Self-concept, attitude toward illness and family functioning in adolescents with type 1 diabetes

Abstract

OBJECTIVES:

The primary objective of the present study was to assess self-concept in adolescents with type 1 diabetes, and to determine whether this is associated with attitudes toward having chronic disease, family functioning or severity of diabetes. The secondary objective was to assess the impact of family income, sex, age and age at diagnosis on adolescent self-concept.

METHODS:

A cross-sectional, self-report survey of 48 adolescents with type 1 diabetes (22 boys and 26 girls; mean ± SD age at time of study 15.2±1.7 years [range 12.2 to 18.0 years]; mean age at diagnosis 9.2±3.3 years [range 1.3 to 14.9 years]) was performed using the Piers-Harris Children’s Self-Concept (PHCSC) scale, second edition; the Child Attitude Toward Illness Scale; and the Family Assessment Measure scale, version III. Demographic information including net family income and a symptom inventory form to assess disease severity was collected.

RESULTS:

Adolescents’ self-concept measured by the PHCSC scale was significantly positively correlated with a more positive attitude toward chronic illness as measured by the Child Attitude Toward Illness Scale. The PHCSC scale was found to have a significant negative correlation with the Family Assessment Measure scale score, indicating that a better self-concept was correlated with enhanced family functioning. Self-concept was not significantly correlated with disease severity, income of family, sex, age at diagnosis, age at time of study, episodes of diabetic ketoacidosis or episodes of hypoglycemia.

CONCLUSIONS:

Adolescents with better self-concept had more positive attitudes toward their chronic illness and enhanced family functioning. Although no correlation with diabetes disease severity was seen in the study population, interventions aimed at improving adolescent self-concept may have a positive impact on diabetes treatment by improving attitude toward living with type 1 diabetes.

Diabetes in childhood has significant effects on the entire family. Paediatric parenting stress in parents of children with type 1 diabetes has been found to increase in frequency with younger children and lower socioeconomic status (1). Better psychosocial well-being and fewer depressive symptoms have been associated with fewer diabetes-specific family conflicts in adolescents with type 1 diabetes (2). Family functioning has been shown to have an impact on metabolic control in adolescents with type 1 diabetes (3). Good metabolic control is associated with parental involvement in helping with insulin injections (4). In addition, glycemic control has been shown to have less deterioration over time in children from families that are more cohesive and have less conflict (5).

Adolescents, particularly younger ones, with type 1 diabetes have reported higher degrees of perceived parental control when compared with healthy adolescents (6). Adolescents who are more worried about their diabetes have reported that they are more depressed and have families that are less warm and caring (7). Families have reported that interactions are more structured, and less cohesive and stimulating when compared with families of healthy adolescents (8).

Self-concept refers to an adolescent’s perception of and feelings about themselves. It has been described as a composite view of oneself that is formed through direct experiences with others and evaluations from others (9). Self-concept has been examined in many chronic diseases. For example, self-concept is positively correlated with adherence to asthma treatment in children (9) and family functioning in adolescents with epilepsy (10). Self-concept appears to be associated with an adolescent’s glycemic control (11). Adolescents with improved control of their diabetes have been shown to have less anxiety and a more positive self-concept compared with adolescents with poorly controlled diabetes (11). In a study (12) of Turkish children 10 to 14 years of age with type 1 diabetes, self-esteem scores were significantly higher than those obtained from a control group of healthy school children, using the Coopersmith self-esteem inventory. The authors speculated that this may have been related to education given about their disease, increased support from family and friends, and participation in social programs.

Because self-concept appears to be an important aspect that is associated with positive diabetes control, identifying associations among adolescents’ self-concept, their attitudes toward illness and disease severity, as well as family functioning will add to current knowledge and may also help in optimizing diabetes treatment by providing evidence-based practice in psychosocial care and improving treatment adherence.

The primary objective of the present study was to assess adolescent self-concept in patients with type 1 diabetes, and to determine whether their self-concept is associated with an attitude toward having a chronic disease, family functioning or severity of diabetes. The secondary objective was to assess the impact of family income, sex, age and age at diagnosis on adolescent self-concept.

METHODS

Patients

A cross-sectional survey of adolescents followed at the Alberta Children’s Hospital (Calgary, Alberta) diabetes clinic was conducted. Patients were included if their age at the time of study was between 12 and 17 years, and if they had a diagnosis of type 1 diabetes for more than six months. Patients were excluded if they had a diagnosis of another chronic disease, handicap or developmental delay, or if there were any language barriers to understanding or completing the questionnaires. Patients were recruited from May 1, 2006, to July 15, 2006, through a letter of invitation, follow-up phone calls or directly being approached at their diabetes clinic visit.

