Personal Projects and Psychological Well-Being: Emerging Adults With and Without Diabetes

Abstract

Objective

The goal of this study was to examine the nature of the personal projects that emerging adults with and without diabetes were pursuing and the implications of those projects for psychological well-being.

Methods

We asked emerging adults with and without type 1 diabetes to identify five personal projects, rate four dimensions of those projects (importance, typicality, stress, and progress), and complete several well-being measures (depressive symptoms, life purpose, life satisfaction, perceived stress, and resilience) when they were age 19. Those with diabetes also indicated the extent to which diabetes interfered with each of the projects. We followed participants for 1 year to determine the status of projects and reassess project dimensions and psychological well-being.

Results

The kinds of projects identified by the two groups were similar. However, those with diabetes reported lower levels of progress on projects and completed fewer projects 1 year later compared with controls. Project progress, importance, and completion were linked to higher psychological well-being, whereas project stress was linked to lower psychological well-being. However, the most robust cross-sectional and longitudinal predictor of psychological well-being was project typicality (i.e., the extent to which projects were typical of participants). The pursuit of more typical projects was linked to higher psychological well-being. These findings were largely similar for emerging adults with and without diabetes. Diabetes interference with projects revealed some links to psychological well-being.

Conclusions

These results suggest that personal project engagement and completion is linked to the overall mental health of emerging adults with and without diabetes.

There is a large literature on psychosocial factors involved in psychological, behavioral, and physical health in the area of youth with type 1 diabetes (see Helgeson, Naqvi, Van Vleet, & Zajdel, in press; Wiebe, Helgeson, & Berg, 2016, for reviews). Much of that research focuses on sources of stress, depression, and difficulties with self-care behavior and glycemic control. These topics, of course, are important in their own right. However, research seems to have neglected the fact that youth with type 1 diabetes grow up and face the same tasks as youth without diabetes in determining their educational, vocational, and relational pursuits. These normative tasks and goals that youth with diabetes face are rarely the focus of research.

Goal-setting and goal-striving research may be useful in understanding the psychological health of youth with type 1 diabetes. Goal-setting involves determining which goals to pursue, and goal-striving involves the planning and execution of behaviors that lead to those goals (Mann, de Ridder, & Fujita, 2013). In the health arena, research has shown that setting goals and meeting goals is related to good psychological, relational, and physical well-being among newlywed couples (Jakubiak & Feeney, 2016) and to health behavior change across an array of situations, such as diet, exercise, and smoking cessation (Strecher et al., 1995). In fact, goal-setting is often an integral aspect of interventions aimed at changing health behaviors (Levack et al., 2006; Ries et al., 2014).

There is less research on goal-setting and goal-striving among adolescents and young adults. We followed a cohort of youth with diabetes and a comparison group of healthy controls over adolescence and emerging adulthood to identify the goals that both groups have set for themselves and to understand the implications of those goals for their well-being. The approach we used was based on personal project analysis (PPA). Little (1993), a personality psychologist, stated that there are a variety of ways to examine personal goal-directed activity and that identifying personal actions or projects was one of them. PPA is a methodology in which individuals are asked to report their current or ongoing projects in an open-ended format and then to rate each project on a set of dimensions, such as importance, progress, typicality, and stress (Little, 1993). Personal projects can be as mundane as wanting to be more organized or as grand as wanting to become a physician. Although the framework is termed PPA and respondents are asked to identify projects, Little and researchers who use PPA often use the term “project” and “goal” interchangeably. For clarity, we use the term “project” throughout.

It is not the content of people’s personal projects as the status of those projects that is related to well-being. Little (1987) suggests that the extent to which projects reflect one’s own identity or are typical of the self is a central dimension of personal projects. He states that the lack of typical projects is associated with depression, whereas the presence of typical projects is related to life satisfaction. A meta-analysis of personal project dimensions and well-being showed that project efficacy (i.e., feeling capable of executing the project) and project stress (i.e., stress associated with engaging in the project) are important dimensions that connect with well-being (Wilson, 1990, as cited in Little, 1998). Consistent with this claim, Nurmi, Salmela-Aro, and Aunola (2009) examined three personal projects of college students in Finland and found that high levels of project stress and low levels of project progress were linked to more depressive symptoms.

