Association between life events and later depression in the population-based Heinz Nixdorf Recall study-The role of sex and optimism.

BACKGROUND: The association between life event stress and depressive symptoms has not been analyzed in the general population before.
METHODS: In the population-based Heinz Nixdorf Recall study, we assessed the association of 1.) the presence of important life events and 2.) life event stress, with the amount of depressive symptoms in univariable linear regressions and in multivariable regressions adjusted for age and sex (model 1) and age, sex and optimism as important determinants of coping with life events (model 2). Presence of life events and life event stress were assessed with the Social Readjustment Rating Scale (SRRS), optimism with the Life Orientation Test-Revised (LOT-R), and depressive symptoms with the 15-item Center for Epidemiological Studies Depression Scale (CES-D).
RESULTS: Of the total cohort of 4,814 participants, 1,120 had experienced important life events during the previous 6 months. Presence of important life events was significantly associated with higher CES-D scores (B = 2.6, 95%CI = 2.2 to 3.0, p < .001; model 2) compared to absence of life events. Associations were stronger for women than for men and for pessimists than for optimists. Among the participants with important life events, median (Q1; Q3) stress-score was 45.0 (39.0; 63.0). Stress-scores >Q3 were significantly associated with higher CES-D scores (2.2, 1.1 to 3.3, < .001) with a stronger association in pessimists than in optimists.
CONCLUSIONS: Experiencing life-changing events is associated with depression. Women and individuals with pessimistic personality are especially vulnerable which should be considered in prevention strategies.

1. Introduction

It has consistently been demonstrated that negative life events often precede the onset of major depressive episodes [1]. Complex study designs such as longitudinal analyses of population-based samples including dizygotic and monozygotic twins have made it possible to conclude that life events actually trigger, that is causally influence depressive reactions, instead of just being symptoms of depression [2].

Previous research investigating the association between life events and depression was mostly performed in case-control studies using fixed lists of life events. These studies showed that patients diagnosed with depression reported more life events and especially more undesirable life events than non-depressed controls from the general population [3], hospital staff [4], or patients with nonpsychiatric disease [5].

Fewer studies were performed in general population samples. These studies mostly compared depressive participants identified by scores above the cut-off in established depression screenings such as the Center for Epidemiologic Studies Depression Scale (CES-D) or Beck Depression Inventory (BDI) with non-depressive participants identified by scores below the cut-off in these depression screenings. So far, the largest population-based study on the cross-sectional association between life events and depression was performed within the Outcome of Depression in Europe Network. In 8,787 participants from 5 European countries, depressive participants (defined by BDI score >12), reported a higher number of life events (assessed with the List of Threatening Experiences including 12 negative life events) than non-depressive participants [6]. Also, in a smaller sample of the Kuopio Depression Study including 1,339 participants from the Finnish general population, depressive participants (defined as BDI score >9) reported a higher number of life events (assessed with a list of 12 negative life events similar to the List of Threatening Experiences) than non-depressive participants [7]. The minority of population-based studies assessed depression with structured diagnostic interviews because–compared with self-report screenings–structured diagnostic interviews require a lot more time and psychiatrically trained staff and thus are often not feasible in large population-based studies designed to address multiple research questions. Nonetheless, Assari and Lankarini assessed 12 months major depressive episode with a modified version of the World Mental Health Composite International Diagnostic Interview, a fully structured diagnostic interview, in a large population-based sample of American adults comprising 5,008 Blacks (African-Americans or Caribbean Blacks), and 891 Non-Hispanic Whites from the National Survey of American Life. They observed a significant association between the mean number of life events (assessed with a list of 8 stressful life events adequate for multiethnic samples) and major depressive episode independent of race [8]. Similarly, analyses of the Americans’ Changing Lives study showed that the experience of an event from a list of 11 stressful life events was significantly associated with major depressive episode, also assessed with structured diagnostic interview in a population-based sample of 1,024 men and 1,800 women. When the life events were analyzed separately, especially death of a spouse or child, death of a friend or relative, and divorce or marital/love problem were significantly associated with major depressive episode [9].

Even though the above-mentioned studies showed that depressives more often report life events than non-depressed controls, not every individual who experiences major life events becomes depressed [10]. Research aiming to understand why some persons develop depression whereas others do not has directed attention to (a) whether life events differ in their stress-inducing properties and (b) whether life events have a differential impact across subgroups of individuals. Early studies could show that certain life events, such as death of a close person or change in relationships, are consistently judged to be more stressful than others [11, 12]. There is an ongoing debate about how to measure the impact of life events. The Social Readjustment Rating Scale (SRRS) developed by Holmes & Rahe (1967) is a widely used tool in research on the relationship between life events and various types of illness. The SRRS is based on the observation that life events per se are stressful because they require adjustment of an individual’s life regardless of the desirability of the event [13]. Therefore, both desirable (positive) and undesirable (negative) life events are combined in the life stress-score, which has the advantage that it represents a continuous measure offering better statistical properties than presence/absence of short lists of life events. The SRRS has not been used in population-based studies on the association between life events and depression yet. So far, it has only been used in 3 studies with depressive patients [14-16]. Two of these studies including 90 major depressive patients and 121 controls, and 79 major depressive patients and 102 controls, respectively, showed that depressives had higher stress-scores than controls [14, 16]. One large study including 10,257 depressive patients observed higher stress-scores to be associated with a higher number of depressive episodes and depression severity [15].

There is still limited knowledge on moderators of the life events–depression association [8]. Various factors in addition to life events have been shown to increase the risk of depression and should be analyzed as possible moderating factors in life event research [1]. Despite considerable age- and sex differences in the prevalence of depression, only few studies analyzed the influence of sex [6, 8, 9] and age [17-19] on the association between life events and depression. Regarding the influence of sex, previous evidence is heterogenous with the analysis of Outcome of Depression in Europe Network data observing no significant interaction between sex and negative life events [6], the National Survey of American Life study stronger associations for men than for women [8] and the Americans’ Changing Lives study stronger associations for women than for men [9]. Previous evidence mostly does not support a significant moderating effect of age on the association between number of life events or presence of specific life events and depression, except for the events of maternal loss or being unmarried, which put younger persons at higher risk [17-19]. Due to this scarce and heterogenous evidence on factors moderating the association between life events and depression, we decided to focus on age- and sex differences in the association between life events and depression and also analyze the influence of optimism, which has been shown to be associated with a more favorable way of coping with life events [20-22], leading to lower rates of depression [22]. Further, no population-based study has analyzed life events using a continuous life event stress-score yet. From a clinical point of view, subclinical depression is especially important because it is associated with an increased risk of developing major depression and because it already represents a condition with significant psychological difficulties and need for treatment. To close the gaps of knowledge, we quantified life events with the SRSS, which offers a continuous score of life-event stress and analyzed its relationship with depressive symptoms in a large random sample of the adult German population. In addition, we analyzed a possible effect modification by age, sex, and optimistic personality.

