Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Dec 1.
Published in final edited form as: J Affect Disord. 2021 Aug 13;295:123–130. doi: 10.1016/j.jad.2021.08.014

Age Effects on Clinical and Neurocognitive Risk Factors for Suicide Attempt in Depression — Findings from the AFSP Lifespan study

M Buerke a,+, H Galfalvy b,c, J Keilp b,c, A Sheftall d, A Burke b,c, J Bridge d, J Mann b,c, K Szanto a
PMCID: PMC8551053  NIHMSID: NIHMS1735012  PMID: 34425314

Abstract

Background:

Studies of risk factors for suicidal behavior are typically restricted to narrow age ranges, making it difficult to determine if they have the same relevance or potency across the full adult lifespan.

Methods:

This study examined selected clinical and neurocognitive risk factors for suicidal behavior – borderline personality traits, aggression, depressive rumination, memory performance, and language fluency– in a multi-site sample (N=309, ages 16–80) of depressed patients with a recent (last 5 years) suicide attempt or no history of attempt, and demographically similar non-psychiatric controls. We examined cross-sectional age and attempter/non-attempter differences on these risk factors, and whether certain risk factors were more prominent discriminators of past suicide attempt earlier or later in the lifespan. Correlations with age were computed, and logistic regression was used to classify attempter status based on each risk factor and its interaction with age.

Results:

Nearly all risk factors were negatively correlated with age. Borderline traits, aggression, memory, and category fluency each predicted attempter status (p < 0.05), but these effects were not different across ages. In contrast, the association between rumination and suicide attempt status differed across the lifespan, becoming a stronger discriminator of past suicidal behavior at older ages.

Limitations:

The cross-sectional design limits our developmental findings.

Conclusions:

Despite age-related changes in symptom severity or neurocognitive performance, key risk factors for suicidal behavior previously identified in studies with more restricted age-ranges are salient throughout the adult lifespan. In contrast, depressive rumination may be particularly salient in later life.

Keywords: suicidal behavior, age effects, depressive rumination, aggression, borderline personality, neurocognitive deficits

Introduction:

A variety of clinical and neurocognitive risk factors for suicidal behavior have been identified in adults, but little is known about possible associations with age on their relevance or importance. Most research on these risk factors has focused on select age groups, such as adolescents and young adults, middle-aged adults, or the elderly. Nonetheless, broad conclusions have been drawn about risk for suicidal behavior that typically do not take the contribution of aging into account (McGirr et al., 2008). While some risk factors have been found to be consistently salient across different age cohorts, such as history of suicide attempt, current suicidal ideation, and the presence of mood disorders (Steele et al., 2018), others may differ across the lifespan. For instance, substance abuse is a consistent risk factor in younger to middle adulthood, while comorbid medical illnesses and dementing processes are significant risk factors in geriatric populations (Steele et al., 2018).

This study examined risk factors frequently associated with suicidal behavior across an age range encompassing most of the adult lifespan. The objective was to evaluate the relationship of each factor with age and to determine whether age enhances or diminishes that factor’s ability to distinguish those who have attempted suicide from those without a history of attempt. Borderline personality traits, trait aggression, depressive rumination, memory performance, and language fluency were selected as representative personality, behavioral, and neurocognitive characteristics that have distinguished past suicide attempters in previous studies within varying age cohorts.

Borderline personality disorder, which is marked by impulsivity, interpersonal dysfunction, and affective instability, has been consistently linked to both suicidal ideation (Rizk et al., 2019) and behaviors (Lesage et al., 1994; Pompili et al., 2005; Yen et al., 2004), with suicide rates as high as 10% in those meeting diagnostic criteria (Paris & Zweig-Frank, 2001). Developmentally, borderline symptoms usually manifest in late adolescence, worsen in young adulthood, and diminish with age (Gunderson et al., 2011; Paris & Zweig-Frank, 2001; Sansone & Wiederman, 2014; Zanarini et al., 2010). Some studies have found borderline traits are more strongly associated with risk for suicidal behavior in younger compared to older adults (Qin, 2011; Stepp & Pilkonis, 2008), but studies of borderline personality disorder in late-life are limited (Videler et al., 2019).

Aggression, a trait common in borderline personality disorder (Látalová & Praško, 2010) but distinct from the disorder itself (Keilp et al., 2006), has consistently been related to suicidal behavior (Gvion & Apter, 2011). While aggressive traits had been thought to be relatively stable across the lifespan, more recent evidence suggests an age-related decline (see review Piquero et al., 2012). It is unclear how such an age-related decline in aggression might impact its relevance as a risk factor for suicidal behavior at older ages.

Depressive rumination plays a role in both worsening depression and suicidal ideation (Nolen-Hoeksema et al., 2008; Morrison & O’Connor, 2008; Rogers & Joiner, 2017). Less known, however, is the relationship between overall depressive rumination and suicidal behavior, with a recent meta-analysis (Rogers & Joiner, 2017) identifying only three samples that studied this relationship, two of which were in college students, and all of which were in younger adults. In addition, while substantial evidence indicates depressive rumination decreases across the lifespan in community samples (De France & Hollenstein, 2019; Emery et al., 2020; Ricarte et al., 2016; Ricarte Trives et al., 2016; Sütterlin et al., 2012), its importance to the risk of suicidal behavior at older ages is unclear.

Neuropsychological impairments have been associated with suicidal behavior via performance on neurocognitive tasks measuring memory, language fluency, and various aspects of executive functioning and decision making (Richard-Devantoy et al., 2014; Saffer & Klonsky, 2018). These have primarily been associated with suicidal behavior in middle-aged samples (Bartfai et al., 1990; Keilp et al., 2001, 2013, 2014), however, deficits associated with neurocognitive decline and possible prodromal symptoms of dementia may contribute to risk for suicidal behavior more strongly in older adults (Gujral et al., 2020). Two types of deficits commonly associated with late-life cognitive decline, specifically deficits in episodic memory performance and language fluency, are known to be related to suicidal behavior in younger adult and middle-aged samples (Richard-Devantoy et al., 2014). These measures are also sensitive indicators of early dementia-like changes in older individuals (Devanand et al., 1997; Tabert et al., 2006), which may play a role in risk for suicidal behavior in this population (Gujral et al., 2020).

