Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Anxiety Disord. Author manuscript; available in PMC 2010 March 30.
Published in final edited form as:
PMCID: PMC2847490

Positive affect regulation in anxiety disorders


Although individual differences exist in how people respond to positive affect (PA), little research addresses PA regulation in people with anxiety disorders. The goal of this study was to provide information about responses to PA in people with symptoms of social phobia, generalized anxiety disorder, panic disorder, agoraphobia, and obsessive-compulsive disorder. The tendency to dampen PA and the ability to savor PA were examined in an undergraduate sample. Analyses examined the unique links between these reactions and symptoms of anxiety disorders, controlling for a history of depression. Given the high comorbidity of depression and anxiety, exploratory analyses further controlled for generalized anxiety disorder. Results demonstrated that one or both measures of affect regulation made a unique and substantial contribution to predicting each anxiety disorder except agoraphobia, above and beyond prediction afforded by symptoms of depression and generalized anxiety disorder. Clinical implications and areas for future research are discussed.

Keywords: Anxiety disorders, Positive affect, Affect regulation, Dampening, Savoring

Robust individual differences exist in how people respond to positive affect (PA). Many people use strategies to enhance and sustain positive affective states; these include thinking about positive self-qualities, reflecting on how good life is, or focusing on the experience that triggered the PA. Somewhat counterintuively, though, it has also become clear that some people react in ways that are likely to dampen PA (Feldman, Joormann, & Johnson, 2008). An example of a dampening response to PA would be to reflect that “This will never last.” Responses that sustain and amplify PA are related to higher self-esteem and confidence (Larsen & Prizmic, 2004; Martin & Tesser, 1996), whereas those that diminish PA are tied to lower self-esteem (Feldman et al., 2008). It has been suggested that people with low self-esteem may try to diminish their positive feelings because they do not believe they deserve to experience PA (Parrott, 1993; Wood, Heimpel, & Michela, 2003).

Dampening responses to PA may also be related to psychopathologies (Clark & Watson, 1991). The syndrome that has received the most attention in this regard is depression, which is known to be characterized by low levels of PA. People with depression report that they avoid PA, engage in less cognitive elaboration of positive mood states, savor their positive experiences less, and are more likely to dampen their positive moods, compared to persons who are less depressed (Bryant, 2003; Feldman et al., 2008; Min’er & Dejun, 2001). Thus, depression seems to be linked to reactions that dampen PA that naturally arises.

Less is known about how people with anxiety disorders respond to PA. However, indirect evidence suggests that a similar tendency may be at work in at least some anxiety disorders. For example, it is known that PA is low among people with social anxiety disorder, above and beyond what can be attributed to co-occurring depression (Brown, Chorpita, & Barlow, 1998; Kashdan, 2002, 2004, 2007). When people with social anxiety have opportunities to pursue activities that could generate PA, they seem not to exploit those opportunities, but instead are preoccupied by attempts to conceal or suppress their socially anxious feelings (Kashdan & Steger, 2006). Thus, the positive feelings fail to emerge. Also consistent with dampening of PA, social phobia and generalized anxiety disorder have both been linked to elevated fear of positive emotions and lower expression of positive emotions, compared to persons without these disorders (Roemer, Salters, Raffa, & Orsillo, 2005; Turk, Heimberg, Luterek, Mennin, & Fresco, 2005). Such responses to PA among persons with anxiety disorders do not appear to be explained by depression (Kashdan & Steger, 2006), suggesting that minimizing of PA may be an element in anxiety disorders even without comorbid depression.

These findings pertaining to anxiety suggest maladaptive responses to PA in social phobia and generalized anxiety disorder. Although other anxiety disorders are characterized by low levels of PA, little is known about how PA is regulated in these other anxiety disorders (Brown et al., 1998; Watson, Clark, & Carey, 1988). Thus, the goal of the study reported here was to provide information about responses to PA among people with symptoms of various kinds of anxiety disorders. Two measures of PA regulation were used: a measure designed to assess tendencies to amplify or to dampen PA and a measure designed to assess the ability to savor and sustain PA. Symptoms of social phobia, generalized anxiety disorder, panic disorder, agoraphobia, and obsessive-compulsive disorder were assessed. A measure of lifetime depression was included to provide additional evidence on the relationship between PA regulation and depression and to test whether any associations that emerged for anxiety symptoms did not depend on symptoms of comorbid depression. Finally, given that generalized anxiety disorder is highly comorbid with both depression and other anxiety disorders and shares similarities in emotion dysregulation with depression (Mennin, Holaway, Fresco, & Heimberg, 2007; Watson, 2005), exploratory analyses of maladaptive responses to PA were conducted controlling for generalized anxiety to determine whether any relationship between other anxiety disorders and PA regulation measures was unique.