2. Methods

2.1 Participants

Data was drawn from the baseline examination of the Heinz Nixdorf Recall study, a prospective population-based study focusing on risk factors for cardiovascular disease and mortality. A random sample of men and women aged 45–75 years were enrolled via mandatory citizen registries in Essen, Bochum, and Mülheim/Ruhr between December 2000 and August 2003 and received 2 follow-up examinations after 5 and 10 years. The study design has been described in detail elsewhere [23]. The total cohort included 4814 participants (50.2% female) with a mean age of 59.6 years and a standard deviation of 7.8 years. The present analysis only uses the baseline data. The study was approved by the ethical committee of the University of Duisburg-Essen, Germany. All participants gave written informed consent and all methods were carried out in accordance with the relevant guidelines and regulations of the ethical committee and the 1964 Helsinki declaration and its later amendments.

2.2 Measures

2.2.1 Social Readjustment Rating Scale (SRRS) for the assessment of life events and life event stress

Within a self-administered paper-pencil questionnaire, participants answered the question whether they had experienced any life events during the previous 6 months which were very important or life-changing (such as death, severe illness of a close person, career change, separation, move). For this, they had to tick a box for “No” or “Yes”. If the participants answered that they had experienced an important or life-changing event during the previous 6 months, that is if they ticked “Yes”, they were asked to describe the event in an open response format, meaning that they could freely write down these events on several lines. One rater evaluated all participant responses and scored them according to the Social Readjustment Rating Scale by Holmes and Rahe [11]. A small sample (n = 20 participants who reported life events) was rated by a second rater to determine interrater reliability. The SRRS represents the checklist approach and consists of 43 stressful life events generated based on clinical research to characterize the events that most often occurred to patients before seeking treatment. The SRRS is usually presented as a checklist and participants check these events if they have experienced them previously. However, the open response format used in our study offers more flexibility and is less time-consuming for the participants. Based on normative data, weights (life change units, LCUs) were assigned to each life event by the rater ranging from 11 to 100 to describe life event severity. Life event severity is characterized by the stress elicited by life events due to the required changes in usual activities (readjustment). In the process of the construction of the SRRS by Holmes and Rahe, the LCU for each event was determined by average ratings of a large group of subjects who rated all events regarding the amount of social readjustment that the experience of each life event required. The event “marriage” was assigned a value of 500 as an arbitrary anchor point, events needing more readjustment should be given higher ratings and events needing less readjustment lower ratings. Mean values were obtained for each event and divided by the constant of 10 to achieve a handy average amount of social readjustment required by the events. These values are termed LCUs and are summed up to a total life stress-score. LCUs of all life events experienced by each study participants were summed up to create a total score of life stress. Higher total scores represent a higher amount and duration of change in the participant’s accustomed pattern of life resulting from various life events, regardless of their desirability. A high interrater reliability for the SRRS score was achieved (interrater correlation = 0.86).

2.2.2 Life Orientation Test–Revised (LOT-R) for the assessment of optimism

Optimism was assessed with the Life Orientation Test-Revised [24]. Optimism is regarded as a stable personality trait. Optimists expect things to go well and believe that future outcomes will be good rather than bad [25]. Optimism has been associated with a more favorable way of coping with negative life events [21, 22] and better adjustment towards important life transitions [20], leading to lower rates of depression and better overall mental and physical health [22]. The LOT-R consists of 10 items, 3 assessing optimism, 3 assessing pessimism and 4 fillers. The filler items were not included in the present study. Each item is scaled on a 5-point Likert scale with responses ranging from 0”strongly agree” to 4 “strongly disagree”. After reversing the responses on the statements assessing pessimism, all responses to the 6 items are summed up. The total LOT-R score can range from 0 to 24 with higher values representing a higher disposition of optimism.

2.2.3 Center for Epidemiological Studies Depression Scale (CES-D) for the assessment of depressive symptoms

Depressive symptoms over the preceding week were assessed by self-administered questionnaire through the 15-item Center for Epidemiologic Studies Depression scale (CES-D) [26]. The CES-D is a screening tool for measuring depressive symptoms, which was specially designed for the use in non-clinical epidemiological populations but also validated in different psychiatric and psychosomatic clinical samples. Each item asks for feelings or behaviors during the previous week and is scaled on a 4-point scale with responses ranging from 0”rarely or none of the time (less than 1 day)” to 4 “most or all of the time (5–7 days)”. Possible scores for the 15-item version range from 0 to 45, with higher levels indicating more or more frequent depressive symptoms. The CES-D is considered an indicator of a probable depressive episode but is not equivalent to a face-to-face physician diagnosis. The assessment of signs of depression is particularly important, since they show a higher prevalence compared to the clinical depression diagnosis according to Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Statistical Classification of Diseases and Related Health Problems (ICD). Further, signs of depression represent a frequent comorbidity and exert an important influence on the recovery from somatic and psychosomatic diseases.

2.3 Statistical analysis

Continuous data are presented as mean (standard deviation) for normally distributed or median (Q1,Q3) for non-normally distributed data, categorical data are shown as number (%). Statistical comparisons between two independent groups (e.g., men vs women), were done by student’s t-test for normally distributed continuous data, by Mann-Whitney test for non-normally distributed continuous data, and by Chi-square test or Fisher’s exact test for categorical data. Associations between life events (categorical as yes vs no or—in case participants reported important life events -, as SRRS-Score >Q3 vs ≤Q3) and depressive symptoms (continuous CES-D score) were analyzed by unadjusted linear regression analyses and multivariable linear regression analyses adjusted for age (>65 vs ≤65 years) and sex (male vs female, model 1), and additionally for optimism (LOT-R-Score >Q3 vs ≤Q3, model 2). Interactions between life events and adjusting variables were analyzed as multiplicative interactions. All hypothesis tests used two-sided tests, and p-values < .05 were considered statistically significant. Missing values were excluded listwise. All analyses were done with IBM SPSS Statistics 21 for Windows (IBM Corporation, Armonk, NY, USA).