Measurements

Adolescents were asked to complete the demographic information and symptom inventory with the assistance of their parent(s). The surveys used to assess self-concept (the Piers-Harris Children’s Self-Concept [PHCSC] scale, second edition), attitude toward illness (the Child Attitude Toward Illness Scale [CATIS]) and family functioning (the Family Assessment Measure, version III [FAM III] scale) were completed by the adolescent alone to avoid any potential bias from their parents. Questionnaires were completed immediately following a clinic visit with the patient’s usual endocrinologist, so the information was reflective of the most recent metabolic control, assessment of complications and laboratory investigations.

Demographic information

Patients were asked to complete a demographic information form, including their net family income level, and a symptom inventory. Demographic information was collected to be able to describe the study patient population and assess whether net family income affected any of the measures. The symptom inventory was designed to assess the impact that diabetes was having on the patient, and included episodes of diabetic ketoacidosis (DKA) not including initial diagnosis, episodes of severe hypoglycemia (decreased level of consciousness, glucagon use required or seizures) and presence of diabetes complications (microalbuminuria, nephropathy, retinopathy or neuropathy).

Disease severity

To get an idea of the magnitude of impact that type 1 diabetes was having on the patient’s daily life, a self-reported disease severity was determined based on several factors. Using a self-reported symptom inventory, patients were asked to rate their diabetes as well-controlled (hemoglobin A1c level less than or equal to 8.5% with no episodes of DKA, no episodes of severe hypoglycemia and no complications of diabetes), moderately controlled (hemoglobin A1c level less than or equal to 8.5% with an episode of DKA and/or an episode of severe hypoglycemia) or poorly controlled (hemoglobin A1c level greater than 8.5% or any level of hemoglobin A1c associated with diabetes complications and/or multiple episodes of DKA or hypoglycemia). Glycemic control, as measured by the hemoglobin A1c level, was not used as the sole indicator of disease severity because mild hyperglycemia is usually not symptomatic and does not affect the adolescent’s current daily life. A hemoglobin A1c level of less than 8.5% was chosen to include those children who were achieving optimal glycemic control (less than 7% as defined by the Canadian Diabetes Association – clinical practice guidelines [13]) as well as those who were achieving suboptimal control with regard to risk of future long-term complications of diabetes, but who would not be frequently symptomatic from their hyperglycemia.

The PHCSC scale

The PHCSC scale (14) was used to assess self-concept including the following six domains – physical appearance and attributes, freedom from anxiety, intellectual and school status, behavioural adjustment, happiness and satisfaction, and popularity. It is a widely used measure of psychological health in children and adolescents, and has acceptable reliability and validity. The PHCSC scale has been validated for children and adolescents seven to 18 years of age, is written at a third-grade reading level and contains 60 statements that are answered with a yes or no response. A higher score reflects a more positive response.

The CATIS

The CATIS (15,16) is a short self-report scale that has been used to assess adolescent’s attitudes about having a chronic medical condition. It contains 13 questions and is designed to be completed by children and adolescents eight years of age or older. A higher score reflects a more positive attitude, while lower scores reflect more negative attitudes. The CATIS has been mainly used in children who have epilepsy or asthma, but it was developed to allow insertion of different chronic medical conditions into the questions.

The FAM III scale

The FAM III scale (17, 18) is a well-validated instrument used to assess family functioning that can be completed by family members older than 10 years of age. Patients were asked to complete the general scale, which consists of 50 statements presented in the form of a Likert scale. In addition to an overall rating score, the clinical parameters assessed include task accomplishment, role performance, communication, affective expression, involvement, control, values and norms. Raw scores are converted to T scores. Each subscale has a mean T score of 50, with the normal range being between 40 and 60. Unlike the PHCSC scale and the CATIS, a higher score (ie, a T score over 60) reflects poor family functioning.

Data analysis

Statistical analysis was carried out using SPSS version 14.0 (SPSS Inc, USA) for Windows (Microsoft Corporation, USA). Descriptive data were presented as mean ± SD. To assess whether adolescents with type 1 diabetes have a poorer self-concept than the normative population, the total PHCSC scale score as well as the six domain scores were compared with the normative means using a t test. To assess the impact of some diabetes-specific variables, attitude toward illness, family function and family income on self-concept, regression analysis was used to examine the relationship among the total scores on the PHCSC scale, demographic variables, disease severity, the CATIS and the FAM III scale. Pearson correlation coefficients are also reported.

Ethical approval

The present study was approved by the University of Calgary (Calgary, Alberta) Ethics Committee of Human Experimentation, in accordance with the Declaration of Helsinki (Conjoint Health Research Ethics Board of the Faculty of Medicine, University of Calgary, Calgary, Alberta). The research project was explained to both the adolescent and the parent(s). Parental consent and child assent was obtained, and families were invited to mail back completed surveys or return them after their clinic visit.

