Abstract
Aims
To examine the association of midlife report of crisis following parental death (CFPD) during childhood and adolescence, with dementia at old age.
Methods
In 1965, 9362 male participants of the Israel Ischemic Heart disease (IIHD) study were asked whether they have experienced CFPD (paternal or maternal) during the following ages: 0–6. 7–12. 13–18 or >18 years. Dementia was assessed over three decades later in 1889 survivors of the original cohort, 1,652 of whom were assessed for CFPD in 1965.
Results
Controlling for age, the estimated odds for dementia relative to individuals who reported crisis following paternal parental death (CFPR-P) at the age of 18 and above, were 3.06 (95%CI 1.42–6.61), 2.15 (95% CI 0.87–5.31) and 2.35 (95%CI 1.05–5.28) for those who reported CFPD-P at the ages of 0–6, 7–12 and 13–18 respectively. Odds for dementia were 0.60 (95% CI 0.32–1.11) for participants who reported CFPD-P at ages of 18 and above, compared to participants who did not report such a crisis. Similar results were obtained for the association of crisis reported following maternal parental death (CFPD-M) at different age groups and dementia.
Conclusions
CFPD during childhood is associated with increased risk for dementia in males who survived until old age.
Keywords: crisis, parental death, childhood, dementia
Introduction
Increasing evidence demonstrates that exposure to certain factors across lifespan (including early childhood and intrauterine phases) is associated with increased risk for dementia 1, 2. Among these are psychological factors such as proneness to stress and depression 3, 4. The association of these factors with dementia may be affected by subjects' age at time of psychological assessment 4,' proximity to the stressogenic event or by the way psychological stress is measured, i.e. proneness to stress versus exposure to objective stress.
Death of a parent during childhood is a significant stressful event, associated with morbidity 5, 6, mortality 7, health related behaviors 8 and depression 9, 10 during adulthood, each associated with late life cognitive performance and with dementia 3, 11.
The importance of a life course approach has been recognized in other chronic diseases, enabling detection of periods of vulnerability to certain risk factors12. Since the impact of a certain risk factor 13 as well as recommended treatments may differ according to individual's age at exposure 14, preventive interventions considering a life course approach may help to identify optimal timing for intervention or modification
In this study, the relationship of age at crisis following childhood and adolescence parental loss with dementia at old age was assessed.
Methods
The IIHD project
The sample of the Israeli Ischemic Heart Disease (IIHD) Project was chosen by stratified sampling of civil servants and municipal employees aged 40 and above in 1963. Sampling procedures, mortality and follow-up have been described in detail elsewhere 15. Of 11,876 approached, 10,232 (86.2%) men agreed to participate and were enrolled in 1963. A more detailed description is available at the supplemental online version of ADAD.
Diagnosis of dementia
Dementia was diagnosed in 1999/2000 using a two-step procedure as described in detail elsewhere 15. A more detailed description is also available on line at the supplemental online version of ADAD.
According to the Israel Mortality Registry, 7136 men died by the beginning of the study in 1999; another 306 men died before being approached for a phone interview and 13 subjects were lost in the matching process. Detailed description about subjects assessed in 1999 is available online at as well as in previous publications15. The follow up study characterized the presence or absence of dementia in 1889 subjects: 307 were demented, 175 had cognitive impairment not sufficient to meet diagnostic criteria for dementia (cognitively impaired not demented-CIND9), and 1407 elderly were cognitively normal [No Cognitive Impairment (NCI)]. Since subjects diagnosed with CIND were not included in the analysis, the final sample for logistic regression models included 1417 subjects.
Definition of Crisis Following Parental Death (CFPD)
In 1965, participants were asked whether they had experienced a serious crisis, specifically, crisis following maternal or paternal death, in childhood, adolescence or as adults. If affirmative, they were further asked about the age range at which they had experienced the crisis, i.e.: 0–6, 7–12, 13–18 or >18 years. In the questionnaire, there was no specification as to the definition of crisis, and participants' answers were based on their subjective definition of crisis. The aim of the questionnaire was to detect CFPD rather than record age of subjects at death of parents, therefore, the group of subjects classified in 1965 as "no CFPD" included a minority of subjects whose parents were still alive as well as subjects whose parents died but they did not report crisis following their death. The number of participants who suffered parental death but did not report it as a crisis is not known. However, we assume that the number of subjects whose parents were still alive was small since the subjects' age in 1963 was 40 and above and life expectancy in Israel in 1965 was 70 28. Moreover, many of the subjects immigrated to Israel after they and there families had been persecuted further increasing the odds that their parents had died by 1963–1965.
Assessment of anxiety
assessment of anxiety in 1963 is described in detail at the supplemental online version of ADAD as well as in previous publications 15.
Socioeconomic status index (SES)
assessment of SES in 1963 is described in detail at the supplemental online version of ADAD as well as in previous publications 15.
Statistical analysis
An order-directed score test of trend developed by Cuzick 16, an extension of the Wilcoxon rank-sum test, was used to examine the departure of sample results, consistent with a ‘dose-response’ association between age-at-parental death crisis report categories, and late-life dementia prevalence, from a null hypothesis of an independence of dementia rates from the above categories.
Logistic regression was used to estimate the covariate-adjusted odds ratios for dementia prevalence, associated with having been at each of three different childhood and adolescence age groups during CFPD (0–6, 7–12, 13–18), compared to the group that reported CFPD after the age of 18. Logistic regression was also used to estimate the covariate-adjusted odds ratios for dementia in subjects who reported CFPD after the age of 18 compared to subjects who did not report CFPD. Adjustment was made for age, socioeconomic status, and order of birth and country of birth.
The study was approved by the Sheba Medical Center ethics committee.
Results
In 1965, 9715 participants of the IIHD who were recruited in 1963, returned for examination. Of these, 9362 answered questions relating to CFPD. Dementia evaluation was performed in 1889 survivors of the original cohort, 1652 of whom were interviewed in midlife about CFPD. CFPD-P at each of three different childhood and adolescence age groups (0–6, 7–12, 13–18) was reported among 2252 individuals participating in the original cohort (24.0%) and among 241 survivors in 1999/2000 who were evaluated for dementia (14.6%). CFPD-M at each of three different childhood and adolescent age groups was reported in 1488 individuals participating in the original cohort (15.9%) and 169 survivors who were evaluated for dementia (10.2%) in 1999/2000. One hundred and eight (6.5%) survivors reported CFPD-P at the age of 18 or above and 71 (4.3%) reported CFPD-M at the age of 18 or above.
Description of study participants with CFPD data
Socio-demographic characteristics
Report of CFPD-M at any age was associated with lower midlife socioeconomic status. Of those who did not report CFPD-M, 22.9% were in the lowest socioeconomic status group in midlife compared to 35.8%, 32.0%, 30.6% and 23.6% in those who reported CFPD-M at the age groups of 0–6, 7–12, 13–17 and >18 respectively), and subjects who reported CFPD-M were also less likely to be in high socioeconomic groups in midlife compared to those who did not report CFPD-M (Chi2 =36.89, DF=12, P<0.0005)(table 1). Similar results were obtained when CFPD-P was examined (Chi2=42.49, DF=12, P<0.0005).
Table 1.
Socio-demographic variables in midlife according to age during CFPD*
| Age at CFPD* | Age (sd) | SES^ | Order of birth | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1^ | 2 | 3 | 4 | 5^ | Total | First born |
Middle born |
Last born |
|||
| Age at CFPD-P* | No crisis | 49.1 (6.8) | 1,554 (22.0) | 1,647 (23.4) | 2,307 (32.6) | 798 (11.2) | 766 (10.8) | 7,072 (100) | 2,059 (29.1) | 3,673 (51.9) | 1,340 (19.0) |
| 0–6 | 49.6 (5.9) | 239 (36.4) | 154 (23.4) | 174 (26.4) | 53 (8.0) | 38 (5.8) | 658 (100) | 156 (23.7) | 211 (31.9) | 294 (44.4) | |
| 7–12 | 50.3 (6.8) | 160 (34.6) | 109 (23.4) | 127 (27.4) | 29 (6.4) | 38 (8.2) | 463 (100) | 114 (24.7) | 219 (47.5) | 128 (27.8) | |
| 13–18 | 49.0 (6.9) | 177 (31.3) | 140 (24.7) | 165 (29.2) | 46 (8.1) | 38 (6.7) | 566 (100) | 119 (20.9) | 283 (49.8) | 165 (29.3) | |
| > 18 | 48.6 (6.7) | 109 (20.1) | 135 (25.0) | 201 (37.1) | 54 (10.0) | 42 (7.8) | 541 (100) | 143 (26.5) | 276 (51.2) | 120 (22.3) | |
| Total | 2,239 (24.1) | 2,185 (23.5) | 2,974 (32.0) | 980 (10.5) | 922 (9.9) | 9,300 (100) | 2,591 (27.8) | 4,662 (50.0) | 2,047 (22.2) | ||
| Age at CFPD-M* | No crisis | 49.1 (6.8) | 1,794 (22.9) | 1,839 (23.5) | 2,529 (32.3) | 849 (10.8) | 817 (10.5) | 7,828 (100) | 2,271 (29.0) | 3,949 (50.5) | 1,607 (20.5) |
| 0–6 | 49.3 (6.1) | 125 (35.8) | 80 (22.9) | 101 (28.9) | 26 (7.5) | 17 (4.9) | 349 (100) | 81 (23.2) | 155 (44.4) | 113 (32.4) | |
| 7–12 | 50.7 (6.7) | 100 (33.0) | 82 (27.0) | 90 (29.8) | 15 (4.9) | 17 (5.3) | 304 (100) | 64 (21.0) | 163 (53.5) | 77 (25.5) | |
| 13–18 | 49.7 (7.2) | 114 (30.6) | 83 (22.3) | 97 (26.1) | 41 (11.0) | 37 (10.0) | 372 (100) | 73 (19.4) | 174 (46.4) | 128 (34.2) | |
| > 18 | 49.1 (6.6) | 106 (23.6) | 101 (22.5) | 158 (35.2) | 49 (10.9) | 35 (7.8) | 449 (100) | 102 (22.9) | 221 (49.5) | 123 (27.6) | |
| Total | 2,239 (24.0) | 2,185 (23.5) | 2,975 (32.0) | 980 (10.5) | 923 (10.0) | 9,302 (100) | 2,591 (100) | 4,662 (100) | 2,048 (100) | ||
CFPD= Crisis Following Parental Death, CFPD-P= Crisis Following Parental Death- Paternal, CFPD-M= Crisis Following Parental Death-Maternal.
SES= Socio Economic Status; 1= lowest SES, 5= highest SES
Relatively fewer subjects who reported CFPD-P at age 18 and above were born in Middle East (Arab countries, Turkey and Iran) or North Africa (table 2). Subjects who reported CFPD at the age of 0–6 included a higher percentage of last born children (tables 1–2).
Table 2.
Country of origin according to age during CFPD*
| Age at CFPD* |
Country of origin | |||||||
|---|---|---|---|---|---|---|---|---|
| Israel | East Europe |
Central Europe |
Balkan | Middle East |
North Africa |
Total | ||
| Age at CFPD-P* | No crisis | 938 (13.8) | 1,425 (20.0) | 1,038 (14.6) | 1,306 (18.4) | 1,525 (21.4) | 833 (11.8) | 7,110 (100) |
| 0–6 | 121 (18.2) | 91 (13.7) | 52 (7.8) | 84 (12.6) | 241 (36.2) | 77 (11.5) | 666 (100) | |
| 7–12 | 63 (13.5) | 70 (15.0) | 43 (9.3) | 64 (13.8) | 160 (34.4) | 65 (14.0) | 465 (100) | |
| 13–18 | 74 (12.9) | 102 (17.8) | 69 (12.0) | 83 (14.5) | 156 (27.3) | 88 (15.5) | 572 (100) | |
| ≥18 | 79 (14.5) | 100 (18.4) | 90 (16.6) | 97 (17.8) | 112 (20.8) | 65 (11.9) | 543 (100) | |
| Total | 1,320 (14.1) | 1,788 (19.1) | 1,292 (13.8) | 1,634 (17.5) | 2,194 (3.5) | 1,128 (12.0) | 9,356 (100) | |
| Age at CFPD-M* | No crisis | 1,108 (14.1) | 1,494 (18.9) | 1,101 (14.0) | 1,395 (17.7) | 1,851 (23.5) | 925 (11.8) | 7,874 (100) |
| 0–6 | 65 (18.4) | 48 (13.6) | 28 (8.0) | 50 (14.2) | 110 (31.3) | 51 (14.5) | 352 (100) | |
| 7–12 | 37 (12.1) | 58 (18.9) | 27 (8.8) | 42 (13.6) | 97 (31.6) | 46 (15.0) | 307 (100) | |
| 13–18 | 45 (12.0) | 96 (25.5) | 51 (13.5) | 54 (14.3) | 77 (20.5) | 53 (14.2) | 376 (100) | |
| ≥18 | 66 (14.7) | 92 (20.5) | 85 (19.0) | 92 (20.5) | 59 (13.2) | 55 (12.1) | 449 (100) | |
| Total | 1,321 (14.1) | 1,788 (19.1) | 1,292 (13.8) | 1,633 (17.4) | 2,194 (23.5) | 1,130 (12.1) | 9,358 (100) | |
Midlife cardiovascular characteristics
As presented in table 3, men who reported CFPD (maternal or paternal) at different age groups did not differ in midlife cardiovascular profile from men who did not report CFPD. There were some differences in midlife diabetes rates, but no consistency was observed in the relationship between diabetes rates and CFPD at different ages (table 3).
Table 3.
Baseline cardiovascular characteristics according to age during CFPD*
| Age at CFPD* | Systolic BP* mmHg) (sd) |
Diastolic BP* mmHg) (sd) |
Cholesterol (mg/dl) (sd) |
Height (cm) (sd) |
Weight (Kg) (sd) |
BMI (Kg/m2) (sd) |
% diabetes |
% ever smoked |
|
|---|---|---|---|---|---|---|---|---|---|
| Age at CFPD-P* | None | 135 (20) | 86 (12) | 209 (40) | 167.3 (6.6) | 72.0 (10.6) | 25.7 (3.3) | 4.3 | 68 |
| 0–6 | 86 (12) | 85 (11) | 207 (39) | 165.8 (6.6) | 69.7 (11.2) | 25.3 (3.6) | 4.1 | 68 | |
| 7–12 | 136 (21) | 86 (12) | 207 (39) | 166.4 (6.7) | 70.5 (10.7) | 25.4 (3.3) | 7.3 | 67 | |
| 13–18 | 136 (20) | 87 (12) | 209 (40) | 167.0 (6.7) | 71.8 (11.2) | 25.7 (3.3) | 5.8 | 72 | |
| >18 | 134 (18) | 87 (11) | 209 (39) | 167.1 (6.4) | 72.3 (10.0) | 25.8 (3.1) | 5.7 | 69 | |
| Age at CFPD-M* | None | 135 (20) | 84 (11) | 209 (40) | 167.3 (6.6) | 72.0 (10.7) | 25.7 (3.3) | 4.5 | 68 |
| 0–6 | 135 (20) | 84 (11) | 205 (39) | 166.0 (6.5) | 70.1 (11.2) | 25.4 (3.6) | 3.7 | 70 | |
| 7–12 | 136 (21) | 84 (11) | 207 (43) | 165.1 (6.9) | 69.5 (10.9) | 25.5 (3.5) | 4.6 | 67 | |
| 13–18 | 135 (20) | 84 (11) | 211 (40) | 166.6 (7.0) | 71.0 (11.3) | 25.5 (3.4) | 6.1 | 67 | |
| >18 | 133 (18) | 83 (10) | 210 (41) | 166.8 (6.5) | 72.1 (10.1) | 25.9 (3.2) | 5.8 | 71 | |
BP= Blood pressure
Midlife anxiety
Anxiety rates in midlife, 27 years or longer after parental death did not differ between groups. An anxiety score of 3 (highest) was recorded in 4.3% of those who did not report CFPD-P, 4.2%, 4.1% and 4.8% in those who reported CFPD-P at ages 0–6, 7–12 and 13–18 respectively. This anxiety score was recorded in 4.2% of those who did not report CFPD-M, 5.1%, 4.2% and 3.7% in those who reported CFPD-M at ages 0–6, 7–12 and 13–18 respectively.
Survival in 1999/2000 by CFPD group
Death rates per 100,000 years by 1999/2000 were 521 among participants who did not report CFPD-P and 513, 548, 538 and 525 among those who reported CFPD-P at ages 0–6, 7–12, 13–18 and ≥18 respectively. Death rates per 100,000 years by 1999 were 522 among participants who did not report CFPD-M and 533, 536, 523 and 524 among those who reported CFPD-M at ages 0–6, 7–12–13–18 and ≥ respectively. Thus, death rates by 1999 per 10,000 person years did not differ between groups according to age at reported CFPD (maternal or paternal).
Associations of parental death at different age groups with dementia at old age
Relationships of age at CFPD-P with dementia: dementia prevalence rates in subjects who survived by 1999/2000 were 17.3% for those who did not report CFPD-P and 29.8%, 21.4%, 20.9% and 11.1% respectively for subjects who suffered CFPD-P at the ages of 0–6, 7–12, 13–18 and > 18. Controlling for age, the odds for dementia, relative to individuals who reported crisis following paternal death at ages above 18 years, were 3.06 (95%CI 1.42–6.61), 2.15 (95% CI 0.87–5.31) and 2.35 (95%CI 1.05–5.28) for those who reported CFPD-P at the ages 0–6, 7–12 and 13–18 years (table 4). Non parametric trend test yielded z=−3.22 (p= 0.001). Adding order of birth, SES and country of birth to the logistic regression did not change results materially (table 4). Age adjusted odds ratio for dementia were 0.60 (95% CI=0.32–1.11) for those who reported CFPD-P at the age of 18 and above compared to all the participants who reported no such crisis at any age.
Table 4.
The association of age during CFPD* (at age groups younger than 18 compared to CFPD* at age ≥18**) and dementia at old age
| Age at CFPD |
n | OR for dementia adjusted for age (95% CI) |
OR for dementia adjusted for age and birth order (95% CI) |
OR for dementia adjusted for age, birth order, country of birth and SES) (95% CI) |
|
|---|---|---|---|---|---|
| Age at CFP-DP* | 0–6 | 94 | 3.06 (1.42–6.61) | 3.04 (1.41–6.56) | 3.02 (1.35–6.73) |
| 7–12 | 56 | 2.15 (0.87–5.31) | 2.17 (0.88–5.36) | 2.58 (1.00–6.58) | |
| 13–18 | 91 | 2.35 (1.05–5.28) | 2.32 (1.03–5.22) | 2.30 (0.99–5.36) | |
| Age at CFP-DM* | 0–6 | 59 | 3.56 (1.25–9.02) | 3.41 (1.26–9.21) | 3.52 (1.23–0.01) |
| 7–12 | 41 | 3.96 (1.38–11.33) | 4.48 (1.53–3.09) | 4.22 (1.36–13.1) | |
| 13–18 | 69 | 2.65 (0.99–7.08) | 2.53 (0.95–6.8) | 2.61 (0.93–7.40) |
number of subjects who reported crisis following parental death after the age of 18 was 108 following paternal death and 71 following maternal death
Relationship of age at crisis, following maternal death, with dementia: dementia prevalence rates in subjects who survived by 1999 were 17.4% for those who did not suffer CFPD-M and 27.1%, 29.3%, 23.2% and 9.9% respectively for subjects who suffered CFPD-M at the ages of 0–6, 7–12, 13–18 and > 18 years. Controlling for age, the odds for dementia relative to individuals who reported crisis following maternal death at the age of 18 and above were 3.56 (95% CI 1.25–9.02), 3.96 (95% CI 1.38–11.33) and 2.65 (95% CI 0.99–7.08) for those who reported CFPD-M at the ages of 0–6, 7–12 and 13–18 (table 4). Trend test yielded z=−3.22 (p=0.001). Adjusting for order of birth, SES and country of birth did not change results materially (table 4). Similar to the results for paternal death, age adjusted odds for dementia were 0.52 (95% CI=0.24–1.15) for those who reported CFPD-M at ages above18 relative compared to all the participants who reported no such crisis at any age.
Discussion
The present study demonstrated that reported CFPD during childhood and adolescence is associated with higher risk for dementia at old age compared to report of such crisis at ages of 18 and above as well as no report of such crisis. Odds ratio for dementia were highest for those who reported CFPD-P at the age of 0–6 and for those who reported CFPD-M at the age of 7–12 and decreased thereafter with increasing age of crisis. In subjects who reported CFPD at ages above 18, risk for dementia at old age among survivors was reduced compared to subjects whose parents died but did not report such a crisis or those whose parents were still alive in 1965. Considering participants' ages, life expectancy at the time (70 years) and deaths in wars, pogroms and the Holocaust, the majority of subjects must have lost their parents by the time of the study.
Exposure to trauma and parental loss at young age have been reported occasionally to affect morbidity later in life5, 6 and mortality 7. Higher rates of cardiovascular morbidity, smoking rates and type II diabetes, all risk factors for dementia 11were found in individuals who were exposed as children to traumatic experiences, even 60 years after their occurrence 8, 17. Parental death in children and adolescents may also be associated with decreased family income, marked distress, depressive symptoms, social withdrawal and academic difficulties, of significance long after parental death9, 10, 18. Low education, low income in early life and short stature 19, 20 have been associated with late life dementia. In our study reported CFPD at different ages and absence of CFPD were not associated with midlife cardiovascular profile, diabetes rates or smoking status, whereas early CFPD were associated with lower socioeconomic status in midlife. Survivors' dementia remained associated with reported CFPD in multivariate analysis. Depression, a possible mediator between early life CFPD and late life cognition, was not measured; however, anxiety (measured as tendency to experience tension, fear and/or sleeping problems) which is closely associated with depression21was measured. Groups of subjects stratified by age at reported CFPD did not differ in anxiety score.
The tendency for increasing risk for dementia with decreasing age at CFPD may be, at least partially, explained by the fact that in parentally bereaved children and adolescents, child's age is significantly correlated with mental health outcomes, younger age being associated with increased mental health problems22. A child that loses a parent may have less developed internal coping mechanisms as well as overwhelmed and thus non efficient familial support system. Adults who lose their parents may be less overwhelmed and more resilient to such a trauma.
The association found in our study of decreased dementia prevalence in survivors who reported CFPD at ages of above 18, may be explained by the possibility that in adults, tendency to report such a crisis is an indicator of more effective coping skills, such that enable help seeking behavior, appreciation of life, strengthening of emotional bonds and growth from traumatic events23.
Report of crisis following parental death at different ages was not associated with decreased survival by 1999/2000, the time of dementia assessment; reducing the likelihood of survival bias effect. Additionally, baseline variables related to high risk of mortality, mainly smoking, diabetes and hypertension did not differ between groups of subjects stratified according to age at reported CFPD for the entire IIHD cohort, nor in those who survived until follow up in 1999/2000.
Consistent with our findings, death of father at early life was associated with increased risk for dementia in later life in the Cache County Study24 in which paternal death prior to the age of 5 was associated with a 3 fold increased risk for dementia (OR = 3.0; 95% CI: 1.1–8.4). This risk ratio was similar to the dementia prevalence ratio found in survivors of our study. In contrast, maternal death at an early age was not associated with dementia. The discrepant results may result from report of parental death, rather than reported CFPD, with dementia, in Cache. The life period in which subjects were interviewed regarding parental death differed between studies (midlife in our study versus old age in the Cache County Study) In the Cache County Study, subjects whose parents died after the age of 19 were grouped with subjects who still reported parents living as adults, and thus comparisons to the decreased risk associated with reported crisis due to parental loss at an age > 18 found in our study is not possible. Also, circumstances of parent's deaths were perhaps not as violent in Cache County as they were in Jewish migrants to Israel from countries undergoing great turmoil.
The main limitation of this study, as described previously, is the lack of information on the incidence of dementia in the subjects from the original IIHD study reported dead before the follow up study was conducted. Another limitation pertains to the fact that only males were included in the current study. Additionally, only CIND subjects who were identified in the second phase of cognitive assessment were excluded, and there is a possibility that some CIND subjects received a TICS-m score above the 27 cutoff score, thus being classified as cognitively normal. Although the number of subjects participating in the sensitivity sample was relatively small, this sample was randomly selected from the group of subjects whose TICS was 28–29, i.e. subjects at the highest risk of being erroneously classified as cognitively normal. Indeed assessment of these subjects by a clinician at their residence demonstrated the TICS cutoff point (>27) to be highly sensitive, even to CIND. Another limitation was at home assessment of dementia based on subject and informant's report (DQ) combined with the MMSE, rather than on a broad neuropsychological battery. In this regard, it is important to emphasize that that assessment of cognitive status was directly made by a physician with expertise in dementia which enhances the certainty of the assessment.
Lack of data regarding the numbers and midlife health and psychological characteristics of most of the study subjects, who suffered parental death but did not report crisis following this loss, limits our ability to interpret the results as related to the crisis following parental loss per se compared to loss of the parent.
The strengths of our study are the relatively large sample size, longitudinal design and evaluation of CFPD at midlife, when answers regarding childhood experiences are not expected to be affected by the dementing process. In addition, an abundance of data regarding midlife demographic, psychological, social and biomedical data, all of which may affect late life cognitive status, was available to us. These variables have been controlled for when analyzing the association of age of subject at parental death and dementia risk.
Our findings support the potential contribution of early life events to late-life dementia risk. Future studies should assess possible physiological and environmental mechanisms for this association.
Acknowledgments
Support:
This study was supported by NIA grants K01 AG023515-01 and R01 AG034087, and the Graubard Fund for Dr. Beeri, P01 AG02219 and the Berkman Trust for Dr. Haroutunian, and P50 AG05138 for Dr. Sano.
Appendix
The IIHD project
The sample of the Israeli Ischemic Heart Disease (IIHD) Project was chosen by stratified sampling of civil servants and municipal employees aged 40 and above in 1963. Sampling procedures, mortality and follow-up have been described in detail elsewhere15. Of 11,876 approached, 10,232 (86.2%) men agreed to participate and were enrolled in 1963. Subsequent analyses excluded 173 men who were born outside the six pre-defined geographical areas. Among the remaining 10,059 men, 9715 returned to the second examination cycle in 1965 and 9368 and consented to reply to questions regarding CFPD.
Diagnosis of dementia
Dementia was diagnosed in 1999/2000 using a two-step procedure as described in detail elsewhere 16. All identified living participants from the original cohort who consented to participate in the new assessment, were administered a 20 minute phone interview, including a sociodemographic questionnaire and the Hebrew version of the Modified Telephone Interview for Cognitive Status (TICS-m) 17.
The TICS-m is based on the Mini Mental State Exam (MMSE)18. The scores of the TICS–m were highly correlated with the MMSE scores in clinical studies 19. Using a threshold score of 27 out of 50 TICS-m points, the sensitivity of this test was above 99% (CI 95% 91%–100%) and the specificity was 86% (CI 95% 85%–87%) for dementia against the gold standard clinical diagnosis 17. The second step was a face to face interview for all subjects with a TICS-m score of 27 or lower. The aim of the second phase of the study was to ascertain the diagnosis of dementia among subjects who were identified by the TICS-m as possibly cognitively impaired. The subjects were assessed at their residences by a physician (a neurologist or psychiatrist with expertise and extensive experience in the diagnosis of dementia). Physicians were blind to the TICS-m score. The clinical assessment included the Dementia Questionnaire (DQ)20, 21, MMSE18, Global Deterioration Scale (GDS) 22 and the Hachinski Ischemic Scale (HIS) 23.
For persons with suspected dementia, the DQ assesses, through informants, the symptoms, course, and feature of the dementia, permitting a diagnostic classification for the presence and likely type of dementia. The DQ has previously demonstrated good inter-informant and inter-rater agreement in AD probands and normal elderly control populations for detecting the presence or absence of dementia and also, among identified cases of dementia, the type of dementia 21, 24. Compared with a clinical assessment of dementia, the DQ was found to have excellent sensitivity (100%; 95% CI 73%–100%) and very good specificity (90%; 95% CI 63%–100%) 25, 26. In this study, informants were people, usually first degree relatives who had contact with the participants of at least 10 hours per week.
Dementia was diagnosed using the DSM-IV criteria. Subjects were classified as Cognitively Impaired no Dementia (CIND) if both subject and informant reported a memory problem but they had normal activities of daily living and were not demented. CIND was diagnosed at the second phase of the study, by a physician who assessed the subjects in their residence. It is possible that some of the subjects classified as NCI during the first phase of the study (using the TICS) would have met our criteria for CIND, however, it should be noted that the TICS has demonstrated excellent sensitivity in our 16as well as in other 27 samples.
A second physician, blind to the diagnosis assigned by the physician who did the face-to-face interview, examined the data gathered from each patient and assigned an independent diagnosis. Disagreement was solved by a third physician. In five cases, the patient was approached again for this purpose.
According to the Israel Mortality Registry, 7136 men died by the beginning of the study in 1999; another 306 men died before being approached for a phone interview and 13 subjects were lost in the matching process. Detailed description about subjects assessed in 1999 is available online at as well as at () The remaining 2604 subjects qualified for a telephone interview, and 2038 had phone contact. The telephone screening identified 799 potentially demented subjects for a home interview. Of these, 149 could not be examined. Of the remaining 1239 subjects who were not identified as potentially demented by the TICS-m (i.e. TICS score ≥28), 51 subjects whose TICS score was 28 or 29, and thus with the highest risk of being erroneously classified as cognitively normal, were randomly selected and examined at home for a sensitivity study of the phone interview instruments. From the sensitivity sample examined at home, 50 were cognitively intact and one cognitively impaired but not demented. The remaining 1188 subjects were classified as non-demented. The follow up study characterized the presence or absence of dementia in 1889 subjects: 307 were demented, 175 had cognitive impairment not sufficient to meet diagnostic criteria for dementia (cognitively impaired not demented-CIND9), and 1407 elderly were cognitively normal [No Cognitive Impairment (NCI)]. Since subjects diagnosed with CIND were not included in the analysis, the final sample for logistic regression models included 1417 subjects.
Assessment of anxiety
In 1963 (age > 40) subjects were questioned regarding existence of anxiety symptoms: "Do you consider yourself a tense person?", "Do you suffer from anxiety (fear) for no evident reason?"" Do you suffer form sleep problems?” Each question could be answered "yes" or "no". A total score for anxiety was either "0", "1", "2" or "3", reflecting the number of questions answered positively.
Socioeconomic status index (SES)
This index was based on formal education and type of employment at the time of examination. Formal education included nine levels, ranging from "no formal education" to "completed university education". Employment included five levels ranging from "laborer" to professional. The SES scale ranged from "1": very low SES (three lowest education levels and the lowest employment levels) to "5": very high SES (three highest education levels as well as two highest employment level) 29.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reference List
- 1.Fratiglioni L, Wang HX. Brain reserve hypothesis in dementia. J Alzheimers Dis. 2007 August;12(1):11–22. doi: 10.3233/jad-2007-12103. [DOI] [PubMed] [Google Scholar]
- 2.Whalley LJ, Dick FD, McNeill G. A life-course approach to the aetiology of late-onset dementias. Lancet Neurol. 2006 January;5(1):87–96. doi: 10.1016/S1474-4422(05)70286-6. [DOI] [PubMed] [Google Scholar]
- 3.Ownby RL, Crocco E, Acevedo A, John V, Loewenstein D. Depression and risk for Alzheimer disease: systematic review, meta-analysis, and metaregression analysis. Arch Gen Psychiatry. 2006 May;63(5):530–538. doi: 10.1001/archpsyc.63.5.530. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wilson RS, Evans DA, Bienias JL, Mendes de Leon CF, Schneider JA, Bennett DA. Proneness to psychological distress is associated with risk of Alzheimer's disease. Neurology. 2003 December 9;61(11):1479–1485. doi: 10.1212/01.wnl.0000096167.56734.59. [DOI] [PubMed] [Google Scholar]
- 5.Jones GT, Power C, Macfarlane GJ. Adverse events in childhood and chronic widespread pain in adult life: Results from the 1958 British Birth Cohort Study. Pain. 2009 May;143(1–2):92–96. doi: 10.1016/j.pain.2009.02.003. [DOI] [PubMed] [Google Scholar]
- 6.Smith KR, Mineau GP, Garibotti G, Kerber R. Effects of childhood and middle-adulthood family conditions on later-life mortality: evidence from the Utah Population Database, 1850–2002. Soc Sci Med. 2009 May;68(9):1649–1658. doi: 10.1016/j.socscimed.2009.02.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Brown DW, Anda RF, Tiemeier H, et al. Adverse childhood experiences and the risk of premature mortality. Am J Prev Med. 2009 November;37(5):389–396. doi: 10.1016/j.amepre.2009.06.021. [DOI] [PubMed] [Google Scholar]
- 8.Roberts ME, Fuemmeler BF, McClernon FJ, Beckham JC. Association between trauma exposure and smoking in a population-based sample of young adults. J Adolesc Health. 2008 March;42(3):266–274. doi: 10.1016/j.jadohealth.2007.08.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jacobs JR, Bovasso GB. Re-examining the long-term effects of experiencing parental death in childhood on adult psychopathology. J Nerv Ment Dis. 2009 January;197(1):24–27. doi: 10.1097/NMD.0b013e3181927723. [DOI] [PubMed] [Google Scholar]
- 10.Kivela SL, Luukinen H, Koski K, Viramo P, Pahkala K. Early loss of mother or father predicts depression in old age. Int J Geriatr Psychiatry. 1998 August;13(8):527–530. doi: 10.1002/(sici)1099-1166(199808)13:8<527::aid-gps814>3.0.co;2-8. [DOI] [PubMed] [Google Scholar]
- 11.Harwood DG, Kalechstein A, Barker WW, et al. The effect of alcohol and tobacco consumption, and apolipoprotein E genotype, on the age of onset in Alzheimer's disease. Int J Geriatr Psychiatry. 2009 September 14; doi: 10.1002/gps.2372. [DOI] [PubMed] [Google Scholar]
- 12.Ben-Shlomo Y, Kuh D. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. Int J Epidemiol. 2002 April;31(2):285–293. [PubMed] [Google Scholar]
- 13.Xu W, Qiu C, Gatz M, Pederson NL, Johansson B, Fratiglioni L. Mid- and late-life diabetes in relation to the risk of dementia: a population-based twin study. Diabetes. 2010;58(1):71–77. doi: 10.2337/db08-0586. Ref Type: Generic. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Triplitt C. Cardiac risk factors and hypoglycemia in an elderly patient: how good is good enough? Consult Pharm. 2010 June;25 Suppl B:19–27. [PubMed] [Google Scholar]
- 15.Ravona-Springer R, Beeri MS, Goldbourt U. Repetitive thinking as a psychological cognitive style in midlife is associated with lower risk for dementia three decades later. Dement Geriatr Cogn Disord. 2009;28(6):513–520. doi: 10.1159/000257089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cusick J. A Wilcoxon-type test for trend. Stat Med. 2010;14(4):445–446. doi: 10.1002/sim.4780140409. Ref Type: Generic. [DOI] [PubMed] [Google Scholar]
- 17.Alastalo H, Raikkonen K, Pesonen AK, et al. Cardiovascular health of Finnish war evacuees 60 years later. Ann Med. 2009;41(1):66–72. doi: 10.1080/07853890802301983. [DOI] [PubMed] [Google Scholar]
- 18.Cerel J, Fristad MA, Verducci J, Weller RA, Weller EB. Childhood bereavement: psychopathology in the 2 years postparental death. J Am Acad Child Adolesc Psychiatry. 2006 June;45(6):681–690. doi: 10.1097/01.chi.0000215327.58799.05. [DOI] [PubMed] [Google Scholar]
- 19.Beeri MS, Davidson M, Silverman JM, Noy S, Schmeidler J, Goldbourt U. Relationship between body height and dementia. Am J Geriatr Psychiatry. 2005 February;13(2):116–123. doi: 10.1176/appi.ajgp.13.2.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Borenstein AR, Copenhaver CI, Mortimer JA. Early-life risk factors for Alzheimer disease. Alzheimer Dis Assoc Disord. 2006 January;20(1):63–72. doi: 10.1097/01.wad.0000201854.62116.d7. [DOI] [PubMed] [Google Scholar]
- 21.Moffitt TE, Harrington H, Caspi A, et al. Depression and generalized anxiety disorder: cumulative and sequential comorbidity in a birth cohort followed prospectively to age 32 years. Arch Gen Psychiatry. 2007 June;64(6):651–660. doi: 10.1001/archpsyc.64.6.651. [DOI] [PubMed] [Google Scholar]
- 22.Lin KK, Sandler IN, Ayers TS, Wolchik SA, Luecken LJ. Resilience in parentally bereaved children and adolescents seeking preventive services. J Clin Child Adolesc Psychol. 2004 December;33(4):673–683. doi: 10.1207/s15374424jccp3304_3. [DOI] [PubMed] [Google Scholar]
- 23.Wolchik SA, Coxe S, Tein JY, Sandler IN, Ayers TS. Six-year longitudinal predictors of posttraumatic growth in parentally bereaved adolescents and young adults. Omega (Westport) 2008;58(2):107–128. doi: 10.2190/OM.58.2.b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Norton MC, Ostbye T, Smith KR, Munger RG, Tschanz JT. Early parental death and late-life dementia risk: findings from the Cache County Study. Age Ageing. 2009 May;38(3):340–343. doi: 10.1093/ageing/afp023. [DOI] [PMC free article] [PubMed] [Google Scholar]