Additive,Curvilinear,and Interactive Relations of Anxiety and Depression With Indicators of Psychosocial Functioning

Abstract

This study examines additive, curvilinear, and interactive relations of anxiety and depression with several subjective indicators of intrapersonal (i.e., hope, self-compassion, shame) and interpersonal (i.e., social connectedness, quality of social relationships) functioning in a sample of adults (N = 547, M_age = 43.37 ± 12.02, female = 56.88%) seeking treatment for psychological difficulties. Results of complementary analyses were largely consistent with the hypothesis that increasing levels of anxiety and depression would correspond with worse psychosocial functioning, although nonlinear relations indicated that the effect of depression progressively attenuated at higher levels of symptom severity. Whereas the findings generally supported additive effects of anxiety and depression, the hypothesis that there would be synergistic effects of anxiety and depression was not supported. Supplementary group comparisons revealed that the functional implications of subsyndromal combinations of anxiety and depression may be comparable to those associated with symptoms that meet more traditional standards (i.e., syndromal or dimensional definitions) of comorbid anxiety–depression. The findings offer further insight into the complex relations of anxiety and depression with psychosocial functioning and emphasize the importance of detecting and offering appropriate treatments for anxiety and depression symptoms that coexist at subsyndromal levels.

Keywords

Anxiety depression comorbidity impairment psychosocial functioning

Introduction

Mental illness is one of the largest contributors to the global burden of disease and accounts for 32.4% of disability worldwide (Vigo et al., 2016). Of the mental illness categories identified among the top 20 contributors to the global burden of disease (see Vos et al., 2015), those ranked highest are major depression (second) and anxiety disorders (ninth). Global prevalence estimates (e.g., Baxter et al., 2014) of major depressive disorder (4.4%) and anxiety disorders (4%) underscore the magnitude of the burden that each of these categories of mental illness carries. Yet, anxiety and depression are highly comorbid (Kroenke et al., 2016). A recent cross-national epidemiological analysis involving 74,045 adults from 24 countries estimated the 12-month prevalence of major depressive disorder comorbid with at least one 12-month anxiety disorder to be 41.6% (Kessler et al., 2015). Simultaneous co-occurrence may reflect a core transdiagnostic mechanism that underlies the cognitive vulnerabilities of anxiety and depression (Hong & Cheung, 2015). Psychometric analysis has found support for a bifactor structure of anxiety and depression, which suggests an underlying dimension may be common to both psychopathologies (Kroenke et al., 2016). An alternative view is that anxiety represents a vulnerability phenotype that increases the risk of stress-induced depression (Weger & Sandi, 2018). This proposition is consistent with epidemiological evidence that has found a majority of individuals with anxious major depressive disorder tend to report earlier onset of anxiety disorders than major depressive disorder (Kessler et al., 2015).

Alongside progress that has been made toward understanding the causal determinants of comorbid anxiety–depression, researchers have sought to determine whether the debilitating effects of either disorder occurring in isolation differs from co-occurring anxiety and depression. In a review of neurobiological evidence from 24 studies, Ionescu, Niciu, Mathews, et al. (2013) assert that anxiety and depression comorbidity is neurobiologically different from depression without anxiety and is associated with distinct psychosocial outcomes. Evidence from a range of studies (e.g., Fusar-Poli et al., 2014; Saris et al., 2017; Zhou et al., 2017) suggests that relations of comorbid anxiety and depression with diminished psychological and interpersonal functioning are amplified when compared to the effects of anxiety or depression alone.

Although anxiety and depression comorbidity has generally been linked to less favorable psychosocial functioning, there are several reasons for further examining anxiety and depression in combination. First, existing research in this area has often relied on modeling approaches that emphasize linear additive effects of anxiety and depression (e.g., Brown et al., 2010), thereby overlooking the salience of nonlinear and multiplicative effects that could provide useful insight into the complex interrelations of anxiety and depression. Second, researchers have typically applied syndromal (i.e., concurrent formal diagnosis of major depressive disorder and one or more anxiety disorder) or dimensional (i.e., diagnosis of major depressive disorder based on formal diagnostic criteria, along with concurrent anxiety symptoms that meet a given threshold on psychometrically validated measures) definitions to classify comorbid anxiety and depression (Ionescu, Niciu, Henter, et al., 2013). While a finite approach to classifying comorbidity is advantageous because it aligns more closely with formal diagnostic criteria that may be of clinical significance, overlapping symptoms of anxiety and depression frequently coexist at subsyndromal levels and can have similar effects on psychosocial functioning compared to symptoms that meet diagnosable levels (Wanders et al., 2016). Therefore, consideration of subsyndromal levels of anxiety and depression may be important for early detection of individuals at risk of symptom progression, formulating prognoses, and effectively treating specific configurations of anxiety and depression. In the current study, a multi-analytic approach is used to examine relations of anxiety and depression with several subjective markers capturing overarching intrapersonal and interpersonal domains of psychosocial functioning.

Anxiety, depression, and subjective indicators of psychosocial functioning

At the intrapersonal level, criterion variables of interest included hope, self-compassion, and shame. The concept of hope refers to a future-oriented thought process involving a person’s self-perceived capacity to find practical means (i.e., pathways) of reaching their goals and the motivation (i.e., agency) to attain them (Snyder, 2005). According to hope theory (for an overview, see Snyder et al., 2002), the pathways and agency components represent distinct processes that are both necessary ingredients for successfully yielding hopeful thinking. An abundance of research (for reviews, see Gallagher & Lopez, 2018) has linked hope with lower levels of internalizing symptoms (e.g., anxiety, depression), although evidence suggests that hope and psychological well-being have a dynamic influence on each other (Gum, 2018). For example, symptoms consistent with anxiety (e.g., persistent worry) or depression (e.g., apathy) may undermine goal-oriented thought processes (Kashdan et al., 2006; Thimm et al., 2013) that could promote adaptive goal pursuits and lead to improved mental health (Snyder et al., 2002).

Self-compassion is a positive self-orientation that involves being kind, loving, and understanding toward the self during instances of difficulty, inadequacy, and disappointment (Neff, 2003a). Conceptualizations of self-compassion assert that it buffers against the negative effects of self-denigrating cognitive-emotional experiences (e.g., self-criticism) that ordinarily accompany failure, thereby supporting psychological well-being (Neff, 2003b). Self-compassion may also be facilitated when lower levels of psychopathology (e.g., anxiety, depression) are experienced, with more severe psychopathology a potential barrier to taking a compassionate stance toward the self (MacBeth & Gumley, 2012). In support of theoretical links between self-compassion and mental health, findings of various studies indicate that self-compassion is inversely associated with symptoms of anxiety and depression (for a review, see Barnard & Curry, 2011).

Compared to the supportive role that hope and self-compassion play in sustaining well-being, shame is often associated with impaired psychological functioning. Shame is a dysphoric state involving perceived inferiority about the global self (Leach, 2017). Particularly when shame reflects a view that the whole self is unalterably flawed, it can have a profound impact on self-worth and debilitating mental health consequences (for an overview, see Cibich et al., 2016). Although experienced shame may heighten vulnerability to anxiety and depression, the onset of such internalizing symptoms might also be interpreted as evidence of a flawed self and therefore precipitate feelings of shame (for meta-analytic reviews, see Cândea & Szentagotai-Tăta, 2018; Kim et al., 2011).

