The preliminary findings of the largest on-going study examining children at risk for bipolar disorder. Can an early symptom profile be detected? The clinical features of hyper-alertness and oversensitivity as early predictors of a bipolar disorder.
Latest News:
- Juvenile Bipolar Research Foundation Genetic Study
- Conferences in Colorado Springs and Ashville, North Carolina featuring Demitri Papolos, M.D.
In the spring issue of 2001, we wrote a newsletter entitled What Can the Amish Teach Us About Early-Onset Bipolar Disorder? It focused on a study of adult members of the Old Order Amish in Lancaster, Pennsylvania who suffered with bipolar I disorder. At the end of the newsletter we informed our readers that the principal authors of the study, Dr. Janice Egeland and her colleagues, were conducting a long-term study with 210 children (many who had a parent with bipolar I disorder). One of the goals of the research was to determine if an early-symptom profile could be detected before the onset of illness in the at-risk youngsters. We promised we would interview Dr. Egeland again and report on her findings.
In July of 2003, the preliminary results of the largest, most methodologically rigorous prospective study of children at risk for bipolar disorder was published in the Journal of the American Academy of Child & Adolescent Psychiatry. It is entitled “Prospective Study of Prodromal Features for Bipolarity in Well Amish Children,” and we think that this study is such a benchmark and that it archives such a treasure trove of information that we wanted all of our readers to know about it.
Why the Amish Are a “Living Laboratory” For Research In Bipolar Disorder
In many ways the Amish community provides a natural laboratory for all genetic and clinical and phenomenological research. It is a well-defined, closed population with little migration into or out of the community. The community can trace its ancestry back to 30 progenitors in Switzerland, and it maintains extensive genealogic records. The Amish community encourages a high birth rate, so a researcher can study large families. It is also important that this community prohibits the use of alcohol and drugs—substances known to complicate prenatal health as well as diagnostic ascertainment and assessment.
Finally, while the Amish have no more bipolar disorder than any other population group, they have always viewed bipolar disorder as a medical condition (“Siss im blut”—it’s in the blood, as they say), and they seek medical care for what they view as medical illnesses.
Dr. Egeland has been conducting genetic and epidemiological studies among the Old Order Amish since 1976 and has had a long-standing, trusting relationship with this community. Her earlier studies have identified families with a high loading of bipolar I disorder under genetic linkage study, and she can now look at the fourth and fifth generation of children as she has known many of their parents since the parents were babies themselves.
The Study Design and the Hypothesis
The Child and Adolescent Research Evaluation (CARE) program of the Amish study was initiated in 1994 and was designed to follow a group of 210 children and adolescents in two samples: a bipolar I sample and a control sample. The bipolar I sample were the children of a parent who was known to have bipolar I disorder. The control sample consisted of children who had a well parent whose sibling had the disorder, and a group of children with a family history negative for any psychiatric illness.
The hypothesis of the study was that there would be a gradation of risk for bipolar disorder: with children of one parent with the illness having the highest risk, followed by children whose parents were well but had a sibling diagnosed with the disorder (nieces and nephews), and by children in families with no history of the illness having the lowest rating of risk.
If this hypothesis were correct, the goal would be not only to gauge the genetic risk factors, but to identify the temperamental features and behaviors that might be predictors of an eventual manifestation of the illness.
At the time of recruitment in 1994, 14 candidate bipolar I families (8 fathers and 6 mothers with the illness) were invited and agreed to participate. A matched control group was assembled with children of same-sex psychiatrically-unaffected parents who had a sibling with the illness. Because it was not possible to obtain a sibling control for all the families, the parent with bipolar disorder was matched by sex, age, and family size to an unrelated Amish man or woman with a family history that was negative for psychiatric illness.
The final sample of 210 children consisted of 100 children from 14 families where a mother or father had bipolar I disorder; 77 children from 9 control families where the parents had a sibling who had bipolar disorder; and 33 children from four control families with a history negative for the illness.
How Was the Information About the Children Collected?
In order to launch the study that would follow children and adolescents over a twelve- to fifteen-year period, Dr. Egeland and her colleagues and a group of child psychiatrists, child development scientists, a pediatrician, and Amish advisors developed a formal schedule of questions that became known as the CARE Interview. This interview covered medical and developmental histories (Part A), a health narrative (Part B), and a third questionnaire with 69 inquiries related to a wide range of symptoms and life events (Part C). Part C’s questions were considered comprehensive enough to reveal potential early or prodromal features of bipolar illness.
