The present study explored concepts of illness held by children with intellectual disability (ID) and compared them with the beliefs of two comparison groups of typically developing children who were matched for cognitive developmental level and chronological age (ns = 11). Content of responses to questions regarding illness causation, symptoms, treatment, and prevention was analyzed; children with ID provided significantly different responses than did children without ID. A positive association between cognitive development and prevention concepts was found for children with ID. Children with ID did not differ from developmentally matched children in sophistication of illness concepts. Thus, it appears important to assess cognitive developmental level of children with ID to provide appropriate health education and services.
Children with intellectual disability (ID) have a higher incidence rate of health disorders than do children without ID (van Schrojenstein Lantman-de Valk et al., 1997; Waldman, Perlman, & Swerdloff, 2001; Yousef, 1993) and are very likely to have at least one additional health-related condition other than intellectual disability (McQueen, Spence, Garner, Pereira, & Winsor, 1987). The illnesses experienced by children with ID range from chronic disorders requiring ongoing care and significant treatment to avoidable diseases that may be prevented or remedied through healthy decisions about activities and lifestyle. Despite this, researchers have found that a low percentage of children with ID participate in health-promoting activities, such as taking their own temperature, exercising regularly, choosing or preparing healthy meals, or using a safety belt, (Bechtel & Schreck, 2003; Steele, 1986), and they access health care providers less often than typically developing children (Horwitz, Kerker, Owens, & Zigler, 2001; Krauss, Gulley, Sciegaj, & Wells, 2003). The disparities among illness prevalence, treatment, and prevention have been noted; the Surgeon General's Conference on Health Disparities and Mental Retardation (Office of the Surgeon General, 2002) highlighted the need for increased knowledge and understanding of health-related topics by individuals with IDs. However, to date, few studies have investigated the concepts of illness held by children with ID. Therefore, the purpose of the present study was to explore the concepts of illness beliefs held by children with ID and compare them with beliefs of typically developing children.
Children's Concepts of Illness
Children's concepts of illness include beliefs about the causes, symptoms, treatments, and prevention of physical illness and form a distinctly different construct than children's knowledge of health facts. This is a notion central to earlier work (Inagaki & Hatano, 1993; McMenamy, Perrin, & Wiser, 2005), which found that children may have explicit knowledge of health processes or information (e.g., “You should blow your nose if you have a cold”) without a conceptualization of illness causation, symptoms, treatment, or prevention. However, much of the previous literature has explored these constructs concurrently. For example, Noland, Riggs, and Hall (1985) found that adolescents with ID have limited knowledge of health behaviors despite 71% of them having previously taken and 80% currently taking a health education course at school. Children with ID scored significantly lower on the health knowledge questionnaire than did the children with learning disabilities in areas including physical fitness, disease, and physical illness.
Several studies have supported the theory that children have a qualitatively different understanding of illness than adults and that children's understanding of illness develops in a sequential, hierarchical manner as they age (for reviews, see Burbach & Peterson, 1986; Tinsley, 2003), paralleling Piaget's cognitive developmental stages (Bibace & Walsh, 1980; Perrin & Gerrity, 1981). In general, studies of children's concepts of illness using Piagetian tasks have found that children in the preoperational stage of cognitive development have confused and often superstitious ideas about the causes of health and illness. As children develop into the concrete operational stage, their beliefs gain some sophistication, and concepts of contamination and internalizing infectious agents are understood to have an effect on the body. At the formal operational stage of cognitive development, children's understanding of illness reaches the greatest level of complexity. Children have an understanding that interactions between the environment and our internal structures, as well as our psychology, develop or maintain illnesses within the body.