Interpersonal functioning was assessed via two criterion variables, namely, social connectedness and perceived quality of social relationships. Social connectedness is an attribute reflecting one’s sense of self in relation to the social world (Lee & Robbins, 2000). In contrast to individuals with higher social connectedness, those on the lower end of the continuum have difficulty identifying with and feeling close to others, experience discomfort in social settings, and feel misunderstood or isolated (Lee et al., 2001). As people are inherently social beings, these kinds of relational challenges may evoke distress and contribute to psychological maladjustment, including depression and anxiety (see Lee & Robbins, 1998; Williams & Galliher, 2006). Similarly, mental health is also affected by the quality of a person’s social relationships. Findings of numerous studies link depression and anxiety with impairments on a range of behavioral (e.g., social support) and affective (e.g., belongingness) indicators of social functioning (e.g., Cacioppo et al., 2010; Saris et al., 2017), the effects of which can further undermine the quality of relationships a person is able to initiate and maintain (see Trompenaars et al., 2007).

The present study

Extending previous research, this study uses complementary analytical approaches to examine the independent and joint effects of anxiety and depression on indicators of intrapersonal and interpersonal functioning. It was expected that (a) outcomes resonating with worse intrapersonal (i.e., lower hope and self-compassion, higher shame) and interpersonal (i.e., lower social connectedness and quality of social relationships) functioning would be found among individuals with more severe symptoms of anxiety and depression, and (b) anxiety and depression would be synergistically associated with psychosocial functioning, such that relations of depression with lower quality psychosocial functioning would be amplified at higher levels of anxiety.

Methods

Participants

The sample (N = 547) comprised of individuals who had elected to attend a short-term residential psychotherapeutic treatment program (see Table 1, for sample characteristic details). Participants ranged between 19 and 73 years of age (M_age = 43.37, standard deviation_age = 12.02), the majority of which were females (56.88%), self-identified as White (92.31%), and were married or in a domestic partnership (51.55%). Most had completed a bachelor’s degree or higher (72.76%), were currently employed (78.90%), and affiliated religiously with Christianity (63.79%).

Table 1.

Sample characteristics (N = 547).

Mean age, years ± SD (range)	43.37 ± 12.02 (19–73)
Sex, n (%)	545
Female	310 (56.88)
Male	235 (43.12)
Race/ethnicity, n (%)	546
White	504 (92.31)
Minority	42 (7.69)
Highest education, n (%)	547
Less than a high school diploma	2 (0.37)
High school diploma or equivalent	118 (21.57)
Associate’s degree	29 (5.30)
Bachelor’s degree	241 (44.06)
Graduate degree (e.g., MA, MD, PhD)	157 (28.70)
Marital status, n (%)	547
Single (never married)	131 (23.95)
Married or in a domestic partnership	282 (51.55)
Widowed	15 (2.74)
Divorced	74 (13.53)
Separated	45 (8.23)
Employment status, n (%)	545
Employed part-time or full-time	430 (78.90)
Unemployed	41 (7.52)
Full-time student	12 (2.20)
Retired	34 (6.24)
Homemaker	28 (5.14)
Religious affiliation, n (%)	533
Agnostic	40 (7.50)
Areligious/nonspiritual	60 (11.26)
Atheistic	12 (2.25)
Buddhist	6 (1.13)
Christian	340 (63.79)
Hindu	2 (0.38)
Jewish	27 (5.07)
Spiritual	46 (8.63)

Note. SD: standard deviation.

Measures

Anxiety

The generalized anxiety disorder seven-item (GAD-7; Spitzer et al., 2006) scale was used to measure the severity of generalized anxiety symptomology experienced by participants during the past two weeks (e.g., “Feeling nervous, anxious or on edge”). A four-point scale (0 = not at all; 3 = nearly every day) is used to rate each item. A recent meta-analysis (Plummer et al., 2015) found that sensitivity (.74) and specificity (.83) for detecting generalized anxiety disorder at a cut-off score of ≥10 were comparable to those reported in the initial validation study. A cut-off score of ≥10 on the GAD-7 has also moderate sensitivity (.66 to .74) and specificity (.80 to .81) for detecting other anxiety-related disorders, including posttraumatic stress disorder, panic disorder, and social anxiety disorder (Kroenke et al., 2007). Categories typically used to delineate severity of anxiety on the GAD-7 (see Spitzer et al., 2006) include minimal (≤4), mild (5–9), moderate (10–14), and severe (≥15). Internal consistency of the GAD-7 in the current sample was ω_t = .92.

Depression

Depression symptoms were measured using the Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001). The nine-item scale assesses the frequency of depressive symptomology experienced during the past two weeks (e.g., “Feeling down, depressed, or hopeless”). Participants use a four-point scale (0 = not at all; 3 = nearly every day) to respond to the items. Several meta-analyses have revealed an optimal sensitivity and specificity criterion of the PHQ-9 for detecting major depressive disorder is ≥10 (see Levis et al., 2019; Mitchell et al., 2016). Commonly used categories to classify depression severity on the PHQ-9 (see Kroenke et al., 2001) include minimal (≤4), mild (5–9), moderate (10–14), moderately severe (15–19), and severe (≥20). In this sample, internal consistency of the PHQ-9 was ω_t = .90.

Hope

Participants completed the Adult State Hope Scale (ASHS; Snyder et al., 1996). The ASHS contains six items designed to measure goal-directed thinking (e.g., “I can think of many ways to reach my current goals”), which are evenly distributed across two subscales (i.e., agency and pathways). Items are rated on an eight-point scale (1 = definitely false; 8 = definitely true), with subscale scores added for a total hope score (Snyder et al., 1996). The convergent, concurrent, and discriminant validity of the ASHS has been supported in several studies (Curry et al., 1997; Martin-Krumm et al., 2015; Snyder et al., 1996). Estimated internal consistency of total ASHS scores reported in prior studies was ≥.71 (Martin-Krumm et al., 2015; Snyder et al., 1996). In this study, the internal consistency of the total ASHS was ω_t = .90.

Self-compassion

The Self-Compassion Scale (SCS; Neff, 2003b) measured the tendency of participants to be caring, kind, and understanding toward themselves during moments of difficulty, inadequacy, or failure. The 26-item scale consists of three sets of subscales with contrasting poles: self-kindness (five items) and self-judgment (five items), common humanity (four items) and isolation (four items), and mindfulness (four items) and overidentified (four items). Participants use a five-point scale (1 = almost never; 5 = almost always) to rate each item (e.g., “I’m kind to myself when I’m experiencing suffering”). Evidence from clinical and nonclinical samples of various countries supports the factorial, convergent, and criterion validity of the SCS (Neff, 2003b; Neff et al., 2017, 2019). Internal consistency values of total SCS scores range from .91 to .95 (Neff, 2003b; Neff et al., 2017), and test–retest reliability estimates over various intervals (e.g., three weeks, four weeks) were ≥.78 (Castilho et al., 2015; Neff, 2003b). A total self-compassion score was used in this study, which was generated by reverse scoring the negative subscale items, calculating subscale means, and adding together the means of all six subscales (ω_t = .95).

Shame

Shame was measured using the 25-item Experience of Shame Scale (ESS; Andrews et al., 2002). The ESS captures various aspects of characterological (12 items), behavioral (nine items), and bodily shame (four items). Using a four-point scale (1 = not at all; 4 = very much), participants rated the items (e.g., “Have you felt ashamed of the sort of person you are?”) with reference to the past month. Findings of several studies support the concurrent, convergent, and incremental predictive validity of the ESS (Andrews et al., 2002; Vizin et al., 2016). Estimated internal consistency and test–retest reliability of total ESS scores over an 11-week interval were .92 and .83, respectively (Andrews et al., 2002; Cheung et al., 2004). In this study, total scores were derived by summing responses to all 25 items (ω_t = .96).

