Evolving notions of childhood chronic illness.

present an analysis of 3 cohorts of children spanning 1988 to 2006 included in the National Longitudinal Survey of Youth (NLSY). The authors report that the prevalence of several categories of chronic illness in childhood is increasing and that these conditions arise, continue, or resolve in a highly dynamic fashion. Both findings have important implications and raise a number of significant questions. Themorethandoubling inprevalenceofchronicconditions reported by the authors is consistent with an increasing body of evidence documenting a historic shift in the epidemiology of child health—from acute to chronic illnesses—that began at least 50 years ago. Indeed, while mortality rates, hospitalizations forcommonacuteconditionsand injuries, andschool absence days due to illness were declining, the prevalence of chronicconditions severeenough tocause some levelof activity limitation more than doubled between the 1960s and the 1980s. Results from the report by Van Cleave et al suggest the upward trend in prevalence of chronic conditions has continued throughthe1990sand into the firstdecadeof thiscentury. Making sense of these findings is not an easy task. Access to and improvements in the delivery of specialized care have resulted in a growing number of survivors of complex congenital disorders, prematurity, and cancer whose survival was almost unimaginable a few decades ago. Better access to care and better tools to diagnose chronic conditions, particularly emotional and behavioral conditions, may also contribute to the upward trend in prevalence. Concepts of health and disease and definitions of what constitutes a childhood chronic illness have also changed. The narrow focus on physical defects during the “crippled children” era of the 1930s through 1960s evolved to include developmental disabilities and other physical conditions in the 1970s through 1980s, with the terminology shifting to “handicapped children.” In 1998, the federal Maternal and Child Health Bureau reset the parameters and repositioned the delivery system infrastructure for childrenwithchronic illnessbypromulgatingan expansivedefinitionofchildrenwithspecialhealthcareneeds that incorporatedphysical,developmental, emotional, andbehavioral conditions. When these broader concepts and definitions were operationalized in new national surveys of children’s health, no longer were 5% to 8% of children designated as chronically ill and disabled, but 16% to 18% were now classifiedashavingspecialneedsduetochronicconditions. While some have claimed that diagnostic or conceptual creep is at play, itseemsmorelikelythatthesenewdefinitionsareresponding to the changing realities of childhood and to the changing epidemiology of childhood chronic disease. Increasing rates of obesity appear to be driving the overall trend in prevalence described by Van Cleave et al, but significantincreaseswerenotedforallphysicalhealthconditions,with lesser increases for asthma and fairly minimal changes in the reported rates of learning and behavioral problems. Although theobesity trendand its antecedentshavebeenwelldescribed, the increasingrateofasthmaandotherphysical ailments is less wellunderstood.It issomewhatsurprisingthatthelearningand behavioral problem category did not show a greater increase inprevalence,givendatafromothercross-sectionalsurveyssuggesting attention-deficit/hyperactivity disorder (ADHD) and other mental disorders are also increasing. The authors also report large dynamic changes in how conditionsmanifestandresolve.Are thesedynamicchanges inonset,duration,andresolutionreal?TheNLSYidentifiedchronic conditionsbyquestioning thechild’sparentwhether thechild had an activity-limiting physical, mental, or emotional health condition that requiredregularhealthcare.Motherswere then asked to identify the condition and provide an estimate of its duration.Conditionswererecordedverbatimandcodedbythe interviewer. It is easy to suspect the reliability of the short sequence of questions and the ability of parents to report consistently over time. Consequently, some of the within-cohort dynamics are likely artifacts of the inherent inconsistencies in parental reports from one reporting interval to the next. Even thoughsomeof the findingsmaybedueat least inpart to measurement error, there are sound reasons from a developmental perspective to expect that conditions in childhood should be dynamic. Other studies have documented developmental fluctuations inconditions likeasthma,whichcanhave an episodic and relapsing course and can be influenced by the onset of puberty or stressful transitions or changes in family environments. As theauthorssuggest, less severeconditions,