Early parental death and psychosocial risk factors for dementia: A case–control study in Europe

To assess the association between early parental death and the risk of dementia in adult life and to examine the risk factors associated with early parental death in people with and without dementia.

Objectives: To assess the association between early parental death and the risk of dementia in adult life and to examine the risk factors associated with early parental death in people with and without dementia.
Methods/design: A population-based case-control study of a sample of 65 997 participants from the Survey of Health, Ageing and Retirement in Europe study. Early parental death was operationalized as parental death at the age of ≤16 years. Main analyses were conducted using bivariate and multivariate logistic regression analyses.
Results: The odds ratio (OR) for dementia in individuals who experienced early parental death (father or mother) at the age of ≤16 years was 1.83 (95%CI 1.61-2.09) and Discussion: Early parental death (≤16 years) was associated with an increased risk of dementia. We discuss the neurobiological markers associated with adverse childhood experiences (ACEs) and dementia as well as interventions to counteract the negative health effects on adults. The early death of parents has been one of the most studied adverse childhood experiences (ACEs) and has been shown to have negative effects on health throughout the life cycle. 6 It has been associated with a higher risk of depression 7,8 and suicide, 9 lower academic achievement, 10 lower occupational position, 11 higher separation rates, 12 and a higher probability of psychosis. 13 In terms of physical health, early parental death has been associated with worse health perception 14 and more somatic symptoms, 15 as well as higher mortality, 16 cardiovascular disease, 4,5,17 self-harm, 18 and an increased risk of dementia. [19][20][21][22][23][24][25][26] Previous studies have identified certain mechanisms involved in the association between early parental death and the risk of dementia and other pathologies associated with older age: stress, 27,28 increased inflammatory processes, 29 alterations of the hypothalamic-pituitaryadrenal axis, 30 and shorter telomeres on chromosomes. 31,32 Research into the relationship between early parental death and dementia and/or severe cognitive impairment in the elderly has confirmed a general increased risk (odds ratios [OR]: 1.52-6.41), [19][20][21][22][23][24][25][26] although the results vary according to gender and age at the time of parental death. In a prospective longitudinal study, Persson and Skoog (1996) 19 found that the death of a parent before the age of 16 was associated with a relative risk (RR) of developing dementia of 6.3 (95% CI: 1.8-21.1). In contrast, in Fu's study (2020), 20 only the death of the mother before the age of 16 was associated with severe cognitive decline in men (OR = 1.52, 95%CI: 1.01-2.28), though not in women.
Other authors have reported different age ranges. Conde-Sala (2003) 21 found that the death of the father before the age of 20 was associated with an increased risk of dementia in women. In Norton et al (2009), 22 the father's death before the age of four resulted in a threefold increase in the dementia risk, without differences for gender and education and after adjustment for APOE-4. With a larger sample, the same authors (Norton et al, 2011) 23 also found a greater risk in the case of the father's death before the age of five and the mother's death between 11 and 17 years. In both cases, female gender, older age and APOE-4 were associated with a greater risk of dementia.
Whalley et al (2013) 24 found an increased risk of dementia with the death of either parent before the age of 11, which was also associated with female gender, older age and APOE-4. In Wei et al (2014), 25 early parental death was associated with an increased risk of dementia in the whole sample. In the study by Ravona-Springer et al (2012) 26 which included data only on men (0-18 years), early parental death, either of the mother or the father, was associated with a higher risk of dementia.
In view of the heterogeneity of the results of previous studies and also bearing in mind the small size of some samples, the present study aimed to explore the association between early parental death and the risk of dementia according to age ranges, gender of parents and participants, schooling and income, using a large representative sample of older Europeans. We hypothesized that individuals with early parental death would have a higher prevalence of dementia.
The specific objectives of the present study were (a) To determine the frequency of early parental death in childhood and adolescence and its association with the risk of dementia in adult life and (b) To examine the risk factors associated with early parental death in people with and without dementia.