Social connectedness

The Social Connectedness Scale—Revised (SCS-R; Lee et al., 2001) was used to assess participants’ sense of interpersonal closeness. The measure consists of 20 items (e.g., “I feel close to people”) that are rated on a six-point scale (1 = strongly disagree; 6 = strongly agree). After reverse scoring the 10 negatively phrased items, responses are summed for a total score. Research supports the convergent and discriminant validity of the SCS-R (Lee et al., 2001, 2008). Prior studies have reported internal consistency of scores for the SCS-R that range from .92 to .94 (Grieve et al., 2013; Lee et al., 2001, 2008). In the current sample, estimated internal consistency of the SCS-R was ω_t = .95.

Social relationship quality

Participants completed the three items comprising the social relationships domain of the World Health Organization Quality of Life—BREF scale (WHOQOL-BREF; The WHOQOL Group, 1998). The items (e.g., “How satisfied are you with your personal relationships?”) measure perceived quality of interpersonal relationships, each of which is rated on a five-point scale (1 = not at all; 5 = completely). Scale scores are calculated by multiplying the mean of the three items by four. The social relationships domain of the WHOQOL-BREF has demonstrated appropriate convergent and discriminant validity in samples from a variety of countries (Hawthorne et al., 2006; Skevington & McCrate, 2012). The estimated internal consistency of scale scores reported in previous research ranges from .66 to .68 (Skevington et al., 2004; The WHOQOL Group, 1998). Internal consistency of the social relationships domain in this study was ω_t = .70.

Physical health quality

Drawing on evidence linking internalizing symptoms and physical health problems (see Scott et al., 2016), the physical health domain of the WHOQOL-BREF scale (The WHOQOL Group, 1998) was used to measure perceived quality of physical health for inclusion as a covariate in the primary analyses. A five-point scale (1 = not at all; 5 = completely) is used to rate each of the seven items (e.g., “How well are you able to get around?”), two of which are reverse coded. Scale scores are derived by multiplying the mean of the seven items by four. Evidence supports the convergent and discriminant validity of the physical health domain (Hawthorne et al., 2006; Skevington & McCrate, 2012). Internal consistency values for the scale were ≥.82 across a number of studies (Skevington et al., 2004; The WHOQOL Group, 1998). In this study, internal consistency for this index was ω_t = .80.

Procedure

This study utilizes baseline data from an ongoing longitudinal intervention project assessing the efficacy of a short-term residential multimodal experiential psychotherapy program. Data were collected between January 2018 and February 2019. Ethical approval to conduct this project was granted by an independent review board (protocol 2017/11/23). Upon arrival at the program premises, program attendees were invited to participate in the study. Interested individuals were directed to a large group room where they were presented with information about the nature of the project and their participation in it. Those who agreed to participate provided written informed consent and then self-completed the battery of measures. Participants first completed a set of sociodemographic items followed by a randomized ordering of the measures.¹

Statistical approach

Data preparation and exploratory analyses

The data were initially screened for missing values. A negligible quantity (0.13%) of item-level data was missing. They were imputed using a nonparametric random forest technique (Stekhoven & Bühlmann, 2012) based on 10,000 iterations (proportion falsely classified = .22). Standardized values of the primary study variables were screened for univariate outliers (z > |3.29|, p < .001), three of which were found (z ≥ |3.51|). Mahalanobis distance, χ²(8) = 26.12, p < .001, detected three multivariate outliers (D² ≥ 29.34). An a priori decision was taken to omit all outliers before proceeding with subsequent analyses. Omega total (ω_t) was used to estimate the internal consistency reliability of all measures (Revelle & Zinbarg, 2009), a procedure that is favored over alternative approaches (e.g., alpha) because it makes fewer assumptions about the internal structure of a scale (Peters, 2014). Internal consistency values for all measures were ≥.70. Univariate skewness (maximum = |0.59|) and kurtosis (maximum = |0.88|) did not reveal any preliminary concerns with normality. Pearson correlations were used to explore bivariate relations among the primary study variables.

Polynomial regression of anxiety and depression

To examine additive and synergistic effects of anxiety and depression, separate polynomial regression models were performed by regressing each of the outcome variables (Z) included in this study (i.e., hope, self-compassion, shame, social connectedness, and quality of social relationships) on anxiety (X), depression (Y), and the interaction term between anxiety and depression (X × Y). Regression models were estimated with quadratic terms (i.e., squared value) of anxiety (X²) and depression (Y²) included, as an interaction effect involving highly correlated variables (as is the case with anxiety and depression) may mask an underlying nonlinear effect (Matuschek & Kliegl, 2018) and lead to spurious conclusions about synergistic effects (i.e., an interaction effect may be observed when there is no true interaction). Anxiety and depression were mean centered prior to model estimation (Schielzeth, 2010). Age, sex, race/ethnicity, and quality of physical health were included as covariates in each model. When evidence of a quadratic effect emerged, the two-line approach (see Simonsohn, 2018) was used to test whether the quadratic function conforms to a U-shape association with an outcome. The approach uses an algorithm to identify an optimal breakpoint along the predictor and interrupted regressions to estimate average slopes, one for low values (i.e., below the breakpoint) and another for high values (i.e., above the breakpoint) of the predictor. The procedure provides an indication of whether the signs of the average slopes derived for low and high values of a predictor vary, thereby benefiting the interpretation of nonlinear effects. For each regression model, visual inspection of quantile-quantile (QQ) plots and residual plots produced via a Wallyplot technique (for an overview, see Ekstrøm, 2014) indicated that the residuals appeared approximately normal and homoscedastic in distribution. There were also no concerns with multicollinearity (i.e., all variance inflation factor values ≤4.14).

Anxiety–depression group comparisons

To further explore relations of anxiety and depression in combination, comparative analyses were used to identify whether distinct subgroups of participants with syndromal or subsyndromal combinations of anxiety and depression symptoms differed on indicators of psychosocial functioning. Applying a linear approach to classifying scores on the PHQ-9 and GAD-7, participants were designated into anxiety–depression groups that varied by the severity of anxiety and depression symptoms. Participants with symptoms that met the thresholds for both anxiety (≥10) and depression (≥10) were classified into the anxiety–depression group (n = 222, 41.04%). Those with symptoms meeting the threshold for depression (≥10) and subsyndromal symptoms of anxiety (5–9) comprised the depression with subsyndromal anxiety group (n = 49, 9.06%). Conversely, the anxiety with subsyndromal depression group (n = 62, 11.46%) consisted of participants with symptoms meeting the threshold for anxiety (≥10) and subsyndromal symptoms of depression (5–9). Subsyndromal symptoms of anxiety (5–9) and depression (5–9) reflected participants with subsyndromal anxiety–depression (n = 78, 14.42%). Because we were most interested in comparing subsyndromal combinations of anxiety and depression to symptomology that aligned with standards commonly used to classify comorbidity, participants who did not meet any of the aforementioned criteria (n = 130, 24.03%) were excluded. The four groups formed the anxiety–depression group variable used in subsequent analysis.

