Introduction: Research on coping in advanced old age is scarce. In the present study, we explored coping patterns in near-centenarians and centenarians, and characteristics associated to using a specific coping pattern. Methods: We analyzed the frequency with which participants (N = 87, MAge = 99.05; SDage = 2.6; age range 95–107) reported using specific coping strategies (i.e., coping strategy use) and the relative preference for specific strategies (i.e., relative coping preferences) in data from the Fordham Centenarian Study. Moreover, we applied cluster analysis to detect coping patterns, and we compared cluster characteristics. Results: Very old individuals reported using emotion control and acceptance the most. Cluster analysis further revealed two distinct groups: The high coping group reported significantly higher coping strategy use than the low coping group (p < 0.001). The two groups also favored different strategies (p < 0.001), with the high coping group showing significantly higher relative preferences than the low coping group for active problem-solving, proactive prevention, and strategic planning (all ps < 0.05). The groups furthermore differed significantly in psychological strengths (i.e., personality, self-efficacy, ps < 0.001) and well-being outcomes (i.e., life satisfaction, p = 0.05). Discussion: This study provides evidence for a general preference for acceptance and emotion control in very old individuals, supporting theories of a developmental coping shift in old age, yet our findings also document the existence of groups with different coping patterns. More frequent coping use, and particularly continued use of active problem-solving and proactive prevention, may enable well-being in very old age.

Exceptional longevity is the result of a fortunate combination of genetic, lifestyle, and person characteristics [1]. Among the latter, an individual’s capacity to deal with difficulties in life may be particularly important. The process of managing stressors, including adaptation to difficult life events and circumstances, is referred to as “coping.” How individuals handle difficulties in life can reduce or increase their negative effects [2], either helping with adaptation or augmenting the risk for adverse short-term (e.g., acute stress experience) or long-term outcomes (e.g., negative physical and mental health trajectories). Research has identified several ways of coping, such as active problem-solving, support seeking, or acceptance, and higher-order coping dimensions have been proposed (e.g., problem-focused vs. emotion-focused; [3, 5]). To further understand interindividual differences as well as more general age-related trends, one needs to consider that the extent to which a given response is useful depends on stressor- and situation-related characteristics: e.g., active problem-solving is more adaptive when a difficulty can be resolved (e.g., unemployment can end by finding a new job), whereas when the problem cannot be eliminated (e.g., disability), adjusting expectations, and managing negative emotions are likely to be more adaptive [6].

One important factor to consider when investigating how very old adults deal with challenges is that – despite being all of the same advanced age – there are strong interindividual differences. Aside from event and situation characteristics that influence the choice and usefulness of coping responses, individuals have different coping styles, determining in what manner they typically cope in stressful situations [4, 5]. For example, the tendency to use active problem-solving has been found to depend on personality [7, 8], as well as on the belief that one is able to control events or manage stressors [9, 11]. Interindividual differences in psychological strengths, coping responses, and well-being outcomes remain important in older adults. A literature review [10] reported consistent findings of relationships between psychological strengths (i.e., personality and control beliefs), coping, and depressive symptomatology in older (60+) adults from clinical and community settings. They also showed that a higher sense of control and internal locus of control as well as more frequent use of problem-focused coping was related to less depressive symptoms.

Age-Associated Shift in Coping Responses in Older Adults

Above and beyond interindividual differences, a more general development in coping occurs throughout the life span [12, 13]. As adulthood can be described as shift in the balance of gains to losses [14, 15], a constant change in the stressors and in the resources available to deal with these stressors characterizes the life span. When reaching advanced old age, individuals face unique challenges in terms of stressor quantity, as losses accumulate across all key domains of functioning, and in terms of quality, as many of these losses cannot be fixed [6, 16]. Specifically, very old adults experience declines in physical, functional, and cognitive health (e.g., multimorbidity, mobility, and sensory impairment; [17]) and are further confronted with social (e.g., loss of children, relatives, and friends; [18]) and psychological losses (e.g., loss of meaningful activities; [19]). These age-related reductions in functional, social, and psychological resources not only constitute a stressor per se but also hamper implementation of strategies that would help deal with the stressor or compensate for loss [20].

When individuals can no longer actively solve problems because of the unchangeable nature of the stressors, resource restrictions, or reduced motivation for changing the situation, a shift in coping is to be expected to match the constraints and preserve well-being [13, 21, 23]. For example, the dual-process model of coping [21] describes that there will be a reduced preference for coping that focuses on solving the problem (i.e., assimilation) and increased preference for coping that aims at regulating emotions and cognitions (i.e., accommodation). In a similar vein, the motivational theory of life-span development [24] proposes a developmental shift from taking direct action to change issues and environment (primary control striving) toward internal, mostly motivational processes (secondary control strategies) with the purpose to reduce age-related losses and to maintain a maximum of control in one’s life.

Several studies support a coping shift in very old adults consistent with these models. For example, Martin and colleagues [25] found that centenarians were less likely to rely on strategic planning and active problem-solving and more likely to prefer cognitive reappraisal and acceptance of problems. Furthermore, there is evidence that, in advanced age, this developmental coping shift yields better outcomes in terms of well-being. For example, Rothermund and Brandstädter [22] found that, when coping with disabilities, problem-focused strategies were associated with positive well-being outcomes until the age of 70. After the age of 70 years, strategies focused on regulating emotion and cognition were associated with positive well-being outcomes and were found to buffer the negative effects of performance loss. Various other studies reported similar benefits of increased flexible goal adjustment as opposed to tenacious goal pursuit, which seems particularly useful for maintaining well-being in older adults with disability or worsening self-rated health [23, 26].

Present Study

In the current study, we investigated coping in centenarians and near-centenarians. Due to considerable loss accumulation, this life period is particularly challenging [16], and coping may become especially important. As only a few studies have investigated coping in very old age [25], our research question pertained to how centenarians and near-centenarians cope, and what are associated characteristics and outcomes.