1. Method

Participants were 248 undergraduate students (54% female) at the University of Miami. Measures were administered in large group sessions in partial fulfillment of a course requirement. Age and ethnicity information was not collected in connection with responses, but the sample presumably did not differ materially from the University of Miami’s student body, which is ethnically diverse (23% Hispanic, 6% African American, 8% Asian, 55% non-Hispanic White, and 7% other).

1.1. Responses to Positive Affect

The Responses to Positive Affect (RPA; Feldman et al., 2008) measure is a 17-item self-report scale that measures the use of strategies to respond to PA. It is modeled after the Response Styles Questionnaire (RSQ, Nolen-Hoeksema & Morrow, 1991). Exploratory and confirmatory factor analysis studies on the RPA revealed a 3-factor solution: (1) focusing on affective and somatic experiences of the PA (Emotion-focus: “I think about how happy I feel.”), (2) focusing on cognitive and goal-oriented facets of the mood such as confidence (Self-focus: “I am living up to my potential.”), and (3) cognitive responses that are likely to counter PA (Dampening: “I don’t deserve this.”). The Emotion and Self-focus scales demonstrated higher correlations with each other and modest associations with the Dampening subscale. A previous study related higher scores on the Emotion-focus and Self-focus positive rumination subscales to greater self-esteem (Feldman et al., 2008). In the current study, we were particularly interested in the Dampening subscale, as this has been found to relate to both current and lifetime depressive symptoms in previous research (Feldman et al., 2008). Internal consistency was good in this sample for all subscales (Emotion-focus, α = .77; Self-focus, α = .77; Dampening, α = .85).

1.2. Savoring Beliefs Inventory

The Savoring Beliefs Inventory (SBI; Bryant, 2003) is a 14-item questionnaire that measures an individual’s beliefs about his or her capacity to savor positive experiences. The SBI has three subscales that assess anticipating upcoming positive events, savoring positive experiences, and reminiscing about past positive experiences. The Savoring the Moment subscale has 8 items and is the focus of this study. This subscale has correlated positively with measures of affect intensity, extraversion, optimism, internal locus of control beliefs, and dimensions of subjective well-being; it has correlated negatively with hopelessness, depression, neuroticism, physical and social anhedonia, guilt and shame; and it was uncorrelated with measures of social desirability (Bryant, 2003). This subscale has adequate internal consistency in past studies (α’s ranging from .69 to .89) as well as in the current study (α = .83).

1.3. Inventory to Diagnose Depression-Lifetime

The Inventory to Diagnose Depression-Lifetime (IDD-L; Zimmerman & Coryell, 1987) is a 45-item self-report scale to assess lifetime history of depressive symptoms. The questions are designed to parallel the symptoms required for DSM-IV diagnoses of major depression. Following endorsement of a symptom, participants are asked whether the symptom lasted for at least two weeks. The DSM symptoms are then summed to provide an index of severity. The IDD-L has been shown to have excellent agreement with structured diagnostic interviews for depression (97%; Zimmerman & Coryell, 1987). The IDD-L demonstrated high internal consistency in past studies (α = .92) and in this study (α = .90).

1.4. Psychiatric Diagnostic Screening Questionnaire

The Psychiatric Diagnostic Screening Questionnaire (PDSQ; Zimmerman & Mattia, 2001) is a 126-item questionnaire that was designed to screen for 13 of the DSM-IV disorders that have been found to be most prevalent in large epidemiological studies (Kessler et al., 1994). PDSQ items assess current symptoms, but the scale cannot be used to assess clinically significant diagnoses because it does not assess for functional impairment. Rather, the PDSQ captures symptom profiles. The subscales used in this study were symptoms of panic disorder (9 items), agoraphobia (11 items), generalized anxiety disorder (10 items), and social phobia (14 items). Items concerning panic disorder refer to the past 2 weeks. Items concerning agoraphobia, generalized anxiety disorder, and social phobia refer to the past six months.

