Rumination and Related Constructs: Causes, Consequences, and Treatment of Thinking Too Much

By Ashley Borders

Rumination and Related Constructs: Causes, Consequences, and Treatment of Thinking Too Much synthesizes existing research relating to rumination. Integrating research and theories from clinical, social, cognitive, and health psychology, it features empirical findings related to why people ruminate, as well as treatments that decrease rumination. The book applies a transdiagnostic approach, looking beyond just depression to emphasize the wide range of clinical outcomes associated with repetitive thought. The book additionally describes research on physiological reactivity to rumination, the expression of rumination, potential benefits of rumination, and much more.

• Summarizes research on the emotional, behavioral, and physical consequences of rumination
• Discusses rumination in conjunction with different psychological disorders
• Integrates existing theories about rumination
• Identifies triggers and personality traits that influence whether people ruminate
• Explores cognitive and neural correlates of rumination
• Reviews established treatments for rumination

• Language: English
• Publisher: Academic Press
• Release date: Apr 11, 2020
• ISBN: 9780128126318

Ashley Borders

Associate Professor of Psychology at The College of New Jersey, and research affiliate with the War Related Illness and Injury Study Center at the VA Medical Center in East Orange, NJ. Her research focuses on exploring the causes and consequences of rumination. Using both cognitive-behavioral and mindfulness-focused therapies, she applies her and others’ empirical findings to help clients suffering from maladaptive rumination. She has published her findings in various peer-reviewed psychology journals. She is a licensed clinical psychologist who received her Ph.D. in clinical psychology from the University of Southern California.

Book preview

process.

Section 1

Consequences of rumination

Outline

Chapter 1 Rumination and mood disorders

Chapter 2 Rumination and anxiety-related disorders

Chapter 3 Rumination, anger, and aggression

Chapter 4 Rumination and dysregulated behaviors

Chapter 5 Rumination and physical functioning

Chapter 6 Consequences of expressed rumination

Chapter 1

Rumination and mood disorders

Abstract

This chapter reviews the literature on depressive rumination, or repetitive thinking about the nature, causes, and consequences of depressive symptoms. After describing the development of this construct and the associated response styles theory, I review the empirical evidence linking depressive rumination to maintained or exacerbated sad mood and depressive symptoms. I then explore empirical evidence for the proposed mechanisms of depressive rumination: negative thinking, impaired problem-solving, behavioral avoidance, and poor social support. After an examination of the nascent field of rumination and bipolar disorder, I explore research on gender differences in depressive rumination. The chapter ends with a discussion of moderators of depressive rumination and a proposed theory that integrates the rumination literature with other known predictors of depression.

Keywords

Rumination; depression; response styles theory; bipolar; gender differences; mechanisms; moderators

Overview

This chapter reviews the literature on depressive rumination, or repetitive thinking about the nature, causes, and consequences of depressive symptoms. After describing the development of this construct and the associated response styles theory, I review the empirical evidence linking depressive rumination to maintained or exacerbated sad mood and depressive symptoms. I then explore empirical evidence for the proposed mechanisms of depressive rumination: negative thinking, impaired problem-solving, behavioral avoidance, and poor social support. After an examination of the nascent field of rumination and bipolar disorder, I explore research on gender differences in depressive rumination. The chapter ends with a discussion of moderators of depressive rumination and a proposed theory that integrates the rumination literature with other known predictors of depression.

Ann was a client of mine for almost 2 years. She came to therapy a year after the death of her husband of 47 years. Now 69, she had never lived on her own and felt scared and adrift. She cried often, slept poorly, and had trouble finding the energy to do much of anything. She used to be a painter but seemed to have lost interest in what was once a great passion. Ann and her husband had no children, and she had not returned friends’ calls for a while, so no one checked in on her regularly. Early in our getting to know each other, she shared that she had birthed a son before her marriage but had given him up for adoption. She now found herself constantly wondering about her son and regretting that hasty decision of her 19-year-old self. When Ann was 16 years old, her mother had been accidentally shot and killed by her father during one of their regular shouting matches. Soon after, Ann left her home in Missouri and moved to Los Angeles, trying to escape her grief by losing herself in the excitement of a Hollywood life. Now, however, she found herself reliving that horrible episode, longing for her mother, and feeling guilty for leaving behind her younger sister.