A multivariate analysis of covariance (MANCOVA) was performed to test for mean differences between the anxiety–depression groups on the criterion variables in this study (i.e., hope, self-compassion, shame, social connectedness, and quality of social relationships). The MANCOVA was run while controlling for the same covariates as those that were included in the polynomial regression analyses. For outcome variables that reached statistical significance, follow-up analysis of covariance (ANCOVA) tests were performed. Prior to interpreting effects, Wallyplots for normality of residuals (i.e., QQ plots) and homogeneity of variance (i.e., residual plots) were inspected. No issues concerning normality and homogeneity of residuals were identified. ANCOVAs that yielded significant effects for the anxiety–depression group variable were followed by planned post hoc contrasts in which the subsyndromal anxiety–depression group was compared to each of the other three groups. To correct for multiple comparisons, a Bonferroni adjustment was applied to post hoc comparisons for each criterion variable (i.e., p values were multiplied by three).

Results

Preliminary analyses indicated that the relation between anxiety and depression was large in effect size (r = .70, p < .001). Relations of anxiety and depression with the outcome variables in this study were in the expected direction and ranged from moderate to large in effect size (r = |.33 to .54|, all p values < .001; see Table 2).

Table 2.

Descriptive statistics, internal consistency estimates, and zero-order correlations among study variables.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
(1) Anxiety
(2) Depression	.70 [.65, .74]
(3) Hope	−.44 [−.50, −.36]	−.54 [−.60, −.47]
(4) Physical health	−.47 [−.53, −.40]	−.67 [−.71, −.62]	.49 [.43, .55]
(5) Self-compassion	−.51 [−.57, −.45]	−.52 [−.58, −.45]	.46 [.40, .53]	.42 [.34, .48]
(6) Shame	.51 [.45, .57]	.49 [.43, .55]	−.38 [−.45, −.31]	−.38 [−.45, −.30]	−.64 [−.69, −.59]
(7) Social connectedness	−.33 [−.41, −.26]	−.38 [−.45, −.31]	.44 [.36, .50]	.31 [.23, .38]	.52 [.46, .58]	−.41 [−.48, −.34]
(8) Social relationship quality	−.34 [−.41, −.26]	−.40 [−.47, −.33]	.36 [.28, .43]	.30 [.22, .37]	.39 [.31, .46]	−.32 [−.39, −.24]	.54 [.48, .60]
M (SD)	10.68 (5.57)	10.68 (6.45)	29.57 (8.99)	14.95 (2.57)	15.07 (4.23)	65.48 (17.51)	78.66 (19.13)	11.20 (3.39)
ω_t	.92	.90	.90	.80	.95	.96	.95	.70

Note. All correlations are statistically significant at p < .001. SD: standard deviation.

Polynomial regression of anxiety and depression

Polynomial regression model results are reported in Table 3. Although there was no evidence of a linear association between anxiety and social relationship quality (p = .078), linear relations of anxiety (all p values ≤ .018) and depression (all p values < .001) with all other criterion variables were in the expected direction and reached statistical significance. Depression (all p values ≤ .007), but not anxiety (all p values ≥ .088), evidenced curvilinear relations with all criterion variables. For hope, self-compassion, social connectedness, and social relationship quality, curvilinear effects were positive. Two-line tests indicated that at lower levels of depression, relations with these criterion variables were negative (z = −3.98 to −5.19, all p values ≤ .001). With the exception of social connectedness (z = 2.00, p = .046), relations of these variables with depression were absent at higher values of depression (z = −1.05 to 0.76, all p values ≥ .294). The curvilinear effect for shame was negative, with positive relations evidenced at lower values of depression (z = 4.53, p < .001) but not at higher values of depression (z = −1.37, p = .172). The results did not support a multiplicative effect of anxiety and depression on any of the criterion variables (all p values ≥ .277).

Table 3.

Polynomial regression summary statistics.

Predictor	Outcome variable
	Hope		Self-compassion		Shame		Social connectedness		Social relationship quality
	B (SE)	β [95% CI]	B (SE)	β [95% CI]	B (SE)	β [95% CI]	B (SE)	β [95% CI]	B (SE)	β [95% CI]
(Constant)	16.05 (2.88)***		11.03 (1.32)***		87.35 (5.47)***		69.59 (6.78)***		11.14 (1.18)***
Age	−0.01 (0.03)	−.02 [−.09, .05]	0.02 (0.01)*	.07 [.00, .14]	−0.26 (0.05)***	−.18 [−.25, −.11]	−0.06 (0.06)	−.04 [−.12, .04]	−0.04 (0.01)**	−.13 [−.21, −.05]
Sex^a	1.51 (0.65)*	.08 [.01, .15]	−0.32 (0.30)	−.04 [−.11, .03]	−1.16 (1.24)	−.03 [−.10, .04]	−4.14 (1.53)**	−.11 [−.19, −.03]	−0.75 (0.27)**	−.11 [−.19, −.03]
Race/ethnicity^b	−0.76 (1.24)	−.02 [−.09, .05]	−0.60 (0.57)	−.04 [−.11, .03]	−0.83 (2.35)	−.01 [−.08, .06]	−2.08 (2.91)	−.03 [−.11, .05]	0.33 (0.51)	.03 [−.05, .10]
Physical health quality	0.87 (0.17)***	.25 [.16, .35]	0.18 (0.08)*	.11 [.02, .20]	−0.42 (0.32)	−.06 [−.15, .03]	0.73 (0.39)	.10 [−.01, .20]	0.07 (0.07)	.06 [−.05, .16]
Anxiety	−0.20 (0.08)*	−.12 [−.22, −.02]	−0.22 (0.04)***	−.29 [−.39, −.19]	0.93 (0.16)***	.30 [.20, .39]	−0.44 (0.19)*	−.13 [−.24, −.02]	−0.06 (0.03)	−.10 [−.21, .01]
Anxiety²	0.00 (0.02)	.00 [−.11, .12]	0.01 (0.01)	.10 [−.01, .20]	0.01 (0.03)	.02 [−.09, .13]	0.03 (0.04)	.06 [−.07, .18]	−0.00 (0.01)	−.03 [−.15, .10]
Depression	−0.44 (0.09)***	−.32 [−.44, −.20]	−0.19 (0.04)***	−.30 [−.42, −.18]	0.77 (0.17)***	.28 [.16, .40]	−0.88 (0.20)***	−.30 [−.43, −.16]	−0.19 (0.04)***	−.37 [−.51, −.24]
Depression²	0.04 (0.01)**	.18 [.07, .28]	0.02 (0.01)**	.17 [.06, .27]	−0.08 (0.02)***	−.20 [−.31, −.09]	0.08 (0.03)**	.18 [.07, .30]	0.01 (0.00)**	.16 [.04, .28]
Anxiety × Depression	−0.02 (0.02)	−.08 [−.22, .06]	−0.00 (0.01)	−.01 [−.14, .13]	−0.00 (0.03)	−.01 [−.15, .13]	−0.00 (0.04)	−.00 [−.16, .16]	0.00 (0.01)	.05 [−.10, .21]

F ₍ _df ₎	31.05_(9, 515)***		34.54_(9, 515)***		35.73_(9, 515)***		15.00_(9, 515)***		15.64_(9, 515)***
Model R²_Adjusted	.34		.37		.37		.19		.20

Note. Anxiety and depression variables mean centered for all analyses. SE: standard error; CI: confidence interval.

Female = 0, male = 1.

Other race/ethnicity = 0, White = 1.

*p < .05. **p < .01. ***p < .001.

