Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Journal of Clinical Child & Adolescent Psychology

ISSN: 1537-4416 (Print) 1537-4424 (Online) Journal homepage: https://www.tandfonline.com/loi/hcap20

Early Detection and Prevention of Mental Health


Problems: Developmental Epidemiology and
Systems of Support

E. Jane Costello

To cite this article: E. Jane Costello (2016) Early Detection and Prevention of Mental Health
Problems: Developmental Epidemiology and Systems of Support, Journal of Clinical Child &
Adolescent Psychology, 45:6, 710-717, DOI: 10.1080/15374416.2016.1236728

To link to this article: https://doi.org/10.1080/15374416.2016.1236728

Published with License by Taylor & Francis


Group© E. Jane Costello

Published online: 18 Nov 2016.

Submit your article to this journal

Article views: 18548

View related articles

View Crossmark data

Citing articles: 20 View citing articles

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=hcap20
Journal of Clinical Child & Adolescent Psychology, 45(6), 710–717, 2016
Published with License by Taylor & Francis Group
ISSN: 1537-4416 print/1537-4424 online
DOI: 10.1080/15374416.2016.1236728

Early Detection and Prevention of Mental Health


Problems: Developmental Epidemiology and Systems
of Support
E. Jane Costello
Department of Psychiatry and Behavioral Sciences, Duke University

This article reviews the role of developmental epidemiology in the prevention of child and
adolescent mental disorders and the implications for systems of support. The article distinguishes
between universal or primary prevention, which operates at the level of the whole community to
limit risk exposure before the onset of symptoms, and secondary or targeted prevention, which
operates by identifying those at high risk of developing a disorder. It discusses different aspects of
time as it relates to risk for onset of disease, such as age at first exposure, duration of exposure, age
at onset of first symptoms, and time until treatment. The study compares universal and targeted
prevention, describing the systems needed to support each, and their unintended consequences.

In this article we discuss how epidemiology, the study of TYPES OF PREVENTION


patterns of illness in time and space, can help to prevent
mental health problems by preventing either their occur- What should we be trying to prevent? Of course we want to
rence or their causes. Preventing something is usually easier prevent disease, but disease has many precursors, consisting
if we can reliably detect the problem, or its cause, or both. both of early symptoms and of risk factors that increase the
We tend to assume, probably with reason, that it is better to likelihood of disease. Public health has tried to create a man-
detect a problem or its cause earlier rather than later in the ageable taxonomy of risk by classifying exposures according
course of the disorder and that that systems of support will to the level of response best designed to prevent either risk or
be strengthened if detection occurs early. A case can be disease (Gordon, 1983; Mrazek & Haggerty, 1994; Rothman
made against this argument for early intervention (think of & Greenland, 1998). At the most extreme level (“indicated”
all those tonsillectomies and umbilical hernia operations prevention) we enforce treatment, or even lock people away, to
later shown to be a complete waste of resources). But until prevent their harming others or themselves. Isolating or quar-
proven otherwise, epidemiology is ethically bound to oper- antining cases for the safety of the community has been used
ate on the assumption that early detection is a service that it for centuries to limit the spread of infectious diseases; psychia-
can and should provide to the health care system. try is one of the few areas of chronic disease medicine that still
has recourse to this type of prevention.
“Targeted” prevention projects are designed for individuals
who have symptoms of a disorder or a clearly recognized
risk factor (e.g., parental bereavement as a risk factor for depres-
sion; Sandler et al., 2010). “Universal” or “primary” prevention
describes public health activities designed to reduce risk for the
© E. Jane Costello
This is an Open Access article. Non-commercial re-use, distribution, and whole population In this article we present the case that most
reproduction in any medium, provided the original work is properly attributed, mental health problems are prevented by aspects of the way we
cited, and is not altered, transformed, or built upon in any way, is permitted. live that have little to do with identification, treatment, or
The moral rights of the named author(s) have been asserted. “targeted prevention” at the level of existing symptoms.
Correspondence should be addressed to E. Jane Costello, Department of
Epidemiology can be used to identify ordinary everyday aspects
Psychiatry and Behavioral Sciences, Suite, 22, 905 W Main St, Durham NC
27701. E-mail: jcostell@psych.duhs.duke.edu of life, like having two parents or being born full-term, which
Color versions of one or more of the figures in the article can be found we don’t think of as mental health interventions but which can
online at www.tandfonline.com/hcap. be shown to be protective. Sometimes, however, the good-
MENTAL HEALTH PROBLEM DETECTION AND PREVENTION 711

