Associations between impulsivity and self-care adherence in individuals diagnosed with Type 2 or prediabetes.

Diabetes is a chronic disease requiring extensive self-care. Different impulsivity constructs, including choice-based and self-report personality measures are associated with decreasing diabetes self-care adherence. However, both choice-based and self-report impulsivity have never been measured for individuals diagnosed with either Type 2 or prediabetes in the same study. The current study examined the relationship between impulsivity and diabetes self-care in 101 adults diagnosed with either Type 2 or prediabetes. Results indicated that increasing self-reported impulsiveness was significantly correlated with decreasing Type 2 diabetic self-care, whereas the choice-based measure was not associated with any self-care measure. No association between impulsivity and self-care was significant for individuals diagnosed with prediabetes. Path analyses showed that self-reported impulsiveness directly and positively predicted problems controlling blood sugar levels in individuals diagnosed with either prediabetes or Type 2 diabetes. However, self-reported impulsiveness only indirectly and negatively predicted exercise and diet adherence via diabetes management self-efficacy for individuals diagnosed with Type 2 diabetes. These results show what specific impulsivity constructs and diabetes management self-efficacy may be incorporated into interventions for increasing specific self-care behaviors.

Introduction

Diabetes is a chronic condition, impacting an estimated 422 million individuals worldwide (~17% global population), whose pancreas either cannot produce any or enough insulin leading to multiple health complications such as hypoglycemia, eye problems, limb amputations, hypertension, increased heart attack and stroke risk, and kidney disease [1]. Type 1 diabetes is congenital and accounts for about 5% of all diabetes diagnoses, whereas Type 2 diabetes is not congenital and accounts for > 90% of all diabetes diagnoses [1]. Prediabetes, where blood sugar levels are elevated for an extended time [2] is a Type 2 diabetes precursor [3] and impacts an even larger number of individuals (33.9% of US adults; [2]). Both Type 2 and prediabetes, are largely preventable through a balanced diet and periodic exercise [1]. Thus, behavioral interventions (e.g., blood glucose monitoring, diet and exercise adherence) are important for reducing and maintaining normal blood sugar levels [4], in addition to pharmacological treatments [1]. The current study focused on individuals diagnosed with Type 2 and prediabetes due to the high prevalence rates and preventability described above. However, self-care plan complexity (e.g., changing diet, increasing or starting exercise, medication regiments), can result in low treatment adherence rates [5]. That is, both behavioral and pharmacological interventions contain adherence requirements that can decrease compliance [5].

Self-efficacy, an individual’s judgment about their capabilities to organize and execute actions, is a significant predictor for diabetes self-care adherence [4, 6]. Those individuals that self-report higher diabetes management self-efficacy (e.g., to what extend do you feel able to choose the correct foods) are more likely to adhere to behavioral and pharmacological interventions [4, 6]. However, difficulty following through with self-care behaviors may be multiply determined. For example, different impulsivity constructs have been linked to multiple detrimental health behaviors (e.g., obesity, substance abuse, [7, 8]). One aspect of impulsivity, delay discounting (DD), is a behavior gratification measure and involves a series of choices used to assess an individual’s preference between smaller-sooner and larger-later options [9]. For diabetes, increased DD may result in individuals engaging in behaviors that have short-term reinforcement properties (e.g., drinking products high in sugar), at the expense of long-term treatment and management goals (i.e., reducing blood sugar levels). As described below, other aspects of impulsivity, such as a lack of inhibition, also have been shown to predict diabetes self-care adherence. However, a lack of inhibition (as measured by the Barratt Impulsiveness Scale-11 [BIS-11]), typically does not significantly correlate with delay of gratification (as measured by DD; [10]). Thus, these impulsivity constructs will be introduced separately in regard to their association with self-efficacy and diabetes self-care adherence.

One group of researchers has examined the relationship between DD and Type 2 diabetes self-management. Reach and colleagues [11] found that increased DD was associated with non-adherence to medication and increased HbA1c (average glycated hemoglobin–a biologic marker for increased blood sugar) levels in adults. Higher HbA1c levels typically reflect worse diabetes management. During a follow-up, Lebeau et al. [12] used a DD task with delays ranging from 3 days to 10 years. Results continued to show a positive association between DD and HbA1c levels. Medication adherence partially mediated the relationship between DD and HbA1c levels. However, they found no relationship between DD rates and diet adherence.

Another study examined the relationship between DD and non-adherence and poor glycemic control within a sample of adolescents with Type 1 diabetes [13]. DD was measured using a computerized procedure with delays of 1 day, 1 week, 1 month, and 6 months and a larger delay reward of $1000. Lansing et al. [13] found that increased DD was associated with worse diabetes self-care in adolescents. The relationship between DD and glycemic management was moderated by direct parent observation of care, suggesting a potential mode of intervention. DD was not related to self-monitoring frequency for blood glucose levels.

More recently researchers measured the relationship between glycemic control, medication adherence and DD in adults diagnosed with prediabetes [14]. Like the results for individuals diagnosed with Type 2 diabetes [11, 12], individuals diagnosed with prediabetes showed positive correlations between DD rates and HbA1c levels. Additionally, there were significant negative relationships between DD rates and medication adherence, diet quality and physical activity. Thus, researchers have consistently found a positive relationship between DD rates and glycemic management across different diabetes diagnoses (Type 1, Type 2, prediabetes) and ages (adolescents and adults). In sum, the results from these studies suggest that impulsivity measured via the DD task are associated with poorer diabetes self-care adherence. However, the findings are less consistent for behaviors related to glycemic control (e.g., diet adherence) for adults.

The BIS-11 [15] is frequently used as a self-report measure of the inhibition construct for impulsivity and is comprised of three subscales: Attentional, Motor, and Nonplanning. Theoretically, all three subscales should be related to diabetes self-care, as adhering to a diabetes treatment regimen requires attention to detail, forethought, and planning. One study examined the relationship between the BIS-11 and body mass index (BMI) in a sample diagnosed with Type 2 diabetes. Raymond and Lovell [16] found that the BIS-11 was positively related to BMI, with food addiction and nonplanning impulsiveness accounting for 38% of the variance in BMI. Although this study did not assess diabetes self-care adherence behavior or glycemic management, the results provide evidence that other impulsiveness measures are important predictors for BMI among adults diagnosed with Type 2 diabetes. Elevated BMI is a risk factor for Type 2 and prediabetes diagnoses [17], making a relationship between impulsiveness and BMI important to consider.

More recently, Hadj-Abo et al. [18] assessed the relationship between self-reported HbA1c levels, diabetes self-management, impulsivity as measure through the BIS-11, diabetes-specific self-efficacy, and need for cognition in individuals diagnosed with Type 2 diabetes. Like the DD studies described above [11, 12, 14], self-reported impulsivity was significantly negatively associated with self-reported HbA1c levels. Diabetes self-management was also significantly negatively related to BIS-11 score; as impulsivity increased, diabetes self-management decreased. An additional mediation analysis showed that self-efficacy fully mediated this relationship, whereby the effect of impulsivity on diabetes self-management was dependent on self-efficacy. This result was like other important health behaviors that have also measured self-efficacy and self-management [4, 6]. However, unlike previous studies, diabetes self-care management was not broken down into the component subscales of self-care (e.g., diet, physical activity, blood sugar control, and foot care) and analyzed separately. This is important as previous studies have shown differential relationships between impulsivity and self-care management [12-14].

Given the large number of individuals diagnosed with diabetes [1] and prediabetes [3], identifying behavioral processes that influence adherence has potential for improving health and preventing Type 2 diabetes in individuals already diagnosed with prediabetes. No study focusing on impulsivity and self-care behavior has included individuals diagnosed with prediabetes and used different impulsivity measures in the same sample. Previous studies have demonstrated that DD and the BIS-11 measure different impulsivity constructs [10, 19, 20]. Therefore, if each impulsivity construct (i.e., DD and BIS-11) is differentially related to self-care adherence in these populations, then only that impulsivity construct would be an additional target for an intervention to improve or reverse negative health outcomes.