Anxiety–depression group comparisons

The results of the MANCOVA, ANCOVA, and post hoc analyses are reported in Table 4. The MANCOVA yielded a significant effect for anxiety–depression group. Follow-up ANCOVAs supported significant differences in the criterion variables as a function of anxiety–depression group. Post hoc comparisons revealed better outcomes on hope, self-compassion, shame, and quality of social relationships among participants in the subsyndromal anxiety–depression group compared to the anxiety–depression group (all p values < .001). The subsyndromal anxiety–depression group also scored significantly higher on hope than the depression with subsyndromal anxiety group (p = .008). However, the subsyndromal anxiety–depression group did not differ significantly from the depression with subsyndromal anxiety group or the anxiety with subsyndromal depression group on any other criterion variable (all p values ≥ .063). The general pattern of findings suggests that insalubrious outcomes are more likely as the severity of mixed anxiety–depression symptoms increases. However, aspects of psychosocial functioning may be similarly impaired among people with subsyndromal combinations of anxiety and depression compared to those with symptoms that coincide with syndromal definitions of comorbid anxiety–depression.

Table 4.

Summary statistics for MANCOVA, ANCOVA, and post hoc comparison analyses.

	Outcome variable
	Hope		Self-compassion		Shame		Social connectedness		Social relationship quality
	ANCOVA										MANCOVA
Variable	F ₍ _df ₎	$η_{p}^{2}$	F ₍ _df ₎	$η_{p}^{2}$	F ₍ _df ₎	$η_{p}^{2}$	F ₍ _df ₎	$η_{p}^{2}$	F ₍ _df ₎	$η_{p}^{2}$	Wilk’s λ	F ₍ _df ₎	$η^{2}$
Age	0.71_(1, 391)	.002	3.68_(1, 391)	.009	13.30_(1, 391)***	.033	1.48_(1, 391)	.004	11.14_(1, 391)***	.028	.93	6.08_(5, 387)***	.073
Sex^a	4.94_(1, 391)*	.012	0.39_(1, 391)	.001	1.25_(1, 391)	.003	3.88_(1, 391)*	.010	1.86_(1, 391)	.005	.96	3.17_(5, 387)**	.039
Race/ethnicity^b	1.36_(1, 391)	.003	2.24_(1, 391)	.006	0.02_(1, 391)	.000	3.06_(1, 391)	.008	.08_(1, 391)	.000	.99	1.17_(5, 387)	.015
Physical health quality	30.45_(1, 391)***	.072	10.76_(1, 391)**	.027	3.16_(1, 391)	.008	6.58_(1, 391)*	.017	4.14_(1, 391)*	.010	.92	6.89_(5, 387)***	.082
Anxiety–depression group	8.66_(3, 391)***	.062	9.09_(3, 391)***	.065	10.22_(3, 391)***	.073	2.65_(3, 391)*	.020	3.62_(3, 391)*	.027	.86	3.85_{(15, 1068.74)}***	.046
	Post hoc comparisons
	M (SD)		M (SD)		M (SD)		M (SD)		M (SD)
Anxiety–depression	24.77 (8.44)^#		12.82 (2.98)^#		75.15 (14.86)^#		71.89 (18.65)^#		9.96 (2.84)^#
Depression with subsyndromal anxiety	27.43 (7.68)^#		14.97 (3.95)		65.08 (14.53)		75.22 (17.69)		10.42 (3.25)
Anxiety with subsyndromal depression	30.68 (6.96)		14.58 (3.44)		67.95 (15.67)		76.47 (18.25)		11.16 (2.99)
Subsyndromal anxiety and depression	32.82 (7.38)		15.75 (3.61)		61.85 (16.21)		81.23 (15.44)		11.68 (3.47)

Note. Post hoc comparisons (using a Bonferroni adjustment) were only performed following statistically significant (p < .05) univariate analyses. SD: standard deviation; MANCOVA: multivariate analysis of covariance; ANCOVA: analysis of covariance.

Female = 0, male = 1.

Other race/ethnicity = 0, White = 1.

*p < .05. **p < .01. ***p < .001. ^#p < .05, compared with subsyndromal anxiety and depression group.

Discussion

In this study, a multi-analytic approach was used to examine relations of combinations of anxiety and depression with subjective intrapersonal and interpersonal indicators of psychosocial functioning. The findings of the complementary analyses were largely consistent with the expectation that increasing levels of anxiety and depression would correspond with worse psychosocial functioning, although nonlinear relations indicated that the effect of depression progressively attenuated at higher levels of symptom severity. Whereas the findings generally supported additive effects of anxiety and depression, the hypothesis that there would be synergistic effects of anxiety and depression was not supported. That is, there was no evidence that links between depression and lower quality psychosocial functioning were amplified at higher levels of anxiety.

Our findings resonate with an abundance of research that has found concurrent anxiety and depression symptoms may be more debilitating compared to the effects of either set of symptoms occurring in isolation (e.g., Lamers et al., 2011; Wanders et al., 2016). Specifically, higher levels of anxiety and depression were independently associated with unfavorable outcomes on a majority of criterion variables. Independent effects resonate with the tripartite model of anxiety and depression (see Clark & Watson, 1991; Watson et al., 1995), which posits that each share a predominant, nonspecific component of general distress, but are differentiated by elements that are specific to anxiety (i.e., physiological hyperarousal) and depression (i.e., lack of positive affect). Because anxiety and depression were simultaneously included in the regression analyses, the general pattern of findings may reflect the unique contributions of features that are specific to anxiety and depression after partialing out the underlying component of general distress that is common to both (Brown et al., 2010).

Nonlinear trends revealed that relations of depression with worse psychosocial functioning tended to dissipate as the severity of depressive symptoms increases. Whereas the effect of unique depressive symptoms may be stronger at lower levels, the negative implications of such symptoms for psychosocial functioning tend to stabilize as symptoms progressively become more severe. Although curvilinear trends may reflect ceiling or floor effects originating from skewed distributions on criterion variables (Le et al., 2011), such methodological artifacts are unlikely given skewness values were within appropriate limits and nonlinear relations were consistent across all outcomes. Considered within the context of the tripartite model, perhaps negative effects of depression-specific symptoms (i.e., low positive affect) on psychosocial functioning weaken progressively as the severity of those symptoms increases. More generally, these findings indicate that the complexities involved in identifying the psychosocial implications of concurrent anxiety and depression go beyond linear combinations of symptoms.

Group comparisons largely reaffirmed prior evidence that has found subsyndromal symptoms of mixed anxiety–depression may be associated with similar levels of impairment when compared to people with symptoms that correspond with common definitions of anxiety–depression comorbidity (see Hettema et al., 2015). These findings advocate an inclusive approach to assessing comorbid symptoms which extends beyond the narrow boundaries of finite approaches that are frequently used. This perspective aligns with a nonreductionist view of mental health and criticisms leveled against the generalized application of diagnostic criteria without accounting for broader personal and contextual factors that might affect a person’s subjective experience (Castiglioni & Laudisa, 2015). As subsyndromal anxiety and depression have each been linked to an increased likelihood of a subsequent depressive or anxiety disorder (Karsten et al., 2011), a nonreductionist approach to comorbid anxiety–depression will likely benefit early detection of individuals at risk of symptom progression and promote prevention research initiatives targeting subsyndromal combinations of anxiety and depression symptoms.