enough system of daily life fails. At that point, epidemiology cannot, for example, randomly assign newborn infants to
can be used to identify who needs care, of what kind, and to nurturing versus nonnurturing mothers. To chart a course
monitor the adequacy and effectiveness of that care. Taken between true causes and confounders, epidemiologists often
together, these functions of epidemiology help to answer ques- have to rely on observational methods or at best quasi-
tions about the extent to which systems of support for the basic experiments (Cook & Shadish, 1994).
aspects of life are cost-effective for preventing mental illness,
compared with the more focused systems of care that we bring
into play when things go wrong. Many of the articles in this TYPES OF PREVENTION FOR CHRONIC DISEASE
special issue take as their focus various emotional and beha-
vioral problems for which specialized systems of care have been In line with the well-established sequence of prevention levels,
devised. Here, the concern is rather how many of those pro- the sequence of interventions runs from health promotion,
blems might never have developed in the first place given a solid through universal prevention, to selective (high-risk) preven-
groundwork of support. It turns out that not only do many kinds tion, to indicated prevention (Weisz, Sandler, Durlak, & Anton,
of primary prevention avoid the costs of screening and the pain 2005). Different writers have slightly different emphases, but
of stigma but they often prevent a wide range of health and the general principles are similar. Briefly, health promotion
developmental problems, in addition to mental illness, and so enables people to increase control over their health and its
can be highly cost-effective. determinants (World Health Organization, 2005); primary or
universal prevention aims to enable people to escape disease
by either avoiding exposure to risk factors or being able to resist
EPIDEMIOLOGY AND PREVENTION it after exposure; secondary or targeted prevention aims to
minimize harm from disease exposure by preventing full-
As its name implies, epidemiology grew out of the struggle to blown disease; and tertiary prevention aims to restore function-
prevent the epidemics of infectious diseases that swept coun- ing and reduce disease-related complications, relapse, or spread
tries and continents well into the first half of the 20th century. to others. A key difference among them is that different levels of
The job for infectious disease epidemiology is to discover prevention have different target populations. Health promotion
patterns of disease in time and space that could explain why and universal prevention focus on the entire population, either
some people become sick and others do not. Understanding throughout the life course (e.g., food safety regulations) or at a
patterns of transmission enables systematic interventions to particular developmental stage (e.g., safe car seats for infants).
break the chain—from stagnant water to malaria, or unpasteur- Targeted prevention has at its focus those identified as at risk for
ized milk to tuberculosis. The century up to World War II was a a disease, either because of symptoms (e.g., depression identi-
time of many famous victories against infectious diseases. fied by screening) or because of some kind of risk exposure
The challenge for the past half century, however, has been (e.g., parental bereavement). Indicated prevention or “Treatment
how to prevent not infectious or communicable diseases so as Prevention” as advocated for the prevention of HIV/AIDS
much as chronic or episodic or noncommunicable diseases, (http://www.who.int/hiv/pub/mtct/programmatic_update_tasp/)
mental illnesses among them. In the case of infectious diseases looks to clinical treatment both to reduce relapse rates and to
we now generally know the germs that cause them and can prevent the spread of the disease to others (e.g., antiretroviral
employ a fairly simple two-pronged attack: to exterminate the treatment for HIV/AIDS). This review concentrates on the pros
germs and to eliminate the environments in which they breed. and cons of the more common universal versus targeted pre-
So the old “systems of care” for infectious diseases—the vention strategies, although we need to remember that psychia-
plague hospitals and tuberculosis infirmaries—have dwindled try still use tertiary “treatment as prevention” in the form of
to a small fraction of the health care system, and funding goes incarceration and compulsory hospitalization of children and
almost entirely to early universal prevention via inoculation adolescents who might harm themselves or others.
against these diseases. Much of the cost of primary prevention
is borne by nonmedical services like the Environmental
Preventing Symptoms or Preventing Exposure
Protection Agency, the schools, and the local water boards.
At present, the picture for mental health is very different. As noted in the discussion of infectious disease epidemiology,
We are still at the stage of discovering the “patterns of once a causal pathway to disease has been identified, preventive
disease in time and space” that may help to identify major strategies may focus on exposure to risk, or on dealing with the
risk exposures. On the whole, the “one bug–one drug” first symptoms of disease (or both). The former (universal
model that defeated many infectious diseases does not fit prevention) covers everyone in the population, whereas the
mental illnesses (syphilis, caused by the spirochete latter (targeted prevention) is restricted to high-risk or potential
Treponema pallidum, is no longer treated as a psychiatric cases. Arguments about the cost-effectiveness of each type of
disorder). This has the corollary that experimental methods prevention strategy have been made for decades (Mrazek &
such as case-control designs for identifying causal factors Haggerty, 1994). It is likely that the balance depends on the type
rarely work either. Even if we have a causal hypothesis we of illness and the population in question (Eaton et al., 2002). For
712 COSTELLO