Cognitive theories propose that the content of people’s thoughts contribute to depressive symptoms. For instance, Beck (1967) noted that depressed people have excessively negative views of themselves, others, and the future. These negative beliefs are usually not accurate but contribute to people feeling hopeless and helpless. Beck also proposed that everyone has deeply held and often unconscious core beliefs about themselves and the world. These core beliefs, or schemas, often originate in childhood and are shaped by important people and/or life events. Thus, her parents’ frequent arguments about their children and her mother’s sudden death may have contributed to Ann’s core beliefs that she was defective and that support and love from others was fleeting and unreliable. Negative schemas like this make people more vulnerable to depression if and when they later encounter negative life events that activate relevant schemas. The activated schemas then serve as filters through which people perceive the world. For Ann, the death of her husband activated her negative schemas, which in turn colored the way she perceived events in her life. For instance, her observation that few friends called her triggered automatic thoughts that she was unlovable, rather than the more accurate conclusion that they might think she wanted to be left alone.

This cognitive theory of depression has been expanded over the years, and many other types of unhelpful thoughts have been identified, but the basic premise remains the same: the specific content of thoughts that arise in response to negative life events explains the onset and maintenance of depression. There is, however, a second group of cognitive models. These models focus less on the content of thought and more on the process of turning attention inward—a construct called self-focused attention.

Beginning in the early 1970s, researchers began conceptualizing self-focused attention as part of a self-evaluative process in which individuals compare their current and desired states (Duval & Wicklund, 1972). Immediately after a negative event, people automatically increase their self-focus and focus on the discrepancy between current and desired states, presumably in order to understand what went wrong and to try to feel better. This is a natural part of self-regulation, or coping (Carver & Scheier, 1998). In healthy functioning individuals, this self-focus should lead to better understanding and emotion modulation, and perhaps even to behavior change. With the passage of time, therefore, the need for self-focused attention decreases in non-depressed individuals. Another way to think of it is as a negative feedback loop: an undesired event triggers the mind to start paying attention more closely to one’s feelings and thoughts. Ideally, this attention leads to successful regulation, which in turn shuts down the self-focused attention.

However, inward self-focus also elicits negative emotions, particularly when a negative discrepancy exists and when one fails to resolve such a discrepancy. When unresolvable discrepancies occur (e.g., the death of Anne’s husband), the adaptive response might be to avoid further self-focus and move on to alternative goals (e.g., finding new friends). In some instances, however, individuals are unwilling or unable to move on and continue focusing their attention internally. This sustained self-focused attention intensifies negative affect and self-criticism, leading to depression (Pyszczynski & Greenberg, 1987). Once in a depressed mood, people may develop a depressive self-focusing style, in which their default mode is inward focus. These individuals now have a faulty negative feedback loop, in which self-focus is never turned off, even when the situation demands attention elsewhere (Ingram, 1990). In Anne’s case, the self-focused attention triggered by the death of her husband contributed to her sustained depressed mood, which in turn elicited more self-focused attention. And now she focused not only on this one negative incident but other negative experiences in her past. She was stuck in a repetitive, internally-focused thought process—in essence, trapped in her own mind.

In sum, two fairly separate cognitive theories proposed that specific types of thought content and the degree to which people engage in an internal attentional process contribute to depression (Robinson & Alloy, 2003). Susan Nolen-Hoeksema proposed and spent years developing a theory that integrated these two cognitive approaches (Nolen-Hoeksema, 1991; Nolen-Hoeksema, Wisco, & Lyubomirsky, 2008). Her definition of depressive rumination combines negative thought content and sustained self-focus.

Response styles theory: definitions and measurement

Interestingly, the impetus for response styles theory (RST) was not merging two separate cognitive theories of depression, but rather explaining gender differences in the prevalence of depression. After reviewing and refuting other proposed explanations (e.g., genetics, hormones, and sex roles) for why women are diagnosed with depression twice as often as men, Nolen-Hoeksema (1987) argued that women and men have different ways of responding to, or coping with, depressed moods. Her basic hypothesis was that women are more likely to engage in a passive, ruminative response style—whereas men are more likely to engage in an active, distracting response style—when experiencing a sad mood. A ruminative response style amplifies and prolongs transient sad moods, making a longer depressive episode more likely. By contrast, engaging in distracting activities dampens and shortens depressed mood.