By extending our focus beyond linear effects, the findings of this study offer an improved understanding of how anxiety and depression symptoms may be nonlinearly associated with intrapersonal and interpersonal indicators of functioning. The findings also underscore the importance of integrating curvilinear effects into models in which interaction effects involving highly correlated variables are estimated (Matuschek & Kliegl, 2018), thereby safeguarding against spurious conclusions (i.e., type I error) about interaction effects (Ganzach, 1997) and enhancing the validity of statistical conclusions (García-Pérez, 2012). Nevertheless, the findings ought to be interpreted with consideration of selected methodological limitations. First, this study is based on cross-sectional data, which prevents interpretations of causality and directionality from being made. It is possible that indicators of psychosocial functioning included in this study as criterion variables precede anxiety and depression in the causal order or that relations among the variables are bidirectional. Longitudinal studies are needed to ascertain how various concurrent combinations of anxiety and depression—both at syndromal and subsyndromal levels—develop, change over time, and are affected by treatment. Second, the sample consisted of a relatively homogenous group of participants, the majority of which were of nonminority race/ethnic status and highly educated. Caution should be applied in generalizing the findings to other populations, as risk factors for, prevalence of, and outcomes linked to anxiety and depression can vary by sociodemographic (e.g., race/ethnicity) characteristics (see Watkins et al., 2015; Zhao et al., 2009). Third, supplementary comparative analyses relied on a linear approach to generate anxiety–depression subgroups using scores on the PHQ-9 and GAD-7. Although this procedure has widely been applied in lieu of formal diagnoses (Mitchell et al., 2016), subgroups were derived using subjective responses to the PHQ-9 and GAD-7 rather than clinical assessments.

Conclusions

In summary, the general pattern of findings suggests that the implications of comorbid anxiety–depression are complex and vary based on linear (i.e., anxiety) and nonlinear (i.e., depression) trends in the progression of symptom severity. The findings also highlight the importance of detecting concurrent anxiety–depression that may exist at subsyndromal levels and offering customized treatments that appropriately align with the severity of impairment linked to combinations of anxiety and depression that do not meet more traditional criteria of comorbidity.

Footnotes

Article Notes

Note

ORCID iD

Richard Gregory Cowden

References

Andrews

Qian

Valentine

J. D.

(2002). Predicting depressive symptoms with a new measure of shame: The Experience of Shame Scale. British Journal of Clinical Psychology, 41, 29–42. https://doi.org/10.1348/014466502163778

Barnard

L. K.

Curry

J. F.

(2011). Self-compassion: Conceptualizations, correlates, & interventions. Review of General Psychology, 15, 289–303. https://doi.org/10.1037/a0025754

Baxter

A. J.

Scott

K. M.

Ferrari

A. J.

Norman

R. E.

Vos

Whiteford

H. A.

(2014). Challenging the myth of an “epidemic” of common mental disorders: Trends in the global prevalence of anxiety and depression between 1990 and 2010. Depression and Anxiety, 31, 506–516. https://doi.org/10.1002/da.22230

Brown

L. F.

Kroenke

Theobald

D. E.

(2010). The association of depression and anxiety with health-related quality of life in cancer patients with depression and/or pain. Psycho-Oncology, 19, 734–741. https://doi.org/10.1002/pon.1627

Cacioppo

J. T.

Hawkley

L. C.

Thisted

R. A.

(2010). Perceived social isolation makes me sad: 5-year cross-lagged analyses of loneliness and depressive symptomatology in the Chicago Health, Aging, and Social Relations Study. Psychology and Aging, 25, 453–463. https://doi.org/10.1037/a0017216

Cândea

D. M.

Szentagotai-Tăta

(2018). Shame-proneness, guilt-proneness and anxiety symptoms: A meta-analysis. Journal of Anxiety Disorders, 58, 78–106. https://doi.org/10.1016/j.janxdis.2018.07.005

Castiglioni

Laudisa

(2015). Toward psychiatry as a ‘human’ science of mind. The case of depressive disorders in DSM-5. Frontiers in Psychology, 5, 1517. https://doi.org/10.3389/fpsyg.2014.01517

Castilho

Pinto-Gouveia

Duarte

(2015). Evaluating the multifactor structure of the long and short versions of the Self‐Compassion Scale in a clinical sample. Journal of Clinical Psychology, 71, 856–870. https://doi.org/10.1002/jclp.22187

Cheung

M. S.-P.

Gilbert

Irons

(2004). An exploration of shame, social rank and rumination in relation to depression. Personality and Individual Differences, 36, 1143–1153. https://doi.org/10.1016/S0191-8869(03)00206-X

10.

Cibich

Woodyatt

Wenzel

(2016). Moving beyond “shame is bad”: How a functional emotion can become problematic. Social and Personality Psychology Compass, 10, 471–483. https://doi.org/10.1111/spc3.12263

11.

Clark

L. A.

Watson

(1991). Tripartite model of anxiety and depression: Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100, 316–336. https://doi.org/10.1037/0021-843X.100.3.316

12.

Curry

L. A.

Snyder

C. R.

Cook

D. L.

Ruby

B. C.

Rehm

(1997). Role of hope in academic and sport achievement. Journal of Personality and Social Psychology, 73, 1257–1267. https://doi.org/10.1037/0022-3514.73.6.1257

13.

Ekstrøm

C. T.

(2014). Teaching ‘instant experience’ with graphical model validation techniques. Teaching Statistics: An International Journal for Teachers, 36, 23–26. https://doi.org/10.1111/test.12027

14.

Fusar-Poli

Nelson

Valmaggia

Yung

A. R.

McGuire

P. K.

(2014). Comorbid depressive and anxiety disorders in 509 individuals with an at-risk mental state: Impact on psychopathology and transition to psychosis. Schizophrenia Bulletin, 40, 120–131. https://doi.org/10.1093/schbul/sbs136

15.

Gallagher

M. W.

Lopez

S. J.

(Eds.). (2018).The Oxford handbook of hope. Oxford University Press.

16.

Ganzach

(1997). Misleading interaction and curvilinear terms. Psychological Methods, 2, 235–247. https://doi.org/10.1037/1082-989X.2.3.235

17.

García-Pérez

M. A.

(2012). Statistical conclusion validity: Some common threats and simple remedies. Frontiers in Psychology, 3, 325. https://doi.org/10.3389/fpsyg.2012.00325

18.

Grieve

Indian

Witteveen

Tolan

G. A.

Marrington

(2013). Face-to-face or Facebook: Can social connectedness be derived online? Computers in Human Behavior, 29, 604–609. https://doi.org/10.1016/j.chb.2012.11.017

19.

Gum

A. M.

(2018). Promoting hope in older adults. In Gallagher

M. W.

Lopez

S. J.

(Eds.), The Oxford handbook of hope (pp. 143–155). Oxford University Press.

20.

Hawthorne

Herrman

Murphy

(2006). Interpreting the WHOQOL-Brèf: Preliminary population norms and effect sizes. Social Indicators Research, 77, 37–59. https://doi.org/10.1007/s11205-005-5552-1

21.

Hettema

J. M.

Aggen

S. H.

Kubarych

T. S.

Neale

M. C.

Kendler

K. S.

(2015). Identification and validation of mixed anxiety-depression. Psychological Medicine, 45, 3075–3084. https://doi.org/10.1017/S0033291715001038

22.

Hong

R. Y.

Cheung

M. W.-L.

(2015). The structure of cognitive vulnerabilities to depression and anxiety: Evidence for a common core etiologic process based on a meta-analytic review. Clinical Psychological Science, 3, 892–912. https://doi.org/10.1177/2167702614553789

23.