Only a few studies have examined personal projects in the context of chronic illness. Personal projects may be particularly important to study in the context of chronic illness, as the illness could affect the kinds of projects pursued and the progress toward completing those projects. People vary in the extent to which they define themselves in terms of their illness and the extent to which they integrate the illness into their lives (Charmaz, 1995; Helgeson & Novak, 2007). It is important for researchers to step outside the health domain and understand the lives of those with chronic illness more holistically.

Some research has compared the nature of projects identified by those with chronic illness to a healthy comparison group and found few differences. In a study that compared the goals of people with multiple sclerosis to controls, there were no group differences in the content of projects identified (Brooke, Desmarais, & Forwell, 2007). In a study that compared adolescents with cancer to their peers, there were no differences in the content of projects identified, but those with cancer identified more intrinsic compared with extrinsic projects relative to healthy peers (Sulkers et al., 2015). When participants in the present study were age 16, we used PPA and found that those with diabetes were more likely to identify projects connected to appearance and those without diabetes were more likely to identify projects connected to self-improvement (Helgeson & Takeda, 2009). There were no group differences on the other seven categories of projects.

Several studies of chronically ill populations have linked project dimensions to well-being. In a study of women with fibromyalgia, women identified a health/fitness goal and a social goal and then rated the effort they put forth, the progress they made, and the extent to which pain and fatigue interfered with each goal on a daily basis for 30 days (Affleck et al., 1998). Greater progress on social goals was related to better mood, and pain/fatigue interference with health goals was related to worse mood on a daily basis. Studies of adults with low back pain (Vroman, Chamberlain, & Warner, 2009), stroke survivors (Davis, Egan, Dubouloz, Kubina, & Kessler, 2013), and the earlier report on this sample of youth with type 1 diabetes (Helgeson & Takeda, 2009) have linked project stressfulness to increased psychological distress and poor psychological well-being. By contrast, project progress has been linked to good psychological outcomes in the study of adults with low back pain (Vroman et al., 2009) and adolescents with type 1 diabetes (Helgeson & Takeda, 2009).

Here, we expanded on previous research—including our own—by examining personal projects at the cusp of adulthood or what has now been termed “emerging adulthood” (Arnett, 2000) rather than middle adolescence as in the earlier study (Helgeson & Takeda, 2009). Emerging adulthood, defined as the period between ages 18 and 25, is an optimal time to study personal projects, as it is a period in life when the possibilities for the future seem the most numerous and identity exploration is at its lifetime peak. Emerging adults are exploring the possibilities for their futures with respect to residential status, school status, vocation, romantic relationships, and identity. Because these tasks are the developmental milestones of emerging adulthood, this is an optimal period of time to study personal projects (Arnett, 2000, 2004). Facing decisions in these life domains provides an opportunity to set goals and identify projects to pursue.

In a previous report on this sample, we compared emerging adults with and without diabetes on psychological well-being indicators and found no group differences in depressive symptoms, but group differences on other well-being measures appeared (Palladino et al., 2013). Specifically, those with diabetes reported lower levels of life satisfaction than those without diabetes. In addition, group differences in life purpose emerged over the year following high school graduation, such that life purpose (i.e., feeling that life is meaningful) increased for controls over the year but remained the same for those with diabetes. An analysis of emerging adults’ personal projects had the potential to shed light on these findings by providing some information about the projects in which the two groups were engaged, whether the projects were equally important to those with diabetes and controls, and whether those with diabetes and controls made equal progress on the projects over the course of the year. Thus, we elicited the projects in which emerging adults with and without diabetes were engaged, examined the outcomes of those projects over the course of the year, and linked project dimensions to well-being.