Deficits in episodic memory have been associated with suicidal behavior in the context of depression (Keilp et al., 2001, 2013; Richard-Devantoy et al., 2014) primarily in studies of middle-aged adults. However, well-known age effects on declarative memory, especially initial encoding (Rönnlund et al., 2005) suggest that this normal age-related decline may interact with psychiatric syndromes to produce an increased risk for suicidal behavior at older ages.

Similarly, letter (phonemic) and category (semantic) fluency have been commonly used as measures of language abilities in the study of suicidal behavior. Suicide attempters typically perform worse on category fluency than psychiatric controls (Richard-Devantoy et al., 2014), with letter fluency generally more affected by depression (Keilp et al., 2013). This pattern is noteworthy because more rapid decline of category fluency relative to letter fluency is also observed in early dementia (Keilp et al., 1999; Rascovsky et al., 2007). Such changes in old age (Elgamal et al., 2011; Lamar & Resnick, 2004; Maylor et al., 2007), in the context of depression, may lead to an increased risk for suicidal behavior.

The goal of this study was to examine the independent associations between age and select clinical and neuropsychological risk factors for suicidal behavior, and to test whether age moderates risk for suicide attempt across the adult age spectrum. We expected all clinical measures and neuropsychological measures to correlate cross-sectionally with age, such that the examined clinical symptoms and personality traits would be less severe while neurocognition would be poorer in later life. We also hypothesized interactions with age, such that borderline traits, higher levels of aggression, and depressive rumination will distinguish suicide attempters at the younger end of the age spectrum from depressed non-attempters, while suicide attempters at older ages would be distinguished by more prominent memory and language fluency deficits.

Methods

Sample:

Three hundred and nine participants aged 16–80 (M age= 43.31, SD = 17.90; 62.46% female) were recruited from three sites: Pittsburgh, PA, Columbus, OH, and New York City, NY as part of a collaborative multi-center study to examine age associations on clinical and neurocognitive risk factors for suicidal behavior. All study procedures were approved by the University of Pittsburgh Institutional Review Board, Abigail Wexner Research Institute at Nationwide Children’s Hospital Institutional Review Board, and the New York State Psychiatric Institute Institutional Review Board. Participants were recruited via advertisement, primary care providers, outpatient clinics, or inpatient psychiatric units. Informed consent was obtained from all participants.

Participants included depressed suicide attempters (ATT; N=110), depressed non-attempters (DNA; N=101), and a non-psychiatric healthy comparison group (HC; N=98). Primary analyses involved classification of depressed attempters and non-attempters to evaluate the significance of the various risk factors described above for distinguishing these groups. Data from the healthy comparison group are presented in Table 1 and Figure 1 as anchors for normal age-related changes.

Table 1.

Demographic and Clinical Characteristics by Group

Demographic and Clinical Characteristics Healthy Control (N=98) Depressed Non-attempter (N=101) Attempter (N=110) P-value Pairwise Comparison
Site (%) 0.433
 NY 30 (30.6) 22 (21.8) 33 (30.0)
 OSU 29 (29.6) 40 (39.6) 33 (30.0)
 PITT 39 (39.8) 39 (38.6) 44 (40.0)
Age (mean SD) 46.26 (18.47) 43.28 (18.21) 40.73 (16.82) 0.084
Sex = Female (%) 62 (63.3) 66 (65.3) 65 (59.1) 0.632
Race (%) 0.066
 Caucasian 67 (68.4) 73 (72.3) 63 (58.3)
 Black or African American 22 (22.4) 15 (14.9) 19 (17.6)
 Asian 5 (5.1) 3 (3.0) 7 (6.5)
 Other 3 (3.0) 5 (5.0) 6 (5.5)
 More than One Race 1 (1.0) 5 (5.0) 13 (12.0)
Ethnicity (Hispanic or Latino %) 8 (8.2) 8 (7.9) 19 (17.3) 0.049 No pairwise differences
Income (per capita) 34761.99 (20521.84) 23202.88 (21102.34) 20084.36 (20450.20) <0.001 ATT< DNA<HC
Premorbid IQ (WAIS-IV) 1 11.67 (2.16) 11.88 (2.40) 11.25 (2.87) 0.198
Education (years) 15.62 (2.18) 15.03 (2.62) 14.21 (2.39) <0.001 ATT< DNA, HC
Physical Illness Burden (CIRSG) 2 3.03 (3.63) 5.85 (5.01) 5.94 (5.36) <0.001 HC< DNA, ATT
Mental Status Exam (MMSE) 3 29.49 (0.83) 29.19 (0.97) 28.88 (1.22) <0.001 ATT< HC
Pharmacological Treatment (ATHF) 4 - 2.05 (1.64) 2.34 (1.65) 0.224
Age of depression onset - 27.05 (17.21) 22.37 (13.92) 0.036 ATT<DNA
HRSD (with suicide item) - 18.12 (4.93) 20.21 (6.35) 0.006 DNA<ATT
HRSD (without suicide item) - 17.33 (4.52) 18.40 (5.61) 0.186
SSI (Current Ideation) - 5.91 (9.55) 16.80 (11.51) <0.001 DNA<ATT
SSI (Worst Point Ideation) - 10.32 (11.14) 27.07 (4.64) <0.001 DNA<ATT
Presence of lifetime Substance use Disorder (%) - 30 (30.0) 40 (38.1) 0.283
Presence of lifetime Anxiety Disorder (%) - 64 (63.4) 80 (75.5) 0.082
PAI Borderline Total 9.67 (5.93) 32.73 (11.31) 39.01 (13.66) <0.001 HC<DNA<ATT
BPAQ Aggression Total 50.75 (14.43) 66.97 (18.34) 74.74 (19.88) <0.001 HC<DNA<ATT
RRS Rumination Total 5.20 (8.86) 34.62 (15.16) 39.31 (13.73) <0.001 HC<DNA<ATT
Buschke SRT Recall Total 108.43 (18.97) 109.27 (22.51) 106.58 (23.12) 0.676
COWAT Letter Fluency Total 44.21 (11.93) 41.01 (12.25) 38.52 (12.17) 0.005 ATT<HC
COWAT Category Fluency Total 21.00 (5.32) 20.28 (4.95) 19.21 (5.03) 0.048 ATT<HC
Open in a new tab

Legend: HC, healthy comparison; DNA, depressed non-attempters; ATT, suicide attempters

1

Wechsler Adult Intelligence Scale

2

Cumulative Illness Rating Scale for Geriatrics

3

Mini Mental Status Examination

4

Antidepressant Treatment History Form

Fig. 1.

Fig. 1.