3. Results

3.1 Presence of important life events

Valid data on the presence of important or life-changing events during the previous 6 months was available for 4,664 (96.9%) of the total cohort of 4,814 participants. Of those 4,664 participants, 1,120 (24.0%) reported that they had experienced such an important life event. Among those 1,120 participants who experienced an important life event during the previous 6 months, 1,080 (96.4%) recalled one or more concrete events in the subsequent free response format (Fig 1).

Fig 1

Flowchart of the Heinz-Nixdorf Recall analyses samples.

SRRS, Social Readjustment Rating Scale.

Most of those 1,080 participants recalled only one important life event which was listed in the SRRS (n = 925, 85.6%), 135 participants (12.5%) recalled two events, and only very few participants reported three (n = 18, 1.7%) or four events (n = 2, 0.2%). Change in health of a family member (n = 310, 24.7%), death of a close family member, (n = 210, 16.7%), and personal injury or illness (n = 124, 9.9%) were the most frequent of all reported life events (Table 1).

Table 1

Frequency of important life events according to the SRRS.

Life event	Life Change Unit	Number (%)
Change in health of family member	44	310 (24.7)
Death of close family member	63	210 (16.7)
Personal injury or illness	53	124 (9.9)
Change in work hours or conditions	20	73 (5.8)
Change in residence	20	69 (5.5)
Death of close friend	37	59 (4.7)
Retirement	45	56 (4.5)
Change in living conditions	25	42 (3.3)
Death of spouse	100	38 (3.0)
Fired at work	47	36 (2.9)
Marital separation	65	33 (2.6)
Change in financial state	38	31 (2.5)
Gain of new family member	39	30 (2.4)
Change in number of arguments with spouse	35	26 (2.1)
Change to different line of work	36	25 (2.0)
Trouble with in-laws	29	25 (2.0)
Divorce	73	22 (1.8)
Son or daughter leaving home	29	22 (1.8)
Change in responsibilities at work	29	9 (0.7)
Trouble with boss	23	5 (0.4)
Wife begin or stop work	26	3 (0.2)
Marriage	50	2 (0.2)
Revision of personal habits	24	2 (0.2)
Jail term	63	1 (0.1)
Marital reconciliation	45	1 (0.1)
Mortgage over $10,000	31	1 (0.1)
Outstanding personal achievement	28	1 (0.1)
Change in number of family get-togethers	15	1 (0.1)

SRRS, Social Readjustment Rating Scale. Frequencies shown in descending order.

3.2 Life event stress

As explained before, the SRRS offers the possibility to create a total score (stress-score), with higher scores representing a higher amount and duration of change in the participant’s accustomed pattern of life resulting from the experience of various life events. The stress-score was not normally distributed but right-skewed, most participants had low scores (S1 Fig in S1 File). Of the 1080 participants, who recalled at least one concrete life event during the previous 6 months, the median stress-score was 45.0 (Q1 = 39.0, Q3 = 63.0), the minimum was 20.0, the maximum 182.0. Stress-scores >Q3 were defined as high life event stress, stress-scores ≤Q3 as low life event stress.

3.3 Optimism

Valid data on the personality trait of optimism assessed via the sum score of the LOT-R was available for 4,687 (97.4%) of the total cohort of 4,814 participants. Total LOT-R scores were not normally distributed but slightly skewed to the left with only few participants exhibiting low optimism scores (S2 Fig in S1 File). The median LOT-R score was 15.0 (Q1 = 13.0, Q3 = 18.0), the minimum was 0.0, the maximum 24.0. LOT-R scores >Q3 were defined as optimism, LOT-R scores ≤Q3 as pessimism.

3.4 Depressive symptoms

Valid data on depressive symptoms assessed via the total score of the 15-item version of the CES-D was available for 4,645 (96.5%) of the total cohort of 4,814 participants. Total CES-D scores were not normally distributed but right-skewed, most participants had low depression scores (S3 Fig in S1 File). The median depression score was 7.0 (Q1 = 4.0, Q3 = 11.0). The minimum depression score was 0.0, the maximum 42.0.

3.5 Association between presence of important life events and depressive symptoms

Participants who reported that they had experienced an important life event during the previous 6 months (n = 1,120) had significantly higher depression scores (Median = 8, Q1 = 5, Q3 = 14) than participants who did not report the experience of an important life event during the previous 6 months (n = 3546, Median = 6, Q1 = 3, Q3 = 10, p < .001; Fig 2).

Fig 2

Box and whisker plots demonstrating Center for Epidemiological Studies Depression Scale (CES-D) depression score for presence and absence of important life events in the total cohort and stratified by age (young ≙ ≤65 years vs old ≙ >65 years), sex (men vs women), and optimism (optimists ≙Life Orientation Test-Revised (LOT-R) score >18 vs pessimists ≙ LOT-R score ≤18).

The horizontal line in each box represents the median with the box representing the interquartile range (Q3-Q1) and the whiskers representing the total range of the data (max-min).

In an unadjusted linear regression, the presence of important life events was associated with a significant increase in CES-D score (B = 2.6, 95%CI = 2.2 to 3.0, p < .001). In multivariable regressions adjusted for age and sex as well as age, sex, and optimism, presence of important life events remained significantly associated with higher depression score (Table 2). Presence of important life events, age, sex, and optimism explained 10.8% of the total variation in depression score.

Table 2

Association between presence of important life events (yes vs no) and depressive symptoms.

	Unadjusted			Model 1			Model 2
	B	95% CI	P	B	95% CI	P	B	95% CI	p
Total cohort	2.6	2.2 to 3.0	< .001	2.6	2.2 to 3.0	< .001	2.5	2.1 to 2.9	< .001
>65 years	2.7	1.9 to 3.6	< .001	2.6	2.1 to 3.0	< .001	2.6	1.8 to 3.4	< .001
≤65 years	2.7	2.2 to 3.1	< .001	2.6	1.8 to 3.5	< .001	2.5	2.0 to 2.9	< .001
Men	1.9	1.4 to 2.5	< .001	1.9	1.4 to 2.5	< .001	1.9	1.3 to 2.4	< .001
Women	3.0	2.4 to 3.6	< .001	3.1	2.5 to 3.7	< .001	3.1	2.5 to 3.7	< .001
Optimists	1.2	0.4 to 1.9	.002	1.1	0.4 to 1.8	.002	1.1	0.4 to 1.8	.002
Pessimists	2.9	2.4 to 3.3	< .001	2.8	2.3 to 3.2	< .001	2.8	2.3 to 3.2	< .001

B = unstandardized regression weights from linear regression analysis, CI = confidence interval, Model 1 adjusted for age (>65 vs ≤65 years) and sex (male vs female), Model 2 adjusted for age, sex, and optimism (score >Q3 vs ≤Q3); in the stratified analyses not adjusted for the stratification variable.

