ORIGINAL ARTICLE
The Journal of Nursing Research h VOL. 00, NO. 0, MONTH 2015
Predictors of Glycemic Control in Adolescents of Various Age Groups With Type 1 Diabetes Shu-Li Lee1 & Fu-Sung Lo2 & Yann-Jinn Lee3 & Bai-Hsiun Chen4 Ruey-Hsia Wang5* 1
MSN, RN, Instructor, College of Nursing, Kaohsiung Medical University, and Doctoral Student, School of Nursing, Kaohsiung Medical University & 2MD, MS, Attending Physician, Division of Endocrinology, Department of Pediatrics, Chang Gung Memorial Hospital & 3MD, Attending Physician, Division of Endocrinology, Department of Pediatrics, Mackay Memorial Hospital & 4MD, PhD, Attending Physician, Department of Pediatrics, Kaohsiung Medical University Hospital & 5PhD, RN, Professor, College of Nursing, Kaohsiung Medical University.
behaviors. In addition, female adolescents aged 16Y18 years should be considered an at-risk group.
ABSTRACT Background: Understanding the predictors of glycemic control in adolescents of various age groups with type 1 diabetes (T1D) is crucial for nurses to cultivate developmental-specific interventions to improve glycemic control in this age group. However, research has rarely addressed this issue, particularly in the context of Asian populations. Purpose: We explored the predictive influence of demographic characteristics, self-care behaviors, family conflict, and parental involvement on glycosylated hemoglobin (HbA1C) levels 6 months after the baseline measurement in adolescents of various age groups with T1D in Taiwan. Methods: A prospective survey design was applied. At baseline, adolescents with T1D completed a self-care behavior scale. Parents or guardians finished scales of parental involvement and family conflict. The HbA1C levels 6 months after baseline measurement were collected from medical records. Two hundred ten adolescentYparent/guardian pairs were enrolled as participants. Multiple stepwise regressions examined the significant predictors of HbA1C levels 6 months after the baseline measurement in the three adolescent age groups: 10Y12, 13Y15, and 16Y18 years. Results: Family conflict was a significant predictor of HbA1C level within the 10Y12 years of age group 6 months after the baseline measurement. Self-care behaviors were a significant predictor of HbA1C level within the 13Y15 years of age group 6 months after the baseline measurement. Being female and self-care behaviors were each significant predictors of HbA1C level in the 16Y18 years of age group 6 months after the baseline measurement. Conclusions/Implications for Practice: Nurses should design specific interventions to improve glycemic control in adolescents of various age groups with T1D that are tailored to their developmental needs. For adolescents with T1D aged 10Y12 years, nurses should actively assess family conflict and provide necessary interventions. For adolescents with T1D aged 13Y18 years, nurses should exert special efforts to improve their self-care
KEY WORDS: type 1 diabetes, glycemic control, adolescent, predictors, age groups.
Introduction Type 1 diabetes (T1D) is one of the most common chronic diseases in children. The incidence of T1D has been increasing annually by 2%Y5% worldwide (Maahs, West, Lawrence, & Mayer-Davis, 2010). Although the incidence and prevalence of T1D in Asia is lower than in Western countries, T1D remains a crucial health concern for Asian children (Maahs et al., 2010). Between 2000 and 2009 in Taiwan, the incidence of T1D in children below 19 years old ranged from 4.21 to 5.79 per 100,000 (Jiang, Chang, Tai, Chen, & Chuang, 2012). Poor glycemic control in children with T1D may cause life-threatening complications such as microvascular and neuropathic complications. Level of glycosylated hemoglobin (HbA1C) is a crucial indicator of glycemic control. The American Diabetes Association (ADA, 2013) recommends that HbA1C levels should be less than 8.0% and 7.5% in children aged 6Y12 and 13Y19 years, respectively. However, earlier studies have indicated that the level of HbA1C in adolescents is higher Accepted for publication: July 23, 2014 *Address correspondence to: Ruey-Hsia Wang, N100, Shih-Chuan 1st Rd., San-Ming District, Kaohsiung City 80708, Taiwan, ROC. Tel: +886 (7) 312Y1101 ext. 2641; E-mail: [email protected] The authors declare no conflicts of interest. Cite this article as: Lee, S. L., Lo, F. S., Lee, Y. J., Chen, B. H., & Wang, R. H. (2015). Predictors of glycemic control in adolescents of various age groups with type 1 diabetes. The Journal of Nursing Research, 00(0), 00Y00. doi:10.1097/jnr.0000000000000096
1 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
than in other age groups (ADA, 2013; Helgeson, Siminerio, Escobar, & Becker, 2009). It is thus vital that nurses understand the key factors of influence on HbA1C levels to improve glycemic control in adolescents with T1D. Demographic characteristics have been associated with glycemic control. Adolescents who were older, were female, had a longer duration of diabetes, were living in non-dualparent families, and were of lower socioeconomic status (SES) were identified as facing a higher risk of poor glycemic control (Neylon, O’Connell, Skinner, & Cameron, 2013). Psychosocial factors are crucial in influencing the glycemic control of adolescents with T1D. Adolescents with TID must perform self-care behaviors, which include insulin injections, blood glucose tests, diet management, appropriate exercise, and hypoglycemia or hyperglycemia management to control their HbA1C levels (ADA, 2013). An integrative literature review indicated that self-care behaviors positively predicted HbA1C levels in children with T1D in longitudinal studies (Guo, Whittemore, & He, 2011). However, poor self-care behaviors in adolescents with T1D have been frequently reported in numerous studies (Guo et al., 2011; Neylon et al., 2013). In Taiwan, adolescents with T1D performed poorly with regard to administering insulin injections, monitoring blood glucose, adhering to meal plans, and engaging in appropriate exercise (Chang, Yeh, Lo, & Shih, 2007). Parents play a fundamental role in facilitating effective glycemic control management in adolescents with T1D because of the complexity of the control process. Most adolescents with T1D thus rely on parental assistance to conduct diabetes-related activities such as monitoring blood glucose and delivering insulin injections (King et al., 2012). A previous cross-sectional study associated less parental assistance with diabetes management and poor glycemic control in adolescents with T1D (Wysocki et al., 2009). Longitudinal studies have also indicated that, when parents share diabetes management responsibilities, there is a positive association with glycemic control in this population (Ingerski, Anderson, Dolan, & Hood, 2010; King et al., 2012). Thus, parental involvement is crucial to the effective control of glycemic levels in adolescents with T1D. Adolescence is a period in life when individuals often seek greater autonomy from their parents, rely less on parental guidance, and challenge parental authority (Garrison & Felice, 2009). These age-related characteristics may result in disagreements between parents and their T1D children with regard to diabetes care responsibilities (Schilling, Knafl, & Grey, 2006). Adolescents with T1D often struggle with relying on parents and making decisions by themselves, which may exacerbate diabetes-specific conflict with parents (Karlsson, Arman, & Wikblad, 2008). A cross-sectional study conducted in a Western population indicated that increased diabetes-specific family conflict resulted in higher HbA1C levels in adolescents with T1D (Williams, Laffel, & Hood, 2009). A longitudinal study also found that higher diabetes-specific family conflict predicted worse adherence to regimens in adolescents with T1D, which negatively influenced glycemic control (Hilliard, Guilfoyle, Dolam, & Hood, 2011). However, the influence of parental
Shu-Li Lee et al.
involvement and family conflict on HbA1C levels in adolescents with T1D has been less frequently reported in Taiwan. Although numerous studies have associated demographic characteristics, self-care behaviors, parental involvement, and family conflict with HbA1C levels in adolescents with TID, certain gaps in this research await further study and clarification. The first of these gaps is highlighted by the fact that adolescence is frequently subdivided into three developmental stages: early adolescence (10Y13 years), middle adolescence (14Y16 years), and late adolescence (17Y20 years; Garrison & Felice, 2009). Adolescents experience cognitive and psychosocial development phenomena that are distinct to their specific stage (Allender, 2002). Early adolescence is a transitional period for adolescents in which they work to establish independence from their parents (Garrison & Felice, 2009). Adolescents in the middle stage tend to strive for greater autonomy and challenge the authority of parents and other adults (Allender, 2002; Garrison & Felice, 2009). During late adolescence, adolescents are focused on the future and experience new behaviors (Garrison & Felice, 2009). Different developmental characteristics may cause different sets of factors to influence HbA1C levels in each of the three developmental stages. However, numerous previous studies of adolescents with T1D have treated adolescence as a single stage (frequently between 10 and 18 or 13 and 18 years old) rather than further distinguishing adolescence into a series of distinct developmental stages (Chang et al., 2007; Ingerski et al., 2010). The second of these gaps is that most of the previous studies of this issue used cross-sectional designs, which are unable to provide information regarding causal relationships between associated factors and HbA1C levels. The third and final of these gaps is that most of the previous studies of this issue were conducted in Western populations with a relatively high prevalence of T1D. The predictors of glycemic control in adolescents with T1D in Taiwan may differ from those identified for Western populations. It is necessary to clarify the predictors of HbA1C level in adolescents in different age groups with T1D so that nurses may use this knowledge to cultivate developmental-specific interventions for adolescents with T1D. HbA1C levels reveal glycemic control efficacy over the previous 3-month period. Following HbA1C levels 3 months after the baseline measurement may inadequately reflect the influence of the target predictors. However, clarifying the predictors of HbA1C level 6 months after the baseline measurement may effectively manifest the causal relationships between predictors and HbA1C levels. Therefore, the purpose of this study is to explore significant predictors of HbA1c level 6 months after the baseline measurement based on demographic characteristics, self-care behaviors, parental involvement, and family conflict in adolescents of different age groups with T1D. We hypothesize that demographic characteristics, self-care behaviors, parental involvement, and family conflict should significantly influence HbA1C levels at 6 months after the baseline measurement and that this influence should be variegated among the three adolescent age groups.
