The Role of the Hostile-World Scenario in Predicting Physical and Mental Health Outcomes in Older Adults

Abstract

Objective: The hostile-world scenario (HWS) denotes a personal belief system regarding threats to one’s physical and mental integrity. We examined whether the HWS predicted health among older adults. Method: The Israeli branch of the Survey of Health, Ageing and Retirement in Europe (SHARE-Israel) provided data on 1,286 participants, aged 50+, interviewed in two waves 4 years apart. A special measure assembled items pertinent to the HWS throughout the SHARE survey. Nine outcomes indicated physical health (e.g., activities of daily living, medical conditions) and mental health (e.g., depressive symptoms, satisfaction with life). Results: The HWS at Wave 1 predicted all physical and mental outcomes at Wave 2, except cognitive functioning, beyond effects of sociodemographics and the respective outcome’s baseline at Wave 1. This predictive effect was stronger among older participants. Discussion: The results support the conception of the HWS as a psychological monitor that senses approaching functional declines in later life.

Keywords

hostile-world scenario physical health mental health older adults longitudinal prediction

The ability of individuals to scan their world for potential threats is an imperative for survival. Large-scale threats to human lives around the globe may include natural disasters, wars, violence, malnutrition, epidemics, and political upheavals. On a more regular basis, people constantly face personal threats of failure, illness, accidents, interpersonal conflicts, dissolutions of close relationships, loss of significant others, aging, and death. Addressing the subjective representations of these harsh realities, the model on the pursuit of happiness in a hostile world (Shmotkin, 2005, 2011; Shmotkin & Shrira, 2012b, 2013) proposed the concept of the hostile-world scenario (HWS), referring to an image of actual or potential self-perceived threats to one’s life or, more broadly, to one’s physical or mental integrity. The HWS functions as a system of appraisals regarding appearances of significantly negative life conditions. Usually, the HWS is activated adaptively and allows the individual to keep a sufficient level of vigilance toward dangers and detrimental circumstances so that a sense of safety and well-being is maintained. However, an excessive activation of the HWS may result in an overwhelming sense of a catastrophic world. The present study sought to explore the role of the HWS in later life, mainly through the prediction of health-related functional outcomes, by analyzing a nationwide, epidemiologically oriented survey data of older adults.

The predictability of functional outcomes is essential for the study of health on both theoretical and practical grounds. Thus, for both theorists and clinicians, the ability to forecast the functional status of individuals by earlier data obtained from the same individuals attests to the integrity and sustainability of personal health systems over time (Gerstorf, Smith, & Baltes, 2006). This predictability is particularly pertinent to older people, whose functional status may undergo destabilization and ultimate disintegration (Shmotkin, Shrira, Eyal, Blumstein, & Shorek, 2014). In this context, longitudinal research has persistently shown that negative conditions can be predictive of later functional impairment. Pertinent examples of such predictors are cumulative adversity of life events, posttraumatic reactions, depressive symptoms, and loneliness, which all predicted subsequent health outcomes among older populations; the predicted outcomes varied and included functional status, self-rated health, injurious falls, accelerated aging, and mortality (Clemson, Kendig, Mackenzie, & Browning, 2015; Lohr et al., 2015; Luo, Hawkley, Waite, & Cacioppo, 2012; Shrira & Litwin, 2014). In these examples, the predictive conditions appear to mingle self-reports of objective occurrences and subjective interpretations, whose core message depicts actual and perceived threats to one’s life or integrity.

In these considerations of how adversity may predict eventual functional outcomes among older adults, we propose the HWS as a concept that integrates the individual’s perceptions of threats within a framework that emphasizes adaptive capability. In our guiding model, the role of the HWS is regarded as complementary to, and interactive with, positively oriented systems that provide individuals with a favorable psychological environment (Shmotkin & Shrira, 2012b; Shrira, Palgi, Ben-Ezra, & Shmotkin, 2011). Such positive systems notably include subjective well-being (people’s evaluations of the satisfaction and pleasantness in their life) and meaning in life (people’s conceptions that they lead a life corresponding to their values and potentials). Subjective well-being and meaning in life act in both separate and conjoint modes of operation, which are beyond this study’s scope (for a review, see Ryan & Deci, 2001; Shmotkin & Shrira, 2013). Yet, the favorable psychological environment that these systems promote is embodied in the current data as positive, or efficient, functioning across major life domains. In epidemiological research, positive functioning markers typically constitute indicators of wellness and livability, as well as warranted targets of prediction (Baltes & Smith, 1997; Ryff & Singer, 1998).

Presumably playing a dynamic role in adaptation, the HWS is not merely an aggregate of personal references to threats confronting the individual. Thus, the HWS has a vulnerable aspect of negative engagement with the representations of one’s actual and potential adversities in life. Such engagement involves a weakening of one’s sense of competence, as happening when HWS themes render the individual into worries, fears, and pessimistic self-views. Typical of such a negative engagement are expressions such as “I often think about my own death” or “I am alarmed to think of how human beings’ cruelty can hurt me.” However, the HWS also has a resilient aspect of positive engagement with the representations of one’s adversities, involving strengthening of one’s sense of competence, as happening when HWS themes stir the individual into proactive behaviors and optimistic self-views. Typical of such a positive engagement are expressions such as “I plan ahead as to how I would handle possible hazards” or “Handling dangerous situations well strengthens me” (the above expressions were included in a questionnaire of the HWS, established, and prepared for publication; see also Shrira, Palgi, Ben-Ezra, & Shmotkin, 2011). Thus, our guiding model posits that the HWS is a system of self-beliefs that provides an array of responses to significant challenges in life and prioritizes certain cognitive, emotional, and behavioral tendencies. These features correspond to evidence on the dynamic and agentic qualities underlying the interrelated systems of subjective well-being (Busseri & Sadava, 2013) and meaning in life (Hicks & King, 2009).

The interplay between positive and negative engagement with one’s representations of adversities may relate to further differentiations within the HWS concept. This was evident when a realized HWS could be operationalized in terms of cumulative adversity (Shmotkin & Litwin, 2009) or negative anchor periods in life (Shmotkin & Shrira, 2012a). While the HWS, as approached in the present study, is a self-reported image of actual (i.e., present) or potential (i.e., future) threats to one’s integrity, the bordering notion of realized HWS is based on a self-report of traumatic, or otherwise adverse, experiences that threatened one’s integrity in the past. Prior research of older adults indicated that effects of the realized HWS may be shaped by characteristics such as the primary target to which the adversity was oriented (self-oriented vs. other-oriented), the timing of the adversity along the life span, and the dose of adverse events that the individual experienced (Keinan, Shrira, & Shmotkin, 2012; Shrira, Shmotkin, & Litwin, 2012). As the realized HWS manifests evident associations of past adversities with present functioning and health (Shrira, 2012, 2014), this study examines whether the HWS, with its focus on threats operating in present life, is associated with future functioning and health.

Aging people constitute a particularly pertinent population for studying the effects of the HWS. Leading into the final stage in a person’s life, old age is typically characterized by an accumulation of adversities such as major health problems, functional impairment, social seclusion, and death of significant others. The realization of these declines and losses often induces depletion in resources and an increase in vulnerability (Baltes, 1997; Hobfoll & Wells, 1998). The challenges of aging may also echo distant memories of adversity, thus further taxing the well-being that old people seek to sustain (Cappeliez & O’Rourke, 2006; Shmotkin & Shrira, 2012a). To offset increasing negative occurrences, older adults are found to exercise an efficient emotional regulation (Urry & Gross, 2010), whereby they intensify beneficial strategies such as prioritizing positivity effects and optimizing their affect (Labouvie-Vief, 2008; Reed & Carstensen, 2012). However, when confrontation with tangible adversities holds survival value, a better adaptation in older people requires that they expand their attention to potential threats (De Raedt, Koster, & Ryckewaert, 2013) and resort to a more combinatory outlook on positive and negative eventualities (Lang, Weiss, Gerstorf, & Wagner, 2013; Palgi & Shmotkin, 2010; Shrira, Palgi, Ben-Ezra, & Shmotkin, 2011).

Threats to one’s integrity are expected in our guiding model to activate the HWS so that it functions to monitor prospective adversity. Through this monitoring, the HWS may play a role in predicting personal eventualities pertinent to the self-beliefs and anticipations of which the HWS is composed. Therefore, the first hypothesis in the present study was that the HWS would predict the individual’s major markers of functioning over time. As the HWS may be more strongly activated in the face of harsher, aging-related threats, such as imminent deterioration and death, a second hypothesis was that the predictive power of the HWS would be stronger for older, than for younger, participants.

The present study analyzed two-wave data extracted from a longitudinal, national survey of older (age 50+) people. To test the study’s hypotheses in this data set, we examined the ability of the HWS to predict subsequent change in an array of functioning markers pertaining to physical health (basic and instrumental activities of daily living, movement capability, physical symptoms, and medical conditions) as well as to mental health (depressive symptoms, satisfaction with life, cognitive functioning, and social activity). As testing of the hypotheses involved prediction of change in functioning over time, the baseline of the functioning markers was adjusted for. Also adjusted for were sociodemographic variables (age, gender, origin, education, income, and marital status), which might represent either resources or deficiencies that could confound the genuine predictive effect of the HWS on subsequent functioning outcomes.

Method

Participants and Procedure

Data were drawn from the Israeli branch of the Survey of Health, Ageing and Retirement in Europe (SHARE-Israel), which presents a national sample of Israelis aged 50 or older and their spouses regardless of age. SHARE was designed as a multidisciplinary survey of the aging population, collecting data on health, socioeconomic status, psychosocial functioning, and social networks. SHARE-Israel conducted a probability sample of households within 150 representative statistical areas delineated by geographical and sociodemographic criteria. The complete Israeli data at Wave 1 (collected during 2005-2006) consisted of 2,598 community (i.e., not institutionalized) dwellers in 1,771 households. Wave 2 data (collected during 2009-2010) consisted of 2,464 participants (1,575 households), out of which 2,114 had participated at Wave 1 and 350 were refreshers introduced for reasons of representation. The participants were interviewed in Hebrew, Russian, or Arabic. The data were collected by a comprehensive computer-assisted personal interview which lasted about 90 min, followed by a supplementary Drop-Off paper questionnaire, which was returned later. Informed consent had been obtained from all participants prior to the interview. SHARE-Israel received ethical approval by the Institutional Review Board of the Hebrew University of Jerusalem.

Some of the items that eventually composed the focal HWS measure in this study were included in the Drop-Off questionnaire at Wave 1. Therefore, out of 1,725 Drop-Off respondents (66% of the total sample at Wave 1), the present analyses involved 1,286 participants who took part in both waves. Attrition analyses presented elsewhere (Shrira, 2014) indicated that respondents, compared with nonrespondents, of the Drop-Off showed lower disability and depressive symptoms as well as higher cognitive functioning. These differences had small effect sizes (Cohen’s d < 0.20). Among Drop-Off respondents, similar differences were found between those who participated at Wave 2 and those who did not, again yielding small effect sizes (Cohen’s d < 0.27).

The present participants had a mean age of 62.67 (SD = 9.78, range = 38-94) and included 57% women. Three origin groups were indicated by the interview’s language, which was Hebrew for 78.1% (mostly veteran Jews), Russian for 2.6% (former USSR immigrants), and Arabic for 19.4% (Israeli Arabs). The mean level of education (coded on an ordinal 7-step scale) was 3.97 (roughly indicating an average of upper secondary schooling; SD = 1.69), and the mean annual household income was 23,879 (adjusted from local currency to Euro; SD = 28,284). Most participants (81.1%) were married (11.7% widowed, 4.4% divorced, 1.8% never married). Further details on SHARE-Israel are presented in various publications (e.g., Litwin, 2009; Shmotkin & Litwin, 2009; Shrira et al., 2012). The complete SHARE documentation is available online (http://www.share-project.org/).

Measurement of the HWS

The current HWS measure was specifically derived from the SHARE-Israel data set (for other HWS operationalizations, see Shenkman & Shmotkin, 2013; Shrira, 2015; Shrira, Palgi, Ben-Ezra, & Shmotkin, 2011). It included 14 items that corresponded to the conceptualization of HWS, namely, representations of perceived threats to one’s physical or mental integrity, whether actual (including threats appearing to linger at present) or potential (see the introduction above). These items (listed in Table 1) were scattered in various sections of the SHARE booklet of questionnaires at Wave 1. Selection of items into the HWS measure was made by consensus of three researchers acquainted with the HWS concept. The original rating scales differed among HWS items (see footnotes in Table 1 for specification); therefore, responses to each item were converted into standardized scores. For each participant, the personal mean of these scores served as a general HWS score. Notably, the measures to which the 14 items originally belonged were not included in the present predictive models.

Table 1.

Factor Analysis of HWS Items (Constrained to Three Factors).

Items	1	2	3
Items	Loss of control	Social vulnerability	Family conflict
1. My age prevents me from doing the things I would like to^a	.730	.160	−.004
2. I feel that what happens to me is out of my control^a	.724	.127	.176
3. Would you say your health is poor, fair, good, very good, or excellent?^b,c	.666	−.095	−.025
4. I feel left out of things^a	.617	.305	.160
5. What are the chances that you will live to be T or more?^c,d,e	.511	.087	−.158
6. I hardly ever expect things to go my way^a	.251	.160	−.150
7. I felt people were unfriendly over the last week^a	.104	.680	−.204
8. I felt that people disliked me over the last week^a	.073	.656	−.303
9. Family responsibilities prevent me from doing what I want to do^a	.129	.579	.261
10. Shortage of money stops me from doing the things I want to do^a	.233	.554	.043
11. I have been seriously disappointed or hurt by someone to whom I gave my trust^a	.049	.410	.113
12. What are the chances that 5 years from now your standard of living will be worse than today?^d	.020	.298	.053
13. How often, if at all, do you experience conflict with your children?^f	−.035	.094	.810
14. How often, if at all, do you experience conflict with your partner/spouse?^f	.017	.033	.744
Eigen value	2.92	1.53	1.35
Variance explained (after rotation)	16.41	14.13	10.95

Note. Presented are factor loadings obtained in a principal components analysis using varimax rotation. The items’ standardized scores were used for the analysis. After a listwise deletion of missing values, the number of cases in the analysis was 575. HWS = hostile-world scenario.

Response originally given on a 3-point scale.

Response originally given on a 5-point scale.

Reverse-coded.

Response originally given as a probability estimation from 0 to 100.

The T (target age) in this item was changed for respondents in different ages, so that T was usually set as an age of 10 to 14 years ahead of the respondent’s actual age.

Response originally given on a 4-point scale.

An exploratory factor analysis (a principal components extraction with a varimax rotation) of the 14 HWS items provided internal validity information about the thematic composition of the current HWS measure (see Table 1). A solution constrained to three factors appeared most feasible according to the following criteria: (a) The factors’ eigenvalues were greater than 1, (b) items within each factor had loadings of at least 0.30 with a difference of at least 0.10 from their loadings on other factors (Item 6 was a mild exception), and (c) each factor included at least two items as specified above. Six items, labeled loss of control, were loaded on Factor 1. These items were related to being restricted by powerful life conditions (e.g., being old, loss of health) or by otherwise perceived lack of control. Six other items, labeled social vulnerability, were loaded on Factor 2. These items were related to being socially unwanted or vulnerable, and also to feeling constrained in the social sphere as by economic deprivation. Finally, two items, labeled family conflict, were loaded on Factor 3. These items were related to conflicts with one’s significant others (children or spouse).

Three of the current HWS items (numbered 5, 11, and 12 in Table 1) had an exceedingly lower response rate (65.6%, 84.1%, and 80.0%, respectively) compared with the usual response rate (range = 91.4%-99.2%) in the other HWS items. We did not reject these particular items because they appeared pertinent to the HWS notion (e.g., “I have been seriously disappointed or hurt by someone to whom I gave my trust”). Yet, to minimize the bias due to missing values, the general HWS score was calculated for participants who completed at least nine out of the 14 HWS items (at least 65% completion). Cronbach’s alpha coefficient was .66 for the 14 items (using listwise deletion, n = 575). Alpha coefficient remained nearly the same (.67) when we applied single imputations based on linear regression analysis.

Regarding the stability of the HWS items over time, information was available for seven items, which were re-administered at Wave 2 (4 years later). Pearson correlations between Wave 1 and Wave 2 administrations of these items were moderate to low. Items 10 and 11 had correlations of .59 and .54, respectively. Items 3, 4, and 5 had correlations of .30, .27, and .25, respectively. Finally, Items 13 and 2 had correlations of .21 and .18, respectively. All correlations were significant below the .001 level.

Outcome Measures

Difficulties in Activities of Daily Living (ADL) were assessed by six basic activities (e.g., eating, dressing, bathing) that participants reported to have difficulty in their performance, each checked as yes or no. These activities were adapted from the ADL scale of Katz, Downs, Cash, and Grotz (1970). The participant’s score was the number of difficulties checked. The correlation of ADL difficulties at Wave 1 with the counterpart score at Wave 2 was .56 (p < .001). Cronbach’s alpha coefficient (calculated as Kuder–Richardson 20 for dichotomous items) at Wave 2 was .91.

Difficulties in Instrumental Activities of Daily Living (IADL) were assessed by seven everyday activities (e.g., preparing a hot meal, shopping in a store, managing money) that participants reported to have difficulty in their performance, each checked as yes or no. These activities were adapted from the IADL scale of Lawton and Brody (1969). The participant’s score was the number of difficulties checked. The correlation of IADL difficulties at Wave 1 with the counterpart score at Wave 2 was .33 (p < .001). Alpha coefficient at Wave 2 was .86.

Movement difficulties were assessed by a list of 10 activities requiring some degree of physical movement and effort (e.g., reaching or extending one’s arms above shoulder level, lifting or carrying weights over 10 pounds/5 kilos such as a heavy bag of groceries). For each item, participants were asked to report whether they had difficulty in its performance, checked as yes or no. The participant’s score was the number of difficulties checked. Five items were adapted from Nagi’s (1976) physical performance scale. The correlation of movement difficulties at Wave 1 with the counterpart score at Wave 2 was .57 (p < .001). Alpha coefficient at Wave 2 was .84.

Physical symptoms were assessed by summing 11 listed physical symptoms that were reported to have been experienced by participants in the past 6 months (e.g., persistent cough, swollen legs). The participant’s score was the number of symptoms checked. The correlation of physical symptoms at Wave 1 with the counterpart score at Wave 2 was .48 (p < .001). Alpha coefficient at Wave 2 was .73.

Medical conditions were assessed by summing 16 listed illnesses that participants reported to have been suffering from (e.g., diabetes, osteoporosis). The participant’s score was the number of conditions checked. The correlation of medical conditions at Wave 1 with the counterpart score at Wave 2 was .52 (p < .001).

Depressive symptoms were assessed by the European Depression Scale (Euro-D; Prince et al., 1999). This measure contains 12 items that specify recent depressive symptoms (e.g., “In the last month, have you felt that you would rather be dead?”), each checked as yes or no. In the present study, completion of at least 10 items was required for scoring. The participant’s score was the number of symptoms checked. The correlation of EURO-D at Wave 1 with the counterpart score at Wave 2 was .49 (p < .001). Alpha coefficient at Wave 2 was .80.

Satisfaction with life was assessed by a single item asking “How satisfied are you with your life in general?” At Wave 1, response was given on a 4-point scale ranging from 1 (very dissatisfied) to 4 (very satisfied); at Wave 2, a different scale was used, ranging from 0 (completely dissatisfied) to 10 (completely satisfied). The correlation of satisfaction with life at Wave 1 with the counterpart score at Wave 2 was .33 (p < .001).

Cognitive functioning was assessed by a performance test on the following five domains: (a) arithmetic ability, measured by four questions (e.g., “If the chance of catching a disease is 10%, how many people out of 1,000 are expected to catch the disease?”); (b) time-orientation, measured by four questions regarding current year, month, day of the month, and day of the week; (c) verbal learning, measured by the number of words immediately recalled from a 10-word list; (d) verbal recall, measured by the number of words recalled 5 min after immediate recall; and (e) word fluency, measured by the number of correct animal names produced within 1 min. The participant’s composite score was the sum of the standardized performance scores on all domains, with a higher score representing better cognitive functioning (for illustrating the use of this SHARE-based measure, see Ayalon & Litwin, 2009). The correlation of cognitive functioning at Wave 1 with the counterpart score at Wave 2 was .50 (p < .001). Alpha coefficient at Wave 2 was .67.

Social activities were assessed by summing eight listed activities (e.g., doing voluntary or charity work, attending sports, participating in any social club) that participants reported to have been taking part in during the past month. The correlation of social activities at Wave 1 with the counterpart score at Wave 2 was .38 (p < .001).

Sociodemographic characteristics were used from data collected at Wave 1. These included age, gender, origin group (veteran Jews, former USSR immigrants coming to Israel since 1989, and Israeli Arabs), education (coded by one of seven levels ranging from pre-primary to second-stage tertiary education, as classified by the United Nations Educational, Scientific and Cultural Organization, 1997), annual household income (adjusted to Euro), and marital status.

Data Analysis

We first handled descriptive and correlative statistics. To test our first hypothesis that the HWS at Wave 1 would have a unique contribution to predicting change in physical and mental health outcomes at Wave 2, a series of hierarchical multiple regression analyses examined the HWS prediction of each of the dependent variables. In each of these analyses, Step 1 included the six sociodemographic characteristics specified above (two dummies for origin group respectively indicated former USSR immigrants and Israeli Arabs vs. veteran Jews; marital status was collapsed into currently unmarried vs. married). Step 2 additionally included the outcome variable’s baseline at Wave 1, and Step 3 additionally included the HWS score. To test our second hypothesis that the HWS would interact with age so that its effect on subsequent change in the outcome variables would be stronger for older people, we conducted another series of hierarchical regression analyses. In these, Step 1 included the aforementioned sociodemographics (excluding age), Step 2 additionally included the outcome variable’s baseline at Wave 1, Step 3 additionally included both age and the HWS score, and Step 4 additionally included the interaction term of Age × HWS. In all regressions, standard scores were used for all continuous (i.e., nondichotomous) variables.

Results

Construct Validity of the HWS Measure

The bivariate relationships of the SHARE-based measure of HWS at Wave 1 indicated no correlation of HWS with age (r = .01, ns) and similar levels of HWS among men (M = 0.02, SD = 0.94) and women (M = −0.15, SD = 1.05). HWS was negatively correlated with education (r = −.28) and annual household income (r = −.16; for both, p < .001). Married participants reported higher HWS (M = −0.00, SD = 0.42) than nonmarried ones (M = −0.11, SD = 0.48): t(1259) = 2.97, p < .01. An analysis of variance among the three origin groups was significant: F(2, 1258) = 54.30, p < .001. Scheffé post hoc tests indicated that all groups differed from each other at the .05 level, with Israeli Arabs reporting the highest HWS (M = 0.20, SD = 0.34), former USSR immigrants reporting intermediate HWS (M = 0.00, SD = 0.44), and veteran Jews reporting the lowest HWS (M = −0.13, SD = 0.43). When all the above sociodemographics were entered into a multiple regression analysis with HWS as the dependent variable, HWS showed no significant associations with age and gender, but retained unique associations with lower education and income, being married, and belonging either to the Israeli Arab or the former USSR immigrant group as compared with veteran Jews (R² = .12, p < .001).

Bivariate correlation analysis revealed that the HWS was significantly associated with lower levels of physical and mental health as indicated by all current markers at both Wave 1 and Wave 2 (for all correlations, p < .001; see Table 2). In absolute values, the correlations of the HWS ranged from .13 to .42 with Wave 1 markers and from .16 to .33 with Wave 2 markers.

Table 2.

Correlations Among the Study Variables at Wave 1 and Wave 2.

Variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
Wave 1
1. HWS	—
2. ADL difficulties	.29***	—
3. IADL difficulties	.16***	.50***	—
4. Movement difficulties	.32***	.68***	.14***	—
5. Physical Symptoms	.31***	.45***	.26***	.55***	—
6. Medical conditions	.16***	.33***	.19***	.41***	.54***	—
7. Depressive symptoms	.42***	.41***	.18***	.44***	.52***	.29***	—
8. Satisfaction with life	−.41***	−.22***	−.11***	−.23***	−.23***	−.16***	−.38***	—
9. Cognitive functioning	−.25***	−.41***	−.21***	−.36***	−.32***	−.20***	−.35***	.22***	—
10. Social activities	−.25***	−.16***	−.10**	−.15***	−.06*	−.05	−.18***	.23***	.25***	—
Wave 2
11. ADL difficulties	.22***	.55***	.38***	.43***	.29***	.20***	.27***	−.14***	−.30***	−.13***	—
12. IADL difficulties	.30***	.52***	.28***	.43***	.34***	.27***	.33***	−.18***	−.41***	−.18***	.62***	—
13. Movement difficulties	.29***	.54***	.25***	.54***	.40***	.32***	.36***	−.17***	−.40***	−.20***	.61***	.71***	—
14. Physical Symptoms	.30***	.43***	.20***	.46***	.47***	.37***	.38***	−.19***	−.34***	−.15***	.44***	.56***	.69***	—
15. Medical conditions	.30***	.34***	.19***	.40***	.41***	.57***	.28***	−.16***	−.28***	−.12***	.36***	.46***	.54***	.62***	—
16. Depressive symptoms	.37***	.42***	.13***	.41***	.42***	.31***	.50***	−.28***	−.42***	−.22***	.39***	.53***	.58***	.61***	.46***	—
17. Satisfaction with life	−.25***	−.33***	−.28***	−.33***	−.34***	−.30***	−.37***	.32***	.27***	.19***	−.33***	−.36***	−.35***	−.35***	−.33***	−.50***	—
18. Cognitive functioning	−.15***	−.38***	−.21***	−.35***	−.33***	−.24***	−.30***	.16***	.61***	.20***	−.33***	−.46***	−.43***	−.41***	−.32***	−.44***	.36**	—
19. Social activities	−.19***	−.14***	−.07*	−.15***	−.07*	−.07*	−.18***	.12***	.21***	.35***	−.14***	−.22***	−.17***	−.10**	−.09*	−.15***	.14**	.19***	—
M	−0.00	0.56	0.08	0.78	1.54	1.66	2.59	3.07	0.02	1.01	0.33	0.93	1.15	2.00	2.16	2.48	7.19	0.00	0.74
SD	0.45	1.36	0.55	1.24	1.78	1.57	2.40	0.71	0.77	1.14	1.12	1.65	1.72	2.11	1.74	2.55	2.93	3.50	0.93

Note. The intercorrelation matrix is based on 1,071 respondents after a listwise deletion of missing values. HWS = hostile-world scenario; ADL = activities of daily living; IADL = instrumental activities of daily living.

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

Testing Hypothesis 1: Prediction of Change in Outcomes at Wave 2 by the HWS at Wave 1

The HWS at Wave 1 showed a significantly unique contribution to predicting subsequent change in all the physical and mental health outcomes at Wave 2, with the exception of cognitive functioning. Illustrating the analyses of physical health outcomes, results of Step 1 in the hierarchical multiple regression predicting medical conditions at Wave 2 (n = 1,243) revealed that the variance explained by sociodemographic variables at Wave 1 reached 15.5% (p < .001). The introduction of Wave 1 baseline of medical conditions in Step 2 provided additional 20.2% of explained variance (p < .001). With this baseline adjusted for, predictors of greater change (increase) in medical conditions were higher age, belonging either to the former USSR immigrant or the Israeli Arab group as compared with veteran Jews, lower education, and lower income. In a third step, Wave 1 HWS was introduced into the regression model and additionally predicted an increase in medical conditions at Wave 2 beyond the predictions by the sociodemographics and baseline level: β = .18, p < .001. The variance explained in the associated change of the outcome was 2.6% (p < .001).

Illustrating the analyses of mental health outcomes, results of Step 1 in the hierarchical multiple regression predicting depressive symptoms at Wave 2 (n = 1,162) revealed that the variance explained by sociodemographic variables at Wave 1 reached 15.1% (p < .001). The introduction of Wave 1 baseline of depressive symptoms in Step 2 provided additional 16.6% of explained variance (p < .001). With this baseline adjusted for, predictors of greater change (increase) in depressive symptoms were higher age, belonging either to the former USSR immigrant or the Israeli Arab group as compared with veteran Jews, lower education, lower income, and currently being unmarried. In a third step, Wave 1 HWS was introduced into the regression model and additionally predicted an increase in depressive symptoms at Wave 2 beyond the predictions by the sociodemographics and baseline level: β = .13, p < .001. The variance explained in the associated change of the outcome was 1.2% (p < .001).

Table 3 summarizes the predictions of the physical and mental health outcomes at Wave 2. As shown, the HWS produced significant predictions, except for cognitive functioning, after adjustment for the predictive effects by the sociodemographics and baseline level of the respective outcome in each particular analysis. In sum, higher HWS predicted an increase, from Wave 1 to Wave 2, in ADL and IADL difficulties, movement difficulties, physical symptoms, medical conditions, and depressive symptoms. In parallel, higher HWS predicted a decrease in satisfaction with life and social activities. The beta coefficients ranged (in absolute values) from .06 to .19. The effect sizes appeared small (the respective variances explained in the associated changes ranged from 0.3% to 3.1%), but were consistent with the study’s first hypothesis. The results maintained their consistent pattern also after considering Bonferroni’s correction for the desired significance threshold in this case of nine equivalent predictions (0.05:9 = 0.0055), with six predictions reaching significance (p < .001) that exceeded the corrected threshold.

Table 3.

Summary of Results Obtained in a Series of Regression Analysis for Predicting Physical and Mental Health Outcomes at Wave 2 by the HWS at Wave 1.

Predicted outcome	β	R² (ΔR²)
Physical health
ADL difficulties	.061*	.337*** (.003*)
IADL difficulties	.190***	.315* (.031*)
Movement difficulties	.091***	.399* (.007*)
Physical symptoms	.127***	.334* (.013*)
Medical conditions	.175***	.383* (.026*)
Mental health
Depressive symptoms	.129***	.327* (.012*)
Satisfaction with life	−.161***	.237* (.019*)
Cognitive functioning	−.031	.422*** (.001)
Social activities	−.085**	.160 (.006)

Note. The table reports separate multiple regression analyses with ns ranging from 1,126 to 1,251. In each analysis, Step 1 included age, gender, origin groups, education, income, and marital status. Step 2 additionally included the outcome variable’s baseline level, and Step 3 additionally included the HWS. Reported are standardized regression coefficients (βs) predicting the respective outcomes by the HWS as well as R² and ΔR² obtained in the last step. HWS = hostile-world scenario; ADL = activities of daily living; IADL = instrumental activities of daily living.

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

Although the inclusion of each outcome’s baseline in the respective predictive models controlled for the shared variance of that baseline with the HWS, it was still possible that distressing conditions associated with the outcome variables had initially induced or aggravated the HWS. Examining this possible confounding, we conducted supplementary regression analyses among restricted subsamples that were not implicated by any clinical manifestation at the baselines of medical conditions and depressive symptoms. Thus, in a regression analysis for 330 participants who reported at Wave 1 no medical condition, the HWS at Wave 1 was still significant in predicting an increase in medical conditions at Wave 2: β = .17, p < .01, with additionally explained variance of 2.4% (p < .01). This analysis adjusted for sociodemographics but did not include a baseline for medical conditions as this variable was invariant (scored 0) at Wave 1 for all participants of this subsample. In another regression analysis for 838 participants who reported at Wave 1 a nonclinical level of depressive symptoms (less than the threshold of four symptoms; Prince et al., 1999), the HWS at Wave 1 was still significant in predicting an increase in depressive symptoms at Wave 2: β = .15, p < .001, with additionally explained variance of 1.7% (p < .001). This analysis adjusted for sociodemographics as well as for the subsample’s baseline of depressive symptoms at Wave 1. Furthermore, in a regression analysis for 266 participants who reported at Wave 1 no medical condition and a nonclinical level of depressive symptoms (adjusting for sociodemographics and Wave 1 baseline of depressive symptoms), the HWS was again significant in predicting an increase in both medical conditions (β = .19, p < .01, with additionally explained variance of 3.0%, p < .01) and depressive symptoms (β = .14, p < .05, with additionally explained variance of 1.5%, p < .05) at Wave 2.

Testing Hypothesis 2: The Interaction of Age With the HWS in Predicting Change in Outcomes

In testing our second hypothesis, age (of participants aged 50 and above) was treated as a continuous variable in a series of hierarchical multiple regression, or otherwise collapsed into a younger and older groups for a comparative purpose. The interaction term of Age × HWS was introduced into Step 4 of the regression analyses after adjustment for sociodemographics and the respective baselines (see “Data Analysis” section). Interactions were probed using Hayes’s (2013) Process computational macro. These analyses yielded a significant interaction effect for four, out of the nine, outcomes. Thus, in predicting ADL difficulties at Time 2, an Age × HWS interaction indicated a B coefficient of 0.08 (p < .001), explaining an additional 0.7% of the variance. When age was 1 SD below the mean (53.1), the association between HWS and ADL was nonsignificant (B = −0.07, p = .38); when age was 1 SD above the mean (72.3), the association between HWS and ADL was significant (B = 0.35, p < .00001). Illustrating this finding, Figure 1 presents the regression lines for younger (age = 50-64) versus older (age 65 and above) participants. As shown, the slope of the predictive association between HWS at Wave 1 and ADL difficulties at Wave 2 was steeper in the older, versus the younger, participants.

Figure 1.

Prediction of ADL difficulties at Wave 2 by the HWS at Wave 1 in younger (age = 50-64) and older (age = 65 and older) participants (standardized scores).

Similarly, in predicting IADL difficulties at Time 2, an Age × HWS interaction indicated a B coefficient of 0.10 (p < .001), explaining an additional 1.1% of the variance. Thus, when age was 1 SD below the mean, the association between HWS and IADL was nonsignificant (B = 0.24, p = .07); when age was 1 SD above the mean, the association between HWS and IADL was significant (B = 1.07, p < .00001). Moreover, in predicting depressive symptoms at Time 2, an Age × HWS interaction indicated a B coefficient of 0.09 (p < .001), explaining an additional 0.9% of the variance. Thus, when age was 1 SD below the mean, the association between HWS and depressive symptoms was nonsignificant (B = 0.17, p = .44); when age was 1 SD above the mean, the association between HWS and depressive symptoms was significant (B = 1.22, p < .00001). Finally, in predicting movement difficulties at Time 2, an Age × HWS interaction indicated a B coefficient of 0.05 (p < .05), explaining an additional 0.2% of the variance. When age was 1 SD below the mean, the association between HWS and movement difficulties was nonsignificant (B = 0.15, p = .23); when age was 1 SD above the mean, the association between HWS and movement difficulties was significant (B = 1.22, p < .00001). As summarized in Table 4, the aforementioned significant interactions highlight a fairly consistent pattern whereby most of the outcomes at Wave 2 (satisfaction with life deviates from this pattern) tend to be better predicted among the older, than the younger, participants.

Table 4.

Standardized Coefficients and Explained Variances for Predicting Physical and Mental Health Outcomes at Wave 2 by the HWS at Wave 1 in Two Age Groups.

Predicted outcome	β		ΔR²
Predicted outcome	Younger	Older	Younger	Older
ADL difficulties	.052	.072	.002	.004
IADL difficulties	.152***	.245***	.021***	.048***
Movement difficulties	.092*	.098*	.007**	.007*
Physical symptoms	.108**	.162***	.010**	.019***
Medical conditions	.168***	.191***	.024***	.030***
Depressive symptoms	.057	.216***	.002	.032***
Life satisfaction	−.188***	−.139*	.026***	.013*
Cognitive functioning	−.006	.038	.000	.001
Social activities	−.079*	−.100*	.005*	.008*

Note. The table reports separate multiple regression analyses with ns ranging 710 to 764 for the younger (age = 50-64) and 415 to 491 for the older (age 65 and above) group. In each analysis, Step 1 included age, gender, origin groups, education, income, and marital status. Step 2 additionally included the outcome variable’s baseline level and Step 3 additionally included the HWS. Reported are standardized regression coefficients (βs) predicting the respective outcomes by HWS as well as the ΔR² obtained in the last step. HWS = hostile-world scenario; ADL = activities of daily living; IADL = instrumental activities of daily living.

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

Discussion

The present study investigated predictions of change in physical and mental health markers in later life, using Shmotkin’s (2005) concept of the HWS. The HWS is an image of perceived threats that individuals maintain in regard to their own physical and mental integrity. The findings indicated that the HWS added a significant contribution to the prediction of late life functioning along a period of 4 years. In addition, the contribution of the HWS to these predictions was more pronounced for older participants. Next, we discuss the findings in view of the prior hypotheses and the HWS concept at large.

Analyzing data from SHARE-Israel, we relied on a multidimensional, longitudinal survey that assessed a nationwide, originally representative, sample of households in the population aged 50 and above. For measuring the HWS, we used selected items that represented part of the themes entailed in the concept of the HWS (e.g., loss of control due to aging-related dysfunction or to other life circumstances, social vulnerability due to interpersonal problems or economic hardship, family conflict). We hypothesized that the HWS, as measured at Wave 1 of the data collection, would serve as a predictor of functioning in vital physical and mental domains as measured at Wave 2. Underlying this hypothesis was our conceptual view that the HWS does not merely reflect current adversity, but presents a system monitoring for prospective adversity. Accordingly, the HWS was hypothesized to exert its predictive effect beyond covariation with sociodemographic variables and with concurrent (baseline) levels of functioning. Furthermore, the predictive effect of the HWS was hypothesized to be of greater magnitude among older participants because old age increases the likelihood of aging-related adversities such as health declines and loss of resources.

The results supported our first hypothesis. Notably, the HWS could predict change occurring along 4 years in an array of vital functioning outcomes. Although small in terms of additionally explained variance, the predictive effects were relatively more pronounced (in a descending order) for IADL, medical conditions and satisfaction with life, and relatively more modest for physical symptoms, depressive symptoms, movement difficulties, social activities, and ADL. In this list of outcomes, the HWS proved to be similarly relevant to functioning markers of physical and mental health. As an exception, the HWS could not significantly predict change in cognitive functioning.

As shown in the bivariate analyses, the HWS correlated with the participants’ social positions (exhibited by sociodemographic characteristics) as well as with all the focal functioning markers at both waves. However, the data support the notion that the predictive power of the HWS is unique, and not confounded by its correlations with the outcome’s variables. Thus, the study’s analytic approach was meant to control for confounded predictions by a combination of the following: (a) examining how the HWS related to subsequent outcomes, measured at a later time; (b) adjusting this over-time relations for the outcome’s baseline, so that the confounding variance of the outcome within the HWS variance, when both were concurrently measured at Wave 1, was partialed out; and (c) examining the consistency of the HWS’s predictive effect across diverse outcomes. Moreover, we conducted supplementary analyses of outcomes representing physical health (medical conditions) and mental health (depressive symptoms) whereby the predictive power of the HWS was tested only in restricted subsamples that were not implicated by any clinically distressing manifestation in the baselines at Wave 1 (participants reporting either no medical condition or nonclinical level of depressive symptoms as well as both of these reports at the time). In these subsamples, despite the big reduction in sample size, the HWS remained a significant predictor of change in medical conditions and depressive symptoms over time. These results suggest that the ability of the HWS to predict adversity (as a prospective decline in vital functioning) does not necessarily depend on distressing conditions at the prediction time.

Altogether, the present findings appear to support the conception of the HWS in Shmotkin’s (2005) model, positing that the HWS is an adaptational system that alerts the individual to imminent threats to one’s integrity. As such, the HWS maintains an interface with most functioning domains. The emphasis of the HWS notion on a subjective image (expressed as self-beliefs) of currently relevant sensations and anticipations (actual and potential threats) expands prior research on life adversities as completed events (realized HWS, in our term), which dwell on the individual’s memories of her or his past (Shrira, 2012, 2014). Further study should explore the interdependence between the HWS and the realized HWS, thus presenting a fuller time perspective on how individuals, mainly in old age, conceive hostile-world contingencies occurring along their timeline.

The present findings raise the possibility that the HWS contains an element of foresight regarding outcomes that involve prospective changes in functioning, notably declines. Yet, we mentioned earlier that the original concept of the HWS also includes a positive engagement aspect (not examined here) which may also involve foresight of preservation, or even improvement, in functioning. The present data, then, suggest that the HWS fulfills a prevention-focused function aimed at inducing due caution and avoiding potential losses (Higgins, 1998). In fact, the literature has already suggested a heuristic phenomenon of foresight, similar to the one we discuss in relation to the HWS. Thus, the mere self-rating of one’s overall health proved to predict crucial outcomes (such as one’s mortality) even better than factual predictors (Idler & Benyamini, 1997). Moreover, self-perceptions of aging as well as perceived control predicted functional health over time (Levy, Slade, & Kasl, 2002). Notably, these three themes, namely, self-beliefs about one’s own health, aging, and control, are inherent in certain themes of the HWS measure. In line with previous discussions on the predictive value of self-rated health (Benyamini, 2011; Jylhä, 2009), it seems that the self-beliefs composing the HWS are sensitive to contextual and internal information that may not be otherwise accessible by more concrete (e.g., sociodemographic) predictors of people’s functioning. For example, in assessing threats to their life and integrity, people may consider their future resilience through their own perspectives of time as well as in relation to self-perceived options of ameliorating, or compensating for, their vulnerability (Shmotkin, 2011; Shmotkin & Shrira, 2013).

While the idea conveyed so far is that the HWS predicts functioning outcomes through a presumed mechanism of self-forecast, other alternative explanations are available. Thus, by imaging major threats to life, the HWS involves stress-related thinking that may amplify, or prolong, deleterious effects on one’s health through somatic channels such as the endocrine and immune systems (O’Connor, Walker, Hendrickx, Talbot, & Schaefer, 2013), especially when the HWS activates cognitions of ruminative and perseverative nature (Brosschot, Gerin, & Thayer, 2006). Still another explanation to the findings is that the HWS may be an agent for self-fulfilling prophecy. This path may occur when the HWS sensitizes disastrous views of one’s life, possibly by cognitive styles such as looming (Riskind, Williams, & Joiner, 2006) or bracing for the worst (Carroll, Sweeny, & Shepperd, 2006). At such overly activated level, the HWS may become a self-defeating mechanism (Baumeister & Scher, 1988) that hampers perceptive and motivational underpinnings of one’s physical and mental functioning. When these alternative paths operate, they may go counter to the expectation that highly activated HWS would eventually be regulated, or reconstructed, by complementary positive systems such as subjective well-being and meaning in life (Shmotkin & Shrira, 2012b).

In support of our second hypothesis, we found age to be a moderating factor of the predictive pattern of the HWS. Specifically, the HWS was found to be a stronger predictor among older, compared with younger, participants with regard to prospective change in ADL, IADL, depressive symptoms, and movement difficulties. This pattern was consistent, although not to a significant degree, also in regard to several other outcomes. This finding corroborates Shrira, Palgi, Ben-Ezra, Spalter, et al.’s (2011) analyses of SHARE’s participants (age 50+) from a dozen countries. They found that for older, compared with younger, participants, the expectation to worsen in their standards of living was more strongly associated with functioning markers (depressive symptoms, physical symptoms, and cognitive functioning), whereas the counterpart expectation to improve was less strongly associated with those markers. In this vein of negative outlook on the future among aging people, one may also note the advantage of pessimism over optimism in predicting prospective mental states (Robinson-Whelen, Kim, MacCallum, & Kiecolt-Glaser, 1997) as well as physical functioning (Brenes, Rapp, Rejeski, & Miller, 2002). These age-bound effects of negatively oriented dispositions probably express the increasing tendency of aging people to manage threats of declines and losses (Baltes, 1997). This tendency means that older people shift from promotion orientation, emphasizing the pursuit of positive outcomes such as growth and fulfillment, into prevention orientation, emphasizing the effort to minimize negative outcomes such as frailty and disability (Lockwood, Chasteen, & Wong, 2005). Overall, the evidence suggests that the current concept of the HWS is advantageous in depicting the manner by which older people manage their self-beliefs about impending adversities and in relating these self-beliefs to levels and changes of their functioning over time. The HWS offers a broader framework than discrete self-beliefs (e.g., self-rated health) as it touches upon themes that spread over an array of vital domains and combine present and future issues.

The present study should be considered in light of its strengths and limitations. This study was conducted in a longitudinal framework, thus shedding more light on the cross-time role of the HWS as an adaptational system. We showed that the HWS was consistently associated with concurrent functioning markers as well as with their change over time. The consistency of the findings on the association between the HWS and functioning across diverse markers is particularly instructive when it appears in an established data set of a multidimensional, nationwide survey.

A major limitation in this study concerns the current operationalization of the HWS. We had to tailor a new measure comprised of items already available in the SHARE’s questionnaire booklet. We could not use larger questionnaires initially constructed for HWS assessment (e.g., Shrira, 2015; Shrira, Palgi, Ben-Ezra, & Shmotkin, 2011). The present data also lack important concomitants that possibly moderate the activity of the HWS such as personality predispositions and styles of coping with stress. Moreover, the predicted outcomes were based on self-report, which were only partly ameliorated by the reliance of the outcome’s measures on factual information.

Other limitations concern issues of longitudinal data and sampling. This study examined only two waves of data collection, conducted 4 years apart, so that the predictive power of the HWS was not ascertained for longer time spans. Although expected in longitudinal surveys of older people, one should also note the attrition between the two waves (443 participants from Wave 1 did not take part in Wave 2) and the rates of selective nonresponse (between 2.7% and 12.4% did not provide full data on particular outcome variables at either or both waves). As participation in such surveys requires a certain level of cognitive functioning, attrition and nonresponse due to cognitive impairment may restrict the sample’s variability (cognitive functioning was an exceptional outcome that was not predicted by the HWS). Moreover, the necessity to restrict the sample only to Drop-Off respondents further limits the results’ generalizability. One more limitation is typical to surveys that use sampling of households, where the concurrent participation of the two spouses introduces interdependent observations in part of the sample. The problems raised by apparent solutions for such databases (especially when the study’s aim is to conduct hypothesis testing at the person level, rather than to obtain population estimates) are beyond this study’s scope. Indeed, most studies in this case rely on the robustness of statistical tests that assume independent observations to withstand partial interdependence in the data.

Finally, caution should be exercised regarding the possible culture-bound specificity of the results. Thus, past studies have indicated that the likelihood of depression among the older population of Israel is nearly twice as high as we might expect from elderly cohorts in other countries (Shmotkin & Litwin, 2009). Moreover, the HWS itself may be expected to be elevated among Israeli elders, due to the particularly stressful experiences that this cohort endured (Litwin, 2009). Therefore, future studies should attempt to replicate our findings in other cohorts and cultural backgrounds.

In conclusion, the present findings support the theoretical concept of the HWS as an internal monitor of threats and adverse circumstances in the individual’s life. While the HWS proves to serve as a consistent concomitant of simultaneous functioning markers in physical and mental health domains, it may also capture some unique information pertinent to prospective changes of functioning over time. Such effects highlight the particular relevance of the HWS concept to adaptation in old age. Although the causal paths pertaining to the predictive power of the HWS are not yet clear, the present results are in line with the presumption of Shmotkin’s (2005) model that positive functioning and well-being should be understood in the larger context of hostile-world contingencies in life.

Footnotes

Acknowledgements

The authors are grateful to Howard Litwin for facilitating the study with the SHARE-Israel data. They thankfully acknowledge the assistance of Shira Goldberg, Aviad Orbach, and Roy Haziza.

Authors’ Note

Sharon Avidor, PhD, is currently affiliated to Ruppin Academic Center, Israel.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a research grant of the Israeli Ministry for Senior Citizens. The development and data collection of the SHARE project in Israel was supported by the U.S. National Institute on Aging (R03AG029258, R01AG031729, R21AG025169) as well as by the National Insurance Institute of Israel, German-Israeli Foundation for Scientific Research and Development (GIF), European Commission through the 7th framework program, Israeli Ministry of Science, and Israeli Ministry for Senior Citizens. The data were collected by the Israeli Gerontological Data Center at the Hebrew University of Jerusalem.

References

Ayalon

Litwin

(2009). What cognitive functions are associated with passive suicidal ideation? Findings from a national sample of community dwelling Israelis. International Journal of Geriatric Psychiatry, 24, 472-478. doi:10.1002/gps.2140

Baltes

P. B.

(1997). On the incomplete architecture of human ontogeny: Selection, optimization and compensation as foundation of developmental theory. American Psychologist, 52, 366-380. doi:10.1037/0003-066X.52.4.366

Baltes

P. B.

Smith

(1997). A systemic-wholistic view of psychological functioning in very old age: Introduction to a collection of articles from the Berlin Aging Study. Psychology and Aging, 12, 395-409. doi:10.1037/0882-7974.12.3.395

Baumeister

R. F.

Scher

S. J.

(1988). Self-defeating behavior patterns among normal individuals: Review and analysis of common self-destructive tendencies. Psychological Bulletin, 104, 3-22. doi:10.1037/0033-2909.104.1.3

Benyamini

(2011). Editorial: Why does self-rated health predict mortality? An update on current knowledge and a research agenda for psychologists. Psychology & Health, 26, 1407-1413. doi:10.1080/08870446.2011.621703

Brenes

G. A.

Rapp

S. R.

Rejeski

W. J.

Miller

M. E.

(2002). Do optimism and pessimism predict physical functioning? Journal of Behavioral Medicine, 25, 219-231. doi:10.1023/A:1015376510466

Brosschot

J. F.

Gerin

Thayer

J. F.

(2006). The perseverative cognition hypothesis: A review of worry, prolonged stress-related physiological activation, and health. Journal of Psychosomatic Research, 60, 113-124. doi:10.1016/j.jpsychores.2005.06.074

Busseri

M. A.

Sadava

S. W.

(2013). Subjective well-being as a dynamic and agentic system: Evidence from a longitudinal study. Journal of Happiness Studies, 14, 1085-1112. doi:10.1007/s10902-012-9368-9

Cappeliez

O’Rourke

(2006). Empirical validation of a model of reminiscence and health in later life. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 61, 237-244. doi:10.1093/geronb/61.4.P237

10.

Carroll

Sweeny

Shepperd

J. A.

(2006). Forsaking optimism. Review of General Psychology, 10, 56-73. doi:10.1037/1089-2680.10.1.56

11.

Clemson

Kendig

Mackenzie

Browning

(2015). Predictors of injurious falls and fear of falling differ: An 11-year longitudinal study of incident events in older people. Journal of Aging and Health, 27, 239-256. doi:10.1177/0898264314546716

12.

De Raedt

Koster

E. H.

Ryckewaert

. (2013). Aging and attentional bias for death related and general threat-related information: Less avoidance in older as compared with middle-aged adults. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 68, 41-48. doi:10.1093/geronb/gbs047

13.

Gerstorf

Smith

Baltes

P. B.

(2006). A systemic-wholistic approach to differential aging: Longitudinal findings from the Berlin Aging Study. Psychology and Aging, 21, 645-663. doi:10.1037/0882-7974.21.4.645

14.

Hayes

A. F.

(2013). An introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.

15.

Hicks

J. A.

King

L. A.

(2009). Meaning in life as a subjective judgment and a lived experience. Social & Personality Psychology Compass, 3, 638-653. doi:10.1111/j.1751-9004.2009.00193.x

16.

Higgins

E. T.

(1998). Promotion and prevention: Regulatory focus as a motivational principle. Advances in Experimental Social Psychology, 30, 1-46. doi:10.1016/S0065-2601(08)60381-0

17.

Hobfoll

S. E.

Wells

J. D.

(1998). Conservation of resources, stress, and aging: Why do some slide and some spring? In Lomranz

(Ed.), Handbook of aging and mental health: An integrative approach (pp. 121-134). New York, NY: Plenum.

18.

Idler

E. L.

Benyamini

(1997). Self-rated health and mortality: A review of 27 community studies. Journal of Health and Social Behavior, 38, 21-37. doi:10.2307/2955359

19.

Jylhä

(2009). What is self-rated health and why does it predict mortality? Towards a unified conceptual model. Social Science & Medicine, 69, 307-316. doi:10.1016/j.socscimed.2009.05.013

20.

Katz

S. C.

Downs

T. D.

Cash

H. R.

Grotz

R. C.

(1970). Progress in development of the index of ADL. The Gerontologist, 10, 20-30. doi:10.1093/geront/10.1_Part_1.20

21.

Keinan

Shrira

Shmotkin

(2012). The association between cumulative adversity and mental health: Considering dose and primary focus of adversity. Quality of Life Research, 21, 1149-1158. doi:10.1007/s11136-011-0035-0

22.

Labouvie-Vief

(2008). Dynamic integration theory: Emotion, cognition, and equilibrium in later life. In Bengtson

V. L.

Gans

Putney

N. M.

Silverstein

(Eds.), Handbook of theories of aging (2nd ed., pp. 277-293). New York, NY: Springer.

23.

Lang

F. R.

Weiss

Gerstorf

Wagner

G. G.

(2013). Forecasting life satisfaction across adulthood: Benefits of seeing a dark future? Psychology and Aging, 28, 249-261. doi:10.1037/a0030797

24.

Lawton

M. P.

Brody

E. M.

(1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. The Gerontologist, 9, 179-186. doi:10.1093/geront/9.3_Part_1.179

25.

Levy

B. R.

Slade

M. D.

Kasl

S. V.

(2002). Longitudinal benefit of positive self-perceptions of aging on functional health. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 57, 409-417. doi:10.1093/geronb/57.5.P409

26.

Litwin

(2009). Understanding aging in a Middle Eastern context: The SHARE-Israel survey of persons aged 50 and older. Journal of Cross-Cultural Gerontology, 24, 49-62. doi:10.1007/s10823-008-9073-7

27.

Lockwood

Chasteen

A. L.

Wong

(2005). Age and regulatory focus determine preferences for health-related role models. Psychology and Aging, 20, 3763-3789. doi:10.1037/0882-7974.20.3.376

28.

Lohr

J. B.

Palmer

B. W.

Eidt

C. A.

Aailaboyina

Mausbach

B. T.

Wolkowitz

O. M.

. . . Jeste

D. V.

(2015). Is post-traumatic stress disorder associated with premature senescence? A review of the literature. American Journal of Geriatric Psychiatry, 23, 709-725. doi:10.1016/j.jagp.2015.04.001

29.

Luo

Hawkley

L. C.

Waite

L. J.

Cacioppo

J. T.

(2012). Loneliness, health, and mortality in old age: A national longitudinal study. Social Science & Medicine, 74, 907-914. doi:10.1016/j.socscimed.2011.11.028

30.

Nagi

S. Z.

(1976). An epidemiology of disability among adults in the United States. Milbank Memorial Fund Quarterly, 54, 439-467. doi:10.2307/3349677

31.

O’Connor

D. B.

Walker

Hendrickx

Talbot

Schaefer

(2013). Stress-related thinking predicts the cortisol awakening response and somatic symptoms in healthy adults. Psychoneuroendocrinology, 38, 438-446. doi:10.1016/j.psyneuen.2012.07.004

32.

Palgi

Shmotkin

(2010). The predicament of time near the end of life: Time perspective trajectories of life satisfaction among the old-old. Aging & Mental Health, 14, 577-586. doi:10.1080/13607860903483086

33.

Prince

M. J.

Reischies

Beekman

A. T. F.

Fuhrer

Jonker

Kivela

S. L.

. . . Copeland

J. R.

(1999). Development of the Euro-D scale: A European Union initiative to compare symptoms of depression in 14 European centres. British Journal of Psychiatry, 174, 330-338. doi:10.1192/bjp.174.4.330

34.

Reed

A. E.

Carstensen

L. L.

(2012). The theory behind the age-related positivity effect. Frontiers in Psychology, 3, Article 339. doi:10.3389/fpsyg.2012.00339

35.

Riskind

J. H.

Williams

N. L.

Joiner

T. E.

Jr. (2006). The looming cognitive style: A cognitive vulnerability for anxiety disorders. Journal of Social & Clinical Psychology, 25, 779-801. doi:10.1521/jscp.2006.25.7.779

36.

Robinson-Whelen

Kim

MacCallum

R. C.

Kiecolt-Glaser

J. K.

(1997). Distinguishing optimism from pessimism in older adults: Is it more important to be optimistic or not to be pessimistic? Journal of Personality and Social Psychology, 73, 1345-1353. doi:10.1037/0022-3514.73.6.1345

37.

Ryan

R. M.

Deci

E. L.

(2001). On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annual Review of Psychology, 52, 141-166. doi:10.1146/annurev.psych.52.1.141

38.

Ryff

C. D.

Singer

(1998). The contours of positive human health. Psychological Inquiry, 9, 1-28. doi:10.1207/s15327965pli0901_1

39.

Shenkman

Shmotkin

(2013). The Hostile-World Scenario among Israeli homosexual adolescents and young adults. Journal of Applied Social Psychology, 43, 1408-1417. doi:10.1111/jasp.12097

40.

Shmotkin

(2005). Happiness in the face of adversity: Reformulating the dynamic and modular bases of subjective well-being. Review of General Psychology, 9, 291-325. doi:10.1037/1089-2680.9.4.291

41.

Shmotkin

(2011). The pursuit of happiness: Alternative conceptions of subjective well-being. In Poon

L. W.

Cohen-Mansfield

(Eds.), Understanding well-being in the oldest old (pp. 27-45). New York, NY: Cambridge University Press. doi:10.1017/CBO9780511920974.004

42.

Shmotkin

Litwin

(2009). Cumulative adversity and depressive symptoms among older adults in Israel: The differential roles of self-oriented versus other-oriented events of potential trauma. Social Psychiatry & Psychiatric Epidemiology, 44, 989-997. doi:10.1007/s00127-009-0020-x

43.

Shmotkin

Shrira

(2012a). Happiness and suffering in the life story: An inquiry into conflicting expectations concerning the association of perceived past with present subjective well-being in old age. Journal of Happiness Studies, 13, 389-409. doi:10.1007/s10902-011-9270-x

44.

Shmotkin

Shrira

(2012b). On the distinction between subjective well-being and meaning in life: Regulatory versus reconstructive functions in the face of a hostile world. In Wong

P. T. P.

(Ed.), The human quest for meaning: Theories, research, and applications (2nd ed., pp. 143-163). New York, NY: Routledge.

45.

Shmotkin

Shrira

(2013). Subjective well-being and meaning in life in a hostile world: Proposing a configurative perspective. In Hicks

J. A.

Routledge

(Eds.), The experience of meaning in life: Classical perspectives, emerging themes, and controversies (pp. 77-86). New York, NY: Springer.

46.

Shmotkin

Shrira

Eyal

Blumstein

Shorek

(2014). The prediction of subjective wellness among the old-old: Implications for the “fourth age” conception. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 69, 719-729. doi:10.1093/geronb/gbt052

47.

Shrira

(2012). The effect of lifetime cumulative adversity on change and chronicity in depressive symptoms and quality of life in older adults. International Psychogeriatrics, 24, 1988-1997. doi:10.1017/S1041610212001123

48.

Shrira

(2014). Greater age-related decline in markers of physical, mental and cognitive health among Israeli older adults exposed to lifetime cumulative adversity. Aging & Mental Health, 18, 610-618. doi:10.1080/13607863.2013.860951

49.

Shrira

(2015). Transmitting the sum of all fears: Iranian nuclear threat salience among offspring of Holocaust survivors. Psychological Trauma: Theory, Research, Practice, and Policy, 7, 364-371. doi:10.1037/tra0000029

50.

Shrira

Litwin

(2014). The effect of lifetime cumulative adversity and depressive symptoms on functional status. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 69, 953-965. doi:10.1093/geronb/gbu056

51.

Shrira

Palgi

Ben-Ezra

Shmotkin

(2011). How subjective well-being and meaning in life interact in the hostile world? The Journal of Positive Psychology, 6, 273-285. doi:10.1080/17439760.2011.577090

52.

Shrira

Palgi

Ben-Ezra

Spalter

Kavé

Shmotkin

(2011). For better and for worse: The relationship between future expectations and functioning in the second half of life. The Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 66, 195-203. doi:10.1093/geronb/gbq103

53.

Shrira

Shmotkin

Litwin

(2012). Potentially traumatic events at different points in the life span and mental health: Findings from SHARE-Israel. The American Journal of Orthopsychiatry, 82, 251-259. doi:10.1111/j.1939-0025.2012.01149.x

54.

United Nations Educational, Scientific and Cultural Organization. (1997). International standard classification of education 1997. Geneva, Switzerland: Author.

55.

Urry

H. L.

Gross

J. J.

(2010). Emotion regulation in older age. Current Directions in Psychological Science, 19, 352-357.