To our knowledge, only two studies (Binnie & Williams, 2002; March, 1991) have investigated concepts of illness causation in individuals with ID, of which only one involved children with ID. Aggregation of the results from these studies shows that individuals with ID have an unsophisticated conceptualization of illness causality. Binnie and Williams compared the concepts-of-illness causation among typically developing, preschool-aged children (age range = 4.1–5.0 years), children with Down syndrome (age range = 9.0–17.5 years), and children with autism (age range = 6.11–12.00 years). Children were read a scenario in which a character was sneezed on by someone ill. Following this, children were asked if the first character would (a) definitely get sick—a partially correct response, (b) maybe get sick—a fully correct response, or (c) definitely not get sick—an incorrect response. All children, regardless of group, endorsed each response equally, indicating random guessing rather than an indication of understanding illness causality.
March (1991) investigated concepts of illness causality with adults ranging in age from 23 to 51 years who were diagnosed with mild to moderate ID and living in the community. Approximately 27% of the participants were not able to communicate any ideas about illness causation, 21% stated that illness was caused by external phenomenon (i.e., sun or God), close to 31% reported that illness was caused by people or objects within close proximity but by “magical” or unknown means, and approximately 19% indicated that other people or things caused illness. Only a single participant provided a response indicating an understanding of a need to internalize illness and that it harms the body. Interestingly, most participants knew about their medication names and uses despite their lack of a sophisticated understanding of illness causation.
To address the paucity of literature exploring the concepts of illness held by children with ID, in the current study, we used previous theoretical models describing how typically developing children understand causation, symptom recognition, treatment, and prevention of illness. Specifically, this study investigated three questions: (a) What types of responses about the cause, symptoms, treatment, and prevention of illness do children with ID provide? (b) What are the relations among cognitive developmental level, age, and concepts of illness for children with ID? and (c) Do children with ID have similar concepts of illness as typically developing children?
Participants included 33 children (11 children with ID and 22 children without ID). Children's ethnicities included Caucasian (n = 22), Latino/a (n = 6), African American (n = 1), and “other” (n = 4). Children with ID (7 boys, 4 girls) had been previously diagnosed with mild or moderate ID through independent diagnosticians and ranged in age from 12 to 17 years old (M = 14.55, SD = 2.02). These children ranged from the preoperational stage of cognitive development to concrete operational, with most falling in the preoperational stage. Children without ID ranged in age from 6 to 18 years old and were divided into two comparison groups in a matched-groups design. The first group matched the children with ID on cognitive developmental level and gender (referred to as the “CDL group”). Therefore, children in the CDL group also ranged from the preoperational to the concrete operational stages of cognitive development. These children were 6 to 11 years old (M = 7.27, SD = 1.42). The second group was paired with children with ID by chronological age and gender (referred to as the “CA group”); thus, the age distribution of children in the CA group was identical to that of children with ID (M = 14.55, SD = 2.02). Cognitive developmental level for this group ranged from the preoperational to formal operational stages, with most falling in the concrete operational stage. This study used a matched-group design to ensure similarity between children with ID and typically developing comparison groups. Matching children by CDL and CA allows investigation of the possible influences of children's concepts of illness, namely cognitive development (comparison with CDL group) or experience with illness (comparison with CA group).
Because children without ID were not administered IQ testing for participation in the previous study (Malcarne, Drahota, & Hamilton, 2005), we decided to measure children's cognitive developmental level to allow for comparison between the present and previous study investigating children's concepts of illness. Piagetian measures of cognitive development and psychometric measures of intelligence have been closely related in children without ID (Pasnak, Willson-Quayle, & Whitten, 1998) and have overlapped with standard intelligence tests (Humphreys & Parsons, 1979) or related to intelligence in notable ways (Stephens, McLaughlin, Miller, & Glass, 1972). Individuals with ID progress through the Piagetian stages of reasoning in the conventional order, albeit more slowly and reaching a lower ceiling (Weisz & Yeates, 1981; Weisz & Zigler, 1979).
Cognitive developmental level
Cognitive developmental levels for all children were evaluated using standardized Piagetian methods assessing children's understanding of conservation, interrelationship among parts, physical causality, and abstract thinking. Perrin and Gerrity (1981) established standardized questions and coding schematics, yielding a single score for each task. An overall score for each participant's cognitive developmental level was computed by averaging each of the individual scores. Perrin and Gerrity found interrater reliabilities ranging from .90 to .98. Interrater reliability in this study was .85 for each of the cognitive developmental tasks.