| Study design
Data for this population-based case-control study came from the seventh wave of the "Survey of Health, Ageing and Retirement in Europe" (SHARE), corresponding to the year 2017 and including people born between 1912 and 1967. The SHARE study provided information on sociodemographic aspects and on physical and mental health status. 33 The seventh wave collected data from people over 50

| Measures
Periods of parental death. The year of parental death and the year of the participants' birth were used to establish participants' age periods at time of parental death: 0 to 5, 6 to 10, 11 to 16, 17 to 20 and >21 years. The analysis of the risk of dementia focused on the period of early parental death, that is, when the participants were aged ≤16 years. Previous authors have established this age as the most relevant for assessing the effect of early parental death (≤16 years) on the health of the elderly. 11,15,19,34 Sociodemographic data. The following data were recorded: age (<65/≥65 years), gender (male/female), marital status (married, widowed, separated-divorced, or never married), mean years of schooling (<11/≥11 years), economic difficulties (yes/no) and income in percentiles (≤50/>50).
Functional state. Five basic daily life activities (dressing, bathing, eating, cutting up food, walking across a room and getting in or out of bed) and five instrumental activities (telephone calls, taking medication, managing money, shopping for groceries and preparing a hot meal) were examined by assessing the presence or absence of the activity (0/≥1 deficit). Four items were used to determine fragility (falling down, fear of falling down, dizziness and fatigue). On all scales, the higher the score, the greater the difficulty.
Cognition. The items of recent memory (range = 0-10) and deferred memory (range = 0-10) were used. The variable was dichotomized based on the mean of the scores (<9/≥9).

Diseases.
A total of 18 self-reported diseases were recorded, in response to the following question: "Has a doctor ever told you that

Key Points
• Adverse childhood experiences (ACEs) have negative effects on adults' health.
• Early parental death (≤16 years) is an ACE with a relevant risk for dementia.
• Interventions are needed to counter the negative effects of ACEs on health.
you had [or do you currently have] any of the conditions listed on this sheet? By this we mean that a doctor has told you that you have this condition, and that you are either currently being treated for or bothered by this condition. The conditions are heart attack, hypertension, cholesterol, stroke, diabetes, chronic lung disease, cancer, stomach or duodenal ulcer, Parkinson disease, cataracts, hip or femoral fracture, other fractures, emotional disorders, rheumatoid arthritis, osteoarthritis, chronic kidney disease and dementia." The sum of the 18 diseases was also used as a variable.
Personality traits. The short version of the Big Five Inventory (BFI-10) comprising 10 items was used to identify traits of Extraversion, Agreeableness, Conscientiousness, Neuroticism and Openness. 35

| Statistical analysis
The presence in adulthood of a total of 18 diseases was analysed according to the presence or absence of parental death in each age range. The frequency (mean and SD) was examined, and a Student's ttest was applied. A multivariate binomial logistic regression analysis was performed to analyse the degree of association between the adulthood diseases and parental death (≤16 years).
To establish the periods in which parental death presented significant results with respect to dementia, we performed a frequency analysis using the χ 2 test to calculate the differences in each of the age groups. The analysis was performed for the death of the father, the mother, the father or mother and for the two together. A logistic regression analysis was performed for each of the age groups with parental death, indicating the crude and adjusted values for the variables age, gender and years of schooling.
Finally, a bivariate analysis was performed to identify the degree of association between the different variables and the groups with or without dementia. This was completed with a multivariate binomial logistic regression analysis with dementia status as dependent variable.
We also calculated the effect size, using Cohen's d for the continuous variables (weak <0.5; moderate 0.5-0.8 or strong >0.8) and Cramer's V for the categorical variables (weak <0.10, moderate 0.30 or strong 0.50). 36 For odds ratios, we used the values consistent with

| Sample description
The sample with data on parental death consisted of 65 997 people, of whom 1431 (2.2%) had a diagnosis of dementia. Only 3754 subjects had two living parents, and only six (0.3%) of these had dementia. Data on parental death were not available in 6249 subjects, and 410 (6.6%) had a diagnosis of dementia (see Figure 1).
In all variables, the group with living parents had the most favourable data (younger, more education and cognition, less disease), and the group who had lost their parents had the least favourable.
The complete data of the three groups can be found in Table S1.

| Dementia according to the age of participants at time of parental death
Previous analyses (Table 2) justify focusing the study on the association of early parental deaths on dementia. Table S2 shows the frequencies of the presence/absence of dementia in each of the age groups, differentiating between the deaths of the father, mother, either of the two or both. Looking at the early deaths of the father or mother (third column), it can be seen that the rates of dementia compared to no dementia are higher until the age of 16 years. The data at the age of 0 to 5 years are dementia, n = 95, 6.6%; no dementia, n = 2596, 4.0%; χ 2 = 24.5, P < .001. At participant age 10 to 11, the data are dementia n = 82, 5.7%; no dementia, n = 2137, 3.3%; χ 2 = 25. The data for ages 17, 18, 19 and 20, analysed individually, did not differ significantly and so they were combined in a single age group (17-20 years). No significant differences were found here either: dementia n = 72, 5.0%; no dementia, n = 2844, 4.4%; χ 2 = 1.3, P = 0.254. In the age group of 21 years and over, the proportions were reversed as there were significantly more cases without dementia: dementia n = 1073, 75.0%; no dementia, n = 53 984, 83.6%; χ 2 = 75.3, P < .001. Table 3 shows the binary logistic regressions with the odds ratios for the different age groups. The crude ORs of the deaths of the father or mother together (third column), for age groups 0 to 5, 6 to 10, 11 to 16 and 0 to 16 were adjusted with the variables age, gender and years of schooling they continued to be significant, although their values were slightly lower: 1.32, 1.51, 1.56 and 1.54 (P < .05), respectively. In the data for "both parents" (fourth column), the results were not significant due to the small sample size, and older age it was the variable that made the father's death non-significant in the adjusted model (0-5b years). The complete data are shown in Table 3.

| Factors associated with early parental death in people with and without dementia
In the bivariate analysis, the most notable differences between participants with and without dementia with regard to early parental death

| Early parental death and diseases
Many studies have linked parental death with depression. 7,8 However, our data also confirm the negative effects on adult physical health. 6,[14][15][16][17][18] The study by Neeleman et al (2002) 15 analysed 20 chronic diseases and found a higher risk in adults whose parents had died before they were 16 years old.

| Early parental death and dementia
In our study, in all age groups up to 16 years, parental death (death of the father, mother or either father or mother) is associated with an increased risk of dementia, with moderate odds ratios. These results corroborate those of other studies 19,25 ; however, in other studies a greater risk was only observed in some age groups [22][23][24]26 or for the death of the mother 20 or the father, 22 with higher odds ratios than in our analysis. This heterogeneity in the results may be due to small sample sizes.
The greater risk was maintained in the models adjusted for age, gender and education. In the multivariate regression, risk associated with early parental death remained significant even after adding the possible confounding variables age, neuroticism, years of schooling and income. Neuroticism has already been documented as a risk factor for dementia 39 and the loss of a parent may be associated with lower academic achievement and income. 10 Some studies have found differences in relation to participant gender 20,23 ; however, gender was not relevant in our analysis.

| Biomarkers of adverse childhood experiences
A recent review 40 identified the biomarkers related to adverse childhood experiences: inflammatory processes (IL-6, interleukin; CRP, Creactive protein), cardio-metabolic systems (blood pressure), genetics (length of telomeres) and endocrine systems (cortisol).
Norton et al (2017) 29 found that people with two or more family deaths in childhood were 79% more likely to have high CRP in old age, emphasizing the relationship between childhood stress and inflammatory processes in old age. Multiple adverse childhood experiences, including parental death, have also been associated with higher blood pressure, 3 higher levels of IL-6 and shorter telomeres. 41  adolescents. In the last case, the programmes incorporate parents or caregivers into the treatment, highlighting the importance of warmth and supportiveness and a positive family upbringing.

| Strengths and limitations
A strength of this study is the sample size (65 997 participants and 1431 cases of dementia). This sample size is the largest in the studies carried out to date on early parental death and dementia. However, the study has some limitations that should be mentioned.
First, we found few data on parents' backgrounds, especially with regard to health and illness. Second, we could not rule out an information bias, because information on medical diagnoses was selfreferred. Third, the number of lost cases with dementia and without data on parental death was relatively high (6.6%). A fourth limitation is the impossibility of analysing the coexistence of other ACEs with early parental death (such as abuse, neglect, drug addiction, etc.).