As evidence for gender differences in response styles, Nolen-Hoeksema (1987) presented college students with a list of things people do when depressed and asked them to rate how likely they would be to engage in each of the activities. Men scored significantly higher than women on I avoid thinking of reasons why I’m depressed, I do something physical, I play a sport, and I take drugs—all activities that distract individuals from sad moods. Women, on the other hand, scored significantly higher on I try to determine why I’m depressed, I talk to other people about my feelings, and I cry to relieve the tension—responses that focus and maintain attention to sad mood.

Building on these results, Nolen-Hoeksema (1991) defined depressive rumination as passive and repetitive thoughts that focus one’s attention on the symptoms, causes, and consequences of depression. She acknowledged that the content of depressive rumination may resemble the negative thoughts that Beck and other cognitive psychologists identified. However, the style, or process, of repetitive thinking makes this construct unique. Similar to rigid self-focused attention, rumination focuses individuals’ attention on their emotional state and therefore inhibits actions or thoughts that might be more adaptive. Unlike theories of self-focused attention, however, Nolen-Hoeksema argued that depressive rumination does not have to be preceded by a negative life event—because sometimes depressed moods do not have a clear triggering cause. Instead, depressive rumination focuses attention on a negative emotional state. She also argued that self-focused attention was conceptualized as an automatic response, whereas depressive rumination is a purposeful coping effort.

In contrast to rumination, distraction involves purposefully turning attention away from depressive symptoms and their causes and consequences (Nolen-Hoeksema, 1991). Effective distractions are pleasant (or at least neutral), absorbing, and reinforcing activities such as going for a run, seeing a movie with friends, or concentrating on work (Nolen-Hoeksema et al., 2008). There are, of course, types of responses that may be distracting but also potentially harm oneself, such as abusing alcohol and drugs or driving recklessly. However, Nolen-Hoeksema focused on the potential of positive or neutral distractions to decrease depressed moods.

The basic hypothesis of RST is that ruminative responses to depressed mood amplify sad moods and lead to longer periods of depression. By contrast, distraction in response to depressed mood decreases sad moods. Nolen-Hoeksema also proposed several mechanisms of depressive rumination, which I cover later in this chapter.

In order to test her hypotheses, Nolen-Hoeksema developed a measure to assess tendencies for depressive rumination called the Ruminative Responses Scale (RRS: Nolen-Hoeksema & Morrow, 1991). This measure was originally one subscale of a larger measure called the Response Style Questionnaire, which also includes subscales measuring distraction, problem-solving, and dangerous activities. The 22-item RRS assesses rumination about depressed moods (When you feel sad, down, or depressed, how often do you …: think about how alone you feel, think ‘Why do I always react this way?,’). This measure has been used in hundreds of published studies and has been adapted for children (Abela, Brozina, & Haigh, 2002). However, the measure has been criticized for its overlap with symptoms of depression (Roberts, Gilboa, & Gotlib, 1998; Treynor, Gonzalez, & Nolen-Hoeksema, 2003). To address this problem, Nolen-Hoeksema and colleagues eliminated those items in the original measure that assess depressive symptoms (Treynor et al., 2003). The remaining 10 items load onto two factors: brooding reflects a passive comparison of one’s current situation with some unachieved standard, whereas reflective pondering assesses self-focused problem-solving efforts to understand and alleviate one’s depressed mood (Treynor et al., 2003, p. 256).¹ As we will see, many studies have compared these two subscales.

There are several ways that the effects of depressive rumination have been studied. Cross-sectional studies examine whether measures of trait rumination correlate with depressive symptoms at that same time point. Prospective longitudinal studies assess whether rumination at an initial assessment predicts depressive symptoms at a later point in time. These types of nonexperimental designs address questions about the onset and duration of diagnosed depression, in addition to self-reported depressive symptoms. By contrast, experimental studies typically manipulate rumination and examine its causal effect on sad mood. These designs cannot provide insight into diagnosed depression or even many of the symptoms of depression, because these symptoms are not likely to change in response to any brief laboratory manipulation. Still, as we will see, experimental research has been able to test whether rumination increases sad mood.