Concepts of physical illness
A semistructured interview, adapted from Perrin and Gerrity (1981), was used to assess children's understanding of physical illnesses. Children were asked several questions to assess their beliefs regarding illness causality (e.g., “What makes someone get sick?”), symptom recognition (e.g., “How does someone know when they are sick?”), illness treatment (e.g., “When someone gets sick, what can they do to get better?”), and prevention (e.g., “What can someone do to keep from getting sick?”). Additional prompts were used to clarify children's responses (e.g., “Can you tell me more about that?”). Interviewers questioned children until they said, “I don't know” to ensure that they provided complete, detailed answers.
Children's responses were categorized according to content (Malcarne, 1993) for each of the four illness concepts. The purpose of this procedure was to provide descriptive information about the types of answers provided by children in each of the three comparison groups. The categories of responses fell within 1 or more of 11 content codes for each question, ranging from biological (e.g., cough) to “don't know” or off-task. Interrater reliability was .85 for each content code.
Level of understanding
Children's responses were scored on a scale ranging from 1 (no response, inappropriate, or off-task answer) to 6, with higher scores representing more sophisticated levels of cognitive understanding. A score of 2 indicated responses that were circular and phenomenistic, or relied on external sources for information about illness, including treatment or prevention. A score of 3 indicated concrete, specific, isolated causes, symptoms, treatments, or preventions of illness but without knowledge of what is done in or to the body. A score of 4 indicated a response with emerging relativism and perspectivism; that is, the child knew that many different illnesses or that specific symptoms, treatment, or prevention act on the body to “make it work right.” A score of 5 indicated a child's understanding of a relationship between the causal, symptom, treatment, or preventive agent and the body's response. Last, a score of 6 indicated that the child understood the process or mechanism underlying the cause of illness and that this was different from the usual condition or how an agent treated or prevented illness in the body. For coding of all illness-concept questions, interrater reliabilities were .85.
Children without ID
Children without ID (e.g., the CDL and CA groups) were recruited from an archival data set measuring parents' and children's concepts of physical and psychological illnesses. Families were contacted through letters sent home with children enrolled in regular education classrooms at participating schools; interested parents contacted research staff. Children were interviewed during after school hours, typically at the child's home, and received $5.00 for participating. An accurate count of the number of parents who received forms describing the study and inviting participation is not possible for a variety of reasons, but primarily because there is no way of knowing how many children reliably transferred the forms from the classroom to their parents. Based on the ratio between the number of forms distributed and collected, we estimated that 25% of parents chose to participate, a sample consisting of 260 children. Given the relatively low response rate and the inability to compare those who chose to participate with those who did not, generalizing these findings should be done so cautiously.
Children with ID
Children with ID were recruited through community agencies serving children and adolescents with developmental disabilities, such as Special Olympics, child summer programs through the local Arc, and inclusion programs through Boys and Girls Clubs. Agencies distributed flyers to parents of children with ID. Interested parents contacted researchers to schedule interviews, which took place either in the children's homes or at the community agency. To reduce participant fatigue, interviews were conducted over several sessions. Children received two movie tickets for their participation. An accurate count of the number of parents who received forms describing the study and inviting participation is not possible because agencies did not disclose the number of children with ID in their programs or how many flyers were distributed to families.
For both children with and without ID, Institutional Review Boards for the affiliated university and community agencies or school districts approved the study protocol. Written informed consent was obtained from the parent or legal guardian, and written assent was obtained from each participating child. Because informed consent was obtained and data were collected only from children who chose to participate, there is no way to compare participants with children who declined participation.
What Types of Responses About Illness Cause, Symptoms, Treatment, and Prevention Did Children With ID Provide?