The PDSQ has demonstrated that it is a valid instrument across several studies. In previous studies, the PDSQ subscales had good to excellent levels of internal consistency, with all subscales demonstrating alphas greater than .80 (Zimmerman & Mattia, 2001). In this study, Cronbach’s α’s ranged from .69 to .84. The subscales of the PDSQ also demonstrate adequate convergent and discriminant validity with other scales. Across all subscales, the mean correlation between PDSQ subscales and measures of the same construct was .66 compared to .25 between PDSQ subscales and measures of other symptom domains. At the item level, the mean item-parent subscale correlation was .59, and the mean item-other subscale correlation was .17 (Zimmerman & Mattia, 2001). The PDSQ subscales had an average specificity of 70%, sensitivity of 87%, and negative predictive value of 97% in conjunction with a structured clinical interview (Zimmerman & Chelminski, 2006). For every PDSQ subscale, people with the relevant DSM-IV diagnoses scored significantly higher on the corresponding subscale than people without diagnoses, and at the item level, 97% of symptoms on the PDSQ were endorsed significantly more frequently by persons with the relevant DSM-IV diagnosis than by those without a diagnosis (Zimmerman & Mattia, 2001).

1.5. Obsessive-Compulsive Inventory-Revised

The Obsessive-Compulsive Inventory-Revised (OCI-R; Foa et al., 2002) is a self-report measure that assesses symptoms of obsessive-compulsive disorder. It contains 18 items that load onto six factors: washing, obsessing, hoarding, ordering, checking, and neutralizing, and provides a more comprehensive screening of obsessive-compulsive disorder (OCD) than the PDSQ. The OCI-R is highly effective in discriminating between people with OCD and people with other anxiety disorders. The internal consistency for the total score was high in past (α = .81, .93) and in the current sample (α = .90).

2. Results

Bivariate correlations among symptom indices are shown in Table 1. As one would expect, symptom indices were positively correlated. The correlations between symptoms of generalized anxiety disorder with other anxiety disorders and depression were particularly robust. The social phobia scale also had robust correlations with the OCD and agoraphobia scales. As expected, the panic disorder and agoraphobia scales were substantially correlated with one another.

Table 1
Correlations among symptom measures.

Correlations among the subscales of the PRS revealed a significant correlation between the Emotion-focus and Self-focus subscales, r (219) = .67, P < .001. The Dampening subscale showed no association with either the Emotion-focus or Self-focus subscale [r (219) = .10 and .12]. Savoring was positively associated with the Emotion-focus subscale [r (214) = .25, P < .001] and inversely related to the Dampening subscale of the PRS [r (214) =−.37, P < .001] but was not significantly related to the Self-focus scale [r (214) = .12, P = .07].

Measures of PA regulation were then correlated with symptom indices (IDD-L and anxiety subscales of the PDSQ). As shown in Table 2, Emotion-focused positive rumination was not related to symptoms of anxiety or depression. Small positive associations emerged between Self-focused positive rumination and agoraphobia and panic disorder. Dampening correlated positively with symptoms of social phobia, generalized anxiety disorder, and panic disorder, and to a smaller degree with agoraphobia, OCD, and IDD-L. Savoring correlated negatively with symptoms of depression and all anxiety disorders.

Table 2
Correlations and partial correlations of positive rumination, dampening, and savoring with depression and anxiety symptoms.

In order to be sure that the associations for anxiety disorders did not depend on associations of anxiety with depression, analyses were then conducted testing whether symptoms of anxiety disorders remained related to PA regulation after controlling for history of depressive symptoms (Table 2, right side). All correlations noted above remained significant and of similar magnitude, with the exception of that between dampening and agoraphobia, which became nonsignificant.

Structural equation modeling was used to determine whether measures of affect regulation made separate contributions to predicting symptoms of anxiety disorders (Fig. 1). Dampening had a significant direct effect on the generalized anxiety disorder (γ= .18, z = 2.69), social phobia (γ = .21, z = 2.95), and panic disorder (γ = .21, z = 3.00) subscales, controlling for savoring and depression. Controlling for dampening and depression, savoring had a significant negative direct effect on social phobia (γ =−.23, z = −3.26) and OCD (γ = −.26, z = −3.46) subscales.

Fig. 1
Structural equation model examining the effects of dampening and savoring on GAD, panic disorder, social phobia, agoraphobia and OCD controlling for symptoms of depression.

To assess the extent to which associations with the generalized anxiety disorder subscale accounted for associations with other anxiety disorders, symptoms of generalized anxiety disorder were entered into the model as an additional control. That is, we tested the links between anxiety symptoms and affect regulation measures, controlling for symptoms of depression and generalized anxiety (Fig. 2). Dampening remained a significant predictor of social phobia (γ = .13, z = 1.97) and panic disorder (γ = .14, z = 2.11) subscales, controlling for savoring, depression, and symptoms of generalized anxiety. Controlling for dampening, depression, and symptoms of generalized anxiety disorder, savoring had a significant negative direct effect on social phobia (γ =−.18, z = −2.71) and OCD (γ = −.21, z = −2.96) subscales.1

Fig. 2
Structural equation model examining the effects of dampening and savoring on panic disorder, social phobia, agoraphobia and OCD controlling for symptoms of depression and GAD.