Given that old age represents a developmental period with increasing interindividual differences due to specific personal characteristics and experiences [14, 27], we expected very old individuals to vary in their reported coping responses. We used exploratory cluster analysis to map different combinations of coping responses to distinguish individuals with specific coping patterns. In addition, following age-associated trends reported in the literature (e.g., [13, 21, 25]), we hypothesized higher reports of coping strategies that aim at regulating emotion and cognition (e.g., acceptance), compared to coping that focuses on solving the problem (e.g., active problem-solving).

Moreover, we investigated relations between patterns of coping revealed by this analysis and three sets of variables: cognitive and health resources (e.g., cognitive capacity, number of health conditions, and functional health), psychological strengths (e.g., personality, control beliefs, self-efficacy), and well-being outcomes (e.g., life satisfaction, depression). We hypothesized that cognitive and health resources as well as psychological strengths influence the selection of coping responses (e.g., [7, 11, 20]), so that patterns of coping responses would be related to differences in these variables. In turn, the selection of coping responses may contribute to well-being outcomes in advanced old age, so that patterns of coping responses would lead to differences in those variables.

Study Participants

We used data from the Fordham Centenarian Study, a population-based, cross-sectional study, which investigated very old individuals aged 95–107 years old living in New York City (see [18] for more detail). Participants were recruited mostly via the NYC voter registry. Study participation contained two in-person interviews conducted at the participant’s residence. Institutional Review Boards (i.e., Fordham University, Jewish Home Lifecare, and Hebrew Home for the Aged) approved the study and all participants signed an informed consent.

In the current study, we performed analysis on a subsample of participants who provided data for the calculation of coping scale scores (i.e., a response for at least one item of each coping scale). Of the original Fordham sample (N = 119), 20 individuals were unable to respond to the items of the coping instrument, either because of cognitive limitations (i.e., short MMSE <10, see [28]), other forms of inability, or refusal. Therefore, there were no coping data for these individuals, and by consequence these participants were not included in our subsample. Another 12 individuals replied to some coping items but did not provide sufficient data. That is, for some coping scales they did not provide any answers, and so we could not calculate a score for each coping scale. Our clustering method requires that each participant has a score for each of the coping scales; hence, these individuals could not be included. The remaining sample consisted of 87 individuals (MAge = 99.05; SDage = 2.6; age range 95–107), all of whom were assumed to have reliably responded to the items of the coping instrument, given their cognitive capacity presenting no or only mild cognitive impairment (i.e., short MMSE >10 [28]; one exception with a score of 9 was able to respond to the coping items and was also included).

Our sample contained 68 (78.2%) females and 19 (21.8%) males and varied in terms of ethnic/racial background (80.7% White; 18.2% Black), reflecting the composition of the NYC older population. The education level was 12.70 (SD = 3.9) years. Most participants lived in private residences (73.6%). A description of the key study variables can be found in Table 1, and a table comparing the study variables between participants included in the present subsample compared to excluded participants can be found in online Supplemental file 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000529896). From this analysis, we concluded that the included participants had significantly higher cognitive capacity levels than the excluded participants, but they were not statistically different on any other variables.

Table 1.

Characteristics of the sample, split by high versus low coping groups

All (N = 87)High coping (n = 56)Low coping (n = 31)Group comp.
M/%SDM/%SDM/%SDp value
Demographic variables 
 Age 99.1 2.6 98.6 2.7 99.9 2.1 0.02 
 Gender: female 78.2  75.0  83.9  0.42 
 Gender: male 21.8  25.0  16.1   
 Education years 12.7 3.9 13.2 3.7 11.8 4.2 0.12 
 Residence: community 73.6  71.4  77.4  0.62 
 Residence: institution 26.4  28.6  22.6   
Cognitive and health resources 
 Cognitive capacity 17.7 2.9 17.9 2.9 17.4 3.0 0.47 
 Number of health conditions 5.0 2.2 5.0 1.9 4.9 2.6 0.79 
 PADL 10.7 3.4 10.9 3.1 10.3 3.8 0.99 
 IADL 9.2 3.9 9.5 3.8 8.6 4.0 0.64 
Psychological strengths 
 Extraversion 3.1 0.9 3.3 0.9 2.9 0.9 0.03 
 Neuroticism 2.8 1.2 2.7 1.1 3.1 1.2 0.04 
 Agreeableness 4.1 0.7 4.1 0.7 3.9 0.9 0.31 
 Openness 3.8 0.8 4.0 0.7 3.5 0.8 0.002 
 Conscientiousness 4.3 0.7 4.4 0.6 4.1 0.8 0.03 
 Internal control 0.7 0.3 0.8 0.3 0.7 0.3 0.13 
 Self-efficacy 2.0 1.2 2.3 1.2 1.3 1.1 <0.001 
Well-being outcomes 
 Life satisfaction 2.2 1.1 2.4 1.1 1.8 1.0 0.02 
 Depressive symptoms 3.8 3.4 3.2 3.2 4.9 3.6 0.09 
All (N = 87)High coping (n = 56)Low coping (n = 31)Group comp.
M/%SDM/%SDM/%SDp value
Demographic variables 
 Age 99.1 2.6 98.6 2.7 99.9 2.1 0.02 
 Gender: female 78.2  75.0  83.9  0.42 
 Gender: male 21.8  25.0  16.1   
 Education years 12.7 3.9 13.2 3.7 11.8 4.2 0.12 
 Residence: community 73.6  71.4  77.4  0.62 
 Residence: institution 26.4  28.6  22.6   
Cognitive and health resources 
 Cognitive capacity 17.7 2.9 17.9 2.9 17.4 3.0 0.47 
 Number of health conditions 5.0 2.2 5.0 1.9 4.9 2.6 0.79 
 PADL 10.7 3.4 10.9 3.1 10.3 3.8 0.99 
 IADL 9.2 3.9 9.5 3.8 8.6 4.0 0.64 
Psychological strengths 
 Extraversion 3.1 0.9 3.3 0.9 2.9 0.9 0.03 
 Neuroticism 2.8 1.2 2.7 1.1 3.1 1.2 0.04 
 Agreeableness 4.1 0.7 4.1 0.7 3.9 0.9 0.31 
 Openness 3.8 0.8 4.0 0.7 3.5 0.8 0.002 
 Conscientiousness 4.3 0.7 4.4 0.6 4.1 0.8 0.03 
 Internal control 0.7 0.3 0.8 0.3 0.7 0.3 0.13 
 Self-efficacy 2.0 1.2 2.3 1.2 1.3 1.1 <0.001 
Well-being outcomes 
 Life satisfaction 2.2 1.1 2.4 1.1 1.8 1.0 0.02 
 Depressive symptoms 3.8 3.4 3.2 3.2 4.9 3.6 0.09 

PADL, personal activity of daily living; IADL, instrumental activity of daily living.