Open in a new tab

Clinical and neuropsychological risk factors for suicidal behavior across the lifespan. Borderline Traits, Aggression, and Rumination were assessed, with higher scores indicating more pathology. Memory and Category and Letter Fluency were also assessed, with higher scores indicating better neuropsychological functioning. Legend: HC, Healthy Comparisons; DNA, Depressed non-attempters; ATT, Suicide attempters. While attempters had more borderline traits, higher levels of aggression, and higher levels of rumination than non-attempters, age moderated the effect of only rumination on attempter status. Attempters also had lower memory and category fluency scores compared to the depressed non-attempter group, but not letter fluency. This was consistent across the lifespan. Healthy Comparisons were included for visualization purposes but were not included in the main analyses.

All depressed participants were in a current major depressive episode, diagnosed via the Structured Clinical Interview for DSM Disorders (SCID; First et al., 1995) and had a Hamilton Depression Rating Scale (HDRS-17; Hamilton, 1960) score ≥ 14 at the time of study entry. The attempter group had a history of at least one suicide attempt, defined as self-injurious behavior with intent to die and some expectation of death (O’Carroll et al., 1996), within five years of study assessment. The depressed non-attempter group had no history of suicide attempt, though current or past suicidal ideation was permitted. The healthy comparison group had no history of psychiatric illness or substance use, assessed via the SCID (First et al., 1995), and no history of suicidal behavior.

Suicide attempt history was assessed by trained clinicians with the Columbia Suicide History Form, and verified using participant report, medical records, and corroboration from family and friends when available. Individuals with a history of neurological disorder, delirium, psychosis, mania, or dementia were excluded from the study. Additional exclusion criteria included a score <24 on the Mini-Mental State Examination (MMSE; Folstein et al., 1975) and electroconvulsive therapy within the past 6 months. Individuals with current substance abuse were excluded, though a history of abuse or dependence with remission for at least three months prior to enrollment was permitted.

Instruments:

Descriptive Measures

Demographic information and clinical history were gathered via structured interview. Estimated intelligence was assessed via the Vocabulary and Matrix Reasoning subtest of the Weschler Adult Intelligence Scale, 4th edition (WAIS-IV; Wechsler, 2008), and current mental status via the MMSE (Folstein et al., 1975). Lifetime somatic illness burden across 13 organ systems was gathered via the Cumulative Illness Rating Scale (CIRS; Linn et al., 1968). Pharmacological treatment exposure was assessed using the Antidepressant Treatment History Form (ATHF; Sackeim, 2001), which assesses psychopharmacological medications taken, dosage, duration of use, and use of augmenting agents.

The HDRS-17 (Hamilton, 1960) was used as the primary measure of depression severity. The suicidal ideation item on the HRSD-17 was excluded from analyses to avoid overinflating suicidal participants’ total scores. Participants were assessed for current or lifetime Axis I disorders including mood, substance use, and anxiety disorders, using the SCID (First et al., 1995).

The medical seriousness for both the most lethal and most recent suicide attempt was characterized using the Beck Lethality Scale (BLS; Beck et al., 1975), which ranges from 0 (none or minimal) to 8 (attempt resulting in death). The subjective intent, seriousness, and planning involved in both the most lethal and most recent attempt were characterized using the Scale of Suicidal Intent (SIS; Beck et al., 1974).

Clinical Risk Factors

Borderline traits were measured using the Personality Assessment Inventory - Borderline subscale (PAI-BOR; Morey, 2014), a self-report questionnaire that includes questions about negative relationships, self-harm, affective instability and identity problems. The Buss-Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992) was used to assess aggressive traits. Depressive rumination was assessed using the Ruminative Response Scale (RRS; Nolen-Hoeksema & Morrow, 1991), a self-report that assesses the degree to which an individual dwells on their feelings of distress and consequences of their distress when depressed.

Neuropsychological Risk Factors

Memory performance was assessed via the Buschke Selective Reminding Test (Buschke SRT; Buschke, 1973), a list-learning task commonly used in dementia research (Masur et al., 1990), as well as in past studies of suicide attempters (Keilp et al., 2001, 2013, 2014). Total memory score was calculated as the number of words immediately recalled during the encoding phase of this task. Language fluency was assessed via the Letter and Animal Naming tasks of the Controlled Oral Word Association Test (COWAT), a subtest of the Multilingual Aphasia Examination (Benton & Hamsher, 1994). Letter fluency required participants to orally generate as many words as possible within a minute that began with a designated letter, for three separate letters. The letter fluency score is the sum of words produced for the three letters. Category Fluency required participants to generate as many exemplars as possible from a single category, in this case animals (beginning with any letter). This category was selected due to its specific sensitivity to both past suicidal behavior (Keilp et al., 2001) and early symptoms of degenerative cognitive loss in old age (Keilp et al., 1999; Rascovsky et al., 2007). The category fluency score was calculated as the total number of animals named within one minute.

Analytic strategy

The three groups were compared based on quantitative and categorical demographic and clinical variables using one-way analysis of variance for continuous variables and chi-square tests, followed by pairwise comparisons. Primary interest throughout the paper was in the comparison of attempters to non-attempter depressed patients, thus no adjustment for multiple comparisons were made in these analyses. Outliers on our predictors of interest were defined as being 1.5 Inter-Quartile Ranges above the 75th or below the 25th percentile, respectively, and were attenuated to the nearest non-outlier value. Healthy comparison participants were included in Table 1 and Figure 1 as a comparison group for normal age-related changes, but were excluded from correlations and the binomial logistic regression analyses as we were most interested in age-related changes between depressed groups with and without a history of suicidal behavior.

Pearson correlations were computed within the combined depressed sample between age and the selected risk variables. These correlations estimate age-related differences in the risk factors in depressed samples with an approximately equal proportion of attempters and non-attempters.

To test our hypothesis of whether age modified the association between clinical or executive functioning variables and suicide attempt status, we fit binomial logistic regression models in the patient sample with attempter status as the outcome [dependent variable], with age, the candidate predictor, and an interaction between age and the candidate predictor. These analyses covaried for sex and site. For these models, age and candidate predictors of interest were first centered, so values that reflected distance from the mean and significant main effects could be interpreted. Statistical significance was set at p < 0.05, two-tailed. For each analysis, we also ran an additional analysis which added depression as a covariate, to determine whether depression severity had any effect on our results. In the last step of the analysis, all significant candidate risk factors were entered as predictors into a single logistic regression model; interactions with age were also entered where they were significant. All analyses were performed in R, version 4.0.2.