example, evidence that low birth weight is a strong predictor of the stigma associated with being picked out for “anger
female adolescent depression (Bohnert & Breslau, 2008; management class” or “fat camp.” For tertiary prevention,
Costello, Worthman, Erkknali, & Angold, 2007; but see someone—parent or teacher usually—has to identify a pro-
Vasiliadis et al., 2008) supports the case for universal preven- blem or symptom and seek help, which brings its own
tion, whereas prevention of depression following bereavement problems. For example, in a study with more than 20,000
makes more sense only after bereavement has occurred (Haine, assessments of children made by the children themselves
Ayers, Sandler, & Wolchik, 2008). and a parent (Costello et al., 1996), when a child reported
Infectious disease epidemiology has mainly concentrated one of nine symptoms of depression, the parent reported the
on universal approaches to prevention: for example, ensur- same symptom between 0% and 29% of the time. This
ing that children are inoculated; cleaning up water, food, means that at best 71% of child-reported symptoms were
and air; and setting safety standards for cars and toys. Even missed; in the case of cognitive problems, psychomotor
the chronic diseases that are the main causes of mortality agitation, and anhedonia, the parent never identified cases
today are being approached using universal prevention reported by the child. Given that even the most highly
methods: exercise, diet, smoking reduction. developed instruments to identify child and adolescent
What are the universal prevention programs that would psychopathology, whether screening questionnaires or
prevent mental illness? Interesting to note, the work to evaluate interviews, are of only moderate test–retest and interrater
these has hardly begun (Durlak, Weissberg, Dymnicki, Taylor, reliability (Angold et al., 2012), the probability that the
& Schellinger, 2011; Koenen et al., 2009). Lists of exposures children in need of care will be correctly identified is low.
that might respond to universal interventions have been pro- Unfortunately, primary care physicians have an even worse
posed (Eaton et al., 2002), but the evidence for them is based record for case-identification (Dulcan et al., 1990).
on common sense more than on research, and there are few If one is making a case for primary or universal preven-
tests of the relative efficacy of universal versus targeted ver- tion, one has to address the question of systems of support.
sions of the same or similar interventions. Several studies have It is a waste of effort, as well as arguably unethical, to
suggested that, just as environmental effects on psychopathol- identify cases for which no treatment is available. The
ogy may have a stronger effect at the most stressed end of the standard systems of care for children (medical and psychia-
distribution (Lee, Kosterman, McCarty, Hill, & Hawkins, tric specialty services, educational, juvenile justice, and
2012), an intervention to relieve a particular stress may also social services; Burns et al., 1995)) cannot provide income
be most effective at that level (Costello, Compton, Keeler, & supplements or a healthy diet, although they can certainly
Angold, 2003). For example, an income supplement led to do much to reduce low birth weight and improve facilities
lower levels of behavioral problems in children whose families for exercise. But, perhaps more important, primary preven-
were moved out of poverty by the supplement but had no effect tion of this kind can help to bring different branches of
on the (already low) levels of behavioral problems in those medicine and service providers together to share resources
who were never poor (Costello et al., 2003). The types of and lobby together for more. For example, reducing the
exposure that might respond to primary prevention include proportion of low birth weight children also lightens the
poor diet, lack of physical exercise (Koenen et al., 2009), burden of special needs children on educational services. On
poverty (Lee et al., 2012), and low birth weight (Bohnert & the other hand, it can often happen that an intervention puts
Breslau, 2008; Sonuga-Barke & Halperin, 2010). All of these demands on one part of the system (e.g., improving antena-
exposures are also associated with a range of other chronic tal care) but brings benefits only later to another part of the
diseases. This raises questions such as, Does a particular pre- system (e.g., reducing special education service needs). So,
ventive intervention prevent more than one disorder? How institutional pressures can work against otherwise rational
does this affect the cost–benefit calculations for the interven- reallocation of resources.
tion? In the example of the income supplement described
earlier, subsequent studies showed that supplementing family
income reduced substance use and abuse several years later, RISK, EXPOSURE, AND THE MEANING OF TIME
when the children had left home (Costello, Erkanli, Copeland,
& Angold, 2010). It also increased the chance of completing Many questions about early detection and prevention can be
high school, reduced minor crime (Akee, Costello, Copeland, answered only by methods that take into account temporal
Keeler, & Angold, 2010), and slowed the rate of obesity characteristics of risk factors, including age at onset and the
problems (Akee, Simeonova, Copeland, Angold, & Costello, “dose” or level of exposure over time. To be most effective,
2013). Thus, a single intervention applied to a whole popula- systems of support need to be sensitive both to children’s
tion had a wide range of positive effects. developmental needs and to the developmental course of
An advantage of universal prevention is precisely that it risk exposure.
is universal; there is no need to screen or select and there- Age at first exposure, time since first exposure, duration
fore there are no screening costs, as there often are for of exposure, and intensity of exposure are all interrelated
secondary prevention. Another advantage is that it avoids aspects of timing that may have different implications for
MENTAL HEALTH PROBLEM DETECTION AND PREVENTION 713