Cognitive capacity was measured with a short version of the MMSE.

PADL and IADL were measured with the OARS questionnaire.

Personality dimensions (extraversion, neuroticism, agreeableness, openness, conscientiousness) were measured with the BFI.

Life satisfaction was measured through the Diener Satisfaction with Life Scale. Depressive symptoms were measured with the GDS.

Missing values for education and self-efficacy (n = 1); life satisfaction (n = 3); depressive symptoms (n = 4); and IADL (n = 4). Cases with missing values were removed from the corresponding MANCOVAs.

Measures

Demographic Characteristics

The demographic characteristics assessed include chronological age, gender, years of education, and residence type (i.e., community or institution).

Coping Questionnaire

We used the Multidimensional Coping Style Inventory (MCSI; [29]) to assess use of specific coping responses. This 40-item coping measure has been developed in the Fordham Centenarian Study in the absence of a comprehensive tool reflecting recent coping models and was recently validated (see below). The 40 items belong to 11 subscales, and most of the subscales are measured with three items (see online Suppl. File 2 for more details). For each item, participants indicate on a 5-point Likert scale the extent to which they use a coping strategy (0 = not at all; 1 = a little; 2 = moderately; 3 = quite a bit; 4 = very often). Specifically, the questionnaire asks participants what they usually do when confronted with a difficult situation (e.g., examples such as interpersonal conflicts, health issues, other problems, are given in the introduction of the questionnaire), and this context is reiterated by using three different sentence stems (e.g., “When things are difficult, …”; “When things become tough, …,” “When faced with difficult problems or tasks, …”). The MCSI subscales include: active problem-solving (e.g., “… do you take action to try to make the situation better?”); proactive prediction (e.g., “… do you tend to see difficult situations arise?”), proactive prevention (e.g., “…do you try to address issues before they become bigger problems?”); strategic planning (e.g., “…do you make a plan to resolve the situation?”); seeking instrumental support (e.g., “…do you ask for advice?”); seeking emotional support (e.g., “…do you talk to people you’re close to about how you feel?”); religious coping (e.g., “…do you rely on God or a higher power?”); positive reappraisal (e.g., “…do you tend to see that there are also positive aspects?”); emotion control (e.g., “…do you manage to keep your feelings from taking over?”); acceptance (e.g., “…do you just try to live with it?”); and distraction (e.g., “…do you do things to keep your mind off a problem, such as watching TV, sleeping, or shopping?”).

The MCSI has been validated in 620 adults, aged 18–88 years, confirming scalar invariance across young and older adults [29]. In the validation study, all reliabilities were good to high [29]. In the present study, reliabilities for seven out of the 11 scales were above 0.6 (i.e., cognitive reappraisal: α = 0.60, to religious coping: α = 0.88), three scales ranged between 0.5 and 0.6 (i.e., proactive prevention: α = 0.58; acceptance: α = 0.57; emotional support: α = 0.55) and one scale was beneath 0.50 (i.e., proactive prediction: α = 0.49; see online Suppl. File 2 for more details, including the number of complete cases included for the reliability analysis on each subscale).

To calculate coping strategy use in each participant, we calculated an average score for each coping subscale – when the participant had a response for at least one item of that scale. Higher scores for a particular subscale indicated higher use of the coping strategy.

Cognitive and Health Resources

Cognitive capacity was evaluated by a short version of the mini-mental state examination (MMSE; [30]). The use of a short MMSE has been shown to reduce bias that stems from sensory and motor impairments while maintaining good psychometric properties and accuracy, making it a useful tool in centenarian studies [28, 31]. This short version has a maximum score of 21 and includes orientation (10 points), registration (3 points), attention/calculation (5 points), and recall (3 points). Higher numbers indicated higher cognitive functioning.

We assessed number of health conditions with a checklist of common age-related health conditions [17] to which participants responded with “yes” (= 1) or “no” (= 0). The list included high blood pressure, heart conditions, diabetes, chronic lung disease, ulcers or other serious stomach issues, cirrhosis or other liver problems, kidney condition, frequent urinary infections, incontinence, prostate problems, problems with vision or hearing, arthritis, osteoporosis, stroke, cancer, pneumonia, falls, and other. We calculated a sum score, with higher numbers indicating higher multimorbidity.

We measured functional health with the Older Americans Resources and Services (OARS) Multidimensional Functional Assessment Questionnaire [32], in which participants indicated how much difficulty they have performing seven personal activities of daily living (PADLs) and seven instrumental activities of daily living (IADLs) using a 3-point rating scale (0 = cannot do without help, to 2 = no difficulty; 0–14 each). We calculated sum scores for PADLs and IADLs, with higher numbers indicating higher functional health.

Psychological Strengths

Personality dimensions were assessed through the short version of the Big Five Inventory (BFI-10; [33]), as well as nine items from the long version, leading to 19 items. Subscales included extraversion (e.g., “Are you a person who is talkative?”), neuroticism (e.g., “…who gets nervous easily?”), openness (e.g., “…who is curious about many different things?”), conscientiousness (e.g., “… who is a reliable worker?”), and agreeableness (e.g., “…who is considerate and kind to almost everyone?”). Participants responded to these items with a 5-point Likert Scale, ranging from 1 (strongly disagree) to 5 (strongly agree). We measured internal control through three items (e.g., “Do you feel able to direct things in your daily life, so that they happen in the way you want them to happen?”), to which participants responded “yes” (= 1) or “no” (= 0). We measured self-efficacy through 4 items (e.g., “Do you meet the goals that you set for yourself?,” “Can you think of many ways to get out of a jam”?), to which participants responded with a 5-point Likert scale, ranging from 0 (not at all) to 4 (very much). For all personality scales, control, and self-efficacy, we calculated an average score per construct, with higher numbers indicating higher levels.