From the previously described literature, it was hypothesized that individuals with higher impulsivity will have fewer numbers of days adhering to self-care activities and more management problems. Higher self-efficacy would be associated with increased self-care and management understanding, as well as fewer management problems and perceived barriers to care. Self-efficacy would mediate the relationship between impulsivity and diabetes self-care, similar to previous research [18]. Lastly, differences in impulsivity (DD and BIS-11), self-efficacy, and self-care between individuals diagnosed with Type 2 diabetes compared to those diagnosed with prediabetes were also tested for. Given the lack of research comparing these two groups, the direction of these differences was not specified a priori.

Materials and methods

Participants

With Institutional Review Board approval from the University of Texas, San Antonio, 110 participants over the age of 30 years who self-reported a diagnosis of either Type 2 diabetes or prediabetes were recruited through Qualtrics (an internet-based survey company) and incentivized through their preferred method (e.g., airline miles, gift cards). Participants over 30 were recruited based on the timelines to develop both Type 2 and prediabetes. That is, it typically takes years of poor diet and sedentary behavior for Type 2 or prediabetes to develop [1]. Informed consent was obtained online for all participants. In this case, participants consented to study involvement by clicking “I accept” to the study protocol outlined on the consent webpage. There was no penalty for not consenting. However, potential participants who did not click “I accept” were redirected back to a landing page. Self-reported diagnoses were mutually exclusive. Nine participants were excluded from analysis for inconsistent responding. Statistical analysis included 101 participants. These 101 participants successfully completed all measures. Therefore, there was no missing data.

Participants’ age ranged between 30 and 84 years (M = 51.03, SD = 12.58) and 53% were female (Table 1). Fifty participants self-reported having a diagnosis of Type 2 diabetes with an average of 9.04 years (SD = 7.4 years) since diagnosis and 51 self-reported being diagnosed with prediabetes with an average of 3.37 years (SD = 3.83). BMI ranged from 17.79 to 65.73 (M = 33.03, SD = 8.73).

Table 1

Descriptive demographic data.

		Type 2	Prediabetes
	N =	50	51
Gender	Males (%)	19 (38)	29 (57)
	Females (%)	31 (62)	21 (43)
Ethnicity	White/European (%)	40 (80)	41 (80)
	African American/Black (%)	4 (8)	3 (6)
	Asian American (%)	2 (4)	4 (2)
	Hispanic/ Latino American (%)	3 (6)	3 (6)
	Other (%)	1 (2)	2 (4)
Education	Some high school (%)	1 (2)	0
	High school graduate/GED (%)	11 (22)	5 (10)
	Some college, no degree (%)	10 (20)	15 (29)
	Trade/technical training (%)	2 (4)	2 (4)
	Associate’s degree (%)	9 (18)	6 (12)
	Bachelor’s degree (%)	12 (24)	16 (31)
	Master’s degree or higher (%)	5 (10)	7 (14)
Employment	Full time (%)	22 (44)	27 (53)
	Part time (%)	4 (8)	7 (14)
	Retired (%)	16 (32)	9 (18)
	Other (%)	8 (16)	8 (16)
Income (USD)	< $19,999 (%)	6 (12)	4 (8)
	$20,000 - $29,999 (%)	7 (14)	9 (18)
	$30,000 - $39,999 (%)	7 (14)	8 (16)
	$40,000 - $49,999 (%)	9 (18)	3 (6)
	$50,000 - $59,999 (%)	3 (6)	7 (14)
	$60,000 + (%)	18 (36)	20 (39)

Measures

All measures were normally distributed unless otherwise noted.

Management & adherence

Summary of Diabetes Self-Care Activities (SDSCA; [21]). The SDSCA includes 20 questions assessing self-care levels over the past 7 days related to diet, exercise, blood sugar testing, foot care, and medication. For example, participants were asked on a scale from 0–7 to indicate on how many of the past seven days they followed a healthful eating plan. Higher scores indicated more days of adherence. The Diet (α = 0.74) and Exercise (α = 0.86) subscales demonstrated adequate reliability. Previous studies have shown that internal consistency measured via inter-item correlations is acceptable and correlations with other standardized diet and exercise measures are significant [22]. In the current study, the Foot Care scale for individuals with Type 2 diabetes did not have adequate reliability (α = 0.24) and was excluded from analysis. Medication and blood sugar testing were composed of individual items.

Diabetes Care Profile (DCP; [23]). Three subscales from the DCP were used: 1) The Control Problem scale (α = 0.94) contained 19 items that assess instances of poor diabetes management and potential reasons why blood sugar levels were not managed. Responses for all three subscales were on a 1–5 scale with participants indicating how often an event occurred in the past year. Higher scores indicated more difficulty managing the condition. The Control Problem scale was not normally distributed. Transformations did not correct the distribution and non-parametric statistical methods were used for this scale, 2) The Barriers to Testing scale was presented to participants who indicated regularly testing their urine and/or blood for sugar. The scale contained 11 items (α = 0.88) assessing how frequently barriers impede their ability to test (e.g., cost of testing). Due to low event occurrence (i.e., positive skew) the Barriers to Testing scale was normalized using a natural logarithm function. 3) The Understanding Scale, answered only by participants with Type 2 diabetes, contained 13 questions (α = 0.94) assessing understanding of diabetes related issues such as foot care. Higher scores for each scale indicated more difficulty managing, more barriers, and a greater understanding of diabetes, respectively. Previous research has shown similarly high Cronbach’s alphas in English-speaking populations [22-24].

Self-Efficacy

The Self-Efficacy for Diabetes (SED; [25]) is an 8-item questionnaire (α = 0.92) that assesses an individual’s confidence completing diabetes management practices (e.g., exercise). Participants responded on a 10-point Likert scale from “not at all confident” to “totally confident”. Higher scores indicated greater self-efficacy. Previous data analyses indicated good internal consistency for both English (α = 0.83) and Spanish-speaking (α = 0.85) patients diagnosed with Type 2 diabetes [25].

Impulsiveness

The Barratt Impulsiveness Scale-11 (BIS-11, [15]) asks participants to answer questions about how often they engaged in each of 30 statements describing impulsive or non-impulsive behaviors to create a total impulsiveness score across three second-order factors; Attentional (8 items; α = 0.74), Motor (11 items; α = 0.75), and Nonplanning (11 items; α = 0.74). Participants were asked to use a 4-point response option scale of ‘never’ to ‘always’ to indicate how often they engaged in the behavior described in each statement. Some items were reverse scored. Higher scores indicated more impulsiveness. Cronbach’s alphas between 0.79 and 0.83 have been reported in the general population [15]. A Cronbach’s alpha = 0.95 was recently reported for individuals diagnosed with Type 2 diabetes [18].

Delay discounting

The delay discounting task (DD, [8]) consisted of 27 items asking individuals to make a choice between a small immediate reward and a larger but delayed reward (e.g., $25 right now or $60 in 14 days). A free parameter k indicated discounting curve steepness which corresponded with the geometric midpoint of the ranges [26]. Higher k-values indicated more discounting. k-values were normalized using a natural logarithm function. Response consistency scores (R2) ranged from 0.81 to 1 (M = 0.96, SD = 0.04). A within subjects ANOVA with a Greenhouse-Geisser correction provided evidence of a magnitude effect (F(1.82, 178.94) = 18.04, p < 0.001), supporting the criterion validity of the measure within this sample. Previous studies have shown the DD task to be highly consistent and valid [26].