prevention. In addition, every attempt to prevent illness is starting school. Once this period of risk is over, the chance
an implicit or explicit test of a causal hypothesis, and the of developing enuresis is very slight. In this case, age at
timing of preventive interventions adds another level of exposure is clearly the critical developmental risk factor,
causal questioning. The kinds of questions we are thinking because it is very rare for a parallel increase in functional
of include the following: enuresis to occur at later times of stress, such as moving to
middle or high school, and there is no delay between the
● Does physical abuse by parental figures cause psychia- stress and the symptoms.
tric disorders in children? Is a single blow a sufficient Timing of exposure has rarely been treated separately in
cause or does abuse have to go on for a period of time, studies of child psychopathology. Brown and Harris (1978),
or happen at a certain level of severity, before it con- in their work on the social origins of depression, argued that
stitutes a risk factor? Are children of different ages or women who lost their mother in the first decade of life were
developmental stages differentially vulnerable to phy- more vulnerable as adults to depressive episodes in the face
sical abuse as a risk factor? What risks are associated of severe life events. However, theirs was a retrospective
with removing children of different ages from home study that did not address the question of whether these
because of physical abuse? women were also at greater risk of depressive episodes dur-
● Why are depressive disorders rare in both prepubertal girls ing later childhood and adolescence. It is not clear whether
and boys but much more common in postpubertal girls? the crucial factor was the length of time since exposure to the
What causes the observed sex difference to develop? Is it risk factor of the mother’s death, or the age of the child at the
associated with hormonal, morphological, or social time of exposure, or some combination of the two.
changes occurring around puberty? Why is earlier-than- Timing of puberty has emerged as an important aspect of
average maturation apparently a positive event for risk in relation to both depression and behavioral problems.
boys but a negative one, associated with increased risk of In a longitudinal study that measured not only age at
behavioral and school problems, for girls? menarche but also morphological development, Tanner sta-
ging (Marshall & Tanner, 1969), and levels of gonadal and
The answers to such questions imply different assump- steroidal hormones, it was clear that it was high levels of
tions about the causal role of timing, intensity, and duration. estrogen and testosterone, not timing of puberty, that pre-
Here we offer a few examples of the impact on psychiatric dicted adolescent depression (Angold, Costello, Erkanli, &
disorder of different aspects of development: (a) age at first Worthman, 1999). However, there are many studies show-
exposure, (b) time since first exposure, (c) duration of ing that girls who are early in developing the morphological
exposure, and (d) intensity of exposure. We then consider signs of puberty, indexed by Tanner stage or menarche, are
the implications for support systems. at risk for behavioral problems, especially if they have
The importance of age at first exposure has been studied unsupportive families (Costello, Sung, Worthman, &
most intensively of all the aspects of risk over time in child Angold, 2007; Ge, Brody, Conger, & Murry, 2002; Ge,
psychopathology because of the theoretical importance Conger, & Elder, 1996; Magnusson, Stattin, & Allen, 1985).
attached to early experiences in the Freudian and other Sometimes duration of exposure may be a more powerful
psychodynamic models of development. For example, predictor of later harm than timing. For example, Copeland,
researchers investigating the role of attachment in children’s Wolke, Angold, and Costello (2013) found that the number of
development have concentrated on the very early months years over which children reported being bullied was a stronger
and years of life as the crucial period during which the predictor of poor adult outcomes than the timing of the experi-
inability to form one or more such relationships may have ence (e.g., childhood, adolescence, or both.) In a longitudinal
damaging effects that last into childhood and perhaps even study from New Zealand, Moffitt (1990) found that children
into adulthood (Sroufe, 1988). The critical date of onset of identified at age 13 as both delinquent and hyperactive had
risk appears to occur after 6 months, but the duration of the experienced significantly more family adversity (poverty, poor
risk period is not yet clear. Hay (1985) presented evidence maternal education and mental health), consistently from the
that maternal depression, which presumably interferes with age of 7, than children who were only delinquent or only
mothers’ ability to form normal relationships with their hyperactive at age 13. The most striking increase in the
infants, affects motor development if it occurs during the antisocial behavior of ADD+delinquent boys diagnosed at age
1st year of life, and language development but not motor 13 occurred between the ages of 5 and 7, when they attained a
development if it occurs during the 2nd year of life. This is mean antisocial rating that was not reached by other delinquent
a case where age at first exposure appears to interact with boys until 6 years later. School entry and reading failure coin-
the developmental processes most salient at a particular age. cided temporally with this exacerbation of antisocial behavior.