The study had three goals. First, we explored whether there were group differences in the nature of projects elicited, engagement in projects, and completion of projects over time. We largely viewed this question as exploratory. However, given the above group differences in psychological well-being, one might predict that those with diabetes will be less engaged in projects and less likely to complete projects than healthy controls. Second, we examined whether project dimensions predicted project completion over the next year and whether those predictors differed by group. The first part of this goal was hypothesis-driven, as the previous literature leads to the prediction that project importance, progress, and typicality would be linked to higher rates of completion, whereas project stress would be linked to lower levels of completion. The second part of this goal is exploratory, as we had no reason to predict that project dimensions would be differentially predictive of completion by group. Third, we examined whether project completion and project dimensions predicted a broad set of psychological well-being outcomes that included depressive symptoms, life purpose, life satisfaction, perceived stress, and resilience. For those with diabetes, we also examined whether the extent to which diabetes interfered with projects predicted the same outcomes. This goal was hypothesis-driven, as previous literature and theory suggested that project importance, progress, and typicality would predict higher levels of psychological well-being, and project stress would predict lower levels of psychological well-being. For those with diabetes, we hypothesized that diabetes interference with projects would predict lower levels of psychological well-being.

Method

Recruitment

Participants were recruited from a previous study (Helgeson, Snyder, Escobar, Siminerio, & Becker, 2007) when they were average age 12. Those with diabetes were recruited from Children’s Hospital of the University of Pittsburgh (n = 132), and healthy controls were recruited from primary care physician offices as well as area malls (n = 131). Recruitment details are provided elsewhere (Helgeson et al., 2007). Sample characteristics at recruitment when participants were average age 12 are shown in Table I. These are the sample characteristics for those who completed the earlier personal project assessment during adolescence. Demographic information for the sample examined in the present study (7 years later; age 19) is also provided in Table I (Time 1). Note that 89% of those with diabetes and 92% of controls were retained from the original study. Comparisons on these demographic and disease variables showed no differential attrition.

Table I.

Demographics of the Sample (Percent, Ms, and SDs)

	Recruitment		Time 1
	Controls	Diabetes	Controls	Diabetes
n	131	132	121	117
Sex	49% male	47% male	47% male	47% male
Age (years)	12.07 ± .69	12.10 ± .77	19.02 ± .49	19.14 ± 0.40
Race	91% White	93% White	93% White	93% White
Length of diabetes (years)		4.91 ± 2.98		12.06 ± 3.08
Parent social statusa	46.40 ± 13.31	41.97 ± 11.05	46.44 ± 13.64	42.45 ± 11.09
Body mass index	21.63 ± 4.37	22.05 ± 4.36	24.82 ± 4.81	26.12 ± 4.01

^aParent social status was measured with the Hollingshead (1975) index, composed of occupational status and education. These data were only collected at recruitment when youth were average age 12.

Procedure

The project was approved by the institutional review boards of Carnegie Mellon Univeristy and University of Pittsburgh Children’s Hospital. One year following high school graduation, when participants were average age 19, they were contacted to see if they would be able to complete an online questionnaire. If agreeable, informed consent was obtained, and a link to the questionnaire was e-mailed to the participant. As shown in Table I, 117 (89%) of those with diabetes and 121 (92%) of those without diabetes completed this questionnaire 7 years after initial recruitment. At this time, 78% of those with diabetes and 74% of controls were in school, and, of those, the vast majority were full-time college students (92% diabetes; 94% controls). The online questionnaire asked participants to identify five personal projects and assessed the well-being variables described below. One year later (Time 2), participants were contacted to complete a second online questionnaire. The second online questionnaire was tailored to the participant, so that the specific personal projects identified 1 year earlier were listed. Among those with diabetes, four did not complete the follow-up questionnaire and an additional three did not respond to the personal projects section, leaving us with a final sample of 96 at Time 2. Among those without diabetes, four participants did not complete the follow-up questionnaire and an additional three did not respond to the personal projects section, leaving us with a final sample of 114 at Time 2. Participants were paid $50 at each assessment.