Content codes were examined using chi-square analysis to determine if significant differences in the frequency of responses provided by the three comparison groups could be found. When assessing the content of responses provided by the separate groups, noteworthy differences emerged (see Table 1).
Cause and symptom recognition categories
Over one third of children with ID provided “I don't know” or off-task responses when asked what caused illness. In contrast, none of the children without ID provided this response, χ2(2, N = 33) = 9.10, p = .01. The prevalence of providing responses that environmental or weather factors cause illness also significantly differed among the groups, χ2(2, N = 33) = 6.60, p < .05. Although approximately 73% of CA children and close to 50% of CDL children stated that environment or weather caused illness, fewer than 20% of children with ID said this.
For symptom recognition, all of the CA children and close to 75% of children with ID cited at least one specific physical symptom indicating illness, whereas just over 50% of the CDL children did, χ2(2, N = 33) = 6.27, p < .05. In addition, CA children most frequently cited a “general feeling state” as indication of illness compared with the other groups, whereas almost 30% of the CDL children and approximately 10% of the children with ID cited this category, χ2(2, N = 33) = 10.21, p = .01.
Treatment and prevention categories
For all of the groups of children, seeking professional help or treatment for illnesses was cited most frequently. Yet, a trend toward a significant difference in the rate of response was found. Although all of the CA children and 75% of the CDL children acknowledged professional help as a treatment for being sick, only 64% of the children with ID did so, χ2(2, N = 33) = 4.71, p < .10.
Last, clear differences in response about illness prevention were found. Close to 50% of the CDL children stated that they did not know or provided an off-task response to the question about illness prevention, whereas less than 20% of children with ID did and no CA children did, χ2(2, N = 33) = 6.89, p < .05. Furthermore, almost 30% of CA children reported that avoiding the intake of negative externals, such as poison or drugs, would prevent illness, whereas no CDL children or children with ID provided this response, χ2(2, N = 33) = 6.60, p < .05.
What Is the Relation Among Cognitive Developmental Level, Age, and Concepts of Illness for Children With ID?
We calculated Pearson correlation coefficients among the four illness concepts, age, and cognitive developmental level. For the total sample, a significant positive correlation was found between cognitive developmental level and age (r = .42, p < .05) as well as each of the four illness concepts: illness causation (r = .61, p < .01), symptom recognition (r = .54, p < .01), treatment (r = .62, p < .01), and prevention (r = .71, p < .01). However, when the three comparison groups were assessed individually, these correlations were not uniformly found.
For children with ID, the only significant correlation found was between cognitive developmental level and illness prevention (r = .71, p ≤ .01), indicating that, as children's level of cognitive development increases, their understanding of how to prevent illnesses is also more sophisticated. This association was also found for the CA children (r = .78, p < .01). Although not significant, for the CDL children, the correlation between cognitive developmental level and illness prevention as well as treatment approached significance (r = .53, p = .09, and r = .55, p = .08, respectively). For CDL children, a significant positive correlation was found between cognitive development and symptom recognition (r = .65, p < .05). These findings contrast with previous results (Bibace & Walsh, 1980; Perrin & Gerrity, 1981), which indicated that as children became more mature in their reasoning abilities, their overall understanding of illness also became more sophisticated. One reason for this may be due to the small sample sizes included in this study.
Do Children With ID Have Similar Concepts of Illness as Typically Developing Children?
To examine group differences, we conducted a multivariate analysis of variance (MANOVA), with group as the independent variable and concepts of illness (causation, symptom recognition, treatment, and prevention) as dependent variables (see Table 2 for means and standard deviations).
A significant effect was found for group, F(8, 56) = 2.94, p < .01 (see Table 3 for MANOVA). Tests of between-subjects effects was significant for illness causation, F(2, 30) = 9.36, p = .001; symptom recognition, F(2, 30) = 8.54, p = .001; treatment, F(2, 30) = 7.61, p < .01; and prevention, F(2, 30) = 5.73, p < .01. Post hoc analyses indicated that the children with ID and children in the CDL group did not significantly differ for any of the four illness concepts (cause: p = .48, symptom recognition: p = .72, treatment: p = .24, prevention: p = .70); however, significant differences were found between CA children and the two other groups—children with ID and CDL: Illness causation, ps < .001 and .01, respectively; symptom recognition, p < .01, and p = .001, respectively; treatment, p = .001, and p < .05, respectively; and prevention, ps = .01, respectively.