In another example of the importance of timing, Rutter These data suggest that the problem behavior of this group,
(1985) pointed out that once children have achieved urinary despite being generally persistent, is responsive to experience.
continence at around age 2, there is a period of risk for The data also reveal a key point of vulnerability that could be a
relapse into incontinence that appears to coincide with target for intervention: reading readiness.
714 COSTELLO

Another example comes from the Great Smoky Figure 1 shows the mean age at onset of a range of
Mountains Study. Children who had been assessed over an disorders, using repeated assessments up to age 21 in a
8-year period were classified into four groups on the basis of representative population sample of children (National
their body mass index (a ratio of weight to height) at each Institute of Medicine, 2009). The mean age at onset of the
assessment: no obesity (72.8%), childhood-only obesity first symptom, in children who would eventually get a
(5.1%), adolescent-only obesity (7.5%), and chronic obesity diagnosis of that disorder, preceded the full diagnosis by
(14.8%). Only the chronically obese group was at increased about two years. Sometimes the first symptom and full
risk of psychiatric disorder (Mustillo et al., 2003). This is an disorder occurred almost simultaneously (attention deficit/
example of duration of exposure as the key risk character- hyperactivity disorder, anxiety disorders), whereas in other
istic. It is also an example of the impact of intensity of disorders (oppositional defiant disorder, conduct disorder,
exposure because no effects were found of overweight that depression) there was a time gap of about three years.
fell just below the threshold of obesity. This implies that for many psychiatric disorders there is
Intensity of exposure to lead (Needleman & Bellinger, time to identify early symptoms and intervene.
1991) provides an example of a definite dose–response It is also the case that many children who experience one
relationship. Needleman and Bellinger (1991) divided den- or two symptoms never develop a full disorder. In the study
tine lead levels into six classes, from less than 5.1 parts per from which the figure comes, for example, there were 10
million to more than 27 parts per million, and showed a times as many observations including one or more symp-
highly significant effect of dose on teachers’ ratings of toms of depression as there were cases of depression
children’s functioning on a wide range of intellectual and (N = 3,924 with one or more symptoms; N = 278 with a
behavioral tasks. Another aspect of intensity is the number full diagnosis). But the mean age at onset of the first
of different risk factors to which a child is exposed (Seifer, symptom was almost two years earlier in the children who
Sameroff, Baldwin, & Balwin, 1989). Most children appear eventually developed a full diagnosis (12.5 vs. 14.4). This
to be able to cope with a single adverse circumstance, but suggests that really early intervention (e.g., when prepuber-
rates of psychopathology rise sharply in children exposed to tal children showed even one symptom of depression) could
several adverse circumstances or events (Seifer et al., 1989). identify the children most likely to go on to a full diagnosis.
However, Rutter (1985) and others have pointed out that
children exposed to one risk factor are at increased risk of
exposure to others (e.g., no father in the home and poverty)
and that the dose–response relationship to an increasing
number of different risk factors is not a simple linear one
These examples show that it is possible to design studies
that at least begin to allow us to tease out the respective roles
played by time since first exposure, age or developmental
stage at first exposure, duration of exposure, and intensity of
exposure. Multistage models of risk, which have been devel-
oped to address the complexities of causality in chronic
disease, are one way of putting the pieces together. Several
such models have been proposed, particularly in the context
of carcinogenesis (Peto, 1984), and have been reviewed in
terms of developmental psychopathology (Pickles, 1993).
Statistical techniques for exploring causality in such multi-
stage models have made great strides recently (Robins,
1997). The challenge is to incorporate all the various aspects
of risk into a single model and distinguish the ones that carry
the tune from those that are just noise.
FIGURE 1 Age at onset of first symptom and of full psychiatric disorder, by
age 21. Data from the Great Smoky Mountains Study. Note. ADHD = attention
EARLY DETECTION OF MENTAL HEALTH deficit/hyperactivity disorder; ODD = oppositional defiance disorder; CD =
PROBLEMS AND SYSTEMS OF SUPPORT conduct disorder. Reproduced with permission from the National Institute of
Medicine (2009). Preventing Mental, Emotional, and Behavioral Disorders
Is early detection in the form of detection of the first Among Young People: Progress and Possibilities. Committee on Prevention of
Mental Disorders and Substance Abuse Among Children, Youth, and Young
symptoms likely to be beneficial? Do our treatments work Adults: Research Advances and Promising Interventions. Washington, DC:
well when employed with symptomatic but subsyndromal The National Academies Press. © [Rightsholder]. Reproduced by permission
cases? Is a system of support that focuses on early interven- of National Academy of Medicine. Permission to reuse must be obtained from
tion a good use of resources? the rightsholder.
MENTAL HEALTH PROBLEM DETECTION AND PREVENTION 715