Analysis

Pearson zero-order correlations were used to determine if increased scores on the DD task and the BIS-11 (i.e., more impulsiveness) were associated with fewer numbers of days adhering to self-care activities and more management problems. Correlations were used to examine the association between self-efficacy to self-care outcomes, barriers, and understanding of management issues. A series of independent sample t-tests were used to compare individuals diagnosed with Type 2 diabetes compared to those diagnosed with prediabetes in DD, BIS-11, self-efficacy, and self-care. Mann-Whitney U tests were performed on variables that were not normally distributed. All correlations and statistical tests were conducted using SPSS (Version 25).

Finally, two fully saturated path models were specified to estimate the direct and indirect effects of DD and BIS-11 on self-care outcomes via diabetes management self-efficacy for individuals diagnosed with either Type 2 diabetes or prediabetes. Previous psychometric evaluations for the BIS-11 suggest that a single factor structure provides a better fit to the data [27, 28]. Therefore, the total BIS-11 score was used in both path analyses. Path analyses were conducted with maximum likelihood estimation. Indirect effects were estimated using bias-corrected bootstrapped 95% confidence intervals (n = 1000), per [29]. Path models were tested using MPlus.

Results

Type 2 diabetes vs. prediabetes

Measurement means, standard deviations, and ranges are presented in Table 2. Independent sample t-tests were used to compare individuals diagnosed with Type 2 diabetes to those diagnosed with prediabetes, except for DCP Control, which was tested with a Mann-Whitney U test. There were no significant differences for DD, any of the BIS-11 subscales or exercise. Individuals diagnosed with Type 2 diabetes reported higher self-efficacy (M = 7.76, SD = 2.97) compared to individuals diagnosed with prediabetes (M = 6.43, SD = 2.17, t(99) = 2.88, p < 0.01). This difference had a medium effect size (d = 0.51). Individuals diagnosed with Type 2 diabetes also reported more days adhering to a diet (M = 4.73, SD = 1.46) compared to individuals diagnosed with prediabetes (M = 4.05, SD = 1.75, t(99) = 2.09, p = 0.039). Like self-efficacy, the effect size for this difference was medium (d = 0.43). A Mann-Whitney U test showed no significant differences between groups on the control scale.

Table 2

Means (SD) and ranges for each measure by diagnosis.

	Type 2 Diabetes		Prediabetes		Independent sample t-tests
Measure	Mean (SD)	Range	Mean (SD)	Range	t(99) =
SDSCA Diet	4.73 (1.46)	1–7	4.05 (1.75)	0–7	2.09*
SDSCA Exercise	3.51 (2.15)	0–7	3.30 (2.27)	0–7	0.468
DCP Understanding	4.10 (0.73)	2.75–5	N/A+	N/A+	N/A+
DCP Control	2.04 (0.84)	1–4.05	1.98 (0.79)	1–4.74	U = 1003
DCP Barriers#	0.22 (0.19)	0.00–0.65	0.30 (0.18)	0.00–0.68	-1.65
SED	7.76 (2.97)	2–10	6.43 (2.17)	2–10	2.88*
BIS-11 Attention	16.74 (4.48)	9–31	15.47 (4.57)	8–26	1.41
BIS-11 Motor	21.70 (4.97)	14–39	22.63 (5.12)	13–40	-0.92
BIS-11 Nonplanning	23.86 (5.49)	13–35	23.29 (5.12)	13–37	0.53
BIS-11 Total	62.30 (11.19)	39–88	61.39 (12.14)	37–89	0.39
Delay Discounting#	-1.71 (0.71)	-3.8 - -0.60	-1.59 (0.72)	-3.8 - -0.60	-0.71

#Transformed variable.

* p < 0.05.

+DCP Understanding was not measured for individuals with prediabetes, as they would not yet be encountering some of the symptoms described (e.g., foot care).

Between and within scale associations

As shown in Table 3, there were no significant Pearson zero-order correlations between DD and any of the three BIS-11 subscales for either individuals diagnosed with Type 2 or prediabetes, replicating previous null results (e.g., [30]) and suggesting that each measure different impulsivity constructs. However, for individuals diagnosed with prediabetes there was a significant positive correlation between DD and total BIS score. All BIS-11 subscales positively correlated with each other (all r’s > 0.374, p < 0.05), except for the BIS Motor and Nonplanning subscales for individuals diagnosed with Type 2 diabetes. Table 3 also shows the relationships between impulsivity and adhering to self-care activities and diabetes management, via correlations between all subscales calculated separately for each diagnosis.

Table 3

Pearson zero-order correlations between impulsivity, self-efficacy, and self-care.

		DD	BIS attn.	BIS motor	BIS nonplan	BIS total	SED
Type 2	SDSCA Blood	-0.053	0.027	0.207	-0.264	-0.004	0.307*
	SDSCA Medication	-0.054	-0.241	-0.263	-0.294*	-0.238	0.022
	SDSCA Diet	-0.069	-0.306*	-0.011	-0.393*	-0.320*	0.646*
	SDSCA Exercise	0.059	-0.188	0.226	-0.439*	-0.190	0.742*
	DCP Understanding	-0.136	0.058	0.004	-0.387*	-0.165	0.533*
	DCP Control	-0.066	0.306*	0.339	0.246	0.411*	-0.343*
	DCP Barriers+	0.087	0.092	0.252	0.342*	0.304	-0.410*
	BIS Attention	-0.077
	BIS Motor	-0.002	0.587*
	BIS Nonplanning	-0.155	0.374*	0.105
	BIS Total	-0.108	0.845*	0.731*	0.687*
	SED	0.063	-0.252	0.032	-0.631*	-0.396*
Prediabetes	SDSCA Diet	-0.133	-0.149	-0.049	-0.265	-0.191	0.410*
	SDSCA Exercise	-0.042	-0.274	-0.061	-0.260	-0.241	0.425*
	DCP Control	0.052	0.367*	0.124	0.321*	0.327*	-0.125
	DCP Barriers+	0.345	0.315	0.336	0.033	0.317	0.229
	BIS Attention	0.239
	BIS Motor	0.329*	0.554*
	BIS Nonplanning	0.116	0.555*	0.385*
	BIS Total	0.279*	0.849*	0.796*	0.801*
	SED	-0.057	0.014	0.128	-0.179	-0.018

+DCP Barriers scale was presented to participants who tested their urine or blood at least 1 day a week (Type 2: n = 39, Prediabetes: n = 24).

* p < 0.05.

For all individuals, DD was not significantly correlated to any self-care or management problems or self-efficacy, as shown in the leftmost column of Table 3. For individuals diagnosed with Type 2 diabetes, there was no significant correlation between impulsivity and adherence to medication or blood sugar testing as measured by the SDSCA, with the exception of the BIS-11 Nonplanning scale, where higher scores were associated with less days adhering to medication recommendations. BIS-11 Attention was negatively related to SDSCA Diet and positively related to DCP Control, with higher attention impulsiveness associated with fewer days adhering to diet and increased difficulty managing diabetes. Higher BIS-11 Nonplanning scores were associated with fewer days adhering to both a diet and an exercise regime (SDSCA Exercise), a worse understanding of diabetes (DCP Understanding), and increased perceived barriers to care (DCP Barriers). Increased self-efficacy was associated with better DCP Understanding, increased blood sugar testing, diet, and exercise adherence (measured via the DCP), and fewer management problems and perceived barriers to care (measured with DCP Control and Barriers, respectively). Self-efficacy was negatively correlated with the BIS Nonplanning subscale. However, this was the only significant correlation between self-efficacy and any of the impulsivity measures.

For individuals diagnosed with prediabetes, increased self-efficacy was associated with increased diet and exercise adherence (via SDSCA), but not with any of the impulsivity measures. For individuals diagnosed with prediabetes, higher BIS-11 Attention scores were associated with increased management difficulty, as measured by the DCP Control scale. Lastly, higher BIS-11 Nonplanning scores were also associated with increased management difficulty, as measured by the DCP Control scale.