Results

Demographic and Clinical Characteristics

Sample characteristics are presented in Table 1. There were no differences between the groups in their distribution across site, sex, age, or race. Ethnicity differed among the three groups; however, no pairwise differences were significant. The attempter group had the lowest level of education and the lowest estimated income, but there were no differences in estimated premorbid intellectual ability. Both depressed groups reported greater physical illness burden than healthy comparisons.

The attempter group had the lowest MMSE scores but differed by less than a single point from healthy comparisons with all scores well within the range of intact global functioning.

Within the depressed groups, attempters had a younger age of depression onset, but the groups were comparable on depression severity when the suicidal thinking item of the HDRS-17 was removed. The attempter group exhibited greater current and worst-point suicidal ideation than the depressed non-attempter group. There were no between-group differences in the presence of comorbid lifetime substance abuse or anxiety disorders.

Attempters had an average of 3.28 (SD: 9.53) lifetime attempts. Their most lethal suicide attempt had high intent (mean intent score of 16.74 (SD: 4.83)) and had a mean lethality of 3.06 (SD: 2.24), where a lethality of 4 or higher is considered to be high lethality and requiring medical intervention/hospitalization.

In uncorrected comparisons of the variables of interest, the attempter group had the highest levels of borderline traits, aggression, and rumination, followed by the depressed non-attempter and healthy comparison groups. Groups did not differ in memory performance in this comparison of raw scores, and attempters performed more poorly than healthy comparisons on both language fluency measures.

Effects of Age on Variables of Interest

In the combined depressed patient sample (suicide attempters and depressed non-attempters), age was negatively associated with all but one of the measures of interest. Borderline traits (r = −0.35, p < 0.001), aggression (r = −0.25, p < 0.001), and rumination (r = −0.35, p < 0.001) all declined with age. For neuropsychological measures, Letter Fluency did not decline significantly with age (r = −0.08, p = 0.273), but Category Fluency did (r = −0.26, p < 0.001). Memory, in turn, showed the strongest negative correlation with age (r = −0.66, p < 0.001). See Figure 1 for a representation of these age effects.

Putative Clinical Risk Factors for Suicide Attempt

Three separate binary logistic regression models were fit using borderline traits, aggression, or rumination to predict group membership (attempter vs. depressed non-attempter) and included age, site, sex, and each variable’s interaction with age. All three models were significant (Table 2). Borderline traits and aggression both significantly predicted attempter status in these models as main effects, with no significant interaction with age (Table 2). Rumination, however, showed both a significant main effect as well as an interaction with age, such that differences in rumination between the groups were greater at older ages (p=0.018; Table 2). In terms of classification odds, at age 25 years a 20-point higher than average rumination score was associated with a 10% greater likelihood of being classified as an attempter (OR=1.10, 95%CI: 0.6–2.0). For a person at the sample mean age of 41.7 years, a 20-point higher than average score approximately doubled the odds of being an attempter (OR=1.89, 95%CI: 1.2–3.0). At age 65, with a 20-point higher than average score, the odds of being an attempter almost quadrupled (OR=3.95, 95%CI: 1.7–9.0). Age, sex, and site were not significant as main effect covariates (as classifiers of attempt status). All interaction and main effect results remained the same when depression severity was added as a covariate to the models. These associations are illustrated in Figure 1.

Table 2.

Clinical Suicide Risk Factors in Depression across the lifespan

Odds Ratio Lower CI Upper CI Z value P-value
Borderline Traits
(Intercept) 1.69 0.80 3.59 1.37 0.1710
Borderline traits 1.05 1.02 1.08 3.47 0.0005
Age 0.99 0.97 1.02 −0.74 0.4600
Site (Pittsburgh) 1.26 0.46 3.41 0.45 0.6498
Site (Columbus) 0.55 0.26 1.17 −1.55 0.1219
Sex (Female) 0.70 0.38 1.31 −1.11 0.2671
Borderline traits × Age 1.00 1.00 1.00 1.26 0.2094
Aggression
(Intercept) 1.34 0.64 2.81 0.78 0.4385
Aggression 1.02 1.01 1.04 2.66 0.0079
Age 0.99 0.96 1.01 −1.24 0.2167
Site (Pittsburgh) 1.38 0.51 3.72 0.64 0.5209
Site (Columbus) 0.61 0.29 1.29 −1.28 0.1996
Sex (Female) 0.81 0.43 1.52 −0.66 0.5121
Aggression × Age 1.00 1.00 1.00 0.74 0.4607
Rumination
(Intercept) 1.71 0.80 3.63 1.39 0.1642
Rumination 1.03 1.01 1.06 2.68 0.0073
Age 0.98 0.96 1.01 −1.46 0.1432
Site (Pittsburgh) 1.70 0.62 4.64 1.03 0.3042
Site (Columbus) 0.56 0.27 1.19 −1.50 0.1339
Sex (Female) 0.63 0.34 1.20 −1.41 0.1594
Rumination × Age 1.00 1.00 1.00 2.37 0.0178
Open in a new tab

Putative Neuropsychological Risk Factors for Suicide Attempt

In separate binary logistic regression models classifying suicide attempters vs. depressed non-attempters, using memory, Letter Fluency, or Category Fluency as predictors and incorporating age, sex, site, and the age interaction (Table 3), both memory and Category fluency contributed significantly to the classification of attempter status as main effects. Letter fluency was not a significant classifier. These results were unchanged when depression severity was considered in the models. Associations of age with neurocognitive scores by group are illustrated in Figure 1.

Table 3.