Well-Being Outcomes

We measured life satisfaction with the 5-item Satisfaction with Life Scale by Diener and colleagues ([34]; e.g., “Do you think that the conditions of your life are excellent?”). Participants replied to each item by using a 5-point Likert scale (0 = not at all, 4 = very much). We calculated average scores, with higher scores indicating more life satisfaction. We assessed depressive symptoms with a 15-item version of the Geriatric Depression Scale (GDS [35]; e.g., “Do you feel your situation is hopeless?”), with participants responding “yes” (= 1) or “no” (= 0). We calculated a sum score, with higher scores indicating more depressive symptoms.

Analysis Strategy

In our analysis, we examined both coping strategy use (i.e., how often a person indicated using a specific coping response) and relative coping preferences (i.e., how inclined a person was to endorse a specific coping response compared to their mean level of coping). We calculated coping strategy use by creating a mean score for the items of each subscale. We computed relative coping preferences as follows: for each participant, we first calculated the mean coping score across all subscales (i.e., reflecting an individual’s mean level of coping) as well as the standard deviation across all subscales (i.e., accounting for individual’s variability in level of coping). To calculate the relative coping preference, we applied the following formula to each coping strategy score: (Strategy coping score – Mean coping score)/SD coping scores. Thus, the resulting scores reflect the relative preference or relative disfavor for a particular coping response compared to the average. For example, if an individual has a relative coping preference of 0 for a specific coping subscale, this reflects that this response is perfectly average for this individual, indicating neither disfavoring nor favoring of this coping strategy. Instead, a negative relative preference score for one subscale reflects that this response is below average – indicating that the person has a disfavor for this specific coping response, whereas a positive relative preference score reflects that this response is above average – indicating that the person favors this specific coping response. Use of these standardized scores allows us to clearly establish whether, in addition to differences in coping strategy use, there are also differences in the patterns of relative coping preferences, while accounting for mean levels of coping and variability in levels of coping within individuals.

Following our objectives, analyses consisted of two steps. In the first step, we explored combinations of coping strategy use by our participants with a clustering method. Past research has used clustering to identify coping profiles based on 13 different coping strategies in younger cancer patients [36]. In older adults, Gerstorf and colleagues [27] used a clustering approach to examine individual differences in cognition, personality, control beliefs, and social integration and to investigate how these profiles were related to well-being and mortality. In our study, we applied model-based clustering with the Mclust package in R [37]. This method models clusters as Gaussian distributions with mean, variance, and probability of cluster membership and provides a statistical basis and objective criteria for selecting the optimal number of clusters. The clustering was done on coping strategy use in order to account for mean levels of coping. We compared the resulting cluster groups in terms of coping strategy use, as well as relative coping preferences. In the second step, we explored whether the groups emerging from the cluster analysis differed in demographic characteristics, cognitive and health resources (e.g., cognitive capacity, number of health conditions, PADL, IADL), psychological strengths (e.g., personality, control beliefs), and well-being outcomes (e.g., life satisfaction, depression).

All statistical comparisons between groups of individuals were made with SPSS. For the demographic variables, we conducted different tests depending on the measurement level of the variables (Fisher's exact test for gender and residence, ANOVA for age, and education years). Age was the only demographic variable that showed a significant difference between groups and therefore was controlled as a covariant in subsequent analyses to exclude the possibility that differences between groups could be explained by differences in age. The subsequent analyses compared groups of individuals on the different sets of variables (i.e., cognitive and health resources, psychological strengths, well-being outcomes) using multivariate analyses of covariance.

Coping Responses and Coping Groups

Figure 1 shows the coping strategy use, and the gray line represents the full subsample, for which we see the following pattern: the highest average score was found for acceptance, M = 2.72 (SD = 0.82), followed in order by emotion control, M = 2.62 (SD = 1.05), positive reappraisal, M = 2.31 (SD = 1.07), religious coping, M = 2.21 (SD = 1.55), and active problem-solving, M = 2.21 (SD = 1.12). The lowest average score was for seeking instrumental support, M = 1.61 (SD = 0.99), followed in order by seeking emotional support, M = 1.64 (SD = 1.08), distraction, M = 1.65 (SD = 0.99), proactive prediction, M = 1.77 (SD = 1.03), strategic planning, M = 1.91 (SD = 1.19), and proactive prevention, M = 2.16 (SD = 1.12).

Fig. 1.

Coping strategy use in the total sample, and in high versus low coping groups. Note: All scores significantly differ between groups (ps < 0.01), except emotional support (p = 0.87), and distraction (p = 0.28). Error bars represent a 95% confidence interval around the mean.

Fig. 1.

Coping strategy use in the total sample, and in high versus low coping groups. Note: All scores significantly differ between groups (ps < 0.01), except emotional support (p = 0.87), and distraction (p = 0.28). Error bars represent a 95% confidence interval around the mean.

Close modal

To identify potential coping patterns, we applied model-based clustering on the 11 coping strategy use scores. The Mclust algorithm compared the Bayesian Information Criterion for a series of clustering models with 1–5 clusters, and found a best solution, as indicated by the highest BIC (−2833.2), for two clusters of type EEI (i.e., diagonal, equal volume, and equal shape). Group membership was determined based on probability (i.e., for what cluster the probability of membership was highest). The average posterior probabilities for the most likely cluster classification were 0.96 and 0.99 for both clusters, which indicates reasonable classification certainty for the two-cluster solution. The first cluster, characterized by high coping strategy use, contained 56 individuals (solid line), and the second cluster, characterized by low coping strategy use, contained 31 individuals (dashed line, Fig. 1).