RESULTS

Forty-eight of 59 (81%) teens who were approached for the study agreed to participate and returned completed questionnaires. Of these, 22 were boys (45.8%) and 26 were girls (54.2%). The mean age at the time of study was 15.2±1.7 years (range 12.2 to 18.0 years) and the mean age at diagnosis was 9.2±3.3 years (range 1.3 to 14.9 years). The mean duration of diabetes at the time of study was 5.9±3.7 years (range 1.0 to 16.2 years).

Table 1 describes the self-reported net family annual income of the patient population studied, which indicates that approximately one-half were in a higher socioeconomic group with a net annual family income of greater than $50,000. One-third of patients preferred not to answer this question. However, the average family income reported appears to be consistent with the average income of families in Calgary (Statistics Canada reported a median family income of $75,400 during 2005 for families residing in Calgary [19]). There was no significant correlation between family functioning (overall score on the FAM III scale) and family income when analyzed, excluding people who declined to report their income.

TABLE 1.

Self-reported net family annual income

Income ($)	Number of families	%
Preferred not to answer	14	29.2
<15,000	1	2.1
15,000–30,000	0	0.0
30,001–50,000	8	16.7
50,001–80,000	17	35.4
80,001–120,000	5	10.4
>120,000	3	6.3
Total	48	100

Table 2 shows the self-reported indications of disease severity (episodes of DKA since diagnosis, episodes of severe hypoglycemia and diabetes complications). Based on these self-reported indications, 21 patients (43.8%) were classified as having well-controlled diabetes, 23 (47.9%) had moderately controlled diabetes and four (8.3%) had poorly controlled diabetes. The mean duration of diabetes was similar for all groups – 6.1±4.7 years (range one to 16.2 years) for the well-controlled group, 5.6±2.85 years (range two to 10.3 years) for the moderately controlled group and 7.0±2.43 years (range four to 10 years) for the poorly controlled group.

TABLE 2.

Self-reported symptom inventory

	Episodes of DKA since diagnosis, n (%)	Episodes of severe hypoglycemia, n (%)	Complications of diabetes, n (%)
None	34 (70.8)	33 (68.8)	48 (100)
1	10 (20.8)	12 (25.0)	0 (0)
>1	4 (8.3)	3 (6.3)	0 (0)
Total	48 (100)	48 (100)	48 (100)

Diabetic ketoacidosis (DKA) was defined as hyperglycemia and ketonuria associated with a pH less than 7.3 and a bicarbonate level less than 15 mmol/L. Severe hypoglycemia was defined as a decreased level of consciousness, glucagon use required or seizure associated with a blood glucose of less than 4 mmol/L. Diabetes complications were defined as a diagnosis of nephropathy (or urine positive for microalbuminuria), retinopathy or neuropathy that is a result of chronic hyperglycemia

The total score as well as the scores for the six subscales of the PHCSC scale are shown in Table 3. When compared against a mean score of 50 from the normative population, adolescents with type 1 diabetes in the present study had a significantly better self-concept in the areas of physical appearance and attributes, behavioural adjustment, happiness and satisfaction, and popularity. The total score for the PHCSC scale was not found to correlate with disease severity, income, sex, age at diagnosis, age at time of study, episodes of DKA or episodes of severe hypoglycemia (Table 4).

TABLE 3.

Piers-Harris Children’s Self-Concept Scale total score and subscale scores

n=48	Mean ± SD	Compared with normative mean P
Total	53.73±8.75	0.005*
Behavioural adjustment	54.44±8.05	0.000*
Intellectual and school status	50.65±8.42	0.598
Physical appearance and attributes	53.42±7.68	0.003*
Freedom from anxiety	51.54±8.50	0.215
Popularity	53.17±8.95	0.018*
Happiness and satisfaction	52.19±6.33	0.021*

^*P<0.05 indicates significance

TABLE 4.

Summary of correlations of Piers-Harris Children’s Self-Concept Scale total score with various factors

	Pearson correlation (r)	Two-tailed P
Disease severity	−0.144	0.329
Income	0.120	0.416
Sex	−0.178	0.225
Age at diagnosis	−0.021	0.885
Age at time of study	−0.165	0.262
Episodes of DKA	0.148	0.316
Episodes of severe hypoglycemia	0.224	0.125
The CATIS score	0.483	0.001*
The FAM III overall score	−0.507	0.000*

^*P<0.05 indicates significance. CATIS Child Attitude Toward Illness Scale; DKA Diabetic ketoacidosis; FAM III Family Assessment Measure, version III

Adolescents’ self-concept as measured by the PHCSC scale was found to be significantly positively correlated with a better attitude toward chronic illness as measured by the CATIS (Table 4). In addition, the PHCSC scale was found to have a significant negative correlation with the overall FAM III score, which indicates that a better self-concept was correlated with more positive family functioning (Table 4).