Neuropsychological Function in Depression and its Association with Suicide Risk across the Lifespan

Odds Ratio Lower CI Upper CI Z value P-value
Memory
(Intercept) 1.90 0.87 4.14 1.61 0.1064
Memory 0.98 0.96 1.00 −2.44 0.0147
Age 0.96 0.94 0.99 −2.58 0.0100
Site (Pittsburgh) 1.00 0.39 2.57 0.00 0.9999
Site (Columbus) 0.52 0.25 1.10 −1.71 0.0879
Sex (Female) 0.78 0.41 1.46 −0.78 0.4351
Memory × Age 1.00 1.00 1.00 1.71 0.0871
Letter Fluency
(Intercept) 1.92 0.89 4.15 1.6689 0.0951
Letter Fluency 0.98 0.95 1.00 −1.6890 0.0912
Age 0.98 0.96 1.01 −1.4333 0.1518
Site (Pittsburgh) 0.93 0.36 2.40 −0.1398 0.8888
Site (Columbus) 0.50 0.23 1.08 −1.7559 0.0791
Sex (Female) 0.65 0.35 1.21 −1.3487 0.1774
 Fluency × Age 1.00 1.00 1.00 0.3638 0.7160
Category Fluency
(Intercept) 1.70 0.80 3.60 1.39 0.1660
Category Fluency 0.93 0.88 1.00 −2.10 0.0358
Age 0.98 0.96 1.00 −1.87 0.0609
Site (Pittsburgh) 1.01 0.40 2.56 0.02 0.9809
Site (Columbus) 0.55 0.26 1.16 −1.57 0.1168
Sex (Female) 0.70 0.38 1.30 −1.12 0.2611
Category Fluency × Age 1.00 1.00 1.00 0.13 0.8932
Open in a new tab

Full Classification Model

Significant predictors from the individual logistic regression models were included in a full classification model to examine the relative contributions of each variable. Classifiers in this model included borderline traits, aggression, depressive rumination, the interaction of rumination with age, category fluency, memory, and age itself. In the context of these other variables, borderline traits and the interaction of rumination and age were the only significant predictors of attempt status. These remained the only significant predictors when depression was added to the model.

Discussion

Clinical syndromes, traits, and symptoms, as well as neurocognitive performance, all change with age. However, little attention has been given to examining whether risk factors for suicidal behavior have the same relevance or potency at different points in the lifespan. In this study, we observed cross-sectional aging correlations on almost all the measures examined, but a significant age-related association with suicide attempt status in only one. Thus, even though borderline traits and aggression declined with increasing age, they remained significant discriminators of suicide attempt status within our depressed patient sample. This was contrary to our initial hypotheses, where we expected the declining severity of these traits to reduce their salience as risk factors. Similarly, declining neurocognitive performance with older age was expected to increase its importance in discriminating past attempt status, but age interactions did not reach significance. Within the limitations of the sampling, and the form of the linear age associations tested in our models, these results give us confidence that these putative risk factors for suicidal behavior, identified in multiple prior studies at various points in the adult lifespan, are meaningful across the full adult lifespan, from late adolescence to later senescence.

Our findings regarding depressive rumination ran counter to our hypotheses in that previous studies suggest that rumination declines with increasing age, at least in the general population (De France & Hollenstein, 2019; Emery et al., 2020; Ricarte et al., 2016; Ricarte Trives et al., 2016; Sütterlin et al., 2012). While we also observed an overall decline in rumination across the lifespan in all depressed participants, aging effects differed by suicide attempt status, and group differences increased with age. This is important for clinical assessment, given the challenges in identifying elderly at high risk for suicide (Szanto et al., 2020) and given that many elderly die as a result of their first suicide attempt (Paraschakis et al., 2012).

The tendency to ruminate has previously been linked to worsening depression, hopelessness, and has predicted current and prospective suicidal ideation (Nolen-Hoeksema et al., 2008; Morrison & O’Connor, 2008; Rogers & Joiner, 2017; Miranda & Nolen-Hoeksema, 2007). While there has been limited research on the relationship between rumination and suicidal behavior itself, evidence suggests that the perceived controllability of ruminative thoughts may contribute to risk (Rogers et al., 2021; Rogers & Joiner, 2017). In older adults, certain forms of repetitive thinking such as rumination have been associated with biomarkers of dementia and greater brain age (Karim et al., 2021; Marchant et al., 2020).

Our findings regarding borderline personality traits and trait aggression reinforce the relevance of these traits as risk factors for suicidal behavior across the adult lifespan. Both borderline personality disorder and impulsive-aggression appear more common in those who have died by suicide at younger vs. older ages (McGirr et al., 2008; Qin, 2011), and individuals with borderline personality disorder tend to exhibit less suicidal behavior as they age (Stepp & Pilkonis, 2008). In our study, while the prevalence of these traits declined with age, the presence of these traits in late life was still indicative of a higher risk for suicide attempt, even though the traits themselves were less prominent. Therefore, while the population decrease in these traits may have led to earlier conclusions about their salience as risk factors for suicidal behavior in later life, our study shows that they remain important at all ages.

Neuropsychological performance on memory and language fluency tasks also declined with age, and previous work suggests it may be particularly relevant as a risk factor for suicidal behavior in older adults as this decline reaches impairment levels, potentially presaging dementia (Gujral et al., 2020). However, their relevance did not change in their classification of suicide attempters across the age range in this sample. This is consistent with previous findings of deficient memory and category fluency performance in younger to middle age, and suggests they remain important across the adult lifespan (Richard-Devantoy et al., 2014). The specific mechanisms by which these neurocognitive difficulties affect the risk for suicidal behavior (e.g. whether they contribute to daily stresses and decreased functional capacity, or whether they contribute to impaired decision-making or impulse control) are less well-known, but they remain clear targets for future research.

This study’s conclusions regarding developmental trends are limited by the cross-sectional nature of its data. Longitudinal studies of individuals at risk for suicidal behavior would provide an additional perspective on these changes. This study also focused on more recent suicidal behavior, which diminishes our ability to distinguish traits and behaviors related to suicide attempters that have occurred farther in the past. There is evidence that these risk factors may differ for those who had made attempts at a different point in their life (Gujral et al., 2020; Kenneally et al., 2018; Perry et al., 2021; Szücs A et al., 2020). However, study findings suggest that several key risk factors for suicidal behavior remain important throughout the adult lifespan, contributing to our ability to develop general models and algorithms that can be predictive over a wide age range. Our findings also highlight the importance of ruminative thinking as a potential risk factor at older ages - at a point in life where this type of thinking typically diminishes – and provides evidence for assessing it more carefully in older depressed patients who might be at risk.

Table 4.

Full classification model of each significant variable and variable interactions with age on suicide attempt status using age, sex, and site as covariates.

Odds Ratio Lower CI Upper CI Z value P-value
Full Model
(Intercept) 1.69 0.73 3.90 1.23 0.2190
Rumination 1.01 0.99 1.04 0.99 0.3219
Age 0.98 0.95 1.01 −1.39 0.1636
Borderline traits 1.05 1.01 1.09 2.36 0.0182
Aggression 1.00 0.98 1.03 0.38 0.7057
Category Fluency 0.93 0.86 1.01 −1.80 0.0723
Memory 0.99 0.97 1.01 −1.03 0.3048
Site (Pittsburgh) 1.71 0.55 5.32 0.93 0.3524
Site (Columbus) 0.52 0.23 1.19 −1.55 0.1220
Sex (Female) 0.68 0.34 1.38 −1.06 0.2878
Rumination × Age 1.00 1.00 1.00 3.05 0.0023
Open in a new tab

Highlights.