Personal Projects

Consistent with the methodology of Little (1993), participants were instructed:

We are interested in studying the kinds of activities and concerns that young adults have. We call these personal projects. All of us have a number of personal projects at any given time that we think about, plan for, carry out and sometimes (though not always) complete. Some projects may be focused on achievement and others on process; they may be things we choose to do or things we have to do; they may be things we are working towards or things we are trying to avoid. Projects may be related to any aspect of your daily life; school, work, home, or leisure. Some examples are: pass a class, cut down on junk food, clean my room, finish school. Think about as many personal projects and activities that you are currently engaging in or considering over the next couple months—these need not be formal projects or even important ones. Take about 5 minutes and think of 5 projects.

Participants were asked to identify up to five personal projects. Projects were later grouped into seven categories by two independent coders (third coder resolved discrepancies): work/school (e.g., get good grades, go to graduate school, and find a better job), relationships (e.g., spend more time with friends, date someone, and reconnect with people), hobbies/leisure (e.g., travel, play more video games, and pursue art), health (e.g., exercise and quit smoking), appearance (e.g., lose weight, build muscle, and change hair), financial (e.g., buy a car and save money), and self-improvement (e.g., be less forgetful and become more organized). Inter-rater reliability was .78.

At Time 1 and Time 2, participants rated each project on the following four dimensions: importance, typicality, stress, and progress. The wording was as follows: thinking about this project, to what extent: (a) is it important to you, (b) is it typical of you, (c) do you feel stressed while doing or thinking of it, and (d) have you made progress on it. The response scale for all items ranged from 1 (not at all) to 6 (very much). With the same response scale, those with diabetes were also asked to indicate how much diabetes interferes with the project. Correlations among the five dimensions are shown in Table II.

Table II.

Project Dimension Means and Intercorrelations

	T1 mean	Importance	Typicality	Stress	Progress	Interfere (diabetes only)
Importance	5.17		.56****	.18***	.47****	.09
Typicality	4.50	.70****		.14**	.57****	.01
Stress	3.36	.32****	.25****		.06	.42****
Progress	4.13	.58****	.53****	.07		.17*
Interfere (diabetes only)	2.17	.10	.10	.69****	−.05

Note. Dimensions were rated on a 1–6 scale; correlations for Time 1 are shown above the diagonal; correlations for Time 2 are shown below the diagonal.

*p < .10; **p < .05; ***p < .01; ****p < .001.

Psychological Well-Being

Depressive Symptoms

Depressive symptoms were measured with the Center for Epidemiological Studies Depression Scale (Radloff, 1977) which consists of 20 summed items, each of which is rated on a 0 (none of the time) to 3 (most of the time) scale over the past week (T1: α = .93; T2: α = .92).

Purpose

The seven-item purpose in life subscale from Ryff’s Six Dimensions of Psychological Well-Being (Ryff & Keyes, 1995) was used to measure life purpose. Participants rated their agreement with items on a scale from 1 (strongly disagree) to 6 (strongly agree), and responses were averaged to create an index in which higher numbers indicated a stronger sense of life purpose (T1: α = .77; T2: α = .79). Sample items included “I have a sense of direction and purpose in life” and “My daily activities often seem trivial and unimportant to me” (the latter reverse scored).

Life Satisfaction

Life satisfaction was measured with the five-item Life Satisfaction Inventory (Diener, Emmons, Larsen, & Griffin, 1985). Items were rated on a 1 (strongly disagree) to 7 (strongly agree) scale, and averaged (T1: α = .92; T2: α = .88). Sample items included “In most ways my life is close to my ideal” and “I am satisfied with my life.”

Perceived Stress

We used the four-item abbreviated Perceived Stress Scale (Cohen, Kamarck, & Mermelstein, 1983). Responses ranged from 1 (never) to 5 (very often). Internal consistencies were good (α = .76 at T1; α = .75 at T2). Sample items included “How often have you felt difficulties were piling up so high that you could not overcome them?” and “How often have you felt that things were going your way?” (the latter reverse scored).