What Types of Responses About Illness Cause, Symptoms, Treatment, and Prevention Did Children With ID Provide?
Generally, the type and rate of responses provided by children with ID did not significantly differ from either the CDL or CA groups. When examining illness causation responses, children with ID provided responses that were related to their internal body (e.g., germs, biology, injury, and taking in negative externals) or responses related to things being done to their body (e.g., inappropriate self-care or getting wet due to rain). They did not provide responses about external situations, outside of their immediate bodies, such as the social environment or emotional/psychological states causing illness. Of note, only children with ID stated that they did not know what caused illness; children from each of the other two groups always provided at least a single response about the cause of illness. When examining the responses from questions regarding symptom recognition, we noted that the responses were varied. Children with ID responded that their internal body, such as general feeling or emotional states, specific physical symptoms, or concrete clues (e.g., their behavior or appearance), indicated when they were ill. Although they also stated that external sources, such as caregivers and doctors, told them when they were ill, it appeared that children with ID were most likely to state that they knew when they were ill based on their awareness of their internal conditions or states.
We examined treatment responses provided by children with ID and noted that responses were related to others (e.g., seeking and receiving help from a professional or friends and family) or themselves (e.g., resting or becoming aware of the need for or making a change within themselves). For illness-prevention categories, children with ID provided responses that included their own preventative care, such as caring for themselves; avoiding infections or people who are sick; and avoiding poison, drugs, or alcohol. Responses also included others acting as preventative agents, such as doctors or friends and family, as well as external agents, such as avoiding rain or other unfavorable weather conditions. However, children with ID also provided responses that indicated that they may not have had control over illness prevention but stating that there was nothing that could be done sometimes or that they did not know what could be done.
Investigation of the rate of responses for each category revealed significant differences between children with and without ID. Children with ID provided significantly more “don't know” and off-task responses about illness causation than children without ID. This is an important distinction from the results found in the literature involving typically developing children in which illness causality has been found to develop before other concepts of illness (cf. Burbach & Peterson, 1986). Furthermore, a trend toward significance was found for treatment, with less children with ID stating that seeking professional help was treatment for being sick than did children without ID. Therefore, we believe that ID or the health education system provided to children with ID affects their understanding of illness causation and treatment.
In addition, for two responses, we found a different order in the prevalence of responses. Children matched for age cited weather or environmental factors as causes of illness, followed by children matched for cognitive development, and, last, children with ID. This order was also found for children who noted existence of a “general feeling state” as recognition of being ill. Recognizing that general, yet abnormal, feeling states signal illness is highly abstract; this result suggests that children with lower levels of cognitive understanding are unable to associate general somatic states (e.g., “just feeling bad”) as evidence of illness.
What Are the Relations Among Cognitive Developmental Level, Age, and Concepts of Illness for Children With ID?
We found that, for the total sample, cognitive developmental level was positively related to age and each of the four concepts of illness (illness causation, symptom recognition, treatment, and prevention). This finding is consistent with previous research reported by Bibace and Walsh (1988) and Perrin and Gerrity (1989), who found that children's cognitive developmental levels were related to beliefs about illness. However, when the groups were assessed individually, we did not find a positive relationship between cognitive development and age, just as we did not find a uniform association between cognitive developmental level and each of the four concepts of illness. For children with ID, increases in cognitive maturity were associated with the ability to better understand illness-prevention strategies. One hypothesis for the lack of significance between cognitive development and illness understanding may be the specific nature of health education programs currently used in special education classes (Jobling, 2001). Despite cognitive ability, children with ID are taught detailed information about symptoms of specific illnesses (e.g., influenza) and their treatments, rather than a more general understanding of illness. Thus, children with ID may have provided more sophisticated answers to questions regarding illness prevention yet were unable to conceptualize overall illness including causation, symptom recognition, and treatment in a sophisticated manner.