  • Depressive rumination distinguished suicidal behavior best at older ages

  • Clinical traits improved and neurocognitive abilities worsened across the lifespan

  • Borderline traits, aggression, memory, and category fluency discriminate attempters across the adult lifespan

Acknowledgements:

The authors would like to acknowledge the clinical staff at the University of Pittsburgh, Columbia University, and the Ohio State University sites for data collection.

Role of Funding source:

This work was supported by the American Foundation for Suicide Prevention [LSRG-0-007-13] as well as the National Institutes of Mental Health [R01MH085651 to KSz and 2P50MH090964 to JMa]. The sponsors played no role in the collection, analysis or interpretation of data; in the writing of the report; or in the decision to submit the article for publication.

Footnotes

Declarations of Interest: None

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References:

  1. Bartfai A, Winborg IM, Nordstrom P, & Asberg M (1990). Suicidal behavior and cognitive flexibility: Design and verbal fluency after attempted suicide. Suicide Life Threat Behav, 20(3), 254–266. [PubMed] [Google Scholar]
  2. Beck AT, Beck R, & Kovacs M (1975). Classification of suicidal behaviors: I. Quantifying intent and medical lethality. American Journal of Psychiatry, 132(3), 285–287. [DOI] [PubMed] [Google Scholar]
  3. Beck AT, Shuyler D, & Herman I (1974). Development of suicidal intent scales. In Beck AT, Resnik HLP, & Lettieri DJ (Eds.), The prediction of suicide (pp. 45–56). Charles Press. [Google Scholar]
  4. Benton AL, & Hamsher K (1994). Multilingual aphasia examination. AJA Associate. [Google Scholar]
  5. Buschke H (1973). Selective reminding for analysis of memory and learning. Journal of Verbal Learning & Verbal Behavior, 12(5), 543–550. 10.1016/S0022-5371(73)80034-9 [DOI] [Google Scholar]
  6. Buss AH, & Perry M (1992). The aggression questionnaire. Journal of Personality and Social Psychology, 63(3), 452. [DOI] [PubMed] [Google Scholar]
  7. De France K, & Hollenstein T (2019). Emotion regulation and relations to well-being across the lifespan. Developmental Psychology, 55(8), 1768–1774. 10.1037/dev0000744 [DOI] [PubMed] [Google Scholar]
  8. Devanand DP, Jacobs DM, Tang MX, Del Castillo-Castaneda C, Sano M, Marder K, Bell K, Bylsma FW, Brandt J, Albert M, & Stern Y (1997). The course of psychopathologic features in mild to moderate Alzheimer disease. Archives of General Psychiatry, 54(3), 257–263. 10.1001/archpsyc.1997.01830150083012 [DOI] [PubMed] [Google Scholar]
  9. Elgamal SA, Roy EA, & Sharratt MT (2011). Age and verbal fluency: The mediating effect of speed of processing. Canadian Geriatrics Journal: CGJ, 14(3), 66–72. 10.5770/cgj.v14i3.17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Emery L, Sorrell A, & Miles C (2020). Age Differences in Negative, but Not Positive, Rumination. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 75(1), 80–84. 10.1093/geronb/gbz109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. First M Sr., Gibbon M, & Williams JBW (1995). Structured clinical interview for DSM-IV Axis I Disorders—Patient Edition (SCID-I/P). Version 2.0. [Google Scholar]
  12. Folstein MF, Folstein SE, & McHugh PR (1975). “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12(3), 189–198. [DOI] [PubMed] [Google Scholar]
  13. Gujral S, Butters MA, Dombrovski AY, & Szanto K (2020). Late-Onset Suicide: A Dementia Prodrome? The American Journal of Geriatric Psychiatry. 10.1016/j.jagp.2020.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gunderson JG, Stout RL, McGlashan TH, Shea MT, Morey LC, Grilo CM, Zanarini MC, Yen S, Markowitz JC, Sanislow C, Ansell E, Pinto A, & Skodol AE (2011). Ten-Year Course of Borderline Personality Disorder: Psychopathology and Function From the Collaborative Longitudinal Personality Disorders Study. Archives of General Psychiatry, 68(8), 827–837. 10.1001/archgenpsychiatry.2011.37 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gvion Y, & Apter A (2011). Aggression, impulsivity, and suicide behavior: A review of the literature. Arch Suicide Res, 15(2), 93–112. 10.1080/13811118.2011.565265 [DOI] [PubMed] [Google Scholar]
  16. Hamilton M (1960). A rating scale for depression. Journal of Neurology, Neurosurgery & Psychiatry, 23, 56–62. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=14399272 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Karim HT, Ly M, Yu G, Krafty R, Tudorascu DL, Aizenstein HJ, & Andreescu C (2021). Aging faster: Worry and rumination in late life are associated with greater brain age. Neurobiology of Aging, 101, 13–21. 10.1016/j.neurobiolaging.2021.01.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Keilp JG, Beers SR, Burke AK, Melhem NM, Oquendo MA, Brent DA, & Mann JJ (2014). Neuropsychological Deficits in Past Suicide Attempters with Varying Levels of Depression Severity. Psychological Medicine, 44(14), 2965–2974. 10.1017/S0033291714000786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Keilp JG, Gorlyn M, Alexander GE, Stern Y, & Prohovnik I (1999). Cerebral blood flow patterns underlying the differential impairment in category vs letter fluency in Alzheimer’s disease. Neuropsychologia, 37(11), 1251–1261. 10.1016/S0028-3932(99)00032-9 [DOI] [PubMed] [Google Scholar]
  20. Keilp JG, Gorlyn M, Oquendo MA, Brodsky B, Ellis SP, Stanley B, & Mann JJ (2006). Aggressiveness, not impulsiveness or hostility, distinguishes suicide attempters with major depression. Psychological Medicine, 36(12), 1779–1788. 10.1017/S0033291706008725 [DOI] [PubMed] [Google Scholar]