Resilience

We used indicators of cognitive adaptation theory (Taylor & Brown, 1988) to measure resilience: self-esteem, mastery, and optimism. Self-esteem was measured with the 10-item Rosenberg (1965) self-esteem scale (T1: α = .91; T2: α = .89). Mastery was measured with the seven-item Pearlin and Schooler (1978) mastery scale (T1: α = .80; T2: α = .75), and optimism was measured with the Life Orientation Test (Scheier & Carver, 1985; T1: α = .78; T2: α = .83). Respondents were asked to indicate the extent to which they agreed with each item on a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). All of these measures are widely used in the literature. Their intercorrelation ranged from .64 to .73 at T1 and .67 to .71 at T2. Consistent with the methodology of previous research (Helgeson, 2003; Ratelle, Vallerand, Chantal, & Provencher, 2004), we standardized the total scores on each of the three scales and averaged the three standardized scales to form a single resilience index.

Overview of the Analysis

All Analyses Were Completed With SPSS Version 24

First, we used chi-square analysis to compare diabetes and control participants on the project categories. Then, we compared the two groups on the four project dimensions over time with a repeated measures analysis of covariance. Diabetes versus controls was the between-groups factor, time was the repeated measure, and covariates were parent social status and current body mass index. These were the only two demographic or disease variables shown in Table I on which the groups differed, so they were statistically controlled in all subsequent analyses. We used an analysis of covariance to examine group differences in project completion.

We used regression analysis to predict project completion by entering all four project dimensions into the equation on the first step of the equation along with group. We entered the group by project dimension interactions on the second step to see if dimensions differentially predicted project completion for diabetes and control groups. Significant group by project dimension interactions was examined by plotting the simple slopes for diabetes and controls.

We used a similar set of regression analyses to predict Time 1 psychological well-being. Again, we entered all four project dimensions into the equation on the first step along with group and then entered the four group by dimension interactions on the second step of the equation. Because only participants with diabetes completed the interference dimension, we examined whether diabetes interference predicted each of the psychological well-being outcomes in separate analyses. We also used regression analyses to predict Time 2 psychological well-being by additionally controlling for Time 1 well-being on the first step of the equation. At Time 2, we examined whether project completion predicted psychological well-being. In a separate analysis, we also examined whether project completion predicted changes in psychological well-being between Time 1 and Time 2.

All persons who identified projects completed all the project dimensions and the well-being measures. Thus, there were no missing data.

Results

Group Comparisons on Nature of Personal Projects

Among those with diabetes, four persons did not list any personal projects, but all controls identified at least one project. Those with diabetes identified an average of 4.51 projects, and those without diabetes identified an average of 4.58 projects.

Frequencies of the extent to which respondents endorsed each of the seven categories of projects are shown in Columns 1 and 2 in Table III for those with and without diabetes. The most frequent category of project identified had to do with work/school. Hobbies, health, and relationships were named by between one-third and one-half of the sample. Only one group difference appeared. As shown in Table III, control participants were more likely than those with diabetes to identify projects that involved finances (31% vs. 18%).

Table III.

Project Categories and Completion Rates by Group

	Project engagement			Project completion
Type of project	Diabetes (%)	Controls (%)	Significance	Diabetes	Controls	Significance
Work/school	95	90	n.s.	.54	.68	* (eta2 = .03)
Hobbies/leisure	50	41	n.s.	.56	.66	n.s.
Health	39	43	n.s.	.58	.60	n.s.
Relationships	35	41	n.s.	.48	.76	* (eta2= .07)
Financial	18	31	*	.53	.57	n.s.
Appearance	18	22	n.s.	.15	.50	*(eta2 =.13)
Self-improvement	28	20	n.s.	.68	.75	n.s.
Total				.65	.74	* (eta2 = .02)

Note. Project engagement numbers do not add up to 100% because participants described up to five projects; total project completion rate includes ALL projects, including those not in the listed categories; n.s. = not significant;.

*p < .05.

Group Comparisons on Engagement in Projects

The repeated measures analysis of covariance revealed a significant group difference on progress, F(1, 217) = 3.89, p = .05, eta² = .02. Controls reported greater project progress (M = 4.36, SE = .08) than those with diabetes (M = 4.12, SE = .09). There were no group differences on project importance, typicality, or stress, but there were main effects of time on typicality of project, F(1, 217) = 3.90, p = .05, eta² = .02, and project stress, F(1, 217) = 4.37, p < .05, eta² = .02, such that both decreased over time. There also were no group by time interactions, such that the group effects reported above persisted over time.