Rumination and sad mood

Around the time she developed RST, Nolen-Hoeksema also developed a rumination manipulation that remains the gold standard in the larger rumination literature. In it, participants are asked to read and think about a series of 45 self-referential statements for 8 minutes (Lyubomirsky & Nolen-Hoeksema, 1993; Nolen-Hoeksema & Morrow, 1993). The statements are emotion-focused, symptom-focused, and self-focused but do not specify any particular emotion (e.g., what your feelings might mean, the kind of person you are, the physical sensations in your body, and why you react the way you do). The control, or distraction condition involves thinking about statements that are externally focused and not related to symptoms or emotions (e.g., a boat slowly crossing the Atlantic, the expression on the face of the Mona Lisa, and the layout of the local post-office). Participants in both conditions are instructed to focus their attention intently on the prompts for 8 minutes (Nolen-Hoeksema & Morrow, 1993). In essence, this manipulation induces self-focused attention by directing participants’ attention to how they are feeling, why, and what this means about them. This mirrors Nolen-Hoeksema’s definition of depressive rumination as self-focused attention specifically about depressive symptoms. The only piece not included in the manipulation is a focus specifically on depression. In fact, statements in both conditions are equally neutral in terms of affective tone. This was done intentionally, in order to separate the effects of rumination and depression.

This manipulation has been used in various ways to show that induced rumination increases sad mood. In an early experiment, Nolen-Hoeksema included participants with high or low scores on a depression self-report scale (Nolen-Hoeksema & Morrow, 1993). They were randomly assigned to either the rumination or distraction manipulation. Dysphoric participants who ruminated showed an increase in their sad mood, whereas dysphoric participants in the distraction condition showed a significant decrease in their sad mood—to the level of sadness reported by nondysphoric participants. Neither rumination nor distraction affected sad mood in nondysphoric participants. These findings have been replicated in other similar studies (Lyubomirsky, Caldwell, & Nolen-Hoeksema, 1998; Lyubomirsky, Kasri, & Zehm, 2003; Lyubomirsky & Nolen-Hoeksema, 1993, 1995; Lyubomirsky, Tucker, Caldwell, & Berg, 1999; Watkins & Teasdale, 2001). When comparing participants diagnosed with major depressive disorder with nondepressed control participants, manipulated rumination also increases sad mood, relative to manipulated distraction (Donaldson & Lam, 2004; Lavender & Watkins, 2004; Park, Goodyer, & Teasdale, 2004). Alternatively, researchers have induced sad mood in participants before the rumination manipulation. In the first experiment using this procedure, participants read a sad story while listening to sad music (Barber’s Adagio for Strings) in the background (Morrow & Nolen-Hoeksema, 1990). As expected, rumination while in a sad mood led to greater sad mood than did distraction. Subsequent versions of the same research design also find that manipulated rumination following a sad mood induction elicits more sad mood than does distraction (Ciesla & Roberts, 2007).

A research team in Germany developed a novel way of using the Nolen-Hoeksema rumination induction in daily life (e.g., Huffziger, Ebner-Priemer, Koudela, Reinhard, & Kuehner, 2012). Healthy undergraduates were given palm pilots for 2 days and were paged 10 times per day during a 12-hour window. At each page, participants were instructed to rate their momentary rumination and mood. On one of the 2 days (i.e., induction day), participants engaged in a 3-minute rumination induction after each of the 10 assessments, following by a second rating of momentary rumination and mood. On the noninduction day, participants simply provided 10 ratings following each page. The rumination induction involved reading and focusing on 10 rumination statements for a total of 3 minutes. At each time point the rumination induction immediately increased momentary rumination, negative mood, and agitation. Greater increases in state rumination were associated with larger increases in negative mood. Unlike in the laboratory experiments previously described, initial level of depressive symptoms did not moderate the effect of induced rumination.

Finally, researchers have studied the effect of naturally occurring rumination on sad mood. In two studies, some participants were given a delay period of 8 minutes after a sad mood induction, presumably during which high trait ruminators would spontaneously ruminate (Ciesla & Roberts, 2007; Conway et al., 2000). As expected, high trait ruminators who were given the opportunity to ruminate while in a sad mood reported more subsequent emotional distress. High ruminators who were not given the 8-minute delay did not report more distress than low trait ruminators in either condition (Conway et al., 2000). Researchers have also used experience sampling designs to study the longitudinal effect of rumination on sad mood. In a month-long study, the more that participants ruminated on a particular day in response to a depressed mood, the longer that mood lasted that day (Nolen-Hoeksema, Morrow, & Fredrickson, 1993). In addition, the amount of rumination on the first day of a moderately-severe depressed mood predicted how long that mood would last. Similarly, the more that participants were ruminating about their problems and feelings at any given time, the more sad they felt at the next page, roughly 90 minutes later (Moberly & Watkins, 2008).