Ionescu

D. F.

Niciu

M. J.

Henter

I. D.

Zarate

C. A.

(2013). Defining anxious depression: A review of the literature. CNS Spectrums, 18, 252–260. https://doi.org/10.1017/S1092852913000114

24.

Ionescu

D. F.

Niciu

M. J.

Mathews

D. C.

Richards

E. M.

Zarate

C. A.

(2013). Neurobiology of anxious depression: A review. Depression and Anxiety, 30, 374–385. https://doi.org/10.1002/da.22095

25.

Karsten

Hartman

Smit

Zitman

Beekman

Cuijpers

van der Does

A. W.

Ormel

Nolen

W. A.

Penninx

(2011). Psychiatric history and subthreshold symptoms as predictors of the occurrence of depressive or anxiety disorder within 2 years. British Journal of Psychiatry, 198, 206–212. https://doi.org/10.1192/bjp.bp.110.080572

26.

Kashdan

T. B.

Julian

Merritt

Uswatte

(2006). Social anxiety and posttraumatic stress in combat veterans: Relations to well-being and character strengths. Behaviour Research and Therapy, 44, 561–583. https://doi.org/10.1016/j.brat.2005.03.010

27.

Kessler

R. C.

Sampson

N. A.

Berglund

Gruber

M. J.

Al-Hamzawi

Andrade

Bunting

Demyttenaere

Florescu

De Girolamo

Gureje

Huang

Karam

Kovess-Masfety

Lee

Levinson

Medina Mora

M. E.

, . . . Wilcox

M. A.

(2015). Anxious and non-anxious major depressive disorder in the World Health Organization World Mental Health Surveys. Epidemiology and Psychiatric Sciences, 24, 210–226. https://doi.org/10.1017/S2045796015000189

28.

Kim

Thibodeau

Jorgensen

R. S.

(2011). Shame, guilt, and depressive symptoms: A meta-analytic review. Psychological Bulletin, 137, 68–96. https://doi.org/10.1037/a0021466

29.

Kroenke

Spitzer

R. L.

Williams

J. B. W.

(2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16, 606–613. https://doi.org/10.1046/j.1525-1497.2001.016009606.x

30.

Kroenke

Spitzer

R. L.

Williams

J. B. W.

Monahan

P. O.

Löwe

(2007). Anxiety disorders in primary care: Prevalence, impairment, comorbidity, and detection. Annals of Internal Medicine, 146, 317–325. https://doi.org/10.7326/0003-4819-146-5-200703060-00004

31.

Kroenke

Zhangsheng

Bair

M. J.

Kean

Stump

Monahan

P. O.

(2016). The Patient Health Questionnaire Anxiety and Depression Scale (PHQ-ADS): Initial validation in three clinical trials. Psychosomatic Medicine, 78, 716–727. https://doi.org/10.1097/PSY.0000000000000322

32.

Lamers

van Oppen

P. C.

Comijs

H. C.

Smit

J. H.

Spinhoven

van Balkom

A. J.

Nolen

W. A.

Zitman

F. G.

, Beekman, & Penninx

B. W. J. H.

(2011). Comorbidity patterns of anxiety and depressive disorders in a large cohort study: The Netherlands Study of Depression and Anxiety (NESDA). Journal of Clinical Psychiatry, 72, 341–348. https://doi.org/10.4088/JCP.10m06176blu

33.

I.-S.

Robbins

S. B.

Ilies

Holland

Westrick

(2011). Too much of a good thing: Curvilinear relationships between personality traits and job performance. Journal of Applied Psychology, 96, 113–133. https://doi.org/10.1037/a0021016

34.

Leach

C. W.

(2017). Understanding shame and guilt. In Woodyatt

Worthington, Jr

E. L.

, Wenzel

Griffin

B. J.

(Eds.), Handbook of the psychology of self-forgiveness (pp. 17–28). Springer.

35.

Lee

R. M.

Dean

B. L.

Jung

K.-R.

(2008). Social connectedness, extraversion, and subjective well-being: Testing a mediation model. Personality and Individual Differences, 45, 414–419. https://doi.org/10.1016/j.paid.2008.05.017

36.

Lee

R. M.

Draper

Lee

(2001). Social connectedness, dysfunctional interpretational behaviors, and psychological distress: Testing a mediator model. Journal of Counseling Psychology, 48, 310–318. https://doi.org/10.1037//0022-0167.48.3.310

37.

Lee

R. M.

Robbins

S. B.

(1998). The relationship between social connectedness and anxiety, self-esteem, and social identity. Journal of Counseling Psychology, 45, 338–345. https://doi.org/10.1037/0022-0167.45.3.338

38.

Lee

R. M.

Robbins

S. B.

(2000). Understanding social connectedness in college women and men. Journal of Counseling and Development, 78, 484–491. https://doi.org/10.1002/j.1556-6676.2000.tb01932.x

39.

Levis

Benedetti . A., & Thombs

B. D.

(2019). Accuracy of Patient Health Questionnaire-9 (PHQ-9) for screening to detect major depression: Individual participant data meta-analysis. British Medical Journal, 365, 11476. https://doi.org/10.1136/bmj.l1476

40.

MacBeth

Gumley

(2012). Exploring compassion: A meta-analysis of the association between self-compassion and psychopathology. Clinical Psychology Review, 32, 545–552. https://doi.org/10.1016/j.cpr.2012.06.003

41.

Martin-Krumm

Delas

Lafrenière

M.-A.

Fenouillet

Lopez

S. J.

(2015). The structure of the State Hope Scale. The Journal of Positive Psychology, 10, 272–281. https://doi.org/10.1080/17439760.2014.936964

42.

Matuschek

Kliegl

(2018). On the ambiguity of interaction and nonlinear main effects in a regime of dependent covariates. Behavior Research Methods, 50, 1882–1894. https://doi.org/10.3758/s13428-017-0956-9

43.

Mitchell

A. J.

Yadegarfar

Gill

Stubbs

(2016). Case finding and screening clinical utility of the Patient Health Questionnaire (PHQ-9 and PHQ-2) for depression in primary care: A diagnostic meta-analysis of 40 studies. BJPsych Open, 2, 127–138. https://doi.org/10.1192/bjpo.bp.115.001685

44.

Neff

K. D.

(2003a). Self-compassion: An alternative conceptualization of a healthy attitude toward oneself. Self and Identity, 2, 85–101. https://doi.org/10.1080/15298860309032

45.

Neff

K. D.

(2003b). The development and validation of a scale to measure self-compassion. Self and Identity, 2, 223–250. https://doi.org/10.1080/15298860309027

46.

Neff

K. D.

Tóth-Király

Yarnell

L. M.

Arimitsu

Castilho

Ghorbani

Guo

H. X.

Hirsch

J. K.

Hupfeld

Hutz

C. S.

Kotsou

Lee

W. K.

Montero-Marin

Sirois

F. M.

de Souza

L. K.

Svendsen

J. L.

Wilkinson

R. B.

Mantzios

(2019). Examining the factor structure of the Self-Compassion Scale in 20 diverse samples: Support for use of a total score and six subscale scores. Psychological Assessment, 31, 27–45. https://doi.org/10.1037/pas0000629

47.

Neff

K. D.

Whittaker

T. A.

Karl

(2017). Examining the factor structure of the Self-Compassion Scale in four distinct populations: Is the use of a total scale score justified? Journal of Personality Assessment, 99, 596–607. https://doi.org/10.1080/00223891.2016.1269334

48.

Peters

G.-J. Y.