Group Comparisons in Project Completion

As shown in Table III, when we examined whether projects were completed over the past year, control participants completed a larger portion of their projects than those with diabetes (74% vs. 65%), F(1, 220) = 4.87, p < .05, eta² = .02. When we examined individual categories of projects, controls were more likely than those with diabetes to complete projects in three categories: work, relationships, and appearance.

Prediction of Completion

When the four projects dimensions at Time 1 were entered simultaneously into a regression analysis to predict project completion at Time 2, only progress emerged as a significant predictor (b = .27, p < .001). Participants who reported more progress at Time 1 were more likely to have completed their project 1 year later. There were no interactions of any project dimension with group. For those with diabetes, we examined whether diabetes interference predicted project completion. It did not.

Prediction of Psychological Well-Being

Time 1 Psychological Well-Being

The results from multiple regression analyses to predict Time 1 well-being are shown in Table IV. When interactions were not significant, we show the final regression coefficients for the first step of the equation.

Table IV.

Prediction of Well-Being With Project Dimensions

	Depressive symptoms		Life purpose		Life satisfaction		Perceived stress		Resilience
	T1	T2	T1	T2	T1	T2	T1	T2	T1	T2
Time 1 DV		.46****		.63****		.58****		.34****		.59****
Parent social status	.05	.04	.05	.05	.03	.04	.07	.02	−.06	.00
Body mass index	−.10	.04	.00	−.03	.06	−.06	−.01	−.04	−.02	.00
Group	.05	.05	−.26	−.01	.46	.05	.08	−.07	−.09	.15
Importance	.18**	−.07	.25**	−.05	.01	−.02	.13*	−.03	.03	.07
Typicality	−.23***	−.14*	.28**	.10	.32**	.24***	−.27***	−.19**	.21**	.38***
Stress	.32****	.19***	−.07	−.05	−.01	−.06	.29****	.15**	−.29****	−.10
Progress	−.17**	.11	−.07	−.04	.02	−.03	−.18**	−.04	.18**	−.07
Importance × group			.02		−.53					.09
Typicality × group			−.44		−.02					−.63**
Stress × group			−.07		−.47**					.13
Progress × group			.58**		.34					.22

*p < .10; **p < .05; ***p < .01; ****p < .001.

Greater importance, less typicality, greater stress, and less progress were associated with higher levels of depressive symptoms. Higher project importance and typicality were linked with greater life purpose. Project progress interacted with group to predict life purpose. Decomposition of the interaction showed that progress was linked to greater purpose in life for those with diabetes (b = .15, p < .05) but not those without diabetes (b = −.04, p = .56), as shown in Supplementary Figure S1.

Project typicality was associated with higher life satisfaction, and project stress interacted with group to predict life satisfaction. Project stress was linked to lower life satisfaction for those with diabetes (b = −.35, p = .001) but not those without diabetes (b = −.02, p = .88), as shown in Supplementary Figure S2.

Lower project typicality, higher project stress, and lower project progress were associated with higher perceived stress. Higher project typicality, lower project stress, and higher project progress were associated with greater resilience. There were no group by project dimensions interactions to predict perceived stress or resilience.

When diabetes interference was examined by itself, it was associated with greater depressive symptoms (b = .22, p < .05) and greater perceived stress (b = .42, p < .001), but not life purpose, life satisfaction, or resilience.

Time 2 Psychological Well-Being

As shown in Table IV, controlling for Time 1 well-being, project stress predicted an increase in depressive symptoms. No dimension predicted changes in life purpose. Project typicality predicted increases in life satisfaction and decreases in perceived stress. Project stress predicted increases in perceived stress. Project typicality predicted increases in resilience and interacted with group. As shown in Supplementary Figure S3, the relation of project typicality to resilience was stronger for those without (b = .20, p = .001) than those with diabetes (b = .05, p = .27).