Path models

Based on the differences between individuals diagnosed with Type 2 diabetes compared to prediabetes, two path models were specified to estimate the direct and indirect effects BIS-11 on exercise adherence, diet adherence, and diabetes control problems via diabetes management self-efficacy (Figs 1 and 2). As DD was uncorrelated with all other study variables for individuals with either prediabetes or Type 2 diabetes, the variable was excluded from the path analyses. Given the relationship between impulsivity and BMI/obesity [31], and the higher rates of BMI in populations with diabetes [2], BMI was specified as a covariate within both models. With BMI excluded as a covariate two relationships changed, as specified in the prediabetes group below.

Fig 1

Path model estimating direct and indirect effects of impulsivity on self-care outcomes among individuals diagnosed with prediabetes, controlling for BMI.

Completely standardized maximum likelihood parameter estimates shown. Unstandardized slopes of the indirect effects along with the bias-corrected 95% bootstrapped confidence intervals are provided in text. *p < .05, **p < .01.

Fig 2

Path model estimating direct and indirect effects of impulsivity on self-care outcomes among individuals diagnosed with Type 2 diabetes, controlling for BMI.

Completely standardized maximum likelihood parameter estimates shown. Unstandardized slopes of the indirect effects along with the bias-corrected 95% bootstrapped confidence intervals are provided in text. *p < .05, **p < .01.

Prediabetes

Among individuals diagnosed with prediabetes, BMI, BIS-11 and self-efficacy accounted for a significant proportion of variability in exercise adherence (R = .282, p = .008) and a significant proportion of variability in diet adherence (R2 = .221, p = .031). The model predictors failed to explain significant variability in control problems (R2 = .138, p = .149).

BMI was significantly, negatively related to self-efficacy (β = -.258, SE = .026, p = .035), but was uncorrelated with both the BIS-11 and all three self-care outcomes. Total BIS-11 was not significantly related to self-efficacy. Self-efficacy was positively associated with both diet (β = .370, SE = .151, p = .025) and exercise adherence (β = .363, SE = .170, p = .011) but unrelated to control problems. Total BIS-11 had no direct relation to diet adherence. However, total BIS-11 had a direct, positive relation to control problems (β = .344, SE = .128, p = .007) and a significant negative relationship to exercise adherence (β = —.263, SE = .026, p = .026). When BMI was excluded as a covariate the relationships between total BIS-11 and control problems (β = .33, p = .068) and exercise adherence (β = —.23, p = .087) were no longer significant.

Total BIS-11 and self-efficacy were both significantly, directly related to the self-care outcomes summarized above in expected directions. Total BIS-11 was directly, positively related to control problems and negatively related to exercise adherence, while self-efficacy positively related to both diet and exercise adherence. Self-efficacy did not mediate the relationship between BIS-11 and self-care outcomes among individuals diagnosed with prediabetes.

Type 2 diabetes

Among individuals diagnosed with Type 2 diabetes, BMI, BIS-11 and self-efficacy explained a significant and relatively large proportion of variability in exercise (R = .587, p < .001) and diet adherence (R = .441, p < .001). Model predictors explained a significant but small proportion of variability in control problems (R = .208, p = .048).

BMI was not related to exercise (β = -.165, SE = .097, p = .088), self-efficacy (β = -.223, SE = .127, p = .080), and uncorrelated with all other variables. Total BIS-11 was significantly negatively associated with self-efficacy (β = −.342, SE = 0.123, p = .006). Self-efficacy was positively associated with both diet (β = .580, SE = .102, p < .001) and exercise adherence (β = .750, SE = .078, p < .001). There was no effect of self-efficacy on control problems. However, total BIS-11 was directly positively related to control problems (β = .365, SE = .133, p = .006). There were significant negative indirect effects of total BIS-11 via self-efficacy on both diet (β = −.198, b = −.026, 95% CI: [−0.060, −0.006]) and exercise adherence (β = −.256, b = −.049, 95% CI: [−0.112, −0.007]). The direct effect of total BIS-11 on diet and exercise adherence was not significant after accounting for this indirect effect.

Total BIS-11 and self-efficacy were associated with self-care outcomes in the expected directions, showing a direct, positive effect on problems, and an indirect, negative effect on exercise and diet adherence via self-efficacy. Self-efficacy mediated the relationship between BIS-11 and self-care outcomes among individuals diagnosed with Type 2 diabetes.

Discussion

The present study tested associations between different impulsivity constructs, self-efficacy and self-care behaviors for participants with either self-reported prediabetes or Type 2 diabetes diagnoses. Three main hypotheses were test. First, individuals with higher impulsivity would have fewer numbers of days adhering to self-care activities and more management problem, Second, higher self-efficacy would be associated with increased self-care and management understanding, as well as fewer management problems and perceived barriers to care. Third, self-efficacy would mediate the relationship between impulsivity and diabetes self-care. Differences in impulsivity (DD and BIS-11), self-efficacy, and self-care between individuals diagnosed with Type 2 diabetes compared to those diagnosed with prediabetes were also tested for.

BIS-11 was significantly related to self-care behaviors for individuals diagnosed with Type 2 diabetes, with higher BIS-11 scores associated with fewer days adhering to diet and exercise and increasing control problems. This result was partially consistent with the first hypothesis, as DD was not significantly related to self-care behaviors. The relationship between diabetes self-care and impulsivity was mediated by self-efficacy (see Fig 2) for both diet and exercise adherence. This result replicated the results of Hadj-Abo et al. [18] and was also consistent with the third hypothesis, which predicted that self-efficacy would mediate the relationship between impulsivity and self-care behaviors. The methodology for measuring self-care behaviors was different from Hadj-Abo et al. [18] who used an aggregate self-care management measure in their mediation analyses. Therefore, the current findings suggest not all self-care behaviors are related to impulsivity measures via the BIS-11 for individuals diagnosed with Type 2 diabetes. In addition to self-care, diabetes control problems were also measured via the DCP. Interestingly, for individuals diagnosed with Type 2 diabetes the significant positive relationship between total impulsivity and control problems was not mediated by self-efficacy. There may be limits to the mediating effect of self-efficacy on problems associated with impulsivity and control problems for individuals diagnosed with Type 2 diabetes. More specifically, scales like the DCP questionnaire that rely on an individual’s beliefs as to why their diabetes management occurs at a certain level may be a less reliable predictor of actual diabetes management, and thus, unrelated to self-efficacy [32].

Previous studies measuring impulsivity, self-efficacy and self-care behavior have not included individuals diagnosed with prediabetes. The current study aimed to address this gap in the literature. For individuals diagnosed with prediabetes, higher BIS-11 attention and Nonplanning scores were associated with increasing management problems. Several other relationships between the BIS-11 and self-care were non-significant but showed a similar pattern to individuals with Type 2 diabetes; higher BIS-11 scores were associated with fewer days adhering to diet and exercise. These results suggest that stronger associations between BIS-11 and diet and exercise adherence may be meaningful markers for individuals transitioning from prediabetes to Type 2 diabetes diagnoses. Perhaps most importantly, self-efficacy was a significant predictor for diet and exercise, whereas BIS-11 was a significant predictor for both exercise and diabetes control problems when BMI was included as a covariate for individuals self-reporting a prediabetes diagnosis. These results partially supported the hypothesis that self-efficacy would be associated with diabetes understanding, management problems, and perceived barriers. Unlike individuals diagnosed with Type 2 diabetes, self-efficacy did not mediate relationships between impulsivity and any of the self-care behaviors for individuals diagnosed with prediabetes. It may be that diabetes specific self-efficacy for individuals diagnosed with prediabetes is not strong enough to impact these important self-care behaviors, assuming higher self-efficacy causes better self-care adherence. Table 2 shows that self-efficacy is on average more than one point (out of 10) lower for individuals diagnosed with prediabetes, relative to those diagnosed with Type 2 diabetes. It may be that individuals diagnosed with Type 2 diabetes have more experience with self-care behaviors, relative to those individuals diagnosed with prediabetes. As described below, this increased experience with self-care and other factors suggests pathways for important clinical interventions.