3. Discussion

This study examined self-reported regulatory responses to PA among people with symptoms of anxiety disorders. Consistent with previous findings, lifetime depressive symptoms were independently related to measures of downward regulation of PA. Even after controlling for these lifetime depressive symptoms, tendencies to endorse dampening of PA were also positively related to symptoms of panic disorder, social phobia, generalized anxiety disorder, and OCD. Similarly, after controlling for lifetime depressive symptoms, endorsement of savoring was inversely related to symptoms of all anxiety disorders, although the correlation with agoraphobia was modest. Dampening made a unique and substantial contribution to predicting generalized anxiety disorder, social phobia, and panic disorder above and beyond prediction afforded by symptoms of depression. Savoring negatively predicted symptoms of social phobia and OCD above and beyond symptoms of depression. A similar pattern emerged after an extreme step of controlling for symptoms of generalized anxiety disorder. Symptoms of social phobia, panic disorder, and OCD each retained some significant contribution from at least one of the maladaptive positive emotion regulation tendencies. Agoraphobia did not demonstrate a unique relationship with either of the emotion regulation tendencies, likely due to the low level of agoraphobia symptoms reported in this sample.

Before considering implications of these findings, several limitations must be noted. First, this study relied entirely on self-report of dampening, savoring, and symptoms. Future studies should use behavioral paradigms to assess responses to PA. Second, this study was carried out on a non-clinical undergraduate sample, and thus the results may not generalize to a sample with diagnosable levels of disorder. It will be important to determine whether the tendencies observed here are present in a sample diagnosed by a structured clinical interview. Finally, the cross-sectional design does not allow us to infer directionality of influence or to examine whether responses to PA may relate to the course of anxiety disorders over time.

Despite these limitations, this study provides preliminary evidence for existence of maladaptive regulatory responses to PA among people with symptoms of anxiety. Without replication of these results, it is premature to develop a specific theory as to why these emotion regulation tendencies demonstrate differential relationships with anxiety disorders; however we offer some potential clinical explanations. It appears that people with panic disorder and social phobia use more elaborative strategies to down-regulate the experience of positive emotions. In panic disorder this may be attributable to a conscious desire to reduce arousal, whether positive or negative, in order to decrease the intensity and discomfort of the visceral response associated with emotions. Similarly people with social phobia may wish to reduce any intense emotions that they may experience in a social situation, perhaps due to fears about appearing intensely emotional in social contexts. They may do this actively by using dampening tendencies or by not amplifying the positive emotion. Finally, people with symptoms of OCD reported that they do not savor positive experiences. It is possible that obsessions and compulsions in people with symptoms of OCD interfere with efforts to dampen PA.

If such findings can be replicated in clinical samples and can be shown to influence the course of symptoms, this pattern could have important clinical implications. Clinicians have successfully used strategies designed to enhance PA, such as behavioral activation approaches, as a treatment for depression (Jacobson & Gortner, 2000). The present results suggest the potential desirability of such strategies in regard to persons with anxiety disorders. People with specific symptoms of anxiety can be taught to identify, tolerate, accept, enjoy, and even enhance PA, which may be an alternative route to focusing on reducing anxiety.

In sum, these findings add to a growing literature documenting difficulties in regulating PA among people with anxiety disorders. Although these findings are preliminary, they warrant further study in a clinical population using both self-report and laboratory paradigms. Longitudinal research could also examine the impact of these strategies on the course of anxiety disorders. Developing a better understanding of these regulatory processes may guide alternative treatment strategies that can benefit people with anxiety disorders.


1It should be noted that these analyses were initially carried out using hierarchical multiple regression. Depression scores were entered in the first block, affect regulation measures were entered in the second block, and generalized anxiety disorder symptoms were entered in the third block. Results did not differ materially from those of the structural equation model.