Differences in coping strategy use across high and low coping groups were statistically confirmed by a MANCOVA controlling for age, F(11, 74) = 21.16, p < 0.001, Wilks' λ = 0.24, partial η2 = 0.76. Multiple follow-up ANCOVAs revealed significantly higher scores in the high coping group compared to the low coping group for all coping strategies, ps < 0.01, except emotional support and distraction, ps > 0.20.

Differences in relative coping preferences across high and low coping groups were revealed by a MANCOVA controlling for age, F(11, 74) = 6.29, p < 0.001, Wilks' Λ = 0.54, partial η2 = 0.46 (see Fig. 2). Multiple follow-up ANCOVAs revealed that the high coping group showed significantly higher relative preferences than the low coping group for specific coping strategies, including active problem-solving, proactive prevention, and strategic planning, ps < 0.05. For seeking emotional support and distraction, the high coping group showed significantly more relative disfavor than the low coping group, ps < 0.01. There was a nonsignificant trend for instrumental support seeking, with the high coping group tending toward more relative disfavor for this coping strategy than the low coping group, p = 0.07. In contrast, both groups did not differ in their relative preferences for proactive prediction, cognitive reappraisal, emotion control, acceptance, and religious coping, ps > 0.20. Moreover, both the high and low coping group preferred acceptance the most compared to all other strategies, M = 0.56 (SD = 0.55) and M = 0.72 (SD = 0.76), respectively, followed by emotion control, M = 0.47 (SD = 0.72) and M = 0.59 (SD = 0.87), after which the relative preferences of both coping groups started diverging.

Fig. 2.

Relative coping preferences in the high and low coping groups. Note: Upward bars reflect favoring a coping subscale relative to the person average; downward bars reflect disfavoring a coping subscale. Error bars represent a 95% confidence interval around the mean. When error bars do not cross the zero line, coping subscales show significant deviations from the person average. Relative coping preferences significantly differed (ps < 0.05) between high and low coping groups for active problem-solving, proactive prevention, strategic planning, emotional support, distraction, and marginally for instrumental support (p = 0.07) and did not differ for proactive prediction (p = 0.79), cognitive reappraisal (p = 0.86), emotion control (p = 0.39), acceptance (p = 0.36), and religious coping (p = 0.24).

Fig. 2.

Relative coping preferences in the high and low coping groups. Note: Upward bars reflect favoring a coping subscale relative to the person average; downward bars reflect disfavoring a coping subscale. Error bars represent a 95% confidence interval around the mean. When error bars do not cross the zero line, coping subscales show significant deviations from the person average. Relative coping preferences significantly differed (ps < 0.05) between high and low coping groups for active problem-solving, proactive prevention, strategic planning, emotional support, distraction, and marginally for instrumental support (p = 0.07) and did not differ for proactive prediction (p = 0.79), cognitive reappraisal (p = 0.86), emotion control (p = 0.39), acceptance (p = 0.36), and religious coping (p = 0.24).

Close modal

Characteristics of Coping Groups

Table 1 displays the demographic characteristics, cognitive and health resources, psychological strengths, and well-being outcomes of the high and low coping groups. Concerning demographic characteristics, the high and low coping groups did not significantly differ in gender composition, years of education, and residence (i.e., community vs. institution), ps > 0.10. However, the low coping group was significantly older than the high coping group, p < 0.05; consequently, we controlled for this variable in all analyses. Concerning cognitive and health resources, a MANCOVA showed no significant differences between groups, F(4, 77) = 0.22, p > 0.90. Concerning psychological strengths, a MANCOVA showed statistically significant differences between groups, F(7, 77) = 4.66, p < 0.001, Wilks' λ = 0.70, partial η2 = 0.30. Multiple follow-up ANCOVAs revealed that the high coping group had significantly higher extraversion, p < 0.05, significantly lower neuroticism, p < 0.05, significantly higher openness, p < 0.01, conscientiousness, p < 0.05, and self-efficacy, p < 0.001. There was no significant difference for agreeableness and internal control, ps > 0.10. Finally, concerning well-being outcomes, a MANCOVA showed statistically significant differences between groups, F(2, 78) = 3.15, p = 0.05, Wilks' λ = 0.93, partial η2 = 0.08. Multiple follow-up ANCOVAs revealed that the high coping group showed significantly higher life satisfaction, p < 0.05, and a nonsignificant trend toward lower depressive symptoms, p = 0.09, compared to the low coping group.

This study investigated coping in very old individuals, an age-group mostly neglected in research. Considering specific coping strategies, near-centenarians and centenarians reported using acceptance the most, closely followed by emotion control, while they reported seeking instrumental support the least. This corroborates the particularities of centenarian coping responses described by Martin and colleagues [25], who also found that centenarians were particularly prone to using acceptance, while reporting low support seeking. They argued that the former could be due to social comparison mechanisms, whereas the latter could be due to a reduced social network. Martin and colleagues [25] also suggested possible cohort effects, with centenarians having grown up in an environment that stresses acceptance and self-reliance. However, the fact that we find the same effect in a different cohort of centenarians rather suggests that the focus on acceptance and self-reliance may be due to factors that are specific to becoming or being a centenarian. The high use of these coping responses may be a byproduct of how emotion regulation evolves in old age: for instance, the socioemotional selectivity theory (SST; [38]) suggests that older adults prioritize emotion regulation and emotional well-being.

The fact that nearly all very old adults preferred acceptance and emotion control to other strategies also confirms our hypothesis of higher reports of coping strategies that aim at regulating emotion and cognition rather than coping that focuses on solving the problem. This is in line with the developmental coping shift outlined in the introduction, conceptualized as a shift from assimilation to accommodation in the dual-process framework [21], or from primary to secondary control striving [24]. Following the dual-process framework, in order to accommodate, one needs to be able to accept limitations and control negative emotions, and thus the high scores for these coping strategies are in line with an accommodative coping mode in very old adults.