For those with diabetes, diabetes interference predicted increases in depressive symptoms (b = .34, p < .001) and perceived stress (b = .19, p = .05).

Finally, project completion was linked to significant increases in life satisfaction (b = .16, p < .01) between Time 1 and Time 2.

Discussion

One of the first goals of this study was to examine the nature of the projects that emerging adults with and without diabetes identified. Not surprisingly, the primary category of projects in which emerging adults engaged had to do with work and school. Vocational and school challenges are ones that Arnett (2000) identified with respect to emerging adulthood. Relationships were another prominent domain identified by participants, which is also consistent with the theoretical view of the issues emerging adults face. However, hobbies were the second most frequently identified projects, which suggests that emerging adults are also engaged in the pursuit of activities that are not necessarily linked to jobs and career plans.

Consistent with previous research on adults with multiple sclerosis (Brooke et al., 2007) and adolescents with cancer (Sulkers et al., 2015), there were few differences in the nature of the projects identified by those with and without diabetes. Group differences appeared on only one of the seven categories, and there is no strong theoretical rationale as to why controls were more likely to identify projects in the finance category compared with those with diabetes. Even with controls for parent social status, this group difference remained. Recall that there were no group differences in the education of participants, as those with and without diabetes were equally likely to be enrolled in college. Interestingly, there also were no group differences in health-related projects.

Our second study goal was to examine whether emerging adults with and without diabetes were equally like to complete their projects. The longitudinal nature of our work enabled us to determine the outcomes of projects identified 1 year earlier, whether there were group differences in project completion, which dimensions predicted completion, and whether completion had implications for well-being. Importantly, controls reported that they completed more of their projects 1 year later compared with those with diabetes. Relatedly, controls also reported that they had made more progress on their projects than those with diabetes from the start. However, there were no group differences in the stress associated with projects or whether the projects reflected them (i.e., typicality). One of the motivations for undertaking this study was to understand more about our previous finding that controls had reported higher life satisfaction and greater purpose in life than those with diabetes. Engagement in and fulfillment of personal projects could provide some insight into this story. To the extent that these projects are personally fulfilling, making progress and completing projects may lead to being more satisfied with one’s life and the direction it is taking.

We directly examined the links of project dimensions and project completion to psychological well-being. Each of the project dimensions had links to one or more of the outcomes, in the directions predicted: project progress was linked to decreased depressive symptoms and perceived stress and higher resilience, project importance was related to decreased depressive symptoms and higher life purpose, project stress was linked to increased depressive symptoms and perceived stress and decreased resilience, and project completion was linked to increases in life satisfaction. However, the one project dimension that revealed the most consistent links to psychological well-being was project typicality, consistent with Little’s (1987) theory. Emerging adults who identified projects that they said were typical of them, that is, personally reflected them, reported higher levels of psychological well-being and reported an increase in psychological well-being over the course of the year. Those who said the projects in which they were engaging were more typical of them reported an increase in life satisfaction and resilience and a decrease in stress over the next year. These findings are consistent with a synthesis of behavior change theories that concluded the pursuit of goals consistent with one’s self-image is a common determinant of goal pursuit (Fishbein et al., 2001).

Although our findings for project typicality are consistent with theory about the importance of pursuing projects consistent with one’s identity, these findings differ from those from the earlier report of these participants when they were adolescents (Helgeson & Takeda, 2009). In the earlier report, the project dimensions that revealed the most robust relations to well-being were project progress and project stress. The discrepancy in findings between the two studies may have to do with differences in the nature of projects pursued. Perhaps middle adolescents (average age 14) are more likely to be pursuing projects that are less typical of themselves and more a response to the concerns of others and their social environment. In fact, the mean score for the typicality dimension in the previous report was 3.69 compared with 4.50 in the present study. The most frequently identified category of projects that middle adolescents identified had to do with academics (e.g., get good grades, and get into college). This is similar to the most frequently identified category in this report—work/school. However, these pursuits may be viewed as less reflective of personal identity at age 14 than age 19. The second most frequent category of projects identified by middle adolescents was self-improvement (e.g., clean my room and listen to parents)—a category that was not as common among these emerging adults.

