• Part III Childhood and Mental Disorders
• Part IV Childhood and Treatment
• Part V Implications
• References
• Index