(2014). The alpha and the omega of scale reliability and validity: Why and how to abandon Cronbach’s alpha and the route towards more comprehensive assessment of scale quality. European Health Psychologist, 16, 56–69. https://doi.org/10.31234/osf.io/h47fv

49.

Plummer

Manea

Trepel

McMillan

(2015). Screening for anxiety disorders with the GAD-7 and GAD-2: A systematic review and diagnostic meta-analysis. General Hospital Psychiatry, 39, 24–31. https://doi.org/10.1016/j.genhosppsych.2015.11.005

50.

Revelle

Zinbarg

R. E.

(2009). Coefficients alpha, beta, omega and the glb: Comments on Sijtsma. Psychometrika, 74, 145–154. https://doi.org/10.1007/s11336-008-9102-z

51.

Saris

Aghajani

van der Werff

van der Wee

Penninx

(2017). Social functioning in patients with depressive and anxiety disorders. Acta Psychiatrica Scandinavica, 136, 352–361. https://doi.org/10.1111/acps.12774

52.

Schielzeth

(2010). Simple means to improve the interpretability of regression coefficients. Methods in Ecology and Evolution, 1, 103–113. https://doi.org/10.1111/j.2041-210X.2010.00012.x

53.

Scott

K. M.

Lim

Al-Hamzawi

Alonso

Bruffaerts

Caldas-de-Almeida

J. M.

Florescu

De Girolamo

De Jonge

Kawakami

Medina-Mora

M. E.

Moskalewicz

Navarro-Mateu

O’Neill

Piazza

Pasoda-Villa

Torres

Kessler

R. C.

(2016). Association of mental disorders with subsequent chronic physical conditions: World mental health surveys from 17 countries. JAMA Psychiatry, 73, 150–158. https://doi.org/10.1001/jamapsychiatry.2015.2688

54.

Simonsohn

(2018). Two lines: A valid alternative to the invalid testing of U-shaped relationships with quadratic regressions. Advances in Methods and Practices in Psychological Science, 1, 538–555. https://doi.org/10.1177/2515245918805755

55.

Skevington

S. M.

Lotfy

O’Connell

(2004). The World Health Organization’s WHOQOL-BREF quality of life assessment: Psychometric properties and results of the international field trial. A report from the WHOQOL Group. Quality of Life Research, 13, 299–310. https://doi.org/10.1023/B:QURE.0000018486.91360.00

56.

Skevington

S. M.

McCrate

F. M.

(2012). Expecting a good quality of life in health: Assessing people with diverse diseases and conditions using the WHOQOL‐BREF. Health Expectations, 15, 49–62. https://doi.org/10.1111/j.1369-7625.2010.00650.x

57.

Snyder

C. R.

(2005). Teaching: The lessons of hope. Journal of Social and Clinical Psychology, 24, 72–84. https://doi.org/10.1521/jscp.24.1.72.59169

58.

Snyder

C. R.

Rand

K. L.

Sigmon

D. R.

(2002). Hope theory: A member of the positive psychology family. In Snyder

C. R.

Lopez

S. J.

(Eds.), Handbook of positive psychology (pp. 257–276). Oxford University Press.

59.

Snyder

C. R.

Sympson

S. C.

Ybasco

F. C.

Borders

T. F.

Babyak

M. A.

Higgins

R. L.

(1996). Development and validation of the State Hope Scale. Journal of Personality and Social Psychology, 70, 321–335. https://doi.org/10.1037/0022-3514.70.2.321

60.

Spitzer

R. L.

Kroenke

Williams

J. B. W.

Löwe

(2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166, 1092–1097. https://doi.org/10.1001/archinte.166.10.1092

61.

Stekhoven

D. J.

Bühlmann

(2012). MissForest—Non-parametric missing value imputation for mixed-type data. Bioinformatics, 28, 112–118. https://doi.org/10.1093/bioinformatics/btr597

62.

The WHOQOL Group (1998). Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychological Medicine, 28, 551–558. https://doi.org/10.1017/S0033291798006667

63.

Thimm

J. C.

Holte

Brennen

Wang

C. E.

(2013). Hope and expectancies for future events in depression. Frontiers in Psychology, 4, 470. https://doi.org/10.3389/fpsyg.2013.00470

64.

Trompenaars

F. J.

Masthoff

E. D.

van Heck

G. L.

de Vries

Hodiamont

P. P. G.

(2007). Relationships between social functioning and quality of life in a population of Dutch adult psychiatric outpatients. International Journal of Social Psychiatry, 53, 36–47. https://doi.org/10.1177/0020764006074281

65.

Vigo

Thornicroft

Atun

(2016). Estimating the true global burden of mental illness. Lancet Psychiatry, 3, 171–178. https://doi.org/10.1016/S2215-0366(15)00505-2

66.

Vizin

Urbán

Zsolt

(2016). Shame, trauma, temperament and psychopathology: Construct validity of the Experience of Shame Scale. Psychiatry Research, 246, 62–69. https://doi.org/10.1016/j.psychres.2016.09.017

67.

Vos

Barber

R. M.

Bell

Bertozzi-Villa

Biryukov

Bolliger

Charlson

Davis

Degenhardt

Dicker

Duan

Erskine

Feigin

V. L.

Ferrari

A. J.

Fitzmaurice

Fleming

Graetz

Guinovart

Haagsma

, … Murray

C. J. L.

(2015). Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: A systematic analysis for the Global Burden of Disease Study 2013. Lancet, 386, 743–800. https://doi.org/10.1016/S0140-6736(15)60692-4

68.

Wanders

Van Loo

Vermunt

Meijer

Hartman

Schoevers

, Wardenaar, K. J., & De Jonge

(2016). Casting wider nets for anxiety and depression: Disability-driven cross-diagnostic subtypes in a large cohort. Psychological Medicine, 46, 3371–3382. https://doi.org/10.1017/S0033291716002221

69.

Watkins

D. C.

Assari

Johnson-Lawrence

V. J.

(2015). Race and ethnic group differences in comorbid major depressive disorder, generalized anxiety disorder, and chronic medical conditions. Journal of Racial and Ethnic Health Disparities, 2, 385–394. https://doi.org/10.1007/s40615-015-0085-z

70.

Watson

Clark

L. A.

Weber

Assenheimer

J. S.

Strauss

M. E.

McCormick

R. A.

(1995). Testing a tripartite model: II. Exploring the symptom structure of anxiety and depression in student, adult, and patient samples. Journal of Abnormal Psychology, 104, 15–25. https://doi.org/10.1037/0021-843X.104.1.15

71.

Weger

Sandi

(2018). High anxiety trait: A vulnerable phenotype for stress-induced depression. Neuroscience and Biobehavioral Reviews, 87, 27–37. https://doi.org/10.1016/j.neubiorev.2018.01.012

72.

Williams

K. L.

Galliher

R. V.

(2006). Predicting depression and self-esteem from social connectedness, support, and competence. Journal of Social and Clinical Psychology, 25, 855–874. https://doi.org/10.1521/jscp.2006.25.8.855

73.

Zhao

Ford

E. S.

Dhingra

Strine

T. W.

Mokdad

A. H.

(2009). Depression and anxiety among US adults: Associations with body mass index. International Journal of Obesity, 33, 257–266. https://doi.org/10.1038/ijo.2008.268

74.

Zhou

Cao

Yang

Guo

Fang

Cheng

(2017). Comorbid generalized anxiety disorder and its association with quality of life in patients with major depressive disorder. Scientific Reports, 7, 40511. https://doi.org/10.1038/srep40511