2 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
Predictors of Glycemic Control in Adolescents
Methods Design This study used a prospective design. At baseline, data on self-care behaviors were collected from the adolescent participants. Considering that parent or guardian responses may be more accurate than those of children, data of adolescent demographic characteristics, diabetes-specific parental involvement, and diabetes-specific family conflict were collected from the parent or guardian participants. HbA1c levels were collected 6 months after the baseline measurement.
Participants Adolescents aged 10Y18 years and experiencing T1D for at least 6 months were recruited from two medical centers in Northern Taiwan. Adolescents who were pregnant or had a severe learning disability or a significant medical disease other than T1D were excluded. Endocrine physicians referred adolescents who met the inclusion criteria to a research assistant when these adolescents visited the clinic. Two hundred twenty-nine adolescentYparent/guardian pairs were invited to participate in this study. Two hundred ten adolescentYparent/ guardian pairs (response rate: 91.70%) enrolled as participants. In Taiwan, adolescents aged 10Y12, 13Y15, and 16Y18 years are enrolled in elementary, middle, and high school, respectively. We grouped adolescents into 10- to 12-, 13- to 15-, and 16- to 18-year age groups to represent early, middle, and late adolescence, respectively. Of the 210 adolescents participating in this study, 79 (37.6%), 80 (38.1%), and 51 (24.3%) were in the early, middle, and late adolescence groups, respectively.
Measurement Structured, anonymous questionnaires and the medical records for the adolescent participants were used to collect data.
Questionnaire Adolescents and parents or guardians completed the questionnaires independently. Each questionnaire took approximately 15 minutes to complete.
Demographic characteristics The questionnaire given to the parent or guardian participants collected data on adolescents’ age, gender, duration of diabetes, family structure, educational level (scored from 1 to 5), and occupation level of parents (scored from 1 to 5). Family structure was divided into two possible formats: dualparent family and non-dual-parent family (single-parent or no-parent family). Educational level was multiplied by 4, and occupation level was multiplied by 7 to generate SES scores. The highest SES score for the parents was family SES. Scores ranging from 41 to 55, 30 to 40, and 11 to 29 were coded high, middle, and low SES, respectively (Lin, 2005). The high and middle SES scores were grouped together as ‘‘higher SES’’ for further study.
VOL. 00, NO. 0, MONTH 2015
Diabetes-specific self-care behavior scale for adolescents The nine-item scale developed by Tang, Chen, and Wang (2013) was used to measure the degree to which blood glucose was monitored, insulin doses were adjusted, diet was regulated, exercise was modified, and hypoglycemia or hyperglycemia was managed. Each item was rated from 0 = always to 4 = never. The total scale ranged from 0 to 36. Higher scores indicated enhanced self-care behaviors. In this study, the Cronbach’s " was .70. The intraclass correlation coefficient indicated that testYretest reliability was .72. Diabetes-specific parental involvement scale for parents We adapted the 15-item English version of the Diabetes Responsibility and Conflict Scale (DRCS; Rubin, YoungHyman, & Peyrot, 1989) for use in this study. The Englishversion DRCS was developed to separately measure the relative degree of involvement and conflict in diabetes-specific tasks between parents or guardians and adolescents. The original Cronbach’s " of the parental involvement subscale and the family conflict subscale in the 15-item English-version DRCS was .79 and .88, respectively (Rubin et al., 1989). In this study, the 15-item DRCS was translated from English into Chinese by a bilingual expert and then back-translated from Chinese into English by another bilingual expert. A third bilingual expert confirmed that the Chinese and English versions of the DRCS were identical. The research team deleted three items including ‘‘Who tests urine,’’ ‘‘Who talks to adults about your diabetes,’’ and ‘‘Who decides which people know that you are diabetic’’ because of the unsuitability of these issues within the context of Taiwanese culture. Finally, 12 diabetesspecific tasks were used to measure diabetes-specific parental involvement. The example item was ‘‘Who decides what to eat at meals.’’ Each item was rated from 1 = child, all of the time to 5 = parent, all of the time. The range of possible total scores ranged from 12 to 60, with higher scores indicating higher parental involvement. In this study, the Cronbach’s " and intraclass correlation coefficient were .88 and .80, respectively. Diabetes-specific family conflict scale for parents The same diabetes-specific tasks measured in the 12-item diabetesspecific parental involvement scale for parents were also used to assess the frequency of conflict between parents or guardians and adolescents regarding tasks of diabetes-specific care. The question was ‘‘How much you argue/disagree with your child about each of the following aspects of your child’s diabetes care?’’ Each item was rated from 1 = never to 5 = always. The possible total scores ranged from 12 to 60, with higher scores indicating higher family conflict. In this study, the Cronbach’s " and intraclass correlation coefficient were .94 and .78, respectively.
Medical Records A research assistant collected the HbA1C levels of the adolescent participants from medical records 6 months after the baseline measurement. 3
Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
Shu-Li Lee et al.
Validity and Reliability
Statistical Analysis
Seven experts specializing in pediatric diabetes independently examined each scale. The experts rated each item from 1 = not relevant to 4 = very relevant. The number of items rated 3 or 4 were divided by the total number of items in each scale to calculate the content validity index for each scale. The content validity index of each scale ranged from 0.90 to 1.00. Data for all participants were used to examine the Cronbach’s " of each scale. TestYretest reliability was assessed by administering questionnaires to 21 adolescentYparent/guardian pairs over a 2-week interval. The intraclass correlation coefficient was used to calculate testYretest reliability.
Data were analyzed using SPSS software version 19.0 (IBM, Armonk, NY, USA). Descriptive statistics were used to describe all variables. Chi-squared tests, two sample t tests, one-way analysis of variance with Scheffe post hoc tests, and Pearson correlations were used to examine the relationships between demographic characteristics, psychosocial factors, and HbA1C levels. A multiple stepwise regression analysis was used to examine the significant predictors of HbA1C level 6 months after baseline measurement for the three age groups.
Ethical Consideration This study was approved by the institutional review board of a medical university. We informed adolescents and their parents or guardians that refusing to participate in this study would incur no penalty and that they had the right to withdraw from the study at any time. All adolescents and their parents or guardians were required to sign informed assent and consent forms before they participated in this study. The parents or guardians also signed a consent form for their children to participate in the study.
Results Distribution of Demographic Characteristics, Psychosocial Factors, and HbA1C Levels Among the Three Age Groups The mean ages within the 10Y12, 13Y15, and 16Y18 years of age groups were 11.63 (SD = 0.97), 14.54 (SD = 0.82), and 16.83 (SD = 0.63) years, respectively. Table 1 showed that
TABLE 1.
Distribution and Comparison of Demographic Characteristics, Psychosocial Factors, and HbA1c Levels Among the Three Age Groups ffi 10Y12 Year (n = 79) n
Characteristic Demographic characteristics Gender Male Female Duration of diabetes (years): M T SD
%
ffl 13Y15 Year (n = 80) n
%
16Y18 Year (n = 51) n
%
2 2 = 0.44 38 41 5.73
48.10 51.90 3.35
39 41
48.75 51.25
6.40
3.85
22 29 8.28
43.14 56.86 4.45
9 ffia Family structure Dual-parent family Non-dual-parent family SES Higher Lower
Psychosocial factors Self-care behaviors Parental involvement Family conflict HbA1c
2 2/F
73 6
92.40 7.60
76 4
F = 6.99** 2 2 = 3.23
95.00 5.00
44 7
86.27 13.73 2 2 = 1.13
47 32
59.49 40.51
41 39
51.25 48.75
29 22
56.86 43.14
M
SD
M
SD
M
SD
26.15
4.50
25.11
4.58
23.35
5.08
F = 5.56**
9.13
ffi 9 a 31.23 8.22
26.31
7.11
F = 32.10***
23.59
11.75
ffi 9 ffl 9 a 21.66 9.04
22.27
11.16
7.99
1.79
9.07
2.01
37.99
8.54
2.04
9 ffia Note. SES = socioeconomic status. a Scheffe post hoc analysis. **p G .01. ***p G .001.
4 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
F = 0.68 F = 4.94***
Predictors of Glycemic Control in Adolescents
VOL. 00, NO. 0, MONTH 2015
gender, family structure, and SES did not differ significantly among the age groups. Self-care behaviors decreased with increasing age, with the self-care behaviors of the 16Y18 years of age group significantly worse than that of the 10Y12 years of age group. Parental involvement significantly decreased with increasing age. However, family conflict did not significantly differ among the three age groups. The mean HbA1C level for all participants was 8.5% (SD = 1.93%). Table 1 shows that the mean HbA1c levels for the three age groups differed significantly. Scheffe post hoc analysis indicated that the HbA1C level for the 16Y18 years of age group was significantly higher than the level for the 10Y12 years of age group.