This was the first study to measure associations between impulsivity via DD and the BIS-11, and self-care behaviors in individuals diagnosed with prediabetes using an internet-based sample. Although individuals diagnosed with prediabetes were equally impulsive on both measures relative to those diagnosed with Type 2 diabetes, the associations with self-care behaviors were not significant. Prediabetes is not typically associated with noticeable symptoms [3] and individuals have few, if any discriminable signals from their bodies for prediabetes onset. Conversely, many (but not all) individuals diagnosed with Type 2 diabetes typically have noticeable symptoms [1], which may increase motivation to seek a diagnosis and adhere to self-care recommendations. In the current sample, individuals diagnosed with Type 2 diabetes were more likely to adhere to their diet and had higher self-efficacy than individuals diagnosed with prediabetes (see Table 2). Additionally, participants diagnosed with Type 2 diabetes had self-reported being diagnosed almost 3 times as long as those participants with prediabetes. Thus, increased diabetes treatment experience may have contributed to these stronger associations. Future studies using longitudinal designs will be necessary to determine how increasing treatment experience may affect self-efficacy, and/or self-care adherence.

DD was not correlated with self-care behavior regardless of diagnosis nor was it correlated with management problems. The current results failed to replicate the negative relationship between DD and self-care adherence found by Lansing et al. [13] and Reach et al. [11]. However, results replicated Lebeau et al. [12], who also found no relationship between DD and diet. There are a few reasons why the current results did not show a relationship between DD and self-care. First, unlike Reach et al. [11] and Lebeau et al. [12], the main dependent measure for diabetes control was based on self-reports, whereas they obtained quantitative HbA1c levels. Obtaining HbA1c levels for an internet-based sample is logistically difficult. DD is a behavioral measure of impulsivity [9] which previous researchers have shown to be either weakly or completely unrelated to self-report impulsivity measures (e.g., [20]). This also suggests that they measure different impulsivity constructs [19]. The diabetes self-care and management measures were self-reports. This might explain why relationships between the BIS-11 and self-care subscales were found in the current study and Hadj-Abo et al. [18], but not with DD and self-care subscales. It may also explain why Lebeau et al. [12] did find a correlation between DD and HbA1c levels. Self-report measures rely on an individual’s ability to correctly perceive, remember, and report their behavior. Future studies should use more objective measures of diabetes self-care and management, such as HbA1c levels in combination with both DD and self-report impulsivity measures.

Second, treatment regimens for prediabetes and Type 2 diabetes vary considerably for each individual. This variability is dictated by symptomology and demographic variables (e.g., age and family history of Type 2 diabetes). To obtain more precise associations between self-care adherence and psychological constructs like impulsivity and self-efficacy, future studies should obtain prescribed treatment regimes. This would include any changes to treatment based on changes in HbA1c levels. That is, more dynamic models could be created incorporating changes in both disease progression and psychological constructs.

Lastly, the DD measure used in this and other research with individuals diagnosed with diabetes has only focused on the value of money [11-13]. Different commodities have been previously tested within the DD paradigm (e.g., food, sex and health). A recent meta-analysis demonstrated that this commodity effect phenomenon was highly reliable for DD [33]. However, there are also pronounced state-like effects for DD, where non-monetary outcomes are discounted at higher rates than monetary outcomes [34]. It may be that discounting for more relevant commodities (e.g., sugary foods for individuals diagnosed with Type 2 diabetes) may show stronger associations with self-care behaviors related to diet (SDSCA diet), relative to monetary outcomes. Likewise, DD for health outcomes may show a stronger association for DCP understanding, relative to monetary or sugary foods as an outcome.

Consistent with Mohebi et al. [4], having a higher level of self-efficacy was associated with better adherence to diet and exercise for all individuals. For individuals diagnosed with Type 2 diabetes, self-efficacy was correlated with a better understanding of management, fewer management problems, and fewer perceived barriers. This is consistent with research that has shown a positive relationship between behavior change and self-efficacy over a range of health-related behaviors [35], though the causal status of self-efficacy has yet to be empirically demonstrated (e.g., [36]). Although the current study cannot provide evidence either for or against the causal status of self-efficacy, it does demonstrate the consistent significant association self-efficacy has with important self-care behaviors for individuals diagnosed with diabetes. Increasing individual’s self-efficacy may be one mechanism for a clinical intervention using the current and previous [4, 18] results. However, it should be noted that either manipulating self-efficacy or impulsivity is an indirect route for behavior change. A more direct route would be to directly target those deficient self-care behaviors (e.g., diet adherence) though a behavior skills training paradigm which consists of instructions, modeling, rehearsal, and feedback [37]. Research using this approach could not only assess behavior change in real-time, but also measure potential concurrent changes in impulsivity and self-efficacy to help determine the causal role these psychological constructs play in self-care behavior for individuals diagnosed with either prediabetes or Type 2 diabetes. At present, self-efficacy and other psychological constructs associated with diabetes self-care behaviors (e.g., need for cognition; 18) can be used as markers for potential self-care dysfunction. As noted by Hadj-Abo et al. [18], using these psychological constructs as a screening mechanism for self-care dysfunction may be an efficient way to predict who may need additional diabetes self-care training.

This study is not without limitations. First, the sample size used for the path models is relatively small. The small participant sample may have obscured significant associations between impulsivity, self-efficacy, and self-care behaviors due to a lack of statistical power, especially for participants diagnosed with prediabetes. Second, as mentioned above, obtaining HbA1c levels for an internet-based sample is logistically difficult. However, this biological marker is very important in both the diagnoses and treatment for diabetes. In addition, previous studies have shown an association between DD, HbA1c levels, and some self-care behaviors [11, 12, 14]. Measuring HbA1c levels, DD, BIS-11, and self-efficacy all within the same participants would more accurately demonstrate whether diabetes management is differentially associated with distinct impulsivity types. Lastly, this study was cross-sectional. Thus, casual inferences between impulsivity, self-efficacy, and self-care behaviors is not possible. Future studies measuring the association between these constructs should employ a longitudinal design to facilitate these causal interpretations.

Conclusions

Impulsiveness based on the BIS-11 was related to several self-care behaviors, with self-efficacy mediating some of the relationships. DD was not related to diabetes self-care adherence in individuals with self-reported Type 2 diabetes or prediabetes diagnoses. Results highlight the need to systematically measure different aspects of multidimensional constructs like impulsivity. This is especially important when trying to determine associations between these constructs and critical health behaviors, that lead to health consequences like Type 2 diabetes. The null results in regard to DD should be interpreted with caution as DD was only assessed for monetary outcomes and the sample size was small for path model analyses. This is the first study to use a choice-based impulsivity measure in an internet-based sample of adults self-reporting a diagnosis of either Type 2 diabetes or prediabetes. Given the prevalence of Type 2 diabetes, this study should be replicated using a longitudinal design with a larger sample to better understand the relationship between impulsivity and diabetes self-care adherence.

Raw data for all analyzed participants.

(XLS)

Click here for additional data file.