The parents were asked whether their child was “noticeably different” from “other boys and girls” his or her own age. In the Amish community there are such well-defined roles for children, with specific chores expected at various ages that role performance (and any possible impairment) can be detected quickly.
Amish children have no homework after school. They go right home, go upstairs and change their clothes and go out and do their chores. Therefore, if these chores are completed in an inconsistent or spotty fashion, the parents realize that something is wrong. This is important because “role impairment” (functioning) is an element in psychiatric assessment. Hence, parental rating of chores gives a measure of “wellness” for each child annually.
How the Children Were Evaluated for Possible Risk
After a mother answered Part A of the CARE Interview, both parents answered Parts B and C, and the narrative file for each child was presented randomly to the CARE panel. This panel was composed of two board-certified child psychiatrists, a board-certified general psychiatrist, and a clinical psychologist.
All members of the panel were totally blind to the children’s identity or family history.
The panel members independently coded CARE narratives in sets of 10 children and recorded their clinical opinion for risk of developing a bipolar disorder. The options the doctors had to code these well children included high risk, moderate risk, low risk (these codes indicated the highest risk ratings); well with a BP tag; or well with no evidence of risk.
The“ BP tagged” risk category was used for children who were well, but who were manifesting some clinical features that suggested a possible onset of bipolar disorder in years to come, but who did not at this time warrant a risk rating. Risk rating represented a “clinical judgment” based on the substantial clinical experience of the panel, and there had to be consensus about each rating.
Which Group of Symptoms Occurred Most Frequently in Which Group of Children?
When all the data were assembled and the statistical analyses performed, the children of a bipolar parent were reported to have manifested more clinical features on the coding sheet than the control group, and the children of well parents who were siblings of a bipolar I patient manifested more clinical features than the children with no family history of psychiatric illness.
To add in the statistics: When rated for risk, 38% of the bipolar sample (compared to only 17% of the control sample) had high-moderate-low -ratings. Yet the vast majority of children in the control sample who had risk ratings, turned out to be the children who had an aunt or uncle with the illness (83%).
Similar to other studies of genetic risk, the children at highest risk had a parent with the illness. Children with a second-degree relative had a reduced risk but this risk was still higher than the risk for those who came from families negative for any psychiatric disorder.
Because we don’t want anyone to misinterpret the statistics, it is important to point out that the 38% figure is not the genetic risk factor to a child of a parent with bipolar. These Amish children are well with certain symptoms/features, and–depending on how many of the children onset with the illness–these symptoms/features may be suggestive of an early symptom profile for bipolar disorder.
The genetic risk to a child of one parent who has the illness is usually pegged between 20 and 30%. However, no one knows which factors may forestall an illness from developing and which genes might even be protective.
Which Clinical Features or Symptoms/Behaviors Did the At-risk Children Have?
The children who had a parent with bipolar I disorder had a statistically significant higher frequency for 10 clinical features when compared to the control group. Listed alphabetically they are:
- Anxious/worried
- Attention poor/distractible in school
- Energy low
- Excited
- Hyper-alert
- Mood changes/labile
- School role impairment
- Sensitivity
- Somatic complaints
- Stubborn/determined
It is interesting that the temperamental features of sensitivity, hyper-alertness, being anxious/worried or nervous appeared to be continuous as the parents responded to these with remarks such as “always” or “by nature.” However, half or more of the reports about decreased or increased energy and mood were episodic and all but one report on anger/temper showed as periodic rather than continuous.
This differs dramatically from the ultra-ultra rapid-cycling pattern of mood, energy, irritability and temper problems reported in so many non-Amish children, and raises questions about environmental influences on the presentation of symptoms and course of illness.
The Temperamental Features of Being Hyper-alert Or Overly Sensitive
Seventy percent of the children at risk for bipolar disorder had parental reports mentioning how “hyper-alert” and “overly sensitive” the children were. In the retrospective study that Dr. Egeland reported on two years ago, one quarter of the adults with bipolar I disorder had hospital records that noted “overly sensitive compared to others” prior to onset. (That figure may have been higher, but these were chart reviews of first hospitalizations and the symptom profile was not probed systematically upon admission.)
Parents and teachers in the Amish community who identify a child as “overly sensitive” refer to a child who has a heightened sense of awareness. If one observes such children, their “social skin” appears to be overexposed. They may seem “hyper-alert” to the feelings of others—peers and adults alike. It is as though an electrical field surrounds these youngsters and their antennae pick up all possible signals.