EARLY DETECTION, PREVENTION, AND THE Universal Prevention and the Service System
SERVICE SYSTEM
Primary or universal prevention programs generally need a
different set of service systems, most of which have not been
Universal and targeted prevention make very different
established, or exist in bits and pieces across the field of public
demands on those whose job it is to reduce the rate of
health. As with targeted prevention, it is helpful to have a clearly
onset of mental disorders. They also identify different
defined target, which means a taxonomy, such as those of the
kinds of people, with different skills, as the key interveners.
Diagnostic and Statistical Manual of Mental Disorders (5th ed.;
DSM–5; American Psychiatric Association) or International
Classification of Diseases. Second, some method is needed to
Targeted Prevention and the Service System
monitor the effect of a prevention program at the population
In the past two decades a whole new prevention industry has level. These have been set up in some countries, through orga-
emerged in the United States, complete with master’s, nizations such as the Centers for Disease Control and Prevention
doctoral, and postdoctoral training programs; professional in the United States, or using national case registers such as
societies and journals; and expert lobbyists at state and those in force in the Scandinavian countries. Evaluation is often
federal levels. This is understandable, because targeted pre- the weak point of universal prevention programs; for example,
vention is a complicated and expensive activity. First, it is the vast system of health visitors, operating for more than a
necessary to decide what disorders or disabilities are to be century in the United Kingdom with the goal of educating
prevented. This requires a taxonomy agreed to and shared mothers and protecting young children, has hardly been evalu-
across the prevention community. Second, prevention provi- ated at all (Laming, 2009). Another example is the provision of
ders need to work toward one or more recognized prevention specialty mental health and counseling services to school-age
programs or strategies, a process that requires controlled children. There is good evidence that these services reach more
trials, tests of generalizability from one site to another, and children than any other form of mental health care (Brener,
all the accoutrements of clinical research. In fact, targeted Martindale, & Weist, 2001; Burns et al., 1995), but we know
prevention may be even more complex than clinical treatment very little about their effectiveness.
because it requires the interveners to go out and find subjects In other areas of health it is very clear that universal
with whom to work, rather than waiting for them to walk prevention has been extraordinarily effective, for example,
through the clinic door. So, targeted prevention requires in lowering infant and child mortality and morbidity and
methods for identifying potential cases. People normally in improving healthy physical development. There is also indir-
contact with children in the community—parents and tea- ect evidence that mean intelligence levels have increased
chers most obviously—need to be trained and helped to dramatically in the past 100 years, at a far faster speed than
identify subjects for intervention, and if necessary “screen- could be attributable to evolutionary selection (Flynn, 1984).
ing” measures must be developed and tested. “Screening” can No such data are available for mental health. The second step
mean case identification, if the signs are clear enough, but for instituting universal systems of mental health care is
more often the term is used to mean picking possible cases clearly a national (and eventually international) surveillance
out of a general population, with the concomitant risk of false system that can monitor need for services and change in
positives and false negatives. Targeted prevention is often demand.
advocated as being less expensive to implement than It is likely that when and if such systems were to be set
universal prevention, but it is important to bear in mind that up, we should find that some of the universal health care
screening implies at least some level of involvement with programs already in place have a marked effect on chil-
everyone in the target population, which entails its own costs. dren’s mental health. Prenatal care and nutrition is an
Next, to be successful, targeted prevention mandates a obvious example, given the evidence for the damage that
service system that can provide effective interventions, can be caused by low birth weight (Costello, Worthman,
available on a scale that matches the need identified in the et al., 2007; Sommerfelt, Troland, Ellertsen, & Markestad,
screening phase. It is unethical to arouse the awareness of 1996; Szatmari, Saigal, Rosenbaum, Campbell, & King,
needs that cannot be met. 1990). Mandated screening for hearing and visual deficits
The processes involved in targeted prevention are listed here is another example. Programs that affect whole commu-
at some length because of the widely held assumption that nities, such as the removal of lead from the environment,
targeted intervention is more efficient and cost-effective than have effects on both behavioral and cognitive development
primary or universal intervention. This may be true, but targeted (Needleman & Bellinger, 1991).
prevention carries very considerable costs that will have to be It is nevertheless the case that, despite what universal
paid by someone. Also to be considered are the costs to those prevention has done to protect them, far too many children
identified as being “labelled,” correctly or incorrectly, and the have psychiatric disorders. For example, our 20-year pro-
potential damage to those incorrectly missed in the screening spective study of a community sample found that by age
stage. In sum, “targeting” is neither cost-free nor entirely benign. 21 more than 50% of the participants had experienced at
716 COSTELLO