Do Children With ID Have Concepts of Illness That Are Similar to Typically Developing Children?
Children with ID and CDL children did not differ on any of the four illness concepts. For illness causation, responses provided by children in the two groups indicated that the highest stage of their understanding was the contamination stage; they were able to provide concrete, specific, and isolated causes of illness but were unable to recognize how these causes make a person ill. For symptom recognition, the children provided circular or phenomenistic responses stating that a person knows they are sick because other people tell them. Children in both groups were able to provide specific, isolated treatments or agents of treatment but could not understand how those treatments helped an individual get better. Last, children with ID and CDL children did not differ in their understanding of illness prevention and provided circular or phenomenistic responses when explaining how an individual could prevent illness from occurring, such as preventing illness by not having these problems; or they referred to noncausal phenomena as though they were causes of illness.
The importance of understanding concepts-of-illness beliefs in children with ID is threefold. First, understanding the development of children's concepts of illness for children with ID may allow for improved communication between health professionals and children. The most frequently identified barrier to individuals with ID obtaining sufficient health care is communication (Minihan, Dean, & Lyons, 1993). Medical personnel are not universally nor comprehensively trained to provide services to people with ID (Nieman & Markello, 1986) and lack an understanding of the limited cognitive and communicative abilities possessed by these patients (Strauss & Kastner, 1996). Thus, knowing that children with ID can cite fewer causes of illness or understand concepts of illness at their cognitive developmental level rather than age will assist health professionals in communicating and intervening more effectively with children with ID. In addition, developing efficacious health education materials for children with ID requires knowledge of their concepts of illness causation, symptoms, treatment, and prevention. At this time, adults with ID are likely to live independently in the community without intensive support (Office of the Surgeon General, 2002); however, they are poorly prepared for the responsibilities involved with understanding the causes, symptoms, treatments, and prevention of illness and health problems (Jobling, 2001) or independently accessing medical services (Krauss et al., 2003; Office of the Surgeon General, 2002). Developing and implementing use of health education materials that are specifically designed for children with ID are necessary to effect change in the prevalence of illness throughout these individual's lifespan. Last, after understanding the general development of concepts of illness of children with ID, individual differences within the population can be explored in a scientific manner. For example, health locus of control has been demonstrated to relate to children's health behavior and outcomes in the typically developing childhood population (Eiser, Eiser, Gammage, & Morgan, 1989). However, it is unknown if these links exist in the ID population. Only after researchers understand the concepts of illness in children with ID, will we be able to explore individual differences in an informed and empirical manner.
Although this study generated provocative results, it is important to note some limitations. Because of the small sample size, power to detect statistically significant differences was limited. A larger, more diverse sample of children with ID and matched comparison groups may have produced additional and more robust findings. In addition, including a measurement of IQ in future research regarding concepts of illness in children with ID would provide better groups for comparison. Last, we encountered substantial difficulty during the recruitment phase, potentially affecting the generalizability of these findings. Literature addressing methods of involving individuals with and without ID in research (Lennox et al., 2005; Mazzocco & Myers, 2002) has indicated that lack of accessibility to potential participants, an increase in burden to families, uncertainty regarding the benefit to families and communities, and funding constraints significantly impact recruitment opportunities and success. Future research would benefit from involving important components for successful recruitment, including making personal contact with agency personnel or community leaders and securing agreements with key personnel to facilitate access to potential participants by actively distributing recruitment information (Harrington et al., 1997; Lennox et al., 2005).
Amy Drahota, MA (email@example.com), Doctoral Student, University of California, Los Angeles, Graduate School of Education, Los Angeles, CA 90095-1521. Vanessa L. Malcarne, PhD, Professor San Diego State University, Department of Psychology, San Diego, CA