In summary, across several different methodologies, naturally occurring and manipulated depressive rumination contribute to increased and prolonged sad moods. What the studies with manipulated rumination gain in internal validity, those with naturally occurring rumination gain in external validity. Taken together, the evidence is extremely convincing. I want to reiterate that the rumination manipulation actually induces self-focused attention to symptoms, causes, and consequences of individuals’ current mood state. These manipulations only become depressive rumination when paired with a preexisting or induced sad mood. In the absence of such a sad mood, the rumination manipulations have no effect on sad mood. This makes sense, as rumination without something to ruminate about seems more like philosophical reflection.

Rumination and depression

Longitudinal research clearly indicates that rumination predicts worse depressive symptoms. In one study, college students enrolled in an introductory psychology course completed questionnaires assessing depressive rumination and depressive symptoms (Nolen-Hoeksema & Morrow, 1991). In a twist of fate, a 7.1 earthquake struck the area 2 weeks later. The quick-thinking researchers measured depressive rumination and depressive symptoms in these students 10 days and 7 weeks after the earthquake. Students who typically used more depressive rumination, as assessed before the earthquake, showed more severe depressive symptoms 10 days and 7 weeks after the earthquake, even after controlling for initial depressive symptoms. Similar results have been found with individuals who recently experienced the death of a family member or partner (Nolen-Hoeksema, McBride, & Larson, 1997; Nolen-Hoeksema, Parker, & Larson, 1994). Thus, depressive rumination relates to more severe depressive symptoms following a range of negative, stressful events. Since these initial studies, many other researchers have shown that rumination predicts the severity of depressive symptoms across a wide range of time frames, in people with and without initial diagnosed depression, and in children and adolescents as well as adults (for reviews, see Aldao, Nolen-Hoeksema & Schweizer, 2010; Nolen-Hoeksema et al., 2008; Papageorgiou & Wells, 2004; Rood, Roelofs, Bögels, Nolen-Hoeksema & Schouten, 2009). Of particular interest have been the unique effects of brooding versus reflective pondering. In many studies, brooding uniquely predicts depressive symptoms several months to 1 year later, in both adolescents and adults (Barnum, Woody, & Gibb, 2013; Burwell & Shirk, 2007; Hasegawa, Koda, Kondo, Hattori, & Kawaguchi, 2013; Padilla Paredes & Calvete Zumalde, 2015; Raes, Smets, Nelis & Schoofs, 2012; Treynor et al., 2003). By contrast, the effects of reflective pondering are mixed. Some studies found that it uniquely predicted increased depressive symptoms over time, controlling for brooding (Hasegawa et al., 2013; Padilla Paredes & Calvete Zumalde, 2015), whereas others found no such unique effect (Barnum et al., 2013; Pearson, Watkins, & Mullan, 2010; Raes et al., 2012). Moreover, in other studies reflective pondering uniquely predicted less depressive symptoms over time (Arditte & Joormann, 2011; Eisma et al., 2015; Treynor et al., 2003). These mixed results may be explained by moderators. For instance, reflective pondering interacted with age and gender to predict changes in depressive symptoms 1 year later (Verstraeten, Vasey, Raes, & Bijttebier, 2010). Specifically, for younger children, reflective pondering predicted increases in depressive symptoms; for somewhat older children, however, reflective pondering predicted decreased depressive symptoms. Moreover, reflective pondering predicted decreased depressive symptoms in general for boys but not for girls. Cross-sectional studies also found that reflective pondering predicted more depressive symptoms only for participants with poor coping skills (Brennan, Barnhofer, Crane, Duggan, & Williams, 2015; Marroquín, Fontes, Scilletta, & Miranda, 2010). Thus, reflective pondering may be adaptive for some people but not for others. Another possibility is that reflective pondering is indirectly associated with depressive symptoms via increased brooding. In fact, one study found that reflection predicted increased negative rumination over 3 weeks, but negative rumination did not predict subsequent reflection (Takano & Tanno, 2009). The implication is that reflective pondering may make people more likely to start brooding, perhaps if they experience negative events or affect. More research is needed to examine moderators and indirect effects of reflective pondering on depressive symptoms.