  • Brown TA, Chorpita BF, Barlow DH. Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal. Journal of Abnormal Psychology. 1998;107:179–192. [PubMed]
  • Bryant FB. Savoring beliefs inventory (SBI): a scale for measuring beliefs about savouring. Journal of Mental Health. 2003;12:175–196.
  • Clark LA, Watson D. Tripartite model of anxiety and depression: psychometric evidence and taxonomic implications. Journal of Abnormal Psychology. 1991;103:103–116. [PubMed]
  • Feldman GC, Joormann J, Johnson SL. Responses to positive affect: a self-report measure of rumination and dampening. Cognitive Therapy and Research. 2008;32:507–525. [PMC free article] [PubMed]
  • Foa EB, Huppert JD, Leiberg S, Langer R, Kichic R, Hajcak G, et al. The Obsessive-Compulsive Inventory: development and validation of a short version. Psychological Assessment. 2002;14:485–496. [PubMed]
  • Jacobson NS, Gortner ET. Can depression be de-medicalized in the 21st century: scientific revolutions, counter-revolutions, and the magnetic field of normal science. Behaviour Research and Therapy. 2000;38:103–117. [PubMed]
  • Kashdan TB. Social anxiety dimensions, neuroticism, and the contours of positive psychological functioning. Cognitive Therapy and Research. 2002;26:789–810.
  • Kashdan TB. The neglected relationship between social interaction anxiety and hedonic deficits: differentiation from depressive symptoms. Anxiety Disorders. 2004;18:719–730. [PubMed]
  • Kashdan TB, Steger MF. Expanding the topography of social anxiety: an experience-sampling assessment of positive emotions, positive events, and emotion suppression. Psychological Science. 2006;17:120–128. [PubMed]
  • Kashdan TB. Social anxiety spectrum and diminished positive experiences: theoretical synthesis and meta-analysis. Clinical Psychology Review. 2007;27:348–365. [PubMed]
  • Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshelman S, et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Archives of General Psychiatry. 1994;51:8–19. [PubMed]
  • Larsen RJ, Prizmic Z. Affect regulation. In: Vohs KD, Baumeister RF, editors. Handbook of self-regulation: research, theory, and applications. New York: Guilford; 2004. pp. 40–61.
  • Martin LT, Tesser A. Some ruminative thoughts. In: Wyer RS Jr, editor. Ruminative thoughts. Hillsdale, NJ: Erlbaum; 1996. pp. 1–47.
  • Mennin DS, Holaway RM, Fresco DM, Heimberg RG. Delineating components of emotion and its dysregulation in anxiety and mood psychopathology. Behavior Therapy. 2007;38:284–302. [PubMed]
  • Min’er H, Dejun G. Emotion regulation and depression of college students. Chinese Mental Health Journal. 2001;15:438–441.
  • Nolen-Hoeksema S, Morrow J. A prospective study of depression and posttraumatic stress symptoms after a natural disaster: the 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology. 1991;61:115–121. [PubMed]
  • Parrott GW. Beyond hedonism: motives for inhibiting good moods and for maintaining bad moods. In: Wegner DM, Pennebaker JW, editors. Handbook of mental control. Upper Saddle River, NJ: Prentice-Hall; 1993. pp. 278–305.
  • Roemer L, Salters K, Raffa SD, Orsillo SM. Fear and avoidance of internal experiences in GAD: preliminary tests of a conceptual model. Cognitive Therapy and Research. 2005;29:71–88.
  • Turk CL, Heimberg RG, Luterek JA, Mennin DS, Fresco DM. Emotion dysregulation in generalized anxiety disorder: a comparison with social anxiety disorder. Cognitive Therapy and Research. 2005;29:89–106.
  • Watson D. Rethinking the mood and anxiety disorders: a quantitative hierarchical model for DSM–V. Journal of Abnormal Psychology. 2005;114:522–536. [PubMed]
  • Watson D, Clark LA, Carey G. Positive and negative affectivity and their relation to anxiety and depressive disorders. Journal of Abnormal Psychology. 1988;97:346–353. [PubMed]
  • Wood JV, Heimpel SA, Michela JL. Savoring versus dampening: self-esteem differences in regulating positive affect. Journal of Personality and Social Psychology. 2003;85:566–580. [PubMed]
  • Zimmerman M, Chelminski I. A scale to screen for DSM-IV Axis I disorders in psychiatric out-patients: performance of the Psychiatric Diagnostic Screening Questionnaire. Psychological Medicine. 2006;36:1601–1611. [PubMed]
  • Zimmerman M, Coryell W. The inventory to diagnose depression, lifetime version. Acta Psychiatria Scandinavia. 1987;75:495–499. [PubMed]
  • Zimmerman M, Mattia JI. The Psychiatric Diagnostic Screening Questionnaire: development, reliability and validity. Comprehensive Psychiatry. 2001;42:175–189. [PubMed]