The Association Between Demographic Characteristics and Psychosocial Factors and HbA1c Levels 6 Months After Baseline Measurement for Each Age Group Table 2 shows that gender and SES were not significantly associated with HbA1C level in the 10Y12 or 13Y15 years of age groups. The HbA1C levels of female adolescents and of participants with lower SES were significantly higher than those of male adolescents and of participants with higher SES within the 16Y18 years of age group. Family structure and duration of diabetes were not significantly associated with HbA1c levels in any age group. Regarding psychosocial factors, self-care behaviors were significantly associated with HbA1C levels in the 13Y15 and 16Y18 years of age groups. Family conflict was significantly
associated with HbA1C levels in the 10Y12 years of age group. Parental involvement was not significantly associated with HbA1C levels in any age group.
Predictors of HbA1C Level 6 Months After Baseline Measurement in Each Age Group A multiple stepwise regression was used to examine the significant predictors of HbA1C level 6 months after baseline measurement in each age group. Gender, duration of diabetes, family structure, SES, self-care behaviors, parental involvement, and family conflict were used as the independent variables. All variance inflation factors and condition indices were less than 1.05 and 2.69, respectively, indicating no collinearity among the independent variables within each age group. Table 3 showed that family conflict was a significant predictor of HbA1C levels 6 months after baseline measurement in the 10Y12 years of age group, which explained 20.0% of the total variance in HbA1C levels. Greater family conflict at baseline predicted higher HbA1C levels 6 months after baseline measurement. Self-care behaviors was a significant predictor of HbA1C levels 6 months after baseline measurement in the 13Y15 years of age group, which explained 12.0% of the total variance in HbA1C levels. Enhanced self-care behaviors at baseline predicted lower HbA1C levels 6 months after baseline measurement. Female and self-care behaviors were significant predictors of HbA1C level 6 months after baseline measurement in the 16Y18 years of age group, which explained 28.0% of the total variance in
TABLE 2.
Comparison of the Effect of Demographic Characteristics and Psychosocial Factors on HbA1c Levels 6 Months After Baseline Measurement in Each Age Group 10Y12 Year (n = 79) Variable Demographic characteristics Gender Male Female Duration of diabetes (years) Family structure Dual-parent family Non-dual-parent family SES Higher Lower Psychosocial factors Self-care behaviors Parental involvement Family conflict
M
SD
7.71 8.25
1.62 1.92
t/r
13Y15 Year (n = 80) M
SD
8.84 8.24
2.09 1.96
t = 1.35
1.56 3.51
7.67 8.45
1.48 2.11
SD
8.22 9.72
1.50 2.13
8.51 9.03
2.06 1.74
8.17 8.92
1.62 2.36
t/r t = 2.81**
r = .03 t = 0.49
t = 1.93
r = j .03 r = .18 r = .45***
M
t = 1.32
r = .16 t = 1.07 7.87 9.42
t/r
16Y18 Year (n = 51)
r = .15 t = 1.45 8.91 10.09
2.03 1.39
8.58 9.72
1.67 2.28
t = 1.65
t = 2.07*
r = .34** r = .03 r = .16
r = .34* r = .16 r = .19
Note. SES = socioeconomic status. *p G .05. **p G .01. ***p G .001.
5 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
Shu-Li Lee et al.
TABLE 3.
Predictors of HbA1c Level for Each Age Group Age Group
Predictor
10Y12 Year
Family conflict
13Y15 Year
Self-care behaviors
16Y18 Year
Femalea Self-care behaviors
$
Total R 2
t
.45
.20
j.34
.12
j3.23**
4.40***
.41
.14
3.31**
j.37
.28
j3.04**
95% CI [ 0.04,
0.10]
[j0.25, j0.06] [ 0.65,
2.64]
[j0.25, j0.05]
Note. CI = confidence interval. a Reference group: Male. *p G .05. **p G .01. ***p G .001.
HbA1c levels. Both female and self-care behaviors individually accounted for 14.0% of the variance in HbA1C levels. Female and worse self-care behaviors at baseline predicted higher HbA1C levels 6 months after baseline measurement.
Discussion As hypothesized, the significant predictors of HbA1C level in adolescents with T1D were different for the three different age ranges examined. Family conflict was stable across the three age groups in this study, which is consistent with a previous study (Ingerski et al., 2010). As noted, few studies have examined the relationships between family conflict and HbA1c levels in adolescents with T1D at different adolescent age groups. In the United States, parent-reported family conflict was found to predict HbA1C levels 6 months after baseline measurement for adolescents with T1D aged 13Y18 years (Ingerski et al., 2010). In this study, parent-reported family conflict was not a significant predictor of HbA1c levels 6 months after baseline measurement for adolescents with T1D in the 13Y15 or 16Y18 years of age groups. However, family conflict was a significant predictor of HbA1C levels 6 months after baseline measurement in the 10Y12 years of age group. The cause of this divergent finding may be differences in the cultures targeted by the studies. While adolescents aged 10Y12 years begin to struggle for increased independence (Garrison & Felice, 2009), Chinese parents continue to exert more control than Western parents over their children’s lives (Jose, Huntsinger, Huntsinger, & Liaw, 2000). Chinese parents may unconsciously pressure their children when they perceive conflict with children. Therefore, this found that parentreported family conflict was more important than self-care behaviors in predicting HbA1C levels 6 months after baseline measurement for adolescents aged 10Y12 years. Improving communication and problem-solving skills may effectively reduce family conflict (Wysocki et al., 2006). For adolescents aged 10Y12 years with T1D, it is thus important that nurses actively assess family conflict between parents and adolescents to help these adolescents effectively communicate with their parents to solve diabetes-care-related disagreements. Although parental involvement significantly decreased with increasing age, parental involvement was not signifi-
cantly associated with HbA1c levels 6 months after baseline measurement in any of the age groups. The decrease of parental involvement did not necessarily result in poor glycemic control 6 months after baseline measurement in Taiwan. A previous study indicated that perceived parental overinvolvement in diabetes care negatively influenced HbA1c levels in adolescents with T1D (Cameron et al., 2008). Perception of parental overinvolvement may be more crucial than the degree of parental involvement in predicting glycemic control of adolescents with T1D in Taiwan. However, further studies are required to confirm this hypothesis. Self-care behaviors were a significant predictor of HbA1c levels 6 months after baseline measurement for adolescents aged 13Y15 and 16Y18 years. This result was consistent with an integrative review, which concluded that self-care behaviors were specifically crucial for glycemic control among adolescents older than 13 years (Guo et al., 2011). Although middle-aged adolescents pursue greater autonomy (Garrison & Felice, 2009), increasing autonomy does not indicate increased capacity to perform self-care behaviors effectively. For example, only 31% of adolescents aged 16 years were comfortable making decisions on dietary control (Clarke, 2011). A previous study found that self-care patterns were parental dominant, transitional, and adolescent dominant in early, middle, and late adolescence, respectively (Schilling, Grey, & Knafl, 2002). The influence of self-care behaviors on HbA1c levels in adolescents may increase as they assume increasing control on their own diabetes care regimen. Parental encouragement and collaboration with adolescents are crucial for improving the self-care behaviors of this population (Wysocki et al., 2006, 2009). For adolescents aged 13Y18 years, nurses may serve as a coordinator between adolescents and parents to assist adolescents achieve enhanced self-care behaviors. This study found HbA1c levels of female adolescents in the 16Y18 years of age group to be significantly higher than those of male adolescents in the same group. Furthermore, being female accounted for 14% of the total variance in HbA1c levels in the 16Y18 years of age group. Insulin sensitivity decreases in adolescents aged 17Y19 years, with the decrease particularly significant in women (Wilkin et al., 2004). A previous study in Taiwan observed that disturbed eating behaviors contributed to higher HbA1c levels in female adolescents with T1D than in their male counterparts (Hsu,
6 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
Predictors of Glycemic Control in Adolescents
Chen, Huang, Lin, & Lin, 2009). Decreased insulin sensitivity and increased disturbed eating behaviors among female adolescents may support ‘‘being female’’ as a crucial predictor of HbA1c levels in the 16Y18 years of age group. Therefore, female adolescents aged 16Y18 years should be considered an at-risk subgroup of adolescents with T1D. In this study, HbA1c levels in the 16Y18 years of age group were the highest of the three age groups. Self-care behaviors within the 16Y18 years of age group were the poorest among the three age groups. Thus, we propose that poor self-care behaviors are also an important contributing factor in this group. This study explored the influence of individual and family factors on HbA1c levels in adolescents with T1D for three distinct adolescent age groups. However, adolescents experience a strong need for peer acceptance (Garrison & Felice, 2009). Peer relationships have been shown to predict HbA1c levels 4 years after baseline measurement for adolescents with T1D (Helgeson et al., 2009). Depression has also been shown to predict high HbA1C levels in adolescents with T1D (Hilliard, Wu, Rausch, Dolan, & Hood, 2013). We suggest that future studies assess the influence of peer relationships and depression on glycemic control in adolescents of various age groups with T1D.