3.6 Interacting influence of the presence of important life events with age, sex, and optimism on depressive symptoms

The interaction analyses of the presence of important life events with the adjusting variables age, sex, and optimism on depressive symptoms showed a significant interaction between life events and sex (B = -0.8, 95%CI = -1.6 to -0.1, p = .043) and between life events and optimism (B = -0.7, 95%CI = -0.8 to -0.6, p < .001) in the fully adjusted model 2. These interactions revealed the stronger influence of the presence of life events on women compared to men and on pessimists compared to optimists.

In absolute numbers, men who reported to have experienced an important life event during the previous 6 months (21.2%), had significantly higher depression scores (Median = 7, Q1 = 4, Q3 = 11) than men who reported not having experienced such an event (78.8%, Median = 6, Q1 = 3, Q3 = 9, p < .001). However, this difference was more pronounced for women: women who reported to have experienced an important life event during the previous 6 months (n = 26.8%), had significantly higher depression scores (Median = 9, Q1 = 5, Q3 = 15) than women who reported not having experienced such an event (73.2%, Median = 7, Q1 = 4, Q3 = 11, p < .001, Fig 2).

Optimists, who reported to have experienced an important life event during the previous 6 months (22.4%), had significantly higher depression scores (Median = 4, Q1 = 2, Q3 = 8) than optimists who reported not having experienced such an event (77.6%, Median = 4, Q1 = 1, Q3 = 7.0, p = .005). This difference was more pronounced for pessimists: pessimists who reported to have experienced an important life event during the previous 6 months (24.4%), had significantly higher depression scores (Median = 9, Q1 = 5, Q3 = 15) than pessimists who reported not having experienced such an event (75.6%, Median = 7, Q1 = 4, Q3 = 11, p < .001, Fig 2).

Presence of an important life event and depression scores stratified by age, sex, and optimism are shown in S1 Table in S1 File. Presence of all important life events of the SRRS stratified by age, sex, and optimism are shown in S2-S4 Tables in S1 File.

3.7 Association between life event stress and depressive symptoms

Among the 1,120 participants who reported to have experienced an important life event during the previous 6 months, 1,080 reported a concrete life event in the free response format (Fig 1). These concrete life events were scored according to the SRRS [11] and a stress-score was created to describe the readjustment of each individual’s life, which resulted from the experience of these life events.

Participants with high life event stress, defined as stress-score >Q3 (n = 290) had significantly higher depression scores (Median = 10, Q1 = 5, Q3 = 16) than participants with low life event stress, defined as stress-score ≤Q3 (n = 871, Median = 8, Q1 = 4, Q3 = 13, p < .001, Fig 3).

Fig 3

Box and whisker plots demonstrating Center for Epidemiological Studies Depression Scale (CES-D) depression score for Social Readjustment Rating Scale (SRRS) total score >Q3 and ≤Q3 (≙>63 vs ≤63), representing high and low life event stress.

Values are shown for the total cohort and stratified by age (young ≙ ≤65 years vs old ≙ >65 years), sex (men vs women) and optimism (optimists ≙Life Orientation Test-Revised (LOT-R) score >18 vs pessimists ≙ LOT-R score ≤18). The horizontal line in each box represents the median with the box representing the interquartile range (Q3-Q1) and the whiskers representing the total range of the data (max-min).

In an unadjusted linear regression, high life event stress was associated with a significant increase of depression score (B = 2.2, 95%CI = 1.1 to 3.3, p < .001). Even in multivariable regressions adjusted for age and sex as well as age, sex, and optimism, high life event stress was significantly associated with higher depression score (Table 3). Life event stress, age, sex, and optimism explained 10.2% of the total variation in depression score.

Table 3

Association between life event stress (score >Q3 vs ≤Q3) and depressive symptoms.

	Unadjusted			Model 1			Model 2
	B	95% CI	p	B	95% CI	p	B	95% CI	p
Total cohort	2.2	1.1 to 3.3	< .001	1.9	0.8 to 3.1	< .001	1.9	0.8 to 2.9	< .001
>65 years	-0.3	-3.3 to 2.8	.87	-1.2	-4.2 to 1.8	.42	-1.3	-4.2 to 1.6	0.39
≤65 years	2.6	1.4 to 3.8	< .001	2.4	1.2 to 3.6	< .001	2.3	1.1 to 3.4	< .001
Men	2.2	0.6 to 3.9	.007	2.2	0.6 to 3.8	.007	2.2	0.6 to 3.8	.006
Women	1.6	0.1 to 3.1	.032	1.8	0.3 to 3.3	.022	1.6	0.2 to 3.1	.029
Optimists	0.9	-1.0 to 2.9	.34	0.9	-1.0 to 2.9	.34	0.9	-1.0 to 2.9	.34
Pessimists	2.3	1.1 to 3.6	< .001	2.0	0.8 to 3.2	.001	2.0	0.8 to 3.2	.001

B = unstandardized regression weights from linear regression analysis, CI = confidence interval, Model 1 adjusted for age (>65 vs ≤65 years) and sex (male vs female), Model 2 adjusted for age, sex, and optimism (score >Q3 vs ≤Q3); in the stratified analyses not adjusted for the stratification variable.

3.8 Interacting influence of life event stress with age, sex, and optimism on depressive symptoms

The interaction analyses of life event stress with the adjusting variables age, sex, and optimism on depressive symptoms showed a significant interaction between life event stress and optimism (B = -1.0, 95%CI = -1.2 to -0.7, p < .001) in the fully adjusted model 2. This interaction revealed the stronger influence of life event stress on pessimists compared to optimists.

In absolute numbers, optimists with high life event stress (n = 18.7%), had significantly higher depression scores (Median = 6, Q1 = 3, Q3 = 9) than optimists with low life event stress (81.3%, Median = 4, Q1 = 2, Q3 = 8, p < .001). This difference was more pronounced for pessimists: pessimists with high life event stress (19.7%), had significantly higher depression scores (Median = 11, Q1 = 6, Q3 = 18) than pessimists with low life event stress (80.3%, Median = 8, Q1 = 5, Q3 = 15, p < .001, Fig 3).

Life event stress and depression scores stratified by age, sex, and optimism are shown in S1 Table in S1 File.

Exclusion of participants with a history of coronary heart disease (n = 327) did not change our results to a relevant degree (not shown).