According to Dr. Egeland, “They seem to notice everything: how someone is dressed, whether their shoes are shined…they get very close to you and seem to need some physical contact. If another child gets stung by a bee, this child will feel so deeply that she will cry for the injured child.”
It has long been known that people in a manic state are hyper-alert, hyper-vigilant, and hyper-sensitive. According to the authors of the Amish findings, these features of being “overly sensitive” and “hyper-alert” could be early predictors of bipolarity.
Dr. Egeland then mentioned something that struck us when she added: “These children feel things very intensely and they are oversensitive to color.”
Parents who participated in the original survey for The Bipolar Child also mentioned this overall sensitivity; and one area of particular sensitivity was to color. One mother described her young daughter as “very sensitive in the visual realm. She is drawn like a magnet to some designs and colors, beautiful paintings, landscapes, and repelled by others, as strongly as she reacts to odors and tastes.”
When one looks at the art of Peter Paul Rubens, Vincent Van Gogh, Maurice Utrillo, Edvard Munch, and Jackson Pollack (all of whom suffered with manic-depression) it is easy to see this important sensitivity to color.
The Symptoms and Cycling Patterns Of The At-Risk Amish Children In Contrast To Non-Amish Children
More severe symptoms and symptoms of mania tend to manifest later in Amish children—most likely in adolescence. In this population, symptoms were showing up in the prepubertal years, going underground, and reemerging in adolescence. Also of interest, the Amish children of a bipolar I parent were not at a higher risk for patterns of disruptive behaviors, oppositional behaviors, or the hyperactivity so often seen in prepubertal children diagnosed with the disorder in communities outside the Amish culture. The authors write:
It is interesting that in our prospective study, clinical features such as mood, increased and decreased energy, decreased sleep, and anger/temper were noted to occur periodically in 50% or more of the reports for children of a bipolar parent. Other studies have suggested that the most frequent pattern of prodromal symptoms of bipolar disorder is characterized by continuous and chronic manifestations of irritability, mood dysregulation, and rapid cycling with little inter-episode relief.
What accounts for these differences in presentation is not known, but it is interesting to speculate whether the absence of alcoholism within the Amish community may differentially influence the presentation of the illness in comparison with non-Amish families.
A Case Vignette
To demonstrate a possible prodromal syndrome of one of the youngsters in the CARE Program, take a look at Rebecca’s story:
Rebecca was born at home after a long labor and was a fussy, colicky baby. Her infancy, growth and development were normal. At age 6, she was reported to be bold, stubborn, overly attentive, and slower to respond to discipline than her siblings. She was still “hyper-alert” and sensitive when 10- to 11-years old and tended to want to know “everything about everybody.” Becky was said to “worry like an adult about grown up” things. By age 12, she had outgrown these traits and was a good student. Other than headaches and feeling faint, nothing noteworthy was reported for her early teens (ages 13 to14).
Three of the panel rated Becky as well. One child psychiatrist rated her for low bipolar risk: “It is difficult for me to rate her as outside the limits of ‘wellness’ but the combination of symptoms noted make me wonder; they seem to form a ‘mini-cluster.’” After group discussion, the consensus was recorded as “well, with a BP tag.”
In the fall season after her 15th birthday and our CARE update, Rebecca suffered a sudden “breakdown” which lasted three months and required treatment. The updated information was read and independently rated by the clinicians. Becky was moody and sad; had significant weight loss, insomnia, loss of energy and interest, and self reproach; could not focus and think clearly; and was fearful about dying and delusional about death. The CARE panel agreed that she met standard diagnostic criteria for a major depressive disorder. The clinicians upgraded the risk rating from the original “well, with a BP tag,” to a “moderate” risk rating for bipolar problems in the future. This decision was made blind to pedigree identification and the fact that Rebecca’s father is a BPI patient.
Relevant to this pattern was a 3 to4-year prospective study of predictors for those most likely to shift from unipolar to bipolar disorder. Predictors, with high specificity for bipolar, included a depressive cluster with rapid onset of symptoms, a family loading for mood disorder and mood-congruent psychotic features. Becky fits this prediction.