10 Jun 2021

PONE-D-21-06140

Associations between impulsiveness and self-care adherence in individuals diagnosed with type 2 or prediabetes

PLOS ONE

Dear Dr. Romanowich,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Jul 25 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Matthew J. Gullo

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

Additional Editor Comments (if provided):

Concerning Reviewer 1's comments on the BIS-11, psychometric evaluations of the scale tend not to support its subscales and suggest a single factor structure provides a better fit to data (e.g., Steinberg et al., 2013). This could be used to justify your use of a single total score in the path analyses. The construct assessed by BIS-11 is frequently distinguished from that assessed by Delay Discounting. Many researchers view them as assessing different components of impulsivity on theoretical grounds, which may also address one of Reviewer 1's comments (for a review, see Gullo et al., 2014).

Gullo, M. J., Loxton, N. J., and Dawe, S. 2014. “Impulsivity: Four Ways Five Factors Are Not Basic to Addiction.” Addictive Behaviors 39 (11): 1547–56.

Steinberg, L., Sharp, C., Stanford, M. S., and Tharp, A. T. 2013. “New Tricks for an Old Measure: The Development of the Barratt Impulsiveness Scale-Brief (BIS-Brief).” Psychological Assessment 25 (1): 216–26.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Using a cross-sectional design and path analysis, Wainwright et al. examined the relationship between impulsivity, self-efficacy, and self-care adherence and management in adults with a self-reported diagnosis of Type 2 diabetes or prediabetes. The results from this study suggest greater impulsivity may be associated with reduced self-care adherence and that this relationship may be mediated by self-efficacy in participants with Type 2 diabetes. Given diabetes is a chronic condition, that when poorly managed can lead to a variety of serious health complications, investigating mechanisms that can lead to poor management is a relevant addition to this field. While this study highlights several mechanisms that may be important to consider in this context (e.g., impulsivity, self-efficacy), I recommend that this paper not be accepted without major revision.

Overall, this manuscript could be improved with a deeper dive into the research literature to provide concrete justification for the hypotheses and proposed models. In particular, the relationship between impulsivity and self-efficacy was not discussed in enough depth, nor the reason why these variables are important to consider when examining self-care adherence and maintenance in individuals with diabetes or prediabetes. These constructs were not well defined, and it appears some key recent research that would support this study is missing from the Introduction (e.g., Hadj-Abo et al., 2020).

While using advanced statistical analysis (i.e., path analysis) and examining these paths in both Type 2 and prediabetic groups was certainly a strength of this study, the inclusion of certain predictors in these models were not well justified. Given the authors suggested it was important to examine different impulsivity constructs in the Introduction and that the initial correlations suggested not every facet of impulsivity measured by the BIS-11 was associated with the outcome variables, it was unclear why the BIS-11 total score was used as a predictor in the model. Based on the correlations in this study, as well as previous research, it may have been more appropriate to use the non-planning and attention subscales. Additionally, it was also unclear why DD was then included as a predictor given the variable did not correlate with any outcome variables in this study.

While the Discussion section includes a summary of the main findings of this study, the authors could vastly improve this section by including a more in-depth interpretation. For example, Lines 254-258 simply state the study failed to find a relationship between DD and self-care adherence and management problems. While linking these findings to previous research was useful, please include more discussion on why these findings have been so mixed and what the clinical and theoretical implications of this research are. Including clinical implications of all findings from this study in the Discussion would make a valuable addition to the field.

I have included some specific recommendations below:

Abstract

Line 29 – “have never been measured with participants”. This statement is vague and incorrect. Please see Hadj-Abo et al. (2020) who have recently examined these relationships in participants with diabetes.

Line 31 – delete “personality”

Line 27 and throughout – Use the term ‘impulsivity’ instead of ‘impulsiveness’ unless referring specifically to the BIS.

Introduction

Lines 42-45 – As PLOS One is an interdisciplinary journal, consider expanding your definition of diabetes to include a brief discussion on the distinction between Type 1, Type 2, and prediabetes. Based on the information currently available in the introduction, it is not clear why this study has chosen to focus only on Type 2 and prediabetes.

Line 43 – Remove apostrophe from pancreas.

Line 43-44 – Consider including the reference at the end of the sentence to also support the statement of “multiple health complications”. Explicitly listing some of these complications may also help justify the importance of this topic.

Line 47 – Consider a different term for “behavioral treatment”. For example, “behavioral interventions” or “behavioral management strategies” may better capture the actions required to effectively manage diabetes and prevent health complications.

Line 48 – Include reference for “in addition to pharmacological treatments”.

Line 49 – Please expand your definition for “self-care plan complexity”. It is unclear exactly what this term refers to.

Line 51 – Include ‘and’ between “organize execute”

Lines 51-53 – Given that self-efficacy is a core variable in your model, it would be useful to include more detailed discussion on this variable in the Introduction (e.g., be specific about how self-efficacy predicts adherence e.g., does lower or higher self-efficacy predict better adherence? How does lower/higher self-efficacy result in better adherence? What is its relationship with impulsivity?). It would also be helpful to specifically define and discuss diabetes-specific self-efficacy. Reference previous research and also justify why it has been proposed as a mediator between impulsivity and your outcome variables.

Lines 44, Lines 60 etcetera – Inconsistent capitalization of Type 2.

Lines 52-56 – Provide a link between delay discounting and impulsivity so it is clear to the reader why you have introduced DD after mentioning “different impulsiveness constructs” e.g., is DD a behavioural measure of impulsivity? It may also be useful to specifically discuss how DD differs from the facets of impulsivity measured by the BIS-11 and why this distinction is useful.

Line 56-59 – Provide a reference for this statement if available.

Line 61 – Define what HbA1c is for readers who may not be aware of this term.

Lines 60-72 – It would be helpful to provide more discussion as to why DD is associated with some diabetes management behaviours (e.g., non-adherence medication) but not others (e.g., diet adherence). Was this due to methodology, the study sample (e.g., adolescents with Type 1), or particulars of the DD task? It would also be helpful to include studies that have investigated these relationships in participants with prediabetes (if available) and highlight any differences (or expected differences) in findings between these participant groups.

Line 67 – Reword this sentence to state the study examined the relationship between DD and non-adherence and poor glycemic control within a sample of adolescents with Type 1 diabetes (i.e., not between DD and diabetes).

Line 67, 83 etc – Update wording throughout manuscript to ensure person-first language (i.e., adolescents with Type 1 diabetes, not diabetic adolescents). See Dickinson et al. (2017) for guidance on this.

Line 73 – I suggest changing the wording of this sentence to “….is frequently used as a self-report measure of impulsivity”

Lines 73-77 – It would be helpful to mention in this paragraph that BIS-11 includes several subscales measuring different facets of impulsivity. Define these facets and discuss how they may/may not be related to diabetes self-care adherence.

While BMI is a risk factor for diabetes, I believe this section could be much improved by including a summary of previous research specifically investigating the relationship between BIS-11 subscales (and other related measures of impulsivity if applicable) and diabetes related self-care adherence behaviours and outcomes. For example, Hadj-Abo et al., (2020) recently published a study investigating the mediating role of self-efficacy on the relationship between impulsivity and diabetes self-management. This study contains many of the same variables as the current manuscript so the findings from this research would be important to discuss in this Introduction.

Line 82 – Provide a reference for “a similar construct to impulsiveness” and be specific about its relationship with impulsivity. It should be clear why you are discussing this construct in the context of impulsivity, particularly as you have not measured it separately in the current study.

Line 71 & Line 89 – Both studies mentioned in these lines have included an adolescent sample and discussed how the findings are relevant to self-care management for that particular population (i.e., parent observation). Given that research suggests adolescence is a developmental period characterised by increased impulsivity and risk-taking behaviours, it may not be appropriate to extrapolate the findings from these studies to the adult participants used in the current study.

Line 100 – Change “impulsiveness levels” to “impulsivity”.

Materials & Methods

Participants

Line 109 – Provide an explanation as to why only participants over the age of 30 were recruited (particularly as many studies mentioned in the Introduction recruited adolescents)

Line 119 – Demographic Descriptive Table – It would be helpful to split these demographics by participant groups and include a % column.