4. Discussion

4.1 Principal findings

For the first time, we observed a significant association between the presence of important life events and depressive symptoms in the middle-aged to old-aged general population. In participants who experienced important life events, higher life event stress-scores due to the necessary life adjustment after important life events were significantly associated with higher depression scores. Women and individuals with pessimistic personality were more susceptible to the adverse influence of life events on depression than men and individuals with optimistic personality.

4.2 Comparison with other studies

Presence of important life events, life event stress and their association with depressive symptoms have not been assessed in the general adult population before. So far, the largest population-based study on the association between life events and depression was performed within the Outcome of Depression in Europe Network [6]. In 8,787 participants from 5 European countries, presence of 12 negative life events during the previous 6 months, included in the List of Threatening Experiences, and depression, defined by BDI score >12, were assessed. A higher number of negative life events was significantly associated with increased rates of depression. In contrast to our study, there was no significant interaction between sex and negative life events with respect to depression. Like in our study, illness of a relative and death of close friends/relatives were experienced most often. While we focused on the idea that not only negative, but also positive life events can have an influence on an individual’s emotional state, because positive events can require changes to an individual’s life which are experienced as stressful, this study only focused on negative life events. Further, it did not assess the stress-potential of each life event like we did, but only analyzed the number of negative life events as 0, 1, 2, ≥3. Also, in a smaller study including 1,339 participants from the Finnish general population, life events were assessed with a list of negative life events similar to the List of Threatening Experiences and depression with BDI score but with a cut-off of 9 [7]. Again, only the number of life events was analyzed with the mean number of life events in the depressive group being significantly higher than in the non-depressive group. Using a structured diagnostic interview to assess depression, the large population-based National Survey of American Life comprising 5,008 Blacks and 891 Non-Hispanic Whites observed a significant association between the mean number of life events (assessed with a list of 8 stressful life events adequate for multiethnic samples during the previous 30 days) and major depressive episodes during the previous year [8]. In contrast to our study, the association between stressful life events and major depressive episodes was stronger for men than for women. Reasons for the divergent results might be 1.) the less comprehensive assessment of life events with a list of only 8 events compared to the free response format in our study and coding according to the SRRS which includes 43 life events, and 2.) the assessment of clinical major depressive episodes compared to our assessment of subclinical depressive symptoms which was developed for the use in population-based studies. Men tend to report different life events than women (S3 Table in S1 File) and as operationalized in the SRRS, different life events have different stress-inducing properties having a different probability to induce clinical vs subclinical depression. Moreover, specific life events are more adverse for one sex than another, i.e., job loss is usually more severe for men than for women while loss of a family member is usually more severe for women than for men [9]. With short life event checklists, there is a higher probability of having an overrepresentation of life events more often experienced by one sex or overrepresentation of life events leading to clinical vs subclinical depression. The population-based Americans’ Changing Lives study including 1,024 men and 1,800 women also showed a significant association between major depressive episode, again assessed with structured diagnostic interview and the experience of an event from a list of 11 stressful life events [9]. In line with our results, this association was stronger for women than for men. Regarding single life events, death of a friend or relative was experienced most often, confirming the observations of our study. In contrast to our study, life event stress was not examined.

To the best of our knowledge, previous evidence on the moderating effect of age on the association between life event stress and depression is scarce. In our study, age did not moderate the association between life events and depression. Subjects ≤65 years reported the presence of an important life event more often than subjects >65 years, however, life event stress and severity of depressive symptoms were not significantly associated with age. Life course epidemiology suggests that life events can have differential effects on depression, depending on the age when they occur. Therefore, future longitudinal studies should investigate age effects in more depth [27]. Previous studies with depressive and non-depressive subjects observed that younger participants reported more life events than older participants and that prevalence of depression decreases in older age [17] One Danish study, which used registry data to assess prevalence of specific life events (vital, marital and employment status) instead of directly asking participants, similar to our results observed no interaction between most of the life events assessed and age regarding the outcome of being admitted for the first time ever at a psychiatric hospital and discharged with a diagnosis of depression. Only for the life event “being unmarried”, a significant interaction was observed with younger persons having a higher risk to become diagnosed with depression [17]. Also a smaller case-control study including 64 depressive and 74 non-depressive patients observed no significant interaction between age and the number of reported life events [18]. For the specific category of loss events, one larger epidemiological study including 3,491 healthy individuals recruited through general practices, found the impact of maternal loss on the risk of developing depression to vary significantly by age, being highest in those younger at the time of loss, but no significant interaction for marital loss [19].

So far, only one smaller population-based Chinese study including 1,147 Hong Kong residents analyzed the influence of coping on the association between life events and mental health, but not specifically depression [28]. This study showed that resilience moderated the association between multiple traumatic life events and mental health (assessed by Short-Form 12 Health Survey). Resilient people were less susceptible to the negative influence of traumatic life events on mental well-being. This observation supports our results of a stronger influence of life events on individuals with pessimistic personality. Resilient people are characterized by optimism, positive coping and hardiness and can therefore be more flexible to cope with life challenges and adapt to life-changing events.

The SRRS [11] to assess life event stress caused by necessary adaptions of the previous life after the experience of important life events was so far applied only 3 times in the context of depression. In 2 case-control studies with 90 major depressive patients and 121 controls, and 79 major depressive patients and 102 controls, respectively, depressives had higher stress-scores than controls [14, 16]. In the smaller case-control study, stress-scores were not significantly associated with depression in multivariable logistic regression models [16]. In a large patient study comprising 10,257 patients with current single or recurrent major depressive episode, higher stress-scores were associated with a higher number of depressive episodes and depression severity [15].

4.3 Limitations

The design of our study possesses strengths and limitations. Important strengths are its large sample size and that it is representative of the general adult German population. Further, life events were comprehensively analyzed with the SRRS and depressive symptoms assessed with the CES-D representing a validated instrument specifically designed for the use in epidemiological studies. An important limitation is the cross-sectional design of the analysis, which does not allow to draw causal conclusions. The associations between life events and depression could be bidirectional, since not only the experience of life events can lead to depression but also depression can lead to the experience of negative life events. Furthermore, the retrospective recall of life events may be biased by the current emotional state or defense mechanisms. Due to the open response format, it is possible that women/men and optimists/pessimists used different criteria to determine what constituted an important life event. Other variables not included in the present study such as chronic stress could contribute to the experience of life events or influence their association with depression.