DISCUSSION

In our study population, self-concept in adolescents with diabetes was overall very good and significantly better than the mean in certain areas (Table 3). Positive self-concept of adolescents with type 1 diabetes was not found to be significantly correlated with disease severity in our study. This is in contrast to a previous study (11), which showed that better self-concept was associated with good diabetes control. However, good diabetes control was defined as a hemoglobin A1c level lower than 10%, which is quite different from the criteria used in our study. These results suggest that interventions to improve self-concept need to be aimed at all adolescents with diabetes regardless of their disease severity. An example of this could be a teen support group or a family support group that includes interventions aimed at facilitating positive self-concept in adolescents. Other ways of trying to help adolescents perceive their disease in a more positive way could include introducing positive role models or mentors, who also have type 1 diabetes, to newly diagnosed adolescents, specific education for adolescents on diabetes management so that they feel in control of their chronic disease (eg, driving safety, alcohol management, sleep patterns, etc) and helping to encourage developmentally appropriate independence.

In our study, self-concept did not correlate with disease severity. It is possible that the adolescents with more severe disease adjusted better to their diabetes than those with less severe disease, who are still struggling to be ‘normal’ adolescents. On the other hand, it is also possible that those with more severe disease have poorly controlled diabetes because they choose to focus on other areas of their life that improve their self-concept rather than worry about managing their diabetes. There are also many complex parent and adolescent factors that were not captured in the present study and could impact on the adjustment to illness, perceived control over illness and treatment adherence. Some of these complex interactions that were not included in the present study were adolescent temperament, parenting style, parental stress levels, adolescent stress levels, sibling interactions or family type (blended, single parent, etc).

There was no correlation found with self-concept and income of family (although one-third of families declined to answer this question), sex, age at diagnosis, age at time of study, episodes of DKA or episodes of hypoglycemia. However, adolescents with better self-concept had better attitudes toward their chronic illness and better family functioning, which is consistent with previous studies (2–5).

The present study may have been limited by the selection bias because the patients that agreed to participate in the cross-sectional survey may have been more likely to be higher functioning families or well-adjusted teens. Only 8.3% of the patients in our study were classified as having poorly controlled diabetes, which could reflect the selection bias seen in recruiting patients attending the diabetes clinic. Often, adolescents who are having significant difficulty with managing their diabetes may be noncompliant and miss their regularly scheduled clinic appointments and, therefore, would not have been approached to be recruited. However, the mean overall hemoglobin A1c level in our clinic population, estimated in November 2007 from a total of 863 patients, was 8.8%. This would suggest that based on our criteria for level of control in the present study, our sample was likely representative of our clinic population.

The study population was relatively homogeneous and only 8.3% of patients enrolled were classified as having poorly controlled diabetes. This may have affected the ability to detect a correlation with disease severity and self-concept. In addition, the present study did not specifically assess diabetes treatment regimen (eg, three times daily insulin injections, basal bolus insulin injections or continuous subcutaneous insulin infusion), which could potentially affect self-concept, attitude toward illness and diabetes control. It was also not feasible in the present study to gather information around diabetes management, such as dietary compliance, parental involvement in diabetes care or medication compliance. All of these factors could have also played a role in adolescent self-concept and attitude toward illness.

All of the instruments used relied on self-report by the patients and parents, which may have limited the findings. Parents and patients may have not been prepared to disclose some more sensitive aspects of their family functioning even though anonymity was ensured. Of the families enrolled in the present study, 29.2% still preferred not to disclose their family income and, similarly, many patients may have been cautious when responding to some of the instruments. It is also interesting to note that a large proportion of the patients in the present study had a high family income level, which may limit the generalizability of these results.

In our study, we did not find any significant correlations among the self-concept of adolescents with type 1 diabetes and disease severity, income of family, sex, age at diagnosis, age at time of study, episodes of DKA or episodes of hypoglycemia. Adolescents with better self-concept had better attitudes toward their chronic illness and better family functioning. Although there was no correlation with diabetes disease severity seen in our population, interventions aimed at improving adolescent self-concept may have a positive impact on diabetes treatment by improving family functioning and attitude toward living with type 1 diabetes. Interventions should be targeted at all adolescents because self-concept is not always correlated with disease severity. Conversely, supporting families and improving family functioning may also help minimize the impact of having a chronic disease on the adolescent and result in better adjustment and self-concept for the adolescent. Future studies should focus on self-concept in a larger population of adolescents with poorly controlled diabetes to see whether there are other specific correlates of self-concept in these patients.