Beyond this more general emphasis on accommodative coping in our sample, our exploratory cluster analysis also revealed empirical evidence for important interindividual differences, in line with our hypotheses. Specifically, our analysis revealed two subgroups distinct by their mean level of coping. There was a group of individuals who reported higher coping strategy use, and a group of individuals who reported lower coping strategy use. This could be due to some centenarians and near-centenarians having more difficulty maintaining multiple effective coping responses, but another possibility is that some individuals choose to spend less overall effort on coping in very old age [39]. It is of note that high and low coping groups also differed in relative coping preferences: whereas both groups significantly preferred emotion control and acceptance, the high coping group in addition reported a significant preference for active problem-solving and proactive prevention not shared by the low coping group. These are coping strategies that are in line with assimilative coping, suggesting that some very old adults show a coping pattern that consists of a combination of assimilative and accommodative coping, contrary to a complete shift from one coping mode to another. Bailly and colleagues [40] identified a comparable coping profile in old and very old individuals, consisting of high flexible goal adjustment in combination with high tenacious goal pursuit.

We also hypothesized that differences in cognitive and health resources would have an impact on patterns of coping responses. Considering that near-centenarians and centenarians have likely experienced many losses, yet may vary in their level of functioning, one could argue that being less able to maintain a wide repertoire of coping including assimilative as well as accommodative strategies would be dependent on differences in cognitive and health resources. Yet, in our study, we found no evidence for notable differences in cognitive and health resources between both coping groups. Therefore, it seems that, even though the decline of such resources may contribute to a global shift in coping in very old age [22], it is not an important factor in explaining differences in the coping patterns we found at the extreme end of the life span.

We further hypothesized that psychological strengths would be associated with patterns of coping responses. Consistent with this, our results show that the higher coping group displayed unique psychological characteristics: higher extraversion, openness, conscientiousness, self-efficacy, and lower neuroticism, in line with previous research supporting the importance of these traits in determining the use of coping responses requiring problem engagement [8, 10]. Centenarians and near-centenarians with these traits may feel more competent and have a natural tendency to actively engage in situations, which could explain why they use more active problem-solving and proactive prevention. Moreover, these traits may reflect lifelong dispositions, with generally increased coping efforts over the course of the life span and persisting into advanced old age. It is of note that both groups differed in self-efficacy, but not in internal control. It is possible that the lack of an effect is due to how we measured this construct, through three dichotomous items, instead of several 5-point Likert items like the other psychological resources. However, although various studies reported links between control beliefs and coping [10], others did not confirm this link but found that the effect of assimilative coping on well-being was stronger for individuals with high control beliefs [41]. An important issue in interpreting the findings is a possible dissociation between efforts used to exert control and control beliefs. Perhaps individuals at age 100 feel confident regarding their capacity to deal with difficulties (i.e., self-efficacy), but feel less confident about controlling their environment more generally (i.e., control beliefs). That the high coping group also strongly disfavored coping that consisted of seeking support or distraction reinforces the notion that the high coping group contains individuals who feel particularly capable of personally dealing with any challenge that comes their way.

Finally, we also hypothesized that coping patterns would be associated with well-being outcomes. Our findings suggested that the high coping group reports more life satisfaction than the low coping group. However, the tendency for lower depressive symptoms in the high coping group in our study was not significant. The fact that we found significantly higher life satisfaction in the high coping group may imply that the combination of assimilative and accommodative coping strategies, rather than a shift to accommodative coping strategies only, protects well-being in this particularly old age-group. This conclusion is in line with the results from Bailly and colleagues [40] showing that tenacious goal pursuit in combination with flexible goal adjustment is associated to higher well-being.

The co-occurrence of assimilative and accommodative coping in some centenarians is coherent with the notion that many of the difficulties encountered in very old age require both accommodative and assimilative coping [39]. On the one hand, accommodative coping efforts involving acceptance and emotion regulation may render an unchangeable issue (e.g., death of the spouse) more bearable. On the other hand, assimilative coping such as problem-oriented coping may allow to deal with accompanying issues (e.g., consequences of the partner loss for everyday management or feelings of loneliness). Furthermore, as proposed by the strength and vulnerability integration model (SAVI; [42]), assimilative strategies such as proactive prevention may also serve to safeguard older adults’ emotional well-being by limiting exposure to situations that are potentially distressing.

That the high and low coping group differed in subjective well-being is particularly interesting in conjunction with the fact that the high coping group did not show significantly higher functioning in terms of cognition or health. Therefore, there is no evidence that using a wider coping repertoire depends on or confers any objective benefits. However, both assimilative and accommodative coping may be important for how very old adults feel, perhaps because they help maintain a sense of life mastery and positive views of the self and personal development in very old age [43]. Our findings suggest that although emotion control and acceptance are essential to deal with the challenges offered by advanced old age, the combination with assimilative strategies such as active problem-solving remains crucial for maintaining well-being [10, 44].

Limitations

Conducting in-person interviews with centenarians is particularly challenging (e.g., issues with recruitment, health issues, cognitive impairment, fatigue), and thus data such as analyzed in the present study are very rare for this population. We experienced common limitations for this type of data, such as the presence of missing data (e.g., the two coping groups had 12% and 14% missing values across the different items that constituted the coping scales). Because of the challenges of collecting this type of data in centenarian participants, the participants that we included in our sample also had significantly higher cognitive capacity than participants who had to be excluded because they were not able to provide responses. Therefore, a limitation is the uncertainty with which our results can be generalized to individuals with low cognitive capacity. Another limitation concerns our instrument for measuring coping. Self-report measurements preclude ascertaining which coping strategies are effectively used by participants and our instrument did not include collecting information about the specific type of stressors that different participants encountered. A final limitation to note is the cross-sectional design of our study, including the absence of a younger comparison group. This limits the interpretation of our findings in terms of how coping may differ across cohorts and change over time. Future directions should replicate the existence of the different coping patterns in other samples of centenarians, as well as confirm the pattern of associations with cognitive and health resources, psychological strengths, and well-being outcomes. In addition, investigating more complex patterns of coping longitudinally, starting in younger ages, could provide hints on how coping patterns may evolve over time.