4.4 Clinical implications

Even if our results are influenced by reverse causality, i.e., depression leading to specific life events, this still has important clinical implications in a way that the experience of life events must be considered more in the prevention and treatment of depressive episodes. Identification of stressful life events and working on strategies to prevent recurrent depressive episodes resulting from the vicious cycle of depression and stressful life events represents an important part of anti-depressive treatment [29]. A study in depressive adolescents could already show that standard therapy was not able to reduce depressive symptoms in the presence of high life event stress. Thus, the authors suggested that adolescents suffering from high life event stress might require specialized therapy [30]. In line with this idea, anti-depressive therapy, which was able to reduce the experience of negative life events, led to long-term improvement in psychological well-being in adult depressive patients [31].

5. Conclusions

This is the first study exploring the presence of a comprehensive selection of life events as well as the stress caused by the necessary adaptation of the previous life due to the experience of these life events in a large population-based study. Our observation that individuals who experienced important life events and high levels of life event stress had higher levels of depressive symptoms and that women and persons with pessimistic personality are especially vulnerable has important clinical implications. Our results implicate, that the occurrence of major life events signals a period of increased risk for new onset depression and recurrent depressive episodes, which has to be considered in preventing the evolution of distress to disorder and in treating depressive disorders. Prevention programs should be developed to help individuals suffering from stressful life events to better cope with the life-changing situation. Women and persons with a pessimistic outlook on life should be especially targeted with preventive therapies.

(DOCX)

Click here for additional data file.

13 Dec 2021

Association between life events and later depression in the population-based Heinz Nixdorf Recall study – the role of sex and optimism

PLOS ONE

Dear Dr. Gronewold,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.   I plus one reviewer

have read your manuscript. We feel that it addresses an interesting question but all sections of the manuscript need to be revised before it can be accepted. In particular the introduction does not provide a sufficient review of previous related research discussing the relationship between depression and life  event experiences. Additional relevant research needs to be described and evaluated in more depth. Reference is also made to the stress arising from the COVID pandemic yet this paper does not address the impact of COVID. This point needs to be made clear or reference to COVID deleted. There is also insufficient justification for examining age as a moderator. The introduction mentions a lack of research on gender effects but the discussion mentions research that addresses this issue. Material in the discussion should also be in the introduction to create a better rationale for the study. See the comments of reviewer one for specific examples. The methods section lacks detail on numerous places. It is not clear when the study took place or how the experience of life events was measured. In the results section the difference on SRRS  and CED scores could analysed for the different subgroups. There are a number of issues concerning the discussion. The stance on causality  between life events and depression is not clear either in introduction or the discussion, but was particularly problematic  The discussion of  related findings is not well-organized, making it difficult to understand why the studies are grouped together. Additional points have been raised by Reviewer1. In your revision address all points raised by Reviewer 1 and the points outlined in this letter.

Please submit your revised manuscript by January 27, 2022. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Barbara Dritschel, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. One of the noted authors is a group or consortium Heinz Nixdorf Recall Study Investigative Group. In addition to naming the author group, please list the individual authors and affiliations within this group in the acknowledgments section of your manuscript. Please also indicate clearly a lead author for this group along with a contact email address.

5. Your abstract cannot contain citations. Please only include citations in the body text of the manuscript, and ensure that they remain in ascending numerical order on first mention.

6. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

7. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper explored the association between the experience of important life events in the previous six months (measured using the Social Readjustment Rating Scale; SRRS) and depressive symptoms in a large population-based sample. In particular, the study examined whether sex and optimism moderated the association between life events and depression. The results showed that (1) people who had experienced important life events in the previous 6 months exhibited a greater number of depressive symptoms than people who had not experienced an important life event in the previous 6 months; (2) in those who had experienced an important life event, people with high SRRS scores exhibited a greater number of depressive symptoms than those with lower scores; (3) the association between SRRS score and depressive symptoms was stronger in women than men; and (4) the association between SRRS score and depressive symptoms was higher in pessimists than optimists. The authors concluded that women and pessimists should be targeted by prevention programmes to help them cope with stressful life events, and that specialized therapy may be required.

Comments/questions

1. The introduction opens with a reference to ongoing COVID pandemic as an example of an important, life-changing event (lines 47-48), which sets the study up as though it will measure the effect of the pandemic on depressive symptoms. However, it appears that the data for this study were in fact collected many years before the pandemic began. It would be good to make clearer in the introduction that this study is not about the pandemic specifically.

2. In general, the introduction does not describe the context for the study in enough detail. Information about previous studies that have examined the association between life events and depression is scarce (refs 5-9 on lines 59-62), giving the impression that there is relatively little existing research in this area. For example, the authors state that previous studies have not generally examined the effect of sex on this relationship (lines 78-80). However, in the discussion, two sex effects in previous studies are mentioned for the first time (lines 332-334 and lines 341-342), and several more details about these and other previous studies are presented. These details were needed in the introduction, to make it clear how the current study builds on previous work.

3. In lines 78-81 and line 90, age is mentioned as a potential moderating factor and a main focus of this study. However, although age was included in the models, it was not discussed in any detail.

4. In the method section the authors state that the data for this study were drawn from a previous study for which participants were recruited around 20 years ago, with follow-ups at 5 and 10 years (lines 94-98). It is not clear when the data for this study were actually collected – at the outset of the original study, in a follow-up examination, or across multiple time points?

5. It is not entirely clear what instructions participants were given for the self-administered questionnaire (lines 104-108). Were the examples of important life events on line 106 provided to participants as examples, or were the participants left to decide on their own what constituted an important life event (this seems to be what is implied by “open response format” on line 108/113-114)? If left to decide on their own, is it possible that women/men and optimists/pessimists used different criteria to determine what constituted an important life event?

6. On lines 112-115 the authors argue that their open response format offers more flexibility than using a checklist of life events, yet participants’ responses were evaluated according to a checklist of events for which normative data on severity were available. It is not clear why the open response format is therefore more flexible. Perhaps the authors could expand on this idea. To what extent were participants left to self-determine what constituted an important life event?

7. Was there a reason why all of the life events were evaluated by only one rater (lines 108-110)? How can the authors be sure that the rater’s evaluations were reliable? Given the flexible open-response format, did any participants report life events that were not on the existing checklist of events? If so, how were these responses dealt with?

8. Since the comparisons of interest in this study were women vs. men and optimists vs. pessimists, it would have been helpful to see some data on the SRRS scores and CES-D scores in these different subgroups. Were optimists and pessimists equally likely to report important life events, for example?

9. The argument is made that both positive and negative life events can be experienced as stressful due to the need to readjust (e.g., lines 72-73, lines 126-127). Following on from point 6 above, is it possible that, for example, optimists were more likely to report positive important life events, and this could explain the weaker relationship with depressive symptoms?