Measures

-Include the rating scale, scoring information and brief description of psychometric properties (with references) for each measure.

-Capitalize subscale titles e.g., the Control Problem subscale

Line 135 – Suggest rewording “the scale was normalized” to something that better represents what happened (e.g., “As the distribution of the Barrers to Testing Scale was positively skewed, the data was transformed using….”).

Analysis

-Describe what software was used for analyses and reference accordingly (e.g., SPSS v.27; R; MPlus).

-Include mention that a Mann-Whitney U test was also performed on variables that were not normally distributed.

-Given the relationship between impulsivity and BMI/obesity (and the higher rates of BMI in populations with diabetes), it may be appropriate to consider including BMI as a covariate on the path models.

Results

-Specify if there was any missing data and the procedures that were followed to account for it.

Line 171 Table 2

-Include M & SD for BIS-11 total score and range of scores for all measures.

-To assist the reader, report the results from the independent sample t-tests within this table (including the non-significant findings) and then comment on these findings in text. Effect sizes for these tests may also be useful for determining how meaningful the significant differences are.

-Clarify why DCP understanding was reported as N/A in the table.

Line 173 – Refer to table 3 in text before commenting on the results.

Lines 173-176 – Clarify if these relationships were found in both groups separately or if those were combined to get these results. If the groups were combined, please justify this action. Were the relationships the same if the groups were separated?

Table 179 Table 3

-Report all correlations (including non-significant relationships such as BIS-11 & DD, and the relationships between self-efficacy and impulsivity measures) within this table.

-Report relationships between impulsivity measures and self-efficacy and comment on these in text.

Lines 183-197 – Inconsistent use of scale names and constructs measured in text. Use names of specific subscales/measures in text to describe the relationships.

Line 194 – Does this refer to the total BIS score? Report these correlations in the table.

Lines 198-205 – As stated earlier, include results of t tests in Table 3 and comment on these tests in text. Include statistics for non-sig. relationships

Path Models

Lines 206-251 - Given the aim of this study was to examine the relationships between different impulsivity constructs, and that different correlation relationships between BIS-11 subscales and self-care/management problems were found, it is unclear why a total BIS-11 score (rather than subscales, specifically attention & non-planning) was used in the path model. In addition, as DD was not correlated with any model variables in this study it is unclear why this variable was included in the path models as a predictor of these outcomes. It may be more appropriate to consider using simple mediation models with the variables that have demonstrated significant relationships.

I am also concerned that the sample size for each group may be too small to test a path model of this size. It is typically recommended to have a minimum 20 participants per parameter for SEM. While it may be a difficult population to recruit, it is important to highlight this limitation in the Discussion or use a different type of analysis.

Line 221 – It is inappropriate to call a non-significant relationship “marginally significant”. Remove this phrase from the manuscript (including from the Abstract and Discussion) and provide a correct interpretation of these relationships.

Discussion

-There is a strong sense of repetition of the results section within this Discussion. While providing a summary of the results is important, please also include an interpretation of these findings in the context of previous research and the population sampled. Including a clinical implications section would be a valuable addition.

-Include a discussion on how the findings from this study supported/did not support the hypotheses.

-Please include a clear limitations and future directions section.

Line 280-281 – Remove text in brackets.

Line 284 – Use third-person language. Check the rest of manuscript for this language as well.

Lines 303-307 – It is unclear why this paragraph has been included here. Is this paragraph being used to explain why there are often mixed findings when studies have used DD with populations with diabetes? Please provide some clarity and supporting references, and an explanation as to why other commodities (e.g., food) may be more appropriate.

Lines 312-313 – “an open question” is quite vague. Please clarify and expand the statement about causal status of self-efficacy. How does this study contribute to this status?

Lines 316-317 – These relationships were also not found in participants with prediabetes.

Line 319 – type 2 diabetes is not a health behaviour.

Lines 319-320 – Provide a brief reason why DD results should be interpreted with caution as this was not immediately clear from the discussion.

Line 320 – Delete “relatively robust”. This is term is quite vague and it is unclear what is meant by this statement.

Line 322 – Rather than just suggesting replication, provide a more specific example e.g., longitudinal evaluation in larger sample etc?

Suggested References

Dickinson, J. K., Guzman, S. J., Maryniuk, M. D., O’Brian, C. A., Kadohiro, J. K., Jackson, R. A., ... & Funnell, M. M. (2017). The use of language in diabetes care and education. The Diabetes Educator, 43(6), 551-564.

Hadj-Abo, A., Enge, S., Rose, J., Kunte, H., & Fleischhauer, M. (2020). Individual differences in impulsivity and need for cognition as potential risk or resilience factors of diabetes self-management and glycemic control. PloS one, 15(1), e0227995

Reviewer #2: Thanks for the opportunity to review this intetesting paper on the relationships between impulsiveness and self-care in tYPE II zdiabetes and prediabetes. Results are intetesting and the paper is overall well written, as well as statisical Analyses are well conducted.

I have only minor suggestione to make.

Specifically, in my opinion the introduction in too long. Please try to reduce it. For example, you could omit the general paragraphs on diabetes and prediabetes. In my opinion, the are well known concepts. However, I'll briefly cite some paper examining the relationship between diabetes and psychopathology (especially the relevance on specific personality disorders in individuale with diabetes or brutale diabetes). Some of these personality disorders are characterized by impulsiveness and nay probably influence the brittleness of diabetes. For details, you could cite PMID: 27524279. Also alexithymia could interfere in interpersonal relationships and partially influence impulsiveness in isubjects with instabile diabete (PMID: 31125008).

These concepts may be also considered in the discussione of your results, especially in relation with the link you found between impulsiveness and self-care.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

20 Aug 2021

The Response to Reviewers file has been uploaded. See attached file for detailed response to all reviewer comments/suggestions.

Submitted filename: PONE-D-21-06140 Response to Reviewers.docx

Click here for additional data file.

4 Nov 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Matthew J. Gullo

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have done an excellent job so far addressing the comments on their first draft - well done! However, there are still a few areas of improvement that require addressing before this article should be accepted for publication. Please find these listed below:

Abstract:

- Lines 27-30: This is a very long sentence and suggests that both impulsivity measures have not been measured in the same study before (regardless of participant sample). This is not the case (when referring to participant samples beyond diabetes). I would suggest deleting “participants, or for” and specifying that delay discounting has not been measured in individuals with Type 2, while neither delay discounting & BIS have been measured in pre-diabetes (if that is the case).

- To clarify your findings, It would be useful to also mention in the abstract that the choice-based personality measure was not associated with outcome variables so as to really support your final sentence in the abstract (that “ how specific impulsivity constructs ….”)

- Lines 37-39: The final sentence of this abstract is somewhat misleading. Your study highlights what specific impulsivity constructs may be more useful to target in interventions but not how these constructs can be incorporated into interventions. I recommend rewording this sentence so it more accurately reflects your study.

Introduction:

- Line 56: Move the word “can” to after “requirements that”

- Line 69: Insert commas as such: “impulsivity, such as a lack of inhibition, also”

- Line 70: Include a comma after the last bracket

- Line 112, include a comma after references

- Lines 58-73: I think the authors have done a wonderful job expanding on the definition of self-efficacy and how this relates to self-care behaviours.

- Line 95-96: I think it might be helpful to include a brief overall summary about what the consistent findings mean in relation to self-care, so it is very clear to the reader. Something like “Overall, the findings of these studies suggest that higher levels of impulsivity, as measured with the DD task, are associated with poorer diabetes self-care adherence. However, the findings…..”

- Lines 98-101: Great job including a summary of the BIS-11 subscales and how they are likely linked to self-care adherence.

- Line 103: specific how BIS-11 was related to BMI e.g., positive, negative relationship (just so it is very clear to the reader).