Only time will tell whether or not Becky will convert to the bipolar I form of the disorder, however, we should keep in mind that Dr. Barbara Geller of Washington University in St. Louis reported on a ten-year follow-up of a group of 72 children who were originally diagnosed with depression before puberty (the average age at diagnosis was 10). By the age of 20, nearly half–48%–had developed the bipolar form of the disorder. Since children with ADHD were excluded from Dr. Geller’s study, it is possible that these rates of switching would have been even higher.
It bears saying that antidepressants should be used with utmost caution in children presenting with a syndrome that seems to present so squarely in the depressive spectrum. All efforts toward unearthing family history and any possible symptoms that may be temperamentally in the bipolar spectrum should be examined repeatedly before consideration of antidepressant treatment.
What Next?
According to Dr. Egeland and the other authors of this article, the research in the CARE study now rests on the ultimate outcome of a bipolar disorder diagnosis for a well child correctly designated “at risk.” The researchers plan to follow the children for 12 to15 years and will be reporting new findings in the literature throughout that period of time. A new article is expected sometime next spring.
In the meantime, genetic markers in one or more chromosomal regions for susceptibility gene(s) have been established in adults with bipolar disorder in the Amish community, and the researchers are collecting DNA from a number of the children in the CARE program. As this program is the only prospective study with the goal of comparing clinical prediction and genetic patterns for bipolar disorder, future reports from the Amish study will no doubt do much to expand our knowledge of the genetics, the early symptoms, and the course of childhood-onset bipolar disorder.
In Conclusion
Amish children live in a completely pacifistic society where anger or violence are never displayed, and where they are expected to be well-behaved, submissive to authority, quiet and non-intrusive around adults. Their opinions are never asked for or expressed. These children have never seen television, the nightly news or scary or gory movies, and they have never played Nintendo. They use no electricity and tend to go to sleep soon after nightfall and arise early with the sun—their sleep patterns are extremely uniform. They also have many brothers and sisters who act as role models, and are surrounded by cousins and peers who follow the traditions of the community closely and thus provide an additional abundance of role models. The social structure that surrounds children in this community is practically impossible to duplicate.
And yet, Amish children who have early symptoms of a possibly evolving illness cannot always conform to the expectations of their culture—anymore than can children suffering with these symptoms in the world outside.
We’ll write again soon. In the meantime, as always, we look forward to hearing from you. May the holidays find harmony within your homes.
Janice Papolos and Demitri Papolos, M.D.
BIBLIOGRAPHY
Biederman, J., Mick, E., Faraone, S. “Pediatric mania: a developmental subtype of bipolar disorder?” Biological Psychiatry 2002; 48:458-466.
Egeland, Janice A., Jon A Shaw, Jean Endicott, et al. “Prospective study of prodromal features for bipolarity in well Amish children.” Journal of the American Academy of Child & Adolescent Psychiatry 2003; 42:786-796.
Egeland, JA, Hostetter AM, et al. “Prodromal symptoms before onset of manic- depressive disorder suggested by first hospital admission histories.” Journal of the American Academy of Child & Adolescent Psychiatry 2000; 39:1245-1252.
Egeland, J. and Allen, C. Telephone interviews of November 4 and 5, 2003.
Egeland, J. Email correspondence of November 4, 2003.
Geller, B., Williams, M. “Prepubertal and early adolescent bipolarity differentiate from ADHD by manic symptoms, grandiose delusions, ultra-rapid or ultradian cycling.” Journal of Affective Disorders 1998; 51:81-91.
Geller, B., Zimmerman, B., et al. “Bipolar disorder at prospective follow-up of adults who had prepubertal major depressive disorder.” American Journal of Psychiatry 2001;158:125-127.
Hershman, JD. and Lieb, J. Manic Depression and Creativity. Amherst, NY: Prometheus Books, 1998.
Papolos, J. and D. “What can the Amish teach us about early–onset bipolar disorder? “ The Bipolar Child Newsletter January 2001, Vol. 6.
http://bipolarchild.com/newsletters/0101.html
Papolos, D. and J. The Bipolar Child, Revised. New York: Broadway Books, 2002.
Strober, M. and Carlson GA. “Bipolar illness in adolescents with major depression: clinical, genetic and psychopharmacologic predictors in a three-four-year prospective follow-up investigation. Archives of General Psychiatry 1982; 39: 549-555
The authors wish to thank Dr. Janice Egeland and Mrs. Cleona Allen for their pioneering work in this field and for their much-appreciated contributions to this newsletter, and a thank you as well to Catherine Schwartz, Karen Williams, and Jeanne Langer.