10. In the abstract, the authors state that they examine the presence of important life events and the RESULTING life event stress (line 26). Similarly, elsewhere the argument is made that higher stress is associated with higher depression scores (e.g., line 310-312). However, since the study did not directly measure stress, it is not clear that the participants in this study experienced stress, only that these life events tend to be stressful. These claims could be worded a little more carefully, to reflect what was actually measured.

11. The authors’ stance on the causality of the relationship between life events and depression was unclear. In lines 52-55 of the introduction, they state that a robust and causal association between stressful life events and depression has become established, whereby stressful life events trigger depression. Later, it is acknowledged that the current study is unable to draw conclusions about causality, and that the authors rely heavily on other studies to conclude that stressful life events cause depression (line 376). The arguments in the conclusion then appear to depend on the causality evidenced by other studies, rather than on the data presented in the current study. Given that this study is apparently the first to measure the moderating effects of sex and optimism, how can we be confident that this assumption of causality is justified?

12. In the discussion section, there is a list (beginning on line 315) of several previous studies that have explored similar research questions, albeit using different scales, sampling methods, etc. Some of the findings of these previous studies were consistent with the current findings, while others were inconsistent. The list format is quite repetitive, and it is difficult to follow the argument throughout – what is the overall message? The argument would be clearer if the findings from the previous studies could be aggregated in a meaningful way, such as by the methods used, the time period sampled, etc.

13. It is not clear why a less comprehensive assessment of life events using a checklist of 8 events vs. free response would potentially reverse the sex effect (lines 334-335). This argument needs to be made more explicit. The same applies to the potential difference between clinical and subclinical depression (lines 336-338).

14. On lines 356-363, there is a section on previous uses of SRRS (refs 11, 12, 13) showing that depressives had higher stress scores (i.e., more/more serious life events) than controls in 2/3 studies, and that stress score associated with depression severity in one study. This information would be better in the introduction (see point 2, above).

15. On lines 378-380, the authors argue that even if the causality is reversed, such that depression leads to the experience of stressful life events, “this still has important clinical implications in a way that the experience of life events must be considered more in the prevention and treatment of depression”. This argument was difficult to follow; if depression causes stressful life events, how could consideration of the experience of life events prevent or treat depression?

16. A new study is introduced in the conclusion on lines 389-396. It would be better to include this argument in the main body of the discussion so that the conclusion remains focused on the current investigation.

Minor points

Line 84-85 reads “From the clinical point of view, especially subclinical depression is important because it is associated with…”; it would read better as “From a clinical point of view, subclinical depression is especially important because….”

Lines 152-155 – this sentence is quite long, and would read better with commas.

Lines 310-311 – “In case important life events were experienced” would read better as “In participants who experienced important life events”

Line 311-312 – “higher stress… was again significantly associated…” – the “again” appears to be redundant here

Line 313 – “Women and individuals with pessimistic personality were more susceptible to the negative influence of life events on depression”. The wording is ambiguous, as it seems to suggest there is a negative association between life events and depression, rather than the positive association that was observed.

Line 317 – “The so far largest population-based study” would read better as “So far, the largest population-based study”

Line 350 – “this study could show….” presumably means “this study showed that….”?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

7 Mar 2022

Gronewold J, et al.

‘Association between life events and later depression in the population-based Heinz Nixdorf Recall study – the role of sex and optimism’ (MS ID: PONE-D-21-02934)

Responses to editor and reviewers

Please note that all new/revised text is highlighted in yellow in the manuscript.

Editor

I plus one reviewer have read your manuscript. We feel that it addresses an interesting question but all sections of the manuscript need to be revised before it can be accepted. In particular the introduction does not provide a sufficient review of previous related research discussing the relationship between depression and life event experiences.

�  We would like to thank the editor and reviewer for the critical evaluation and appreciation. We added a more in-depth revision of previous evidence in the introduction and discussion.

Reference is also made to the stress arising from the COVID pandemic yet this paper does not address the impact of COVID. This point needs to be made clear or reference to COVID deleted.

�  We deleted the reference to COVID.

There is also insufficient justification for examining age as a moderator.

�  In addition to depression risk, also the experience of life events changes throughout life. Previous evidence regarding the moderating effect of age on the association between life event and depression is scarce and showed heterogeneous results with some studies observing no significant moderation while others observed moderation effects for specific life events such as maternal loss or being unmarried being more harmful at younger ages. We now added this information to the introduction and discussion section.

The introduction mentions a lack of research on gender effects but the discussion mentions research that addresses this issue.

�  Research on the influence of sex on the association between life events and depression is scarce and completely lacking for life event stress. We now clarified the current state of research in our revised introduction.

Material in the discussion should also be in the introduction to create a better rationale for the study

�  As stated before, we added a more in-depth review of previous evidence in the introduction and discussion.

The methods section lacks detail on numerous places. It is not clear when the study took place or how the experience of life events was measured.

�  The study is a cross-sectional analysis of baseline data of the Heinz Nixdorf Recall study, which was collected from 2000 to 2003. We added this information to the methods section and also added more details regarding the measurement of the experience of life events.

In the results section the difference on SRRS and CED scores could be analysed for the different subgroups.

�  We now show these results in Supplemental Table S1 and also show presence of single life events stratified by age, sex, and optimism in Supplemental Tables S2-4.

There are a number of issues concerning the discussion. The stance on causality between life events and depression is not clear either in introduction or the discussion, but was particularly problematic.

�  We now clarified this issue in our revised introduction and discussion: previous studies with more complex study designs like longitudinal analyses of population-based samples including dizygotic and monozygotic twins have made it possible to conclude that life events actually trigger, that is causally influence depressive reactions instead of just being symptoms of depression. However, these studies did not include continuous scores of life event stress and did not analyze moderators of the associations between life events and depression. We closed this gap of knowledge in our present study, but since our study is a cross-sectional observational study, we clarified that it is not suited to draw causal conclusions in our revised limitations section.

The discussion of related findings is not well-organized, making it difficult to understand why the studies are grouped together.

�  We reorganized our discussion and grouped studies by methods used to assess life events and depression. It now begins with:

Dalgard et al., who used a list of 12 negative life events and assessed depression with BDI, followed by

Honkalampi et al., who used a similar list of 12 negative life events and also BDI,

followed by the two studies assessing depression with structured diagnostic interview instead of screening, which are

Assari & Lankarani, who used a list 8 stressful life events and

Maciejewski et al., who used a list of 11 negative life events.