Very old people deal with unrivaled challenges, and our results show that they mostly turn to acceptance and emotion control to deal with these challenges, indicative of an accommodative coping mode that is typically described for old age in the literature. In addition, we found a subgroup of centenarians with a high level of coping and an extended coping repertoire including also active problem-solving and proactive prevention, indicative of an assimilative coping mode. Given that these individuals also had higher well-being outcomes suggests that not the complete shift toward one mode of coping (e.g., accommodation), but the maintenance of a balanced coping repertoire combining assimilative and accommodative coping modes is related to successful adaptation in very old age. That being an coping expert was unrelated to health, but to higher self-efficacy, openness, conscientiousness, extraversion, offers suggestions to better understand which conditions may shape coping patterns until very old age and may provide avenues for potential life-span interventions to strengthen individuals’ coping repertoire. It is further of high practical importance to create environments for older people that foster feelings of self-efficacy and encourage the use of problem-engaging strategies, as these are remained essential to maintain well-being until very old age.

We thank the study participants for sharing valuable information with us. We thank Dr. Antonietti and Dr. Hilpert for statistical advice.

This study protocol was reviewed and approved by Institutional Review Boards, including Fordham University (protocol: Fordham Centenarian Study), Jewish Home Lifecare (JHL Protocol #: 2012-04), and Hebrew Home for the Aged (PROTOCOL: 0811I/P078/00. All participants provided written informed consent.

The authors have no conflicts of interest to declare.

This work was made possible with the support of the Swiss National Science Foundation to the SWISS100 Research Group (CRSII5_186239/1). We furthermore thank the Brookdale Foundation Group and Fordham University for enabling the Fordham Centenarian Study.

Dr. Kim Uittenhove initiated and subsequently led the publication project, conducted the analyses, and was responsible for writing the paper. Prof. Dr. Daniela Jopp led the Fordham Centenarian Study that yielded the results for the analysis in the present paper and led the construction and validation of the coping measure. In addition, Prof. Dr. Daniela Jopp collaborated in the writing of the paper, by refining, e.g., the theoretical background and advising on the results section. Dr. Lampraki contributed to the conceptualization and verification of the analysis, the paper structure, and to the writing of the text. Prof. Dr. Boerner collaborated with Prof. Dr. Jopp on the Fordham Centenarian Study and the coping measure on which this study is based on and contributed to the theoretical refinement of the paper.

The data that support the findings of this study are not publicly available due to these data being collected with the help of private funding agencies, and no data sharing policies were put into place. The data are available from Dr. Daniela Jopp upon reasonable request. Researchers interested in the coping instrument may also contact Dr. Daniela Jopp.