- Line 112: insert comma after references.

- Line 116: Due to a run on sentence, delete “like other important heath behaviours”. If you want to keep this in the introduction, I suggest including this comment as a separate sentence.

- Line 117: insert comma after “studies”

- Line 123: replace “for” with “in”

Materials & Methods:

- Line 196: Reword sentence to “Some items were reversed scored.”

- Line 208: Are the authors referring to internal consistency and a valid measure of impulsivity here? Please report alpha values (or equivalent) are per previous measures in this manuscript, and specify what is meant by ‘valid’

Results:

- Please ensure your Tables align with PLOS formatting requirements.

- There is still inconsistent use of scale/subscale/measures in text/tables. E.g., see table 2, Lines 241-245, Lines 348-249.

- Lines 259-273 and Table 3. This section should come before correlations and path models as it provides further justification as to why you have chosen to analyse these groups separately. Table 3 should be introduced in text prior to appearing.

- Table 3: Reconsider the layout of the final column in this table. Not all rows in this table at t-tests, and the header of this column does not represent every value reported underneath it.

- Table 2: Please include the total BIS-11 score in this correlation table and discuss in text.

- Table 2 and Lines 227-256: Based on the authors’ justification in Lines 220-222, it is unclear why the BIS-11 subscale correlations are reported and total score has been omitted here. The lack of relationship between some subscales of this measure for Type 2 (as reported in Line 231) and the differential relationships between these subscales and self-care adherence and self-efficacy appear at odds with this justification. As a reviewer, I am still not convinced that it is appropriate to use the total BIS-11 score in the path analyses. Theoretically, it makes sense as to why significant negative correlations are noted between BIS11 Non-Planning/Attention and these outcomes. Please ensure adequate and clear justification for this decision process is provided, beyond just the one study suggesting a single factor structure.

- Since you have now included BMI as a covariate, please include this in your correlations and t tests and provide some justification for this decision (e.g., based on previous research, significant correlations in the current study etc).

Path Models

- Capitalise Type 2 in Line 279 and check the remaining manuscript for consistency.

- Lines 283 & 287: For consistency, report “controlling for BMI” the same across these figure notes.

- Lines 280 & 289: Delete ^p < .10 as this is not used in the figures.

- Please provide justification for including BMI as a covariate as this appears to be a new addition since the first draft. From your results, it appears that BMI is not correlated with your outcome variables, so again, it is unclear why this covariate was included. Are the path models the same if you do not include BMI as a covariate? If so, report the path models without the covariate and include a note that the same models were run with BMI as a covariate but findings did not differ (if justification for BMI covariate is included).

Discussion

- To reorientate the readers, it is helpful to restate the hypotheses at the beginning of the discussion. Include a brief statement of the hypotheses following the first sentence in the discussion. This should be the first paragraph, then flow onto the findings. Please refer to PLOS guidelines on writing and formatting the Discussion and Conclusion sections of this manuscript - https://plos.org/resource/how-to-write-conclusions/

- Use third-person language (e.g., Lines 332, 336) and check the remaining manuscript for this.

- Line 328: replace ‘predictive’ with ‘correlated’ to more accurately represent the data analysis and findings.

- Lines 328-331: Discussion about DD might be better suited if it were included with the content in Lines 378-406

- Line 335 – replace ‘this’ for ‘the’

- Lines 332-346: Overall, this paragraph would benefit from a re-write. The discussion bounces between the findings of Hadi-Abo and the current study’s results and hypotheses which makes it somewhat difficult to follow. I would suggest this type of structure: State the results. State how the results support/do not support the hypotheses. State how this relates to previous research. Explanation of findings in the context of the hypotheses and previous research. Be careful to avoid run-on sentences and ensure appropriate grammar (particularly the use of commas).

- Line 344-346: This concluding sentence is too casual (i.e., replace “Perhaps there are” with “There may be”) and does not provide an explanation as to why self-efficacy did not mediate. It might be useful to hypothesise why this result occurred based on your knowledge of the literature, the constructs, and this study’s findings.

- Lines 347-348: This sentence requires a slight re-write as it contains grammatical and word choice errors. The connection between this sentence and the current study’s results are unclear, please provide the link so it is clear to the readers. Something along the lines as “The current study aimed to address this gap in the literature…” would likely suffice.

- Lines 347-363: Similar to the feedback provided about Lines 332-346, this paragraph would also benefit from a re-write. At times, it was unclear if the authors were discussing the results related the pre-diabetes, or Type 2, or both (e.g., Lines 354-357)

- Line 360: Please clarify what the authors mean by “not high enough to impact…”. By scoring lower on self-efficacy, this would imply less adherence/self-care (at least based on Type 2 findings). It is unclear why scoring lower on this measure would result in a non-significant relationship. The significant difference between self-efficacy between participant groups implies that Type 2 have higher levels of self-efficacy compared to prediabetes – is this because they have likely spent more time engaging in these behaviours in the past? This would be important to explore in the discussion as it is likely an important target for clinical intervention (at least for Type 2)

- Line 367: As stated in the previous feedback, please remove the term “marginally significant” and provide the correct interpretation of results.

- Line 337: Remove impulsivity. By the definition of personality, impulsivity is unlikely to be a treatment-modifiable factor. Given impulsivity is the same across these participant groups, longitudinal research should focus on factors that are likely to change over time and with treatment (such as self-efficacy and adherence).

- Lines 384-387: This is a run-on sentence. Consider rephrasing and writing in more succinct sentences.

- Line 390: Delete “To date, no study has done this.”

- There are several errors in punctuation, grammar, and sentence/paragraph structure throughout the discussion and other sections of this manuscript. While I have highlighted somes of these errors, it is recommended that the authors carefully review and copyedit this manuscript before submitting the revised draft. This resource may be useful https://plos.org/resource/how-to-edit-your-work/

Reviewer #2: I have no other comment. In my opinion, the manuscript is now ready for publication in PLoSOne.

Only a comment on your answer to my previous suggestion: "...The articles on brittleness of diabetes only focused on those patients diagnosed with Type 1 diabetes. Therefore, we are unsure whether this addition will aid in an understanding of relationships between impulsiveness and Type 2/prediabetes. This information may fit in the clinical implications section where we outline the role psychological constructs may play in remediating self-care adherence. However, this still feels somewhat ad hoc. For these reasons, we have chosen not to include brittleness in the current manuscript revision".

In my opinion, it is not important whether patients were diagnosed with Type I or TYpe II diabetes/prediabetes. I think that brittleness may be due to a bad self-care adherence related to impulsiveness and specific Cluster B personality traits. In studies on brittleness on Type I diabetes it has been shown that there was a relationship between glycaemic instability and these specific cluster B personality traits. I think that the same psychopathological substrate is what you have observed in patients with type II diabetes.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Lorenzo Pelizza

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

19 Dec 2021

See Response to Reviewers file

Submitted filename: PONE-D-21-06140 R1 Response to Reviewers 12-19-21.docx

Click here for additional data file.

2 Feb 2022

Associations between impulsivity and self-care adherence in individuals diagnosed with type 2 or prediabetes

PONE-D-21-06140R2

Dear Dr. Romanowich,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. There is also a minor error that will need to be corrected during copyediting (see below).

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Matthew J. Gullo

Academic Editor

PLOS ONE

Additional Editor Comments:

Please correct this minor error during copyediting: On Line 284-285: Correct the text to specify that two relationships changed when not controlling for BMI (not one).

Reviewers' comments:

1 Mar 2022

PONE-D-21-06140R2

Associations between impulsivity and self-care adherence in individuals diagnosed with Type 2 or prediabetes

Dear Dr. Romanowich:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Assoc. Prof. Matthew J. Gullo

Academic Editor

PLOS ONE