Project dimensions differentially predicted several of the psychological well-being outcomes for those with and without diabetes. However, the direction of the interactions was not consistent across these outcomes. Cross-sectional relations showed that progress was more strongly linked to higher life purpose, and stress was more strongly linked to lower life satisfaction for those with than without diabetes. However, longitudinal relations showed that project typicality was more strongly linked to an increase in resilience for controls than those with diabetes. We considered these analyses exploratory, as we did not have specific hypotheses. Given the inconsistency across dimensions and cross-sectional versus longitudinal analyses, we hesitate to overinterpret these findings.

For those with diabetes, we also explored whether diabetes interference with personal projects had implications for psychological well-being. There was some evidence that this was the case. Diabetes interference was cross-sectionally linked to higher levels of depressive symptoms and perceived stress and predicted an increase in depressive symptoms and perceived stress 1 year later. However, the extent to which diabetes interfered with projects was not related to project completion and was not related to the other well-being indicators. These findings speak to the larger picture that framed this study—which is that diabetes is only one part of the lives of these emerging adults. For emerging adults with diabetes, there is far more evidence that their well-being is a function of whether they are engaging in projects that they see as typical for themselves—which may mean freely chosen—than whether diabetes interferes with these projects. It also is important to note that overall levels of interference were quite low. These findings are somewhat consistent with those of a study of adults that examined the adoption of exercise (Presseau, Sniehotta, Francis, & Gebhardt, 2010). They found that whether personal projects interfered with the adoption of physical exercise did not predict whether participants exercised over 8 weeks.

One limitation of this research is the use of single items to capture project dimensions. We used single-item measures because of time constraints. Although single-item measures typically have reduced reliability compared with multiple-item measures (Diamantopoulos, Sarstedt, Fuchs, Wilczynski, & Kaiser, 2012; Wanous, Reichers, & Hudy, 1997), single-item measures have been shown to be useful under some conditions. Single-item measures are regarded more positively when the sample size is small, the research is exploratory, and the construct measured is concrete, unambiguous, and face valid (Bergkvist & Rossiter, 2007; Diamantopoulos et al., 2012; Wanous et al., 1997). It is the last condition that applies to the personal project dimensions. Although personal project research historically has relied on single-item scales to assess project dimensions, the extent to which these single-item measures correspond to multiple-item measures is not known and is a direction for future research.

These findings have implications for how health-care providers interact with youth with and without diabetes. Given the escalating rate of mental health problems in youth (Mojtabai, Olfson, & Han, 2016), greater attention is needed to overall levels of psychological well-being. Health-care professionals could engage youth in a conversation about their personal goals, whether they are pursuing goals that interest them, and what the barriers and facilitators are toward completion. Because the pursuit of “typical” goals was most strongly connected to well-being—including changes in well-being over time, a discussion of the reasons or motivations behind the specific goals identified could be useful in determining whether they are more or less reflective of the youth’s personal identity.

For those who treat youth with diabetes, it may beneficial to expand discussions outside the realm of diabetes to talk about other goals youth have set for themselves and barriers and facilitators of those goals. Our longitudinal data showed that youth with diabetes are pursuing goals that are less important to them, making less progress on those goals, and ultimately completing fewer of them. Although youth reported low levels of diabetes interference with goals, there is the possibility that youth may not be aware of all the ways in which diabetes affects goal pursuit and progress. In addition, it is possible that the pursuit of other goals interferes with the management of diabetes. Goal-setting is an integral part of diabetes care, consistent with the recommendations of the American Diabetes Association (2018). Perhaps, goal-setting discussions around self-care and glycemic control could be expanded to include other goals youth have for themselves and how diabetes and non-diabetes goals intersect. The overall message for health-care providers is to recognize that young adults with diabetes have relational, vocational, and interpersonal goals outside of diabetes that could benefit from expressions of interest and discussion in terms of overall psychological well-being as well as diabetes health.

Supplementary Data

Supplementary data can be found at: https://academic.oup.com/jpepsy.