1.
Passarino
G
,
De Rango
F
,
Montesanto
A
.
Human longevity: genetics or Lifestyle? It takes two to tango
.
Immun Ageing
.
2016 Dec
13
1
12
6
.
2.
Taylor
SE
,
Stanton
AL
.
Coping resources, coping processes, and mental health
.
Annu Rev Clin Psychol
.
2007 Apr 27
3
1
377
401
.
3.
Carver
CS
,
Scheier
MF
,
Weintraub
JK
.
Assessing coping strategies: a theoretically based approach
.
J Pers Soc Psychol
.
1989 Feb
56
2
267
83
.
4.
Folkman
S
,
Lazarus
RS
.
An analysis of coping in a middle-aged community sample
.
J Health Soc Behav
.
1980 Sep 1
21
3
219
39
.
5.
Skinner
EA
,
Edge
K
,
Altman
J
,
Sherwood
H
.
Searching for the structure of coping: a review and critique of category systems for classifying ways of coping
.
Psychol Bull
.
2003 Mar
129
2
216
69
.
6.
Boerner
K
,
Jopp
D
.
Improvement/maintenance and reorientation as central features of coping with major life change and loss: contributions of three life-span theories
.
Human Development
.
2007
;
50
(
4
):
171
95
.
7.
Carver
CS
,
Connor-Smith
J
.
Personality and coping
.
Annu Rev Psychol
.
2010 Jan 10
61
1
679
704
.
8.
Connor-Smith
JK
,
Flachsbart
C
.
Relations between personality and coping: a meta-analysis
.
J Pers Soc Psychol
.
2007 Dec
93
6
1080
107
.
9.
Bandura
A
.
Self-efficacy: toward a unifying theory of behavioral change
.
Psychol Rev
.
1977 Mar
84
2
191
215
.
10.
Bjørkløf
GH
,
Engedal
K
,
Selbæk
G
,
Kouwenhoven
SE
,
Helvik
AS
.
Coping and depression in old age: a literature review
.
Dement Geriatr Cogn Disord
.
2013
35
3–4
121
54
.
11.
Petrosky
MJ
,
Birkimer
JC
.
The relationship among locus of control, coping styles, and psychological symptom reporting
.
J Clin Psychol
.
1991 May
47
3
336
45
.
12.
Aldwin
CM
,
Yancura
LA
,
Boeninger
DK
.
Coping across the life span
. In:
Freund
A
,
Lamb
M
, editors.
Handbook of lifespan developmental psychology
New York
John Wiley
2010
Vol. 2
.
13.
Brandtstädter
J
.
Goal pursuit and goal adjustment: self-regulation and intentional self-development in changing developmental contexts
.
Adv Life Course Res
.
2009 Mar 1
14
1–2
52
62
.
14.
Baltes
PB
.
On the incomplete architecture of human ontogeny: selection, optimization, and compensation as foundation of developmental theory
.
Am Psychol
.
1997 Apr
52
4
366
80
.
15.
Heckhausen
J
,
Dixon
RA
,
Baltes
PB
.
Gains and losses in development throughout adulthood as perceived by different adult age groups
.
Developmental Psychology
.
1989 Jan
25
1
109
21
.
16.
Baltes
PB
,
Smith
J
.
New frontiers in the future of aging: from successful aging of the young old to the dilemmas of the fourth age
.
Gerontology
.
2003
;
49
(
2
):
123
35
.
17.
Jopp
DS
,
Boerner
K
,
Rott
C
.
Health and disease at age 100: findings from the second heidelberg centenarian study
.
Dtsch Arztebl Int
.
2016 Mar
113
12
203
10
.
18.
Jopp
DS
,
Park
MKS
,
Lehrfeld
J
,
Paggi
ME
.
Physical, cognitive, social and mental health in near-centenarians and centenarians living in New York City: findings from the Fordham Centenarian Study
.
BMC Geriatr
.
2016 Dec
16
1
1
.
19.
Jopp
DS
,
Boerner
K
,
Cimarolli
V
,
Hicks
S
,
Mirpuri
S
,
Paggi
M
.
Challenges experienced at age 100: findings from the Fordham centenarian study
.
J Aging Soc Policy
.
2016 Jul 2
28
3
187
207
.
20.
Freund
AM
,
Baltes
PB
.
Toward a theory of successful aging. Selection, optimization, and compensation
. In:
Fernández-Ballesteros
R
, editor.
GeroPsychology. European perspectives for an aging world
Cambridge
Hogrefe & Huber
2007
. p.
239
54
.
21.
Brandtstädter
J
,
Rothermund
K
.
The life-course dynamics of goal pursuit and goal adjustment: a two-process framework
.
Developmental Rev
.
2002 Mar 1
22
1
117
50
.
22.
Rothermund
K
,
Brandtstädter
J
.
Coping with deficits and losses in later life: from compensatory action to accommodation
.
Psychol Aging
.
2003 Dec
18
4
896
905
.
23.
Hajek
A
,
König
HH
.
Flexible goal adjustment moderates the link between self-rated health and subjective well-being. Findings from the general population
.
Aging Ment Health
.
2021 Jul 3
25
7
1345
50
.
24.
Heckhausen
J
,
Wrosch
C
,
Schulz
R
.
A motivational theory of life-span development
.
Psychol Rev
.
2010 Jan
117
1
32
60
.
25.
Martin
P
,
Rott
C
,
Poon
LW
,
Courtenay
B
,
Lehr
U
.
A molecular view of coping behavior in older adults
.
J Aging Health
.
2001 Feb
13
1
72
91
.
26.
Boerner
K
.
Adaptation to disability among middle-aged and older adults: the role of assimilative and accommodative coping
.
J Gerontol B Psychol Sci Soc Sci
.
2004 Jan 1
59
1
P35
42
.
27.
Gerstorf
D
,
Smith
J
,
Baltes
PB
.
A systemic-wholistic approach to differential aging: longitudinal findings from the Berlin Aging Study
.
Psychol Aging
.
2006 Dec
21
4
645
63
.
28.
Kliegel
M
,
Moor
C
,
Rott
C
.
Cognitive status and development in the oldest old: a longitudinal analysis from the Heidelberg Centenarian Study
.
Arch Gerontol Geriatr
.
2004 Sep 1
39
2
143
56
.
29.
Jopp
DS
,
Boerner
K
,
Lampraki
C
,
Hoffman
A
,
Roquet
A
,
Uittenhove
K
. The Multidimensional Coping Style Inventory (MCSI): Factor structure, measurement invariance, and construct validity.
University of
30.
Folstein
MF
,
Folstein
SE
,
McHugh
PR
.
“Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician
.
J Psychiatr Res
.
1975 Nov 1
12
3
189
98
.
31.
Holtsberg
PA
,
Poon
LW
,
Noble
CA
,
Martin
P
.
Mini-Mental State Exam status of community-dwelling cognitively intact centenarians
.
Int Psychogeriatr
.
1995 Sep
7
3
417
27
.
32.
Fillenbaum
GG
.
Multidimensional functional assessment of older adults: the Duke Older Americans Resources and Services procedures
Hillsdale, NJ
Erlbaum
1988
.
33.
Rammstedt
B
,
John
OP
.
Measuring personality in one minute or less: a 10-item short version of the Big Five Inventory in English and German
.
J Res Personal
.
2007 Feb 1
41
1
203
12
.
34.
Diener
ED
,
Emmons
RA
,
Larsen
RJ
,
Griffin
S
.
The satisfaction with life scale
.
J Personal Assess
.
1985 Feb 1
49
1
71
5
.
35.
Yesavage
JA
,
Sheikh
JI
.
Geriatric Depression Scale (GDS): recent evidence and development of a shorter version
.
Clin Gerontol
.
1986 Jun
5
1–2
165
73
.
36.
Luszczynska
A
,
Gerstorf
D
,
Boehmer
S
,
Knoll
N
,
Schwarzer
R
.
Patients’ coping profiles and partners’ support provision
.
Psychol Health
.
2007 Oct 1
22
7
749
64
.
37.
Scrucca
L
,
Fop
M
,
Murphy
TB
,
Raftery
AE
.
Mclust 5: clustering, classification and density estimation using Gaussian finite mixture models
.
R J
.
2016
;
8
(
1
):
289
317
.
38.
Carstensen
LL
.
Evidence for a life-span theory of socioemotional selectivity
.
Curr Dir Psychol Sci
.
1995 Oct
4
5
151
6
.
39.
Aldwin
CM
,
Sutton
KJ
,
Chiara
G
,
Spiro
A
.
Age differences in stress, coping, and appraisal: findings from the normative aging study
.
J Gerontol B Psychol Sci Soc Sci
.
1996 Jul 1
51
4
P179
88
.
40.
Bailly
N
,
Martinent
G
,
Ferrand
C
,
Gana
K
,
Joulain
M
,
Maintier
C
.
Tenacious goal pursuit and flexible goal adjustment in older people over 5 years: a latent profile transition analysis
.
Age Ageing
.
2016 Mar 1
45
2
287
92
.
41.
Jopp
DS
,
Schmitt
M
.
Dealing with negative life events: differential effects of personal resources, coping strategies, and control beliefs
.
Eur J Ageing
.
2010 Sep
7
3
167
80
.
42.
Charles
ST
.
Strength and vulnerability integration: a model of emotional well-being across adulthood
.
Psychol Bull
.
2010 Nov
136
6
1068
91
.
43.
Hill
RD
.
A positive aging framework for guiding geropsychology interventions
.
Behav Ther
.
2011 Mar 1
42
1
66
77
.
44.
Diener
ED
,
Suh
ME
.
Subjective well-being and age: an international analysis
.
Annu Rev Gerontol Geriatr
.
1997 Jan 1
17
1
304
24
.