Depressive rumination is also related to diagnosed depression as well as the future onset of major depressive episodes. People with current diagnosed depression, as well as a past history of depression, report more depressive rumination tendencies than do people with no depression history (Nolen-Hoeksema, 2000). Depressive rumination also predicts the onset of a future depression diagnosis in adults (Just & Alloy, 1997; Nolen-Hoeksema, 2000). Whether depressive rumination contributes to longer bouts of depression, however, is unclear. Several studies found that in adults who started out with a major depression diagnosis, depressive rumination did not predict diagnosed depression 6–18 months later (Just & Alloy, 1997; Lara, Klein, & Kasch, 2000; Nolen-Hoeksema, 2000). In clinical samples of patients being treated for major depression, depressive rumination by itself did not predict duration of depressive episodes. However, for patients at the end of their depression treatment, depressive rumination did predict reoccurrence of a depressive episode 3 months later (Kuehner & Weber, 1999). Moreover, undergraduate students who had more prolonged previous dysphoria reported more depressive rumination than those with brief previous dysphoria (Roberts et al., 1998). Thus, the evidence suggests that depressive rumination makes future new and recurrent depressive episodes more likely, but it may not relate to how long a depressive episode lasts. Nolen-Hoeksema suggests that rumination contributes to people crossing the threshold from sad mood to diagnosed depression, but that once depressed, other factors determine the duration of the episode (Nolen-Hoeksema et al., 2008).

A small literature examines the role of rumination in seasonal affective disorder, a variant of depression in which people experience depressive episodes starting in the fall or winter and ending in the spring. In this disorder the depression is usually characterized by fatigue, increased sleep, and increased appetite/weight, in addition to typical affective and cognitive symptoms. The dual vulnerability model proposes that seasonal affective disorder results from a physiological vulnerability to experience vegetative symptoms in response to seasonal changes in the physical environment, as well as a psychological vulnerability to develop affective and cognitive depressive symptoms in response to the vegetative symptoms (Enggasser & Young, 2007). In fact, trait depressive rumination measured in the summer or fall prospectively predicts severity of this kind of vegetative, winter depression (Rohan, Sigmon, & Dorhofer, 2003; Whitcomb-Smith et al., 2014). Moreover, this model suggests that rumination (i.e., a psychological vulnerability) combines—or interacts—with vegetative symptoms to predict affective and cognitive symptoms of seasonal depression. In support of this theory, depressive rumination prospectively predicts the severity of cognitive and affective depressive symptoms in individuals with a history of seasonal vegetative symptoms (Enggasser & Young, 2007). Moreover, in participants with clinical and subclinical symptoms of seasonal affective disorder, both trait and state rumination interact with daily or weekly vegetative symptoms to predict daily or weekly cognitive and affective symptoms (Whitcomb-Smith et al., 2014; Young, Reardon, & Azam, 2008).

Mechanisms of depressive rumination

The RST proposes several mechanisms of the effects of depressive rumination on depressive symptoms. Specifically, rumination is theorized to increase negative thinking, interfere with effective problem-solving, promote behavioral avoidance, and decrease social support (Nolen-Hoeksema, 1991; Nolen-Hoeksema et al., 2008). Evidence exists for all of these mechanisms.

Negative thinking

Nolen-Hoeksema proposed that rumination enhances the effects of depressed mood on thought content. In fact, research shows that depressive rumination is associated with several types of negative thinking. In an early experiment, college students read about several hypothetical problematic situations (e.g., being encouraged by friends to run for president of an organization but then losing the election) and were asked to report their thoughts and feelings about the situation, in order to reveal the types of interpretations they made (Lyubomirsky & Nolen-Hoeksema, 1995). Dysphoric participants in the rumination condition made more negative, pessimistic, and biased interpretations (e.g., minimizing successes and overgeneralizing from failures) than did dysphoric individuals in the distraction condition. Similarly, when participants spoke their thoughts aloud during the standard rumination versus distraction manipulation, the thoughts of dysphoric ruminators were more negative, self-critical, pessimistic, and reflected low perceived control (Lyubomirsky et al., 1999). By contrast, the thoughts of nondysphoric individuals who ruminated, as well as participants in the distraction condition, were more positive, optimistic, and less problem-focused. In correlational studies, depressive rumination is associated with tendencies for dysfunctional attitudes, pessimistic inferences, self-criticism, hopelessness, decreased self-confidence, self-esteem, and low perceived control (Abela, Brozina, & Haigh, 2002; Lam, Schuck, Smith, Farmer, & Checkley, 2003; Lo, Ho, & Hollon, 2008; Nolen-Hoeksema et al., 1994; Sarin, Abela, & Auerbach, 2005; Smith, Alloy, & Abramson, 2006; Weir & Jose, 2008). Several of these constructs—for example, pessimism, hopelessness, low self-esteem, and poor perceived control—mediate the effects of depressive rumination on depressive symptoms.

Depressive rumination is also associated with making more negative interpretations of ambiguous information. In one experiment, participants completed benign or negative word fragments at the end of brief scenarios like this: "While doing your calculus homework, you encounter a problem that you couldn’t solve. The next day, you go to class and turn it in. Later, you keep thinking about it and realize that, during class, everyone else seemed in_ec_re (insecure—benign interpretation)/co_fid_nt (confident—negative interpretation)" (Hertel, Maydon, Cottle, & Vrijsen, 2017). High trait ruminators who were induced to ruminate completed negative fragments more quickly than benign fragments, suggesting a negative interpretation bias. No differences emerged for low trait ruminators or participants in a distraction condition. Similarly, undergraduates high in trait depressive rumination showed a negative interpretation bias on a lexical decision task (Mor, Hertel, Ngo, Shachar, & Redak, 2013).

Rumination also makes dysphoric people more pessimistic about their futures. Dysphoric individuals in a rumination condition generated more negative future events that might happen to them in the coming weeks and years than did dysphoric individuals in a distraction condition (Lavender & Watkins, 2004). In a different study, dysphoric participants who ruminated generated as many happy events that might happen to them as did other participants; however, they were less hopeful that these events would actually happen to them (Lyubomirsky & Nolen-Hoeksema, 1995). Finally, dysphoric participants in a self-focus condition rated negative hypothetical events as more likely to happen to themselves and positive hypothetical events as more likely to happen to others, compared to dysphoric participants with an external focus (Pyszczynski & Greenberg, 1987).

Similarly, rumination and self-focused attention make dysphoric people more likely to remember negative events from their past. After a self-focus manipulation in which participants wrote a story with either first-person or third-person pronouns, dysphoric participants in the self-focused condition remembered more negative events that had happened to them recently than did dysphoric participants in the external-focus condition (Pyszczynski, Hamilton, Herring, & Greenberg, 1989). Similar results were found using Nolen-Hoeksema’s depressive rumination manipulation (Lyubomirsky et al., 1998) and using a manipulation that contrasted ruminative with reflective thinking (McFarland & Buehler, 1998). These studies also found that dysphoric individuals in distraction conditions did not recall more negative memories than nondysphoric individuals, suggesting that distraction decreases negative memory bias.

Depressive rumination appears to increase retrieval of a particular type of memory called overgeneral memories. When depressed people are asked to recall specific personal memories, they are more likely to retrieve categorical summaries of repeated events (e.g., I lost tennis matches) instead of individual events at a specific time and place (e.g., I lost to my friend Juan at tennis last Sunday) (Williams et al., 2007). This memory bias is associated with the maintenance of depressive symptoms and suggests a poor prognosis. Researchers therefore have investigated whether ruminating while depressed is associated with overgeneral memories. In both dysphoric and clinically depressed individuals, rumination inductions lead to more overgeneral memories, whereas distraction inductions decrease overgeneral memories (Park et al., 2004; Watkins & Teasdale, 2001; Watkins, Teasdale, & Williams, 2000). Correlational results suggest that brooding in particular is associated with overgeneral memories, particularly in dysphoric individuals (Debeer, Hermans, & Raes, 2009; Raes et al., 2005; Romero, Vazquez, & Sanchez, 2014; but see Hamlat et al., 2015). The idea is that self-focused rumination, particularly when in a depressed mood, makes abstract self-representations more salient (e.g., I am a failure), which encourages elaboration of overgeneral memories (e.g., "I always lose in