These studies were also inserted in the revised introduction section in the same order.

Reviewer #1

1. The introduction opens with a reference to ongoing COVID pandemic as an example of an important, life-changing event (lines 47-48), which sets the study up as though it will measure the effect of the pandemic on depressive symptoms. However, it appears that the data for this study were in fact collected many years before the pandemic began. It would be good to make clearer in the introduction that this study is not about the pandemic specifically.

�  We deleted the reference to COVID since our study is not about the pandemic specifically.

2. In general, the introduction does not describe the context for the study in enough detail. Information about previous studies that have examined the association between life events and depression is scarce (refs 5-9 on lines 59-62), giving the impression that there is relatively little existing research in this area. For example, the authors state that previous studies have not generally examined the effect of sex on this relationship (lines 78-80). However, in the discussion, two sex effects in previous studies are mentioned for the first time (lines 332-334 and lines 341-342), and several more details about these and other previous studies are presented. These details were needed in the introduction, to make it clear how the current study builds on previous work.

�  We now describe the context for our study in more detail in our revised introduction.

3. In lines 78-81 and line 90, age is mentioned as a potential moderating factor and a main focus of this study. However, although age was included in the models, it was not discussed in any detail.

�  To the best of our knowledge, there is no previous evidence on the moderating effect of age on the association between life event stress and depression, but there are a few studies examining the moderating effect of age on the association between number of life events or presence of specific life events and depression, which we now included in the revised introduction and discussion.

4. In the method section the authors state that the data for this study were drawn from a previous study for which participants were recruited around 20 years ago, with follow-ups at 5 and 10 years (lines 94-98). It is not clear when the data for this study were actually collected – at the outset of the original study, in a follow-up examination, or across multiple time points?

�  This study is based on a cross-sectional analysis of baseline data (collected between the years 2000 and 2003 at the outset of the study), which we now clarified in our revised method section.

5. It is not entirely clear what instructions participants were given for the self-administered questionnaire (lines 104-108). Were the examples of important life events on line 106 provided to participants as examples, or were the participants left to decide on their own what constituted an important life event (this seems to be what is implied by “open response format” on line 108/113-114)? If left to decide on their own, is it possible that women/men and optimists/pessimists used different criteria to determine what constituted an important life event?

�  The original instructions were as follows (original in German):

Translated into English, this would be: “Have there been any events in the past 6 months that were particularly important to you or that changed your life? (e.g. death or serious illness of a loved one, serious career change, separation, relocation)” Participants had to tick a box for “No” or “Yes” and if the ticked “Yes” they could freely write down these events on the lines. We now clarified the instructions in our revised method section. Due to the open response format, it is possible that women/men and optimists/pessimists used different criteria to determine what constituted an important life event. We added this to our limitations section.

6. On lines 112-115 the authors argue that their open response format offers more flexibility than using a checklist of life events, yet participants’ responses were evaluated according to a checklist of events for which normative data on severity were available. It is not clear why the open response format is therefore more flexible. Perhaps the authors could expand on this idea. To what extent were participants left to self-determine what constituted an important life event?

�  As explained above, participants could freely write down everything they regarded as an important life event, which we now added to the method section. Not having a checklist with predefined life events is more efficient for the participants since they do not have to go through a long list with lots of events that do not apply and they do not become biased in any direction, which offers them more flexibility for their answers. Regarding the evaluation of participant responses, we chose the SRRS, which is widely used in life event research and offers the possibility of continuous assessment of stress caused by life events. Another possibility would be to ask how much stress the participants experienced by each of the life events they wrote down to get a subjective evaluation of life event stress, however, this might also be influenced by the current emotional state of the participant.

7. Was there a reason why all of the life events were evaluated by only one rater (lines 108-110)? How can the authors be sure that the rater’s evaluations were reliable? Given the flexible open-response format, did any participants report life events that were not on the existing checklist of events? If so, how were these responses dealt with?

�  Due to the large sample size, it was not possible to have all participant responses rated by several raters. We had a small sample (n=20 participants who reported life events) rated by a second rater, which resulted in a high interrater reliability for the SRRS score (interrater correlation = 0.86). Less than 1% of all participant responses regarding concrete life events could not be scored with the SRRS, this was mostly because lack of information (e.g., just the word “father” or only a date was written down). These participants who gave insufficient information to score life events were included in the analysis of the presence of life events but were not included in the analysis of the SRRS score because they had a missing value here. Due to the very low number of missings, we do not believe that this will distort our observations to a relevant degree.

8. Since the comparisons of interest in this study were women vs. men and optimists vs. pessimists, it would have been helpful to see some data on the SRRS scores and CES-D scores in these different subgroups. Were optimists and pessimists equally likely to report important life events, for example?

�  We added this data to the Supplement (Supplemental Table S1) as suggested. As already indicated in the “Interacting influence of the presence of important life events/life event stress with age, sex, and optimism on depressive symptoms” sections, women and pessimists report more important life events and show higher depression scores than men and optimists, however, next to this main effect, the significant interaction term reveals that the influence of life events on depression is stronger in women than in men and stronger in pessimists than in optimists.

9. The argument is made that both positive and negative life events can be experienced as stressful due to the need to readjust (e.g., lines 72-73, lines 126-127). Following on from point 6 above, is it possible that, for example, optimists were more likely to report positive important life events, and this could explain the weaker relationship with depressive symptoms?

�  As shown in the new Supplemental Table S1, pessimists reported slightly more often important life events but had lower stress scores than optimists. When we look at single life events, pessimists report only slightly more often specific negative life events such as marital separation and fired at work. We now show presence of all important life events of the SRRS in the original order stratified by age, sex, and optimism in Supplemental Tables S2-4.

10. In the abstract, the authors state that they examine the presence of important life events and the RESULTING life event stress (line 26). Similarly, elsewhere the argument is made that higher stress is associated with higher depression scores (e.g., line 310-312). However, since the study did not directly measure stress, it is not clear that the participants in this study experienced stress, only that these life events tend to be stressful. These claims could be worded a little more carefully, to reflect what was actually measured.

�  We now deleted the term “resulting” and rephrased the term “stress” as suggested. As explained in section 2.2.1, the SRRS does not measure individual stress responses but is suggested to measure life event stress because the adaptations required after the experience of a life changing event are thought to be stressful for most individuals.

11. The authors’ stance on the causality of the relationship between life events and depression was unclear. In lines 52-55 of the introduction, they state that a robust and causal association between stressful life events and depression has become established, whereby stressful life events trigger depression. Later