least one DSM-IV (American Psychiatric Association, 1994) Comparative study. Journal of the American Academy of Child & Adolescent
psychiatric disorder, and a further 10%–15% had significant Psychiatry, 51(5), 506–517. doi:10.1016/j.jaac.2012.02.020
Bohnert, K. M., & Breslau, N. (2008). Stability of psychiatric outcomes of
functional impairment associated with psychiatric symptoms low birth weight: A longitudinal investigation. Archives of General
(Copeland, Shanahan, Costello, & Angold, 2011). Other Psychiatry, 65(9), 1080–1086. doi:10.1001/archpsyc.65.9.1080
longitudinal studies show similar results(Jaffee, Harrington, Brener, N. D., Martindale, J., & Weist, M. D. (2001). Mental health and
Cohen, & Moffitt, 2005). Barely one in four received any social services: Results from the school health policies and programs
specialty mental health care, and those who did waited 2 or study 2000. Journal of School Health, 71(7), 305–312. doi:10.1111/
josh.2001.71.issue-7
3 years for services (Burns et al., 1995; Costello, He, Brown, G. W., & Harris, T. O. (1978). The social origins of depression: A
Sampson, Kessler, & Merikangas, 2013). study of psychiatric disorder in women. New York, NY: Free Press.
Burns, B. J., Costello, E. J., Angold, A., Tweed, D., Stangl, D., Farmer, E. M.
Z., & Erkanli, A. (1995). Children’s mental health service use across
CONCLUSIONS service sectors. Health Affairs, 14, 147–159. doi:10.1377/hlthaff.14.3.147
Collins, P. Y., Patel, V., Joestl, S. S., March, D., Insel, T. R., Daar, A. S., …
Walport, M. (2011). Grand challenges in global mental health. Nature,
Returning to the title, “Early Detection and Prevention of 475(7354), 27–30. doi:10.1038/475027a
Mental Health Problems: The Role of Developmental Cook, T. D., & Shadish, W. R. (1994). Social experiments: Some developments
Epidemiology in Planning Systems of Support,” the role of over the past fifteen years. Annual Review of Psychology, 45, 545–580.
developmental epidemiology in the prevention of metal dis- doi:10.1146/annurev.ps.45.020194.002553
order has both optimistic and pessimistic sides. In the case of Copeland, W., Shanahan, L., Costello, E. J., & Angold, A. (2011).
Cumulative prevalence of psychiatric disorders by young adulthood: A
targeted prevention we can be optimistic because, as other prospective cohort analysis from the great smoky mountains study.
articles in this issue show, preventive interventions can be Journal of the American Academy of Child & Adolescent Psychiatry,
effective, and the systems needed to bring them to scale— 50(3), 252–261. doi:10.1016/j.jaac.2010.12.014
screening, training of field staff, destigmatization of mental Copeland, W. E., Wolke, D., Angold, A., & Costello, E. J. (2013).
illness—have been worked out at least in principle (of interest, Adult psychiatric outcomes of bullying and being bullied by peers in
childhood and adolescence. JAMA Psychiatry, 70(4), 419–426.
a lot of the work has been done in the context of global mental doi:10.1001/jamapsychiatry.2013.504
health; Collins et al., 2011; Patel, Flisher, Hetrick, & McGorry, Costello, E. J., Angold, A., Burns, B. J., Stangl, D. K., Tweed, D. L.,
2007). Pessimism is induced by the high cost of instrument and Erkanli, A., & Worthman, C. M. (1996). The Great Smoky Mountains
program development and by the likely service gap that will Study of Youth: Goals, designs, methods, and the prevalence of DSM-III-
yawn when screening reveals the true numbers of children in R disorders. Archives of General Psychiatry, 53, 1129–1136.
doi:10.1001/archpsyc.1996.01830120067012
need of clinical care. Costello, E. J., Compton, S. N., Keeler, G., & Angold, A. (2003).
In the case of universal prevention, we can be more opti- Relationships between poverty and psychopathology: A natural experi-
mistic that programs already in the field, and well accepted by ment. Journal of the American Medical Association, 290(15), 2023–
almost everyone, will be shown to yield mental health benefits 2029. doi:10.1001/jama.290.15.2023
that have not yet been evaluated. Pessimistically, we note the Costello, E., Erkanli, A., Copeland, W., & Angold, A. (2010). Association
of family income supplements in adolescence with development of
high rate of (largely untreated) mental disorder in the popula- psychiatric and substance use disorders in adulthood among an
tion and the lack of surveillance systems that can track the American Indian population. Journal of the American Medical
effects of existing or new programs. Association, 303, 1954–1960. doi:10.1001/jama.2010.621
Costello, E. J., He, J. P., Sampson, N. A., Kessler, R. C., & Merikangas, K.
R. (2013). Services for adolescents with psychiatric disorders: 12-Month
data from the National Comorbidity Survey-Adolescent. Psychiatric
REFERENCES Services, 65(3), 359–366.
Costello, E. J., Sung, M., Worthman, C., & Angold, A. (2007). Pubertal
Akee, R., Costello, E., Copeland, W., Keeler, G., & Angold, A. (2010). maturation and the development of alcohol use and abuse. Drug and
Parent’s incomes and children’s outcomes: A quasi-experiment with Alcohol Dependence, 88(Suppl. 1), SS50–SS59. doi:10.1016/j.
casinos on American Indian reservations. American Economic Journal, drugalcdep.2006.12.009
2, 86–115. Costello, E. J., Worthman, C., Erkanli, A., & Angold, A. (2007). Prediction
Akee, R., Simeonova, E., Copeland, W., Angold, A., & Costello, E. J. from low birth weight to female adolescent depression: A test of compet-
(2013). Young adult obesity and household income: Effects of uncondi- ing hypotheses. Archives of General Psychiatry, 64, 338–344.
tional cash transfers. American Economic Journal: Applied Economics, 5 doi:10.1001/archpsyc.64.3.338
(2), 1–28. Dulcan, M. K., Costello, E. J., Costello, A. J., Edelbrock, C., Brent, D., &
American Psychiatric Association. (1994). Diagnostic and statistical man- Janiszewski, S. (1990). The pediatrician as gatekeeper to mental health
ual of mental disorders (4th ed.). Washington, DC: Author. care for children: Do parents’ concerns open the gate? Journal of the
American Psychiatric Association. (2013). Diagnostic and statistical man- American Academy of Child & Adolescent Psychiatry, 29(3), 453–458.
ual of mental disorders (5th ed.). Washington, DC: Author. doi:10.1097/00004583-199005000-00018
Angold, A., Costello, E. J., Erkanli, A., & Worthman, C. M. (1999). Pubertal Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., &
changes in hormone levels and depression in girls. Psychological Schellinger, K. B. (2011). The impact of enhancing students’ social and
Medicine, 29(5), 1043–1053. doi:10.1017/S0033291799008946 emotional learning: A meta-analysis of school-based universal interven-
Angold, A., Erkanli, A., Copeland, W., Goodman, R., Fisher, P. W., & Costello, tions. Child Development, 82(1), 405–432. doi:10.1111/cdev.2011.82.
E. J. (2012). Psychiatric diagnostic interviews for children and adolescents: A issue-1
MENTAL HEALTH PROBLEM DETECTION AND PREVENTION 717