  21. Keilp JG, Gorlyn M, Russell M, Oquendo MA, Burke AK, Harkavy-Friedman J, & Mann JJ (2013). Neuropsychological function and suicidal behavior: Attention control, memory and executive dysfunction in suicide attempt. Psychological Medicine, 43(3), 539–551. 10.1017/S0033291712001419 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Keilp JG, Sackeim HA, Brodsky BS, Oquendo MA, Malone KM, & Mann JJ (2001). Neuropsychological dysfunction in depressed suicide attempters. American Journal of Psychiatry, 158(5), 735–741. 10.1176/appi.ajp.158.5.735 [DOI] [PubMed] [Google Scholar]
  23. Kenneally LB, Szucs A, Szanto K, & Dombrovski AY (2018). Familial and social transmission of suicidal behavior in older adults. Journal of Affective Disorders, 15, 589–586. 10.1016/j.jad.2018.11.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lamar M, & Resnick SM (2004). Aging and prefrontal functions: Dissociating orbitofrontal and dorsolateral abilities. Neurobiol Aging, 25(4), 553–558. 10.1016/j.neurobiolaging.2003.06.005 [DOI] [PubMed] [Google Scholar]
  25. Látalová K, & Praško J (2010). Aggression in Borderline Personality Disorder. Psychiatric Quarterly, 81(3), 239–251. 10.1007/s11126-010-9133-3 [DOI] [PubMed] [Google Scholar]
  26. Lesage AD, Boyer R, Grunberg F, Vanier C, Morissette R, Ménard-Buteau C, & Loyer M (1994). Suicide and mental disorders: A case-control study of young men. The American Journal of Psychiatry, 151(7), 1063–1068. 10.1176/ajp.151.7.1063 [DOI] [PubMed] [Google Scholar]
  27. Linn BS, Linn MW, & Gurel L (1968). Cumulative illness rating scale. J Am Geriatr Soc, 16(5), 622–626. 10.1111/j.1532-5415.1968.tb02103.x [DOI] [PubMed] [Google Scholar]
  28. Marchant NL, Lovland LR, Jones R, Binette AP, Gonneaud J, Arenaza- Urquijo EM, Chételat G, & Villeneuve S (2020). Repetitive negative thinking is associated with amyloid, tau, and cognitive decline. Alzheimer’s & Dementia, 16(7), 1054–1064. 10.1002/alz.12116 [DOI] [PubMed] [Google Scholar]
  29. Masur DM, Fuld PA, Blau AD, Crystal H, & Aronson MK (1990). Predicting development of dementia in the elderly with the Selective Reminding Test. Journal of Clinical and Experimental Neuropsychology, 12(4), 529–538. 10.1080/01688639008400999 [DOI] [PubMed] [Google Scholar]
  30. Maylor EA, Reimers S, Choi J, Collaer ML, Peters M, & Silverman I (2007). Gender and sexual orientation differences in cognition across adulthood: Age is kinder to women than to men regardless of sexual orientation. Arch Sex Behav, 36(2), 235–249. 10.1007/s10508-006-9155-y [DOI] [PubMed] [Google Scholar]
  31. McGirr A, Renaud J, Bureau A, Seguin M, Lesage A, & Turecki G (2008). Impulsive-aggressive behaviours and completed suicide across the life cycle: A predisposition for younger age of suicide. Psychological Medicine, 38(3), 407–417. 10.1017/S0033291707001419 [DOI] [PubMed] [Google Scholar]
  32. Miranda R, & Nolen-Hoeksema S (2007). Brooding and reflection: Rumination predicts suicidal ideation at 1-year follow-up in a community sample. Behaviour Research and Therapy, 45(12), 3088–3095. 10.1016/j.brat.2007.07.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Morey LC (2014). The Personality Assessment Inventory. In Personality assessment, 2nd ed. (pp. 181–228). Routledge/Taylor & Francis Group. [Google Scholar]
  34. Morrison R, & O’Connor RC (2008). A systematic review of the relationship between rumination and suicidality. Suicide and Life-Threatening Behavior, 38(5), 523–538. [DOI] [PubMed] [Google Scholar]
  35. Nolen-Hoeksema S, & Morrow J (1991). A prospective study of depression and posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta Earthquake. Journal of Personality and Social Psychology, 61(1), 115. [DOI] [PubMed] [Google Scholar]
  36. Nolen-Hoeksema S, Wisco BE, & Lyubomirsky S (2008). Rethinking Rumination. Perspectives on Psychological Science, 3(5), 400–424. 10.1111/j.1745-6924.2008.00088.x [DOI] [PubMed] [Google Scholar]
  37. O’Carroll PW, Berman AL, Maris RW, Moscicki EK, Tanney BL, & Silverman MM (1996). Beyond the Tower of Babel: A nomenclature for suicidology. Suicide & Life-Threatening Behavior, 26(3), 237–252. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=8897663 [PubMed] [Google Scholar]
  38. Paraschakis A, Douzenis A, Michopoulos I, Christodoulou C, Vassilopoulou K, Koutsaftis F, & Lykouras L (2012). Late onset suicide: Distinction between “young-old” vs. “old-old” suicide victims. How different populations are they? Archives of Gerontology and Geriatrics, 54(1), 136–139. 10.1016/j.archger.2011.02.011 [DOI] [PubMed] [Google Scholar]
  39. Paris J, & Zweig-Frank H (2001). A 27-year follow-up of patients with borderline personality disorder. Comprehensive Psychiatry, 42(6), 482–487. 10.1053/comp.2001.26271 [DOI] [PubMed] [Google Scholar]
  40. Perry M, Buerke M, Szücs A, Allen TA, Bruine de Bruin W, Szántó K, & Dombrovski AY (2021). A lifetime of challenges: Real-life decision outcomes in early- and late-onset suicide attempters. Journal of Affective Disorders Reports, 4, 100105. 10.1016/j.jadr.2021.100105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Piquero AR, Carriaga ML, Diamond B, Kazemian L, & Farrington DP (2012). Stability in aggression revisited. Aggression and Violent Behavior, 17(4), 365–372. 10.1016/j.avb.2012.04.001 [DOI] [Google Scholar]
  42. Pompili M, Girardi P, Ruberto A, & Tatarelli R (2005). Suicide in borderline personality disorder: A meta-analysis. Nordic Journal of Psychiatry, 59(5), 319–324. 10.1080/08039480500320025 [DOI] [PubMed] [Google Scholar]
  43. Qin P (2011). The impact of psychiatric illness on suicide: Differences by diagnosis of disorders and by sex and age of subjects. Journal of Psychiatric Research, 45(11), 1445–1452. 10.1016/j.jpsychires.2011.06.002 [DOI] [PubMed] [Google Scholar]