Limitations The small sample size used in this study was collected from only two medical centers in Northern Taiwan. Therefore, the generalizability of findings may be limited. We suggest that further studies recruit samples from diverse hospitals. Parent and adolescent perspectives of family conflict and parental involvement may influence HbA1c levels in differing ways (Guo et al., 2011). Further studies may measure family conflict and parental involvement from both parent and adolescent perspectives to clarify the relative importance of parent or adolescent perspectives in predicting HbA1c levels in the Taiwanese context. We did not follow the same participants across distinct age groups. A longitudinal design may be designed to follow changes in self-care behaviors, family conflict, parental involvement, and HbA1C levels in the same adolescents across multiple age groups. This design could examine the influence of the dynamic change in these factors on HbA1c levels across time.
Conclusions Significant predictors of HbA1c level 6 months after the baseline measurement in adolescents with T1D differed for the three different age groups. Therefore, our results confirm that intervention programs designed to improve glycemic control of adolescents with T1D should be tailored to address the factors of influence most relevant to the target age group. For adolescents with T1D aged 10Y12 years, nurses should actively assess family conflict and provide necessary intervention. For adolescents with T1D aged 13Y18 years, nurses should focus on improving patients’ self-care behaviors. Finally, female adolescents aged 16Y18 years should be considered an at-risk subgroup within adolescents with T1D.
VOL. 00, NO. 0, MONTH 2015
References Allender, M. (2002). Adolescent. In C. L. Edelman & C. L. Mandle (Eds.), Health promotion throughout the life span (5th ed., pp. 619Y647). St. Louis, MO: Mosby. American Diabetes Association. (2013). Standards of medical care in diabetesV2013. Diabetes Care, 36(1), s11Ys66. doi:10 .2337/dc13-S011 Cameron, F. J., Skinner, T. C., de Beaufort, C. E., Hoey, H., Swift, P. G. F., Aanstoot, H., I Skovlund, S. E. (2008). Are family factors universally related to metabolic outcomes in adolescents with type 1 diabetes? Diabetic Medicine, 25(4), 463Y468. doi:10.1111/j.1464-5491.2008.02399.x Chang, C. W., Yeh, C. H., Lo, F. S., & Shih, Y. L. (2007). Adherence behaviours in Taiwanese children and adolescents with type 1 diabetes mellitus. Journal of Clinical Nursing, 16(7b), 207Y214. doi:10.1111/j.1365-2702.2005.01541.x Clarke, W. L. (2011). Behavioral challenges in the management of childhood diabetes. Journal of Diabetes Science and Technology, 5(2), 225Y228. doi:10.1177/193229681100500203 Garrison, W., & Felice, M. E. (2009). Adolescence. In W. B. Carey, A. C. Crocker, W. L. Coleman, E. R. Elias, & H. M. Feldman (Eds.), DevelopmentalYbehavioral pediatrics (4th ed., pp. 62Y73). Philadelphia, PA: Saunders. Guo, J., Whittemore, R., & He, G. P. (2011). The relationship between diabetes self-management and metabolic control in youth with type 1 diabetes: An integrative review. Journal of Advanced Nursing, 67(11), 2294Y2310. doi:10.1111/j.13652648.2011.05697.x Helgeson, V. S., Siminerio, L., Escobar, O., & Becker, D. (2009). Predictors of metabolic control among adolescents with diabetes: A 4-year longitudinal study. Journal of Pediatric Psychology, 34(3), 254Y270. doi:10.1093/jpepsy/jsn079 Hilliard, M. E., Guilfoyle, S. M., Dolam, L. M., & Hood, K. K. (2011). Prediction of adolescents’ glycemic control 1 year after diabetes-specific family conflict. Archives of Pediatrics & Adolescent Medicine, 165(7), 624Y629. doi:10.1001/archpediatrics.2011.86 Hilliard, M. E., Wu, Y. P., Rausch, J., Dolan, L. M., & Hood, K. K. (2013). Predictors of deteriorations in diabetes management and control in adolescents with type 1 diabetes. The Journal of Adolescent Health, 52(1), 28Y34. doi:10.1016/j.jadohealth.2012.05.009 Hsu, Y. Y. A., Chen, B. H., Huang, M. C., Lin, S. J., & Lin, M. F. (2009). Disturbed eating behaviors in Taiwanese adolescents with type 1 diabetes mellitus: A comparative study. Pediatric Diabetes, 10(1), 74Y81. doi:10.1111/j.1399-5448.2008.00422.x Ingerski, L. M., Anderson, B. J., Dolan, L. M., & Hood, K. K. (2010). Blood glucose monitoring and glycemic control in adolescence: Contribution of diabetes-specific responsibility and family conflict. Journal Adolescent Health, 47(2), 191Y197. doi:10.1016/j.jadohealth.2010.01.012 Jiang, Y. D., Chang, C. H., Tai, T. Y., Chen, J. F., & Chuang, L. M. (2012). Incidence and prevalence rates of diabetes mellitus in Taiwan: Analysis of the 2000Y2009 nationwide health insurance database. Journal of the Formosan Medical Association, 111(11), 599Y604. doi:10.1016/j.jfma.2012.09.014 Jose, P. E., Huntsinger, C. S., Huntsinger, P. R., & Liaw, F. R. (2000). Parental values and practices relevant to young children’s social development in Taiwan and the United States. Journal of Cross-Cultural Psychology, 31(6), 667Y702. doi:10 .1177/0022022100031006002
7 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
Karlsson, A., Arman, M., & Wikblad, K. (2008). Teenagers with type 1 diabetesVA phenomenological study of the transition towards autonomy in self-management. International Journal of Nursing Studies, 45(4), 562Y570. doi:10.1016/j.ijnurstu.2006 .08.022
Shu-Li Lee et al.
An evolutionary concept analysis. Journal of Advanced Nursing, 37(1), 87Y99. doi:10.1046/j.1365-2648.2002.02061.x Schilling, L. S., Knafl, K. A., & Grey, M. (2006). Changing patterns of self-management in youth with type 1 diabetes. Journal of Pediatric Nursing, 21(6), 412Y424. doi:10.1016/j.pedn.2006.01.034
King, P. S., Berg, C. A., Butner, J., Drew, L. M., Foster, C., Donaldson, D., I Wiebe, D. J. (2012). Longitudinal trajectories of metabolic control across adolescence: Associations with parental involvement, adolescents’ psychosocial maturity, and health care utilization. Journal of Adolescent Health, 50(5), 491Y496. doi:10.1016/j.jadohealth.2011.09.007
Tang, S. M., Chen, S. W., & Wang, R. H. (2013). Establishing a model to assess the effects of school support and self-care behaviors on life satisfaction in adolescents with type 1 diabetes in Taiwan. The Journal of Nursing Research, 21(4), 244Y251. doi:10.1097/jnr.0000000000000008
Lin, S. C. (2005). Social stratification and education. In S. C. Lin (Ed.), Educational sociology (4th ed., pp. 41Y81). Taipei City, Taiwan, ROC: Chuliu. (Original work published in Chinese)
Wilkin, T. J., Voss, L. D., Metcalf, B. S., Mallam, K., Jeffery, A. N., Alba, S., & Murphy, M. J. (2004). Metabolic risk in early childhood: The early bird study. International Journal of Obesity, 28, S64YS69. doi:10.1038/sj.ijo.0802807
Maahs, D. M., West, N. A., Lawrence, J. M., & Mayer-Davis, E. J. (2010). Epidemiology of type 1 diabetes. Endocrinology and Metabolism Clinics of North America, 39(3), 481Y497. doi:10 .1016/j.ecl.2010.05.011
Williams, L. B., Laffel, L. M. B., & Hood, K. K. (2009). Diabetesspecific family conflict and psychological distress in pediatric type 1 diabetes. Diabetes Medicine, 26(9), 908Y914. doi:10.1111/ j.1464-5491.2009.02794.x
Neylon, O. M., O’Connell, M. A., Skinner, T. C., & Cameron, F. J. (2013). Demographic and personal factors associated with metabolic control and self-care in youth with type 1 diabetes: A systematic review. Diabetes/Metabolism Research and Reviews, 29(4), 257Y272. doi:10.1002/dmrr.2392
Wysocki, T., Harris, M. A., Buckloh, L. M., Mertlich, D., Lochrie, A. S., Taylor, A., I White, N. H. (2006). Effects of behavioral family systems therapy for diabetes on adolescents’ family relationships, treatment adherence, and metabolic control. Journal of Pediatric Psychology, 31(9), 928Y938. doi:10.1093/jpepsy/jsj098
Rubin, R. R., Young-Hyman, D., & Peyrot, M. (1989). ParentYchild responsibility and conflict in diabetes care. Diabetes, 38(2, Suppl.), 28A.
Wysocki, T., Nansel, T. R., Holmbeck, G. N., Chen, R., Laffel, L., Anderson, B. J., & Weissberg-Benchell, J. (2009). Collaborative involvement of primary and secondary caregivers: Associations with youths’ diabetes outcomes. Journal of Pediatric Psychology, 34(8), 869Y881. doi:10.1093/jpepsy/jsn136
Schilling, L. S., Grey, M., & Knafl, K. A. (2002). The concept of self-management of type 1 diabetes in children and adolescents:
8 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research h VOL. 00, NO. 0, MONTH 2015
Predictors of Glycemic Control in Adolescents of Various Age Groups With Type 1 Diabetes Shu-Li Lee1 & Fu-Sung Lo2 & Yann-Jinn Lee3 & Bai-Hsiun Chen4 Ruey-Hsia Wang5* 1
MSN, RN, Instructor, College of Nursing, Kaohsiung Medical University, and Doctoral Student, School of Nursing, Kaohsiung Medical University & 2MD, MS, Attending Physician, Division of Endocrinology, Department of Pediatrics, Chang Gung Memorial Hospital & 3MD, Attending Physician, Division of Endocrinology, Department of Pediatrics, Mackay Memorial Hospital & 4MD, PhD, Attending Physician, Department of Pediatrics, Kaohsiung Medical University Hospital & 5PhD, RN, Professor, College of Nursing, Kaohsiung Medical University.
behaviors. In addition, female adolescents aged 16Y18 years should be considered an at-risk group.
ABSTRACT Background: Understanding the predictors of glycemic control in adolescents of various age groups with type 1 diabetes (T1D) is crucial for nurses to cultivate developmental-specific interventions to improve glycemic control in this age group. However, research has rarely addressed this issue, particularly in the context of Asian populations. Purpose: We explored the predictive influence of demographic characteristics, self-care behaviors, family conflict, and parental involvement on glycosylated hemoglobin (HbA1C) levels 6 months after the baseline measurement in adolescents of various age groups with T1D in Taiwan. Methods: A prospective survey design was applied. At baseline, adolescents with T1D completed a self-care behavior scale. Parents or guardians finished scales of parental involvement and family conflict. The HbA1C levels 6 months after baseline measurement were collected from medical records. Two hundred ten adolescentYparent/guardian pairs were enrolled as participants. Multiple stepwise regressions examined the significant predictors of HbA1C levels 6 months after the baseline measurement in the three adolescent age groups: 10Y12, 13Y15, and 16Y18 years. Results: Family conflict was a significant predictor of HbA1C level within the 10Y12 years of age group 6 months after the baseline measurement. Self-care behaviors were a significant predictor of HbA1C level within the 13Y15 years of age group 6 months after the baseline measurement. Being female and self-care behaviors were each significant predictors of HbA1C level in the 16Y18 years of age group 6 months after the baseline measurement. Conclusions/Implications for Practice: Nurses should design specific interventions to improve glycemic control in adolescents of various age groups with T1D that are tailored to their developmental needs. For adolescents with T1D aged 10Y12 years, nurses should actively assess family conflict and provide necessary interventions. For adolescents with T1D aged 13Y18 years, nurses should exert special efforts to improve their self-care
KEY WORDS: type 1 diabetes, glycemic control, adolescent, predictors, age groups.
Introduction Type 1 diabetes (T1D) is one of the most common chronic diseases in children. The incidence of T1D has been increasing annually by 2%Y5% worldwide (Maahs, West, Lawrence, & Mayer-Davis, 2010). Although the incidence and prevalence of T1D in Asia is lower than in Western countries, T1D remains a crucial health concern for Asian children (Maahs et al., 2010). Between 2000 and 2009 in Taiwan, the incidence of T1D in children below 19 years old ranged from 4.21 to 5.79 per 100,000 (Jiang, Chang, Tai, Chen, & Chuang, 2012). Poor glycemic control in children with T1D may cause life-threatening complications such as microvascular and neuropathic complications. Level of glycosylated hemoglobin (HbA1C) is a crucial indicator of glycemic control. The American Diabetes Association (ADA, 2013) recommends that HbA1C levels should be less than 8.0% and 7.5% in children aged 6Y12 and 13Y19 years, respectively. However, earlier studies have indicated that the level of HbA1C in adolescents is higher Accepted for publication: July 23, 2014 *Address correspondence to: Ruey-Hsia Wang, N100, Shih-Chuan 1st Rd., San-Ming District, Kaohsiung City 80708, Taiwan, ROC. Tel: +886 (7) 312Y1101 ext. 2641; E-mail: [email protected] The authors declare no conflicts of interest. Cite this article as: Lee, S. L., Lo, F. S., Lee, Y. J., Chen, B. H., & Wang, R. H. (2015). Predictors of glycemic control in adolescents of various age groups with type 1 diabetes. The Journal of Nursing Research, 00(0), 00Y00. doi:10.1097/jnr.0000000000000096
1 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
than in other age groups (ADA, 2013; Helgeson, Siminerio, Escobar, & Becker, 2009). It is thus vital that nurses understand the key factors of influence on HbA1C levels to improve glycemic control in adolescents with T1D. Demographic characteristics have been associated with glycemic control. Adolescents who were older, were female, had a longer duration of diabetes, were living in non-dualparent families, and were of lower socioeconomic status (SES) were identified as facing a higher risk of poor glycemic control (Neylon, O’Connell, Skinner, & Cameron, 2013). Psychosocial factors are crucial in influencing the glycemic control of adolescents with T1D. Adolescents with TID must perform self-care behaviors, which include insulin injections, blood glucose tests, diet management, appropriate exercise, and hypoglycemia or hyperglycemia management to control their HbA1C levels (ADA, 2013). An integrative literature review indicated that self-care behaviors positively predicted HbA1C levels in children with T1D in longitudinal studies (Guo, Whittemore, & He, 2011). However, poor self-care behaviors in adolescents with T1D have been frequently reported in numerous studies (Guo et al., 2011; Neylon et al., 2013). In Taiwan, adolescents with T1D performed poorly with regard to administering insulin injections, monitoring blood glucose, adhering to meal plans, and engaging in appropriate exercise (Chang, Yeh, Lo, & Shih, 2007). Parents play a fundamental role in facilitating effective glycemic control management in adolescents with T1D because of the complexity of the control process. Most adolescents with T1D thus rely on parental assistance to conduct diabetes-related activities such as monitoring blood glucose and delivering insulin injections (King et al., 2012). A previous cross-sectional study associated less parental assistance with diabetes management and poor glycemic control in adolescents with T1D (Wysocki et al., 2009). Longitudinal studies have also indicated that, when parents share diabetes management responsibilities, there is a positive association with glycemic control in this population (Ingerski, Anderson, Dolan, & Hood, 2010; King et al., 2012). Thus, parental involvement is crucial to the effective control of glycemic levels in adolescents with T1D. Adolescence is a period in life when individuals often seek greater autonomy from their parents, rely less on parental guidance, and challenge parental authority (Garrison & Felice, 2009). These age-related characteristics may result in disagreements between parents and their T1D children with regard to diabetes care responsibilities (Schilling, Knafl, & Grey, 2006). Adolescents with T1D often struggle with relying on parents and making decisions by themselves, which may exacerbate diabetes-specific conflict with parents (Karlsson, Arman, & Wikblad, 2008). A cross-sectional study conducted in a Western population indicated that increased diabetes-specific family conflict resulted in higher HbA1C levels in adolescents with T1D (Williams, Laffel, & Hood, 2009). A longitudinal study also found that higher diabetes-specific family conflict predicted worse adherence to regimens in adolescents with T1D, which negatively influenced glycemic control (Hilliard, Guilfoyle, Dolam, & Hood, 2011). However, the influence of parental
Shu-Li Lee et al.
involvement and family conflict on HbA1C levels in adolescents with T1D has been less frequently reported in Taiwan. Although numerous studies have associated demographic characteristics, self-care behaviors, parental involvement, and family conflict with HbA1C levels in adolescents with TID, certain gaps in this research await further study and clarification. The first of these gaps is highlighted by the fact that adolescence is frequently subdivided into three developmental stages: early adolescence (10Y13 years), middle adolescence (14Y16 years), and late adolescence (17Y20 years; Garrison & Felice, 2009). Adolescents experience cognitive and psychosocial development phenomena that are distinct to their specific stage (Allender, 2002). Early adolescence is a transitional period for adolescents in which they work to establish independence from their parents (Garrison & Felice, 2009). Adolescents in the middle stage tend to strive for greater autonomy and challenge the authority of parents and other adults (Allender, 2002; Garrison & Felice, 2009). During late adolescence, adolescents are focused on the future and experience new behaviors (Garrison & Felice, 2009). Different developmental characteristics may cause different sets of factors to influence HbA1C levels in each of the three developmental stages. However, numerous previous studies of adolescents with T1D have treated adolescence as a single stage (frequently between 10 and 18 or 13 and 18 years old) rather than further distinguishing adolescence into a series of distinct developmental stages (Chang et al., 2007; Ingerski et al., 2010). The second of these gaps is that most of the previous studies of this issue used cross-sectional designs, which are unable to provide information regarding causal relationships between associated factors and HbA1C levels. The third and final of these gaps is that most of the previous studies of this issue were conducted in Western populations with a relatively high prevalence of T1D. The predictors of glycemic control in adolescents with T1D in Taiwan may differ from those identified for Western populations. It is necessary to clarify the predictors of HbA1C level in adolescents in different age groups with T1D so that nurses may use this knowledge to cultivate developmental-specific interventions for adolescents with T1D. HbA1C levels reveal glycemic control efficacy over the previous 3-month period. Following HbA1C levels 3 months after the baseline measurement may inadequately reflect the influence of the target predictors. However, clarifying the predictors of HbA1C level 6 months after the baseline measurement may effectively manifest the causal relationships between predictors and HbA1C levels. Therefore, the purpose of this study is to explore significant predictors of HbA1c level 6 months after the baseline measurement based on demographic characteristics, self-care behaviors, parental involvement, and family conflict in adolescents of different age groups with T1D. We hypothesize that demographic characteristics, self-care behaviors, parental involvement, and family conflict should significantly influence HbA1C levels at 6 months after the baseline measurement and that this influence should be variegated among the three adolescent age groups.
2 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
Predictors of Glycemic Control in Adolescents
Methods Design This study used a prospective design. At baseline, data on self-care behaviors were collected from the adolescent participants. Considering that parent or guardian responses may be more accurate than those of children, data of adolescent demographic characteristics, diabetes-specific parental involvement, and diabetes-specific family conflict were collected from the parent or guardian participants. HbA1c levels were collected 6 months after the baseline measurement.
Participants Adolescents aged 10Y18 years and experiencing T1D for at least 6 months were recruited from two medical centers in Northern Taiwan. Adolescents who were pregnant or had a severe learning disability or a significant medical disease other than T1D were excluded. Endocrine physicians referred adolescents who met the inclusion criteria to a research assistant when these adolescents visited the clinic. Two hundred twenty-nine adolescentYparent/guardian pairs were invited to participate in this study. Two hundred ten adolescentYparent/ guardian pairs (response rate: 91.70%) enrolled as participants. In Taiwan, adolescents aged 10Y12, 13Y15, and 16Y18 years are enrolled in elementary, middle, and high school, respectively. We grouped adolescents into 10- to 12-, 13- to 15-, and 16- to 18-year age groups to represent early, middle, and late adolescence, respectively. Of the 210 adolescents participating in this study, 79 (37.6%), 80 (38.1%), and 51 (24.3%) were in the early, middle, and late adolescence groups, respectively.
Measurement Structured, anonymous questionnaires and the medical records for the adolescent participants were used to collect data.
Questionnaire Adolescents and parents or guardians completed the questionnaires independently. Each questionnaire took approximately 15 minutes to complete.
Demographic characteristics The questionnaire given to the parent or guardian participants collected data on adolescents’ age, gender, duration of diabetes, family structure, educational level (scored from 1 to 5), and occupation level of parents (scored from 1 to 5). Family structure was divided into two possible formats: dualparent family and non-dual-parent family (single-parent or no-parent family). Educational level was multiplied by 4, and occupation level was multiplied by 7 to generate SES scores. The highest SES score for the parents was family SES. Scores ranging from 41 to 55, 30 to 40, and 11 to 29 were coded high, middle, and low SES, respectively (Lin, 2005). The high and middle SES scores were grouped together as ‘‘higher SES’’ for further study.
VOL. 00, NO. 0, MONTH 2015
Diabetes-specific self-care behavior scale for adolescents The nine-item scale developed by Tang, Chen, and Wang (2013) was used to measure the degree to which blood glucose was monitored, insulin doses were adjusted, diet was regulated, exercise was modified, and hypoglycemia or hyperglycemia was managed. Each item was rated from 0 = always to 4 = never. The total scale ranged from 0 to 36. Higher scores indicated enhanced self-care behaviors. In this study, the Cronbach’s " was .70. The intraclass correlation coefficient indicated that testYretest reliability was .72. Diabetes-specific parental involvement scale for parents We adapted the 15-item English version of the Diabetes Responsibility and Conflict Scale (DRCS; Rubin, YoungHyman, & Peyrot, 1989) for use in this study. The Englishversion DRCS was developed to separately measure the relative degree of involvement and conflict in diabetes-specific tasks between parents or guardians and adolescents. The original Cronbach’s " of the parental involvement subscale and the family conflict subscale in the 15-item English-version DRCS was .79 and .88, respectively (Rubin et al., 1989). In this study, the 15-item DRCS was translated from English into Chinese by a bilingual expert and then back-translated from Chinese into English by another bilingual expert. A third bilingual expert confirmed that the Chinese and English versions of the DRCS were identical. The research team deleted three items including ‘‘Who tests urine,’’ ‘‘Who talks to adults about your diabetes,’’ and ‘‘Who decides which people know that you are diabetic’’ because of the unsuitability of these issues within the context of Taiwanese culture. Finally, 12 diabetesspecific tasks were used to measure diabetes-specific parental involvement. The example item was ‘‘Who decides what to eat at meals.’’ Each item was rated from 1 = child, all of the time to 5 = parent, all of the time. The range of possible total scores ranged from 12 to 60, with higher scores indicating higher parental involvement. In this study, the Cronbach’s " and intraclass correlation coefficient were .88 and .80, respectively. Diabetes-specific family conflict scale for parents The same diabetes-specific tasks measured in the 12-item diabetesspecific parental involvement scale for parents were also used to assess the frequency of conflict between parents or guardians and adolescents regarding tasks of diabetes-specific care. The question was ‘‘How much you argue/disagree with your child about each of the following aspects of your child’s diabetes care?’’ Each item was rated from 1 = never to 5 = always. The possible total scores ranged from 12 to 60, with higher scores indicating higher family conflict. In this study, the Cronbach’s " and intraclass correlation coefficient were .94 and .78, respectively.
Medical Records A research assistant collected the HbA1C levels of the adolescent participants from medical records 6 months after the baseline measurement. 3
Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
Shu-Li Lee et al.
Validity and Reliability
Statistical Analysis
Seven experts specializing in pediatric diabetes independently examined each scale. The experts rated each item from 1 = not relevant to 4 = very relevant. The number of items rated 3 or 4 were divided by the total number of items in each scale to calculate the content validity index for each scale. The content validity index of each scale ranged from 0.90 to 1.00. Data for all participants were used to examine the Cronbach’s " of each scale. TestYretest reliability was assessed by administering questionnaires to 21 adolescentYparent/guardian pairs over a 2-week interval. The intraclass correlation coefficient was used to calculate testYretest reliability.
Data were analyzed using SPSS software version 19.0 (IBM, Armonk, NY, USA). Descriptive statistics were used to describe all variables. Chi-squared tests, two sample t tests, one-way analysis of variance with Scheffe post hoc tests, and Pearson correlations were used to examine the relationships between demographic characteristics, psychosocial factors, and HbA1C levels. A multiple stepwise regression analysis was used to examine the significant predictors of HbA1C level 6 months after baseline measurement for the three age groups.
Ethical Consideration This study was approved by the institutional review board of a medical university. We informed adolescents and their parents or guardians that refusing to participate in this study would incur no penalty and that they had the right to withdraw from the study at any time. All adolescents and their parents or guardians were required to sign informed assent and consent forms before they participated in this study. The parents or guardians also signed a consent form for their children to participate in the study.
Results Distribution of Demographic Characteristics, Psychosocial Factors, and HbA1C Levels Among the Three Age Groups The mean ages within the 10Y12, 13Y15, and 16Y18 years of age groups were 11.63 (SD = 0.97), 14.54 (SD = 0.82), and 16.83 (SD = 0.63) years, respectively. Table 1 showed that
TABLE 1.
Distribution and Comparison of Demographic Characteristics, Psychosocial Factors, and HbA1c Levels Among the Three Age Groups ffi 10Y12 Year (n = 79) n
Characteristic Demographic characteristics Gender Male Female Duration of diabetes (years): M T SD
%
ffl 13Y15 Year (n = 80) n
%
16Y18 Year (n = 51) n
%
2 2 = 0.44 38 41 5.73
48.10 51.90 3.35
39 41
48.75 51.25
6.40
3.85
22 29 8.28
43.14 56.86 4.45
9 ffia Family structure Dual-parent family Non-dual-parent family SES Higher Lower
Psychosocial factors Self-care behaviors Parental involvement Family conflict HbA1c
2 2/F
73 6
92.40 7.60
76 4
F = 6.99** 2 2 = 3.23
95.00 5.00
44 7
86.27 13.73 2 2 = 1.13
47 32
59.49 40.51
41 39
51.25 48.75
29 22
56.86 43.14
M
SD
M
SD
M
SD
26.15
4.50
25.11
4.58
23.35
5.08
F = 5.56**
9.13
ffi 9 a 31.23 8.22
26.31
7.11
F = 32.10***
23.59
11.75
ffi 9 ffl 9 a 21.66 9.04
22.27
11.16
7.99
1.79
9.07
2.01
37.99
8.54
2.04
9 ffia Note. SES = socioeconomic status. a Scheffe post hoc analysis. **p G .01. ***p G .001.
4 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
F = 0.68 F = 4.94***
Predictors of Glycemic Control in Adolescents
VOL. 00, NO. 0, MONTH 2015
gender, family structure, and SES did not differ significantly among the age groups. Self-care behaviors decreased with increasing age, with the self-care behaviors of the 16Y18 years of age group significantly worse than that of the 10Y12 years of age group. Parental involvement significantly decreased with increasing age. However, family conflict did not significantly differ among the three age groups. The mean HbA1C level for all participants was 8.5% (SD = 1.93%). Table 1 shows that the mean HbA1c levels for the three age groups differed significantly. Scheffe post hoc analysis indicated that the HbA1C level for the 16Y18 years of age group was significantly higher than the level for the 10Y12 years of age group.
The Association Between Demographic Characteristics and Psychosocial Factors and HbA1c Levels 6 Months After Baseline Measurement for Each Age Group Table 2 shows that gender and SES were not significantly associated with HbA1C level in the 10Y12 or 13Y15 years of age groups. The HbA1C levels of female adolescents and of participants with lower SES were significantly higher than those of male adolescents and of participants with higher SES within the 16Y18 years of age group. Family structure and duration of diabetes were not significantly associated with HbA1c levels in any age group. Regarding psychosocial factors, self-care behaviors were significantly associated with HbA1C levels in the 13Y15 and 16Y18 years of age groups. Family conflict was significantly
associated with HbA1C levels in the 10Y12 years of age group. Parental involvement was not significantly associated with HbA1C levels in any age group.
Predictors of HbA1C Level 6 Months After Baseline Measurement in Each Age Group A multiple stepwise regression was used to examine the significant predictors of HbA1C level 6 months after baseline measurement in each age group. Gender, duration of diabetes, family structure, SES, self-care behaviors, parental involvement, and family conflict were used as the independent variables. All variance inflation factors and condition indices were less than 1.05 and 2.69, respectively, indicating no collinearity among the independent variables within each age group. Table 3 showed that family conflict was a significant predictor of HbA1C levels 6 months after baseline measurement in the 10Y12 years of age group, which explained 20.0% of the total variance in HbA1C levels. Greater family conflict at baseline predicted higher HbA1C levels 6 months after baseline measurement. Self-care behaviors was a significant predictor of HbA1C levels 6 months after baseline measurement in the 13Y15 years of age group, which explained 12.0% of the total variance in HbA1C levels. Enhanced self-care behaviors at baseline predicted lower HbA1C levels 6 months after baseline measurement. Female and self-care behaviors were significant predictors of HbA1C level 6 months after baseline measurement in the 16Y18 years of age group, which explained 28.0% of the total variance in
TABLE 2.
Comparison of the Effect of Demographic Characteristics and Psychosocial Factors on HbA1c Levels 6 Months After Baseline Measurement in Each Age Group 10Y12 Year (n = 79) Variable Demographic characteristics Gender Male Female Duration of diabetes (years) Family structure Dual-parent family Non-dual-parent family SES Higher Lower Psychosocial factors Self-care behaviors Parental involvement Family conflict
M
SD
7.71 8.25
1.62 1.92
t/r
13Y15 Year (n = 80) M
SD
8.84 8.24
2.09 1.96
t = 1.35
1.56 3.51
7.67 8.45
1.48 2.11
SD
8.22 9.72
1.50 2.13
8.51 9.03
2.06 1.74
8.17 8.92
1.62 2.36
t/r t = 2.81**
r = .03 t = 0.49
t = 1.93
r = j .03 r = .18 r = .45***
M
t = 1.32
r = .16 t = 1.07 7.87 9.42
t/r
16Y18 Year (n = 51)
r = .15 t = 1.45 8.91 10.09
2.03 1.39
8.58 9.72
1.67 2.28
t = 1.65
t = 2.07*
r = .34** r = .03 r = .16
r = .34* r = .16 r = .19
Note. SES = socioeconomic status. *p G .05. **p G .01. ***p G .001.
5 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
Shu-Li Lee et al.
TABLE 3.
Predictors of HbA1c Level for Each Age Group Age Group
Predictor
10Y12 Year
Family conflict
13Y15 Year
Self-care behaviors
16Y18 Year
Femalea Self-care behaviors
$
Total R 2
t
.45
.20
j.34
.12
j3.23**
4.40***
.41
.14
3.31**
j.37
.28
j3.04**
95% CI [ 0.04,
0.10]
[j0.25, j0.06] [ 0.65,
2.64]
[j0.25, j0.05]
Note. CI = confidence interval. a Reference group: Male. *p G .05. **p G .01. ***p G .001.
HbA1c levels. Both female and self-care behaviors individually accounted for 14.0% of the variance in HbA1C levels. Female and worse self-care behaviors at baseline predicted higher HbA1C levels 6 months after baseline measurement.
Discussion As hypothesized, the significant predictors of HbA1C level in adolescents with T1D were different for the three different age ranges examined. Family conflict was stable across the three age groups in this study, which is consistent with a previous study (Ingerski et al., 2010). As noted, few studies have examined the relationships between family conflict and HbA1c levels in adolescents with T1D at different adolescent age groups. In the United States, parent-reported family conflict was found to predict HbA1C levels 6 months after baseline measurement for adolescents with T1D aged 13Y18 years (Ingerski et al., 2010). In this study, parent-reported family conflict was not a significant predictor of HbA1c levels 6 months after baseline measurement for adolescents with T1D in the 13Y15 or 16Y18 years of age groups. However, family conflict was a significant predictor of HbA1C levels 6 months after baseline measurement in the 10Y12 years of age group. The cause of this divergent finding may be differences in the cultures targeted by the studies. While adolescents aged 10Y12 years begin to struggle for increased independence (Garrison & Felice, 2009), Chinese parents continue to exert more control than Western parents over their children’s lives (Jose, Huntsinger, Huntsinger, & Liaw, 2000). Chinese parents may unconsciously pressure their children when they perceive conflict with children. Therefore, this found that parentreported family conflict was more important than self-care behaviors in predicting HbA1C levels 6 months after baseline measurement for adolescents aged 10Y12 years. Improving communication and problem-solving skills may effectively reduce family conflict (Wysocki et al., 2006). For adolescents aged 10Y12 years with T1D, it is thus important that nurses actively assess family conflict between parents and adolescents to help these adolescents effectively communicate with their parents to solve diabetes-care-related disagreements. Although parental involvement significantly decreased with increasing age, parental involvement was not signifi-
cantly associated with HbA1c levels 6 months after baseline measurement in any of the age groups. The decrease of parental involvement did not necessarily result in poor glycemic control 6 months after baseline measurement in Taiwan. A previous study indicated that perceived parental overinvolvement in diabetes care negatively influenced HbA1c levels in adolescents with T1D (Cameron et al., 2008). Perception of parental overinvolvement may be more crucial than the degree of parental involvement in predicting glycemic control of adolescents with T1D in Taiwan. However, further studies are required to confirm this hypothesis. Self-care behaviors were a significant predictor of HbA1c levels 6 months after baseline measurement for adolescents aged 13Y15 and 16Y18 years. This result was consistent with an integrative review, which concluded that self-care behaviors were specifically crucial for glycemic control among adolescents older than 13 years (Guo et al., 2011). Although middle-aged adolescents pursue greater autonomy (Garrison & Felice, 2009), increasing autonomy does not indicate increased capacity to perform self-care behaviors effectively. For example, only 31% of adolescents aged 16 years were comfortable making decisions on dietary control (Clarke, 2011). A previous study found that self-care patterns were parental dominant, transitional, and adolescent dominant in early, middle, and late adolescence, respectively (Schilling, Grey, & Knafl, 2002). The influence of self-care behaviors on HbA1c levels in adolescents may increase as they assume increasing control on their own diabetes care regimen. Parental encouragement and collaboration with adolescents are crucial for improving the self-care behaviors of this population (Wysocki et al., 2006, 2009). For adolescents aged 13Y18 years, nurses may serve as a coordinator between adolescents and parents to assist adolescents achieve enhanced self-care behaviors. This study found HbA1c levels of female adolescents in the 16Y18 years of age group to be significantly higher than those of male adolescents in the same group. Furthermore, being female accounted for 14% of the total variance in HbA1c levels in the 16Y18 years of age group. Insulin sensitivity decreases in adolescents aged 17Y19 years, with the decrease particularly significant in women (Wilkin et al., 2004). A previous study in Taiwan observed that disturbed eating behaviors contributed to higher HbA1c levels in female adolescents with T1D than in their male counterparts (Hsu,
6 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
Predictors of Glycemic Control in Adolescents
Chen, Huang, Lin, & Lin, 2009). Decreased insulin sensitivity and increased disturbed eating behaviors among female adolescents may support ‘‘being female’’ as a crucial predictor of HbA1c levels in the 16Y18 years of age group. Therefore, female adolescents aged 16Y18 years should be considered an at-risk subgroup of adolescents with T1D. In this study, HbA1c levels in the 16Y18 years of age group were the highest of the three age groups. Self-care behaviors within the 16Y18 years of age group were the poorest among the three age groups. Thus, we propose that poor self-care behaviors are also an important contributing factor in this group. This study explored the influence of individual and family factors on HbA1c levels in adolescents with T1D for three distinct adolescent age groups. However, adolescents experience a strong need for peer acceptance (Garrison & Felice, 2009). Peer relationships have been shown to predict HbA1c levels 4 years after baseline measurement for adolescents with T1D (Helgeson et al., 2009). Depression has also been shown to predict high HbA1C levels in adolescents with T1D (Hilliard, Wu, Rausch, Dolan, & Hood, 2013). We suggest that future studies assess the influence of peer relationships and depression on glycemic control in adolescents of various age groups with T1D.
Limitations The small sample size used in this study was collected from only two medical centers in Northern Taiwan. Therefore, the generalizability of findings may be limited. We suggest that further studies recruit samples from diverse hospitals. Parent and adolescent perspectives of family conflict and parental involvement may influence HbA1c levels in differing ways (Guo et al., 2011). Further studies may measure family conflict and parental involvement from both parent and adolescent perspectives to clarify the relative importance of parent or adolescent perspectives in predicting HbA1c levels in the Taiwanese context. We did not follow the same participants across distinct age groups. A longitudinal design may be designed to follow changes in self-care behaviors, family conflict, parental involvement, and HbA1C levels in the same adolescents across multiple age groups. This design could examine the influence of the dynamic change in these factors on HbA1c levels across time.
Conclusions Significant predictors of HbA1c level 6 months after the baseline measurement in adolescents with T1D differed for the three different age groups. Therefore, our results confirm that intervention programs designed to improve glycemic control of adolescents with T1D should be tailored to address the factors of influence most relevant to the target age group. For adolescents with T1D aged 10Y12 years, nurses should actively assess family conflict and provide necessary intervention. For adolescents with T1D aged 13Y18 years, nurses should focus on improving patients’ self-care behaviors. Finally, female adolescents aged 16Y18 years should be considered an at-risk subgroup within adolescents with T1D.
VOL. 00, NO. 0, MONTH 2015
References Allender, M. (2002). Adolescent. In C. L. Edelman & C. L. Mandle (Eds.), Health promotion throughout the life span (5th ed., pp. 619Y647). St. Louis, MO: Mosby. American Diabetes Association. (2013). Standards of medical care in diabetesV2013. Diabetes Care, 36(1), s11Ys66. doi:10 .2337/dc13-S011 Cameron, F. J., Skinner, T. C., de Beaufort, C. E., Hoey, H., Swift, P. G. F., Aanstoot, H., I Skovlund, S. E. (2008). Are family factors universally related to metabolic outcomes in adolescents with type 1 diabetes? Diabetic Medicine, 25(4), 463Y468. doi:10.1111/j.1464-5491.2008.02399.x Chang, C. W., Yeh, C. H., Lo, F. S., & Shih, Y. L. (2007). Adherence behaviours in Taiwanese children and adolescents with type 1 diabetes mellitus. Journal of Clinical Nursing, 16(7b), 207Y214. doi:10.1111/j.1365-2702.2005.01541.x Clarke, W. L. (2011). Behavioral challenges in the management of childhood diabetes. Journal of Diabetes Science and Technology, 5(2), 225Y228. doi:10.1177/193229681100500203 Garrison, W., & Felice, M. E. (2009). Adolescence. In W. B. Carey, A. C. Crocker, W. L. Coleman, E. R. Elias, & H. M. Feldman (Eds.), DevelopmentalYbehavioral pediatrics (4th ed., pp. 62Y73). Philadelphia, PA: Saunders. Guo, J., Whittemore, R., & He, G. P. (2011). The relationship between diabetes self-management and metabolic control in youth with type 1 diabetes: An integrative review. Journal of Advanced Nursing, 67(11), 2294Y2310. doi:10.1111/j.13652648.2011.05697.x Helgeson, V. S., Siminerio, L., Escobar, O., & Becker, D. (2009). Predictors of metabolic control among adolescents with diabetes: A 4-year longitudinal study. Journal of Pediatric Psychology, 34(3), 254Y270. doi:10.1093/jpepsy/jsn079 Hilliard, M. E., Guilfoyle, S. M., Dolam, L. M., & Hood, K. K. (2011). Prediction of adolescents’ glycemic control 1 year after diabetes-specific family conflict. Archives of Pediatrics & Adolescent Medicine, 165(7), 624Y629. doi:10.1001/archpediatrics.2011.86 Hilliard, M. E., Wu, Y. P., Rausch, J., Dolan, L. M., & Hood, K. K. (2013). Predictors of deteriorations in diabetes management and control in adolescents with type 1 diabetes. The Journal of Adolescent Health, 52(1), 28Y34. doi:10.1016/j.jadohealth.2012.05.009 Hsu, Y. Y. A., Chen, B. H., Huang, M. C., Lin, S. J., & Lin, M. F. (2009). Disturbed eating behaviors in Taiwanese adolescents with type 1 diabetes mellitus: A comparative study. Pediatric Diabetes, 10(1), 74Y81. doi:10.1111/j.1399-5448.2008.00422.x Ingerski, L. M., Anderson, B. J., Dolan, L. M., & Hood, K. K. (2010). Blood glucose monitoring and glycemic control in adolescence: Contribution of diabetes-specific responsibility and family conflict. Journal Adolescent Health, 47(2), 191Y197. doi:10.1016/j.jadohealth.2010.01.012 Jiang, Y. D., Chang, C. H., Tai, T. Y., Chen, J. F., & Chuang, L. M. (2012). Incidence and prevalence rates of diabetes mellitus in Taiwan: Analysis of the 2000Y2009 nationwide health insurance database. Journal of the Formosan Medical Association, 111(11), 599Y604. doi:10.1016/j.jfma.2012.09.014 Jose, P. E., Huntsinger, C. S., Huntsinger, P. R., & Liaw, F. R. (2000). Parental values and practices relevant to young children’s social development in Taiwan and the United States. Journal of Cross-Cultural Psychology, 31(6), 667Y702. doi:10 .1177/0022022100031006002
7 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
The Journal of Nursing Research
Karlsson, A., Arman, M., & Wikblad, K. (2008). Teenagers with type 1 diabetesVA phenomenological study of the transition towards autonomy in self-management. International Journal of Nursing Studies, 45(4), 562Y570. doi:10.1016/j.ijnurstu.2006 .08.022
Shu-Li Lee et al.
An evolutionary concept analysis. Journal of Advanced Nursing, 37(1), 87Y99. doi:10.1046/j.1365-2648.2002.02061.x Schilling, L. S., Knafl, K. A., & Grey, M. (2006). Changing patterns of self-management in youth with type 1 diabetes. Journal of Pediatric Nursing, 21(6), 412Y424. doi:10.1016/j.pedn.2006.01.034
King, P. S., Berg, C. A., Butner, J., Drew, L. M., Foster, C., Donaldson, D., I Wiebe, D. J. (2012). Longitudinal trajectories of metabolic control across adolescence: Associations with parental involvement, adolescents’ psychosocial maturity, and health care utilization. Journal of Adolescent Health, 50(5), 491Y496. doi:10.1016/j.jadohealth.2011.09.007
Tang, S. M., Chen, S. W., & Wang, R. H. (2013). Establishing a model to assess the effects of school support and self-care behaviors on life satisfaction in adolescents with type 1 diabetes in Taiwan. The Journal of Nursing Research, 21(4), 244Y251. doi:10.1097/jnr.0000000000000008
Lin, S. C. (2005). Social stratification and education. In S. C. Lin (Ed.), Educational sociology (4th ed., pp. 41Y81). Taipei City, Taiwan, ROC: Chuliu. (Original work published in Chinese)
Wilkin, T. J., Voss, L. D., Metcalf, B. S., Mallam, K., Jeffery, A. N., Alba, S., & Murphy, M. J. (2004). Metabolic risk in early childhood: The early bird study. International Journal of Obesity, 28, S64YS69. doi:10.1038/sj.ijo.0802807
Maahs, D. M., West, N. A., Lawrence, J. M., & Mayer-Davis, E. J. (2010). Epidemiology of type 1 diabetes. Endocrinology and Metabolism Clinics of North America, 39(3), 481Y497. doi:10 .1016/j.ecl.2010.05.011
Williams, L. B., Laffel, L. M. B., & Hood, K. K. (2009). Diabetesspecific family conflict and psychological distress in pediatric type 1 diabetes. Diabetes Medicine, 26(9), 908Y914. doi:10.1111/ j.1464-5491.2009.02794.x
Neylon, O. M., O’Connell, M. A., Skinner, T. C., & Cameron, F. J. (2013). Demographic and personal factors associated with metabolic control and self-care in youth with type 1 diabetes: A systematic review. Diabetes/Metabolism Research and Reviews, 29(4), 257Y272. doi:10.1002/dmrr.2392
Wysocki, T., Harris, M. A., Buckloh, L. M., Mertlich, D., Lochrie, A. S., Taylor, A., I White, N. H. (2006). Effects of behavioral family systems therapy for diabetes on adolescents’ family relationships, treatment adherence, and metabolic control. Journal of Pediatric Psychology, 31(9), 928Y938. doi:10.1093/jpepsy/jsj098
Rubin, R. R., Young-Hyman, D., & Peyrot, M. (1989). ParentYchild responsibility and conflict in diabetes care. Diabetes, 38(2, Suppl.), 28A.
Wysocki, T., Nansel, T. R., Holmbeck, G. N., Chen, R., Laffel, L., Anderson, B. J., & Weissberg-Benchell, J. (2009). Collaborative involvement of primary and secondary caregivers: Associations with youths’ diabetes outcomes. Journal of Pediatric Psychology, 34(8), 869Y881. doi:10.1093/jpepsy/jsn136
Schilling, L. S., Grey, M., & Knafl, K. A. (2002). The concept of self-management of type 1 diabetes in children and adolescents:
8 Copyright © 2015 Taiwan Nurses Association. Unauthorized reproduction of this article is prohibited.