Eaton, W. W., Addington, A. M., Bass, J., Forman, V., Gilbert, S., Hayden, K., Needleman, H. L., & Bellinger, D. (1991). The health effects of low level
& Mielke, M. (2002). Risk factors for major mental disorders: a review of exposure to lead. Annual Review of Public Health, 12, 111–140.
the epidemiologic literature. Baltimore, MD: Department of Mental doi:10.1146/annurev.pu.12.050191.000551
Hygiene, Bloomberg School of Public Health, Johns Hopkins University. Patel, V., Flisher, A. J., Hetrick, S., & McGorry, P. (2007). Mental health of
Flynn, J. R. (1984). The mean IQ of Americans: Massive gains 1932 to 1978. young people: A global public-health challenge. The Lancet, 369, 1302–
Psychological Bulletin, 95(1), 29–51. doi:10.1037/0033-2909.95.1.29 1313. doi:10.1016/S0140-6736(07)60368-7
Ge, X., Brody, G., Conger, R., Simons, R. L., & Murry, V. M. (2002). Contextual Peto, J. (1984). Early- and late-stage carcinogenesis in mouse skin and in
amplification of pubertal transition effects on deviant peer affiliation and man. In M. Börzsönyi, N. E. Day, K. Lapis, & H. Yamasaki (Eds.),
externalizing behavior among African American children. Developmental Models, mechanisms and etiology of tumour promotion (IARCScientific
Psychology, 38(1), 42–54. doi:10.1037/0012-1649.38.1.42 Publications No. 56) (pp. 359–371). Lyon, France: International Agency
Ge, X., Conger, R. D., & Elder, G. H. (1996). Coming of age too early: for Research on Cancer.
Pubertal influences on girls’ vulnerability to psychological distress. Child Pickles, A. (1993). Stages, precursors and causes in development. In D. F.
Development, 67(6), 3386–3400. doi:10.2307/1131784 Hay & A. Angold (Eds.), Precursors and causes in development and
Gordon, R. S. (1983). An operational classification of disease prevention. psychopathology (pp. 23–49). Chichester, UK: Wiley and Sons.
Public Health Reports, 98, 107–109. Robins, J. (1997). Causal inference from complex longitudinal data. In M.
Haine, R. A., Ayers, T. S., Sandler, I. N., & Wolchik, S. A. (2008). Berkane (Ed.), Latent variable modeling and applications to causality.
Evidence-based practices for parentally bereaved children and their lecture notes in statistics (pp. 69–117). New York, NY: Springer Verlag.
families. Professional Psychology: Research and Practice, 39(2), 113– Rothman, K. J., & Greenland, S. (1998). Modern epidemiology (2nd ed.).
121. doi:10.1037/0735-7028.39.2.113 Philadelphia, PA: Lippincott-Raven.
Hay, D. F. (1985). Learning to form relationships in infancy: Parallel Rutter, M. (1985). Resilience in the face of adversity. Protective factors and
attainments with parents and peers. Developmental Review, 5, 122–161. resistance to psychiatric disorder. The British Journal of Psychiatry, 147,
doi:10.1016/0273-2297(85)90007-3 598–611. doi:10.1192/bjp.147.6.598
Jaffee, S., Harrington, H., Cohen, P., & Moffitt, T. E. (2005). Cumulative Sandler, I., Ayers, T. S., Tein, J.-Y., Wolchik, S., Millsap, R., Khoo, S. T.,
prevalence of psychiatric disorder in youths. Journal of the American … Coxe, S. (2010). Six-year follow-up of a preventive intervention for
Academy of Child & Adolescent Psychiatry, 44(5), 406–407. parentally bereaved youths: A randomized controlled trial. Archives of
doi:10.1097/01.chi.0000155317.38265.61 Pediatrics & Adolescent Medicine, 164(10), 907–914. doi:10.1001/
Koenen, K. C., Moffitt, T. E., Roberts, A. L., Martin, L. T., Kubzansky, L., archpediatrics.2010.173
Harrington, H., … Caspi, A. (2009). Childhood IQ and adult mental dis- Seifer, R., Sameroff, A. J., Baldwin, C. P., & Balwin, A. (1989, April). Risk
orders: A test of the cognitive reserve hypothesis. The American Journal of and protective factors between 4 and 13 years of age. Paper presented at
Psychiatry, 166(1), 50–57. doi:10.1176/appi.ajp.2008.08030343 the annual meeting of the Society for Research in Child Development,
Laming, H. B. (2009). The protection of children in England: A progress San Francisco, CA.
report (Vol. 330). London, UK: The Stationery Office. Sommerfelt, K., Troland, K., Ellertsen, B., & Markestad, T. (1996).
Lee, J. O., Kosterman, R., McCarty, C. A., Hill, K. G., & Hawkins, J. D. Behavioral problems in low-birthweight preschoolers. Developmental
(2012). Can patterns of alcohol use disorder in young adulthood help Medicine and Child Neurology, 38(10), 927–940. doi:10.1111/j.1469-
explain gender differences in depression? Comprehensive Psychiatry, 53 8749.1996.tb15049.x
(8), 1071–1077. doi:10.1016/j.comppsych.2012.03.012 Sonuga-Barke, E. J., & Halperin, J. M. (2010). Developmental
Magnusson, D., Stattin, H., & Allen, V. L. (1985). Biological maturation phenotypes and causal pathways in attention deficit/hyperactivity
and social development: A longitudinal study of some adjustment pro- disorder: Potential targets for early intervention? Journal of Child
cesses from mid- adolescence to adulthood. Journal of Youth and Psychology and Psychiatry, 51(4), 368–389. doi:10.1111/j.1469-
Adolescence, 14, 267–283. doi:10.1007/BF02089234 7610.2009.02195.x
Marshall, W. A., & Tanner, J. M. (1969). Variations in pattern of pubertal Sroufe, L. A. (1988). The role of infant–caregiver attachment in develop-
changes in girls. Archives of Disease in Childhood, 44, 291–303. ment. In J. Belsky & T. Nezworski (Eds.), Clinical implications of
doi:10.1136/adc.44.235.291 attachment (pp. 18–38). Hillsdale, NJ: Erlbaum.
Moffitt, T. E. (1990). Juvenile delinquency and attention deficit disorder: Szatmari, P., Saigal, S., Rosenbaum, P., Campbell, D., & King, S. (1990).
Boys' developmental trajectories from age 3 to age 15. Child Psychiatric disorders at five years among children with birthweights <
Development, 61(3), 893–910. 1000g: A regional perspective. Developmental Medicine and Child
Mrazek, P. J., & Haggerty, R. J. (Eds.). (1994). Reducing risks for mental Neurology, 32(11), 954–962.
disorders: frontiers for prevention research. Washington, DC: National Vasiliadis, H. M., Gilman, S., & Buka, S. (2008). Fetal growth restriction
Academy Press. and the development of major depression. Acta Psychiatrica
Mustillo, S., Worthman, C., Erkanli, A., Keeler, G., Angold, A., & Costello, Scandinavica, 117(4), 306–312.
E. J. (2003). Obesity and psychiatric disorder: Developmental trajec- Weisz, J. R., Sandler, I. N., Durlak, J. A., & Anton, B. S. (2005). Promoting
tories. Pediatrics, 111, 851–859. doi:10.1542/peds.111.4.851 and protecting youth mental health through evidence-based prevention
National Institute of Medicine (2009). Preventing Mental, Emotional, and and treatment. American Psychologist, 60(6), 628–648. doi:10.1037/
Behavioral Disorders Among Young People: Progress and Possibilities. 0003-066X.60.6.628
Committee on Prevention of Mental Disorders and Substance Abuse World Health Organization. (2005). The Bangkok Charter for health
Among Children, Youth, and Young Adults: Research Advances and promotion in a globalized world. Health Promotion Journal of
Promising Interventions. Washington, DC: The National Academies Press. Australia, 16(3), 168–171.

You might also like