  44. Rascovsky K, Salmon DP, Hansen LA, Thal LJ, & Galasko D (2007). Disparate letter and semantic category fluency deficits in autopsy-confirmed frontotemporal dementia and Alzheimer’s disease. Neuropsychology, 21(1), 20–30. 10.1037/0894-4105.21.1.20 [DOI] [PubMed] [Google Scholar]
  45. Ricarte J, Ros L, Serrano JP, Martínez-Lorca M, & Latorre JM (2016). Age differences in rumination and autobiographical retrieval. Aging & Mental Health, 20(10), 1063–1069. 10.1080/13607863.2015.1060944 [DOI] [PubMed] [Google Scholar]
  46. Ricarte Trives JJ, Navarro Bravo B, Latorre Postigo JM, Ros Segura L, & Watkins E (2016). Age and Gender Differences in Emotion Regulation Strategies: Autobiographical Memory, Rumination, Problem Solving and Distraction. The Spanish Journal of Psychology, 19, E43. 10.1017/sjp.2016.46 [DOI] [PubMed] [Google Scholar]
  47. Richard-Devantoy S, Berlim MT, & Jollant F (2014). A meta-analysis of neuropsychological markers of vulnerability to suicidal behavior in mood disorders. Psychological Medicine, 44(8), 1663–1673. 10.1017/S0033291713002304 [DOI] [PubMed] [Google Scholar]
  48. Rizk MM, Choo T-H, Galfalvy H, Biggs E, Brodsky BS, Oquendo MA, Mann JJ, & Stanley B (2019). Variability in Suicidal Ideation is Associated with Affective Instability in Suicide Attempters with Borderline Personality Disorder. Psychiatry, 82(2), 173–178. 10.1080/00332747.2019.1600219 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Rogers ML, Gorday JY, & Joiner TE (2021). Examination of Characteristics of Ruminative Thinking as Unique Predictors of Suicide-Related Outcomes. Journal of Psychiatric Research. 10.1016/j.jpsychires.2021.05.001 [DOI] [PubMed] [Google Scholar]
  50. Rogers ML, & Joiner TE (2017). Rumination, Suicidal Ideation, and Suicide Attempts: A Meta-Analytic Review. Review of General Psychology, 21(2), 132–142. 10.1037/gpr0000101 [DOI] [Google Scholar]
  51. Rönnlund M, Nyberg L, Bäckman L, & Nilsson L-G (2005). Stability, growth, and decline in adult life span development of declarative memory: Cross-sectional and longitudinal data from a population-based study. Psychology and Aging, 20(1), 3–18. 10.1037/0882-7974.20.1.3 [DOI] [PubMed] [Google Scholar]
  52. Sackeim HA (2001). The definition and meaning of treatment-resistant depression. Journal of Clinical Psychiatry, 62 Suppl 16, 10–17. http://www.ncbi.nlm.nih.gov/pubmed/11480879 [PubMed] [Google Scholar]
  53. Saffer BY, & Klonsky ED (2018). Do neurocognitive abilities distinguish suicide attempters from suicide ideators? A systematic review of an emerging research area. Clinical Psychology: Science and Practice, 25(1), e12227. 10.1111/cpsp.12227 [DOI] [Google Scholar]
  54. Sansone RA, & Wiederman MW (2014). Sex and age differences in symptoms in borderline personality symptomatology. International Journal of Psychiatry in Clinical Practice, 18(2), 145–149. 10.3109/13651501.2013.865755 [DOI] [PubMed] [Google Scholar]
  55. Steele IH, Thrower N, Noroian P, & Saleh FM (2018). Understanding Suicide Across the Lifespan: A United States Perspective of Suicide Risk Factors, Assessment & Management. Journal of Forensic Sciences, 63(1), 162–171. 10.1111/1556-4029.13519 [DOI] [PubMed] [Google Scholar]
  56. Stepp SD, & Pilkonis PA (2008). Age-related differences in individual DSM criteria for borderline personality disorder. Journal of Personality Disorders, 22(4), 427. 10.1521/pedi.2008.22.4.427 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Sütterlin S, Paap MCS, Babic S, Kübler A, & Vögele C (2012). Rumination and age: Some things get better. Journal of Aging Research, 2012, 267327. 10.1155/2012/267327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Szanto K, Galfalvy H, Kenneally L, Almasi R, & Dombrovski AY (2020). Predictors of serious suicidal behavior in late-life depression. European Neuropsychopharmacology, 40, 85–98. 10.1016/j.euroneuro.2020.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Szücs A, Szanto K, Wright AGC, & Dombrovski AY. (2020). Personality of late- and early-onset elderly suicide attempters. International Journal of Geriatric Psychiatry, 35(4), 384–395. 10.1002/gps.5254 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Tabert MH, Manly JJ, Liu X, Pelton GH, Rosenblum S, Jacobs M, Zamora D, Goodkind M, Bell K, Stern Y, & Devanand DP (2006). Neuropsychological prediction of conversion to Alzheimer disease in patients with mild cognitive impairment. Archives of General Psychiatry, 63(8), 916–924. 10.1001/archpsyc.63.8.916 [DOI] [PubMed] [Google Scholar]
  61. Videler AC, Hutsebaut J, Schulkens JEM, Sobczak S, & van Alphen SPJ (2019). A Life Span Perspective on Borderline Personality Disorder. Current Psychiatry Reports, 21(7), 51. 10.1007/s11920-019-1040-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wechsler D (2008). Wechsler Adult Intelligence Scale (4th ed.). Pearson Assessment. [Google Scholar]
  63. Yen S, Shea MT, Sanislow CA, Grilo CM, Skodol AE, Gunderson JG, McGlashan TH, Zanarini MC, & Morey LC (2004). Borderline Personality Disorder Criteria Associated With Prospectively Observed Suicidal Behavior. American Journal of Psychiatry, 161(7), 1296–1298. 10.1176/appi.ajp.161.7.1296 [DOI] [PubMed] [Google Scholar]
  64. Zanarini MC, Frankenburg FR, Reich DB, & Fitzmaurice G (2010). Time to attainment of recovery from borderline personality disorder and stability of recovery: A 10-year prospective follow-up study. American Journal of Psychiatry, 167(6), 663–667. http://ajp.psychiatryonline.org/doi/pdf/10.1176/appi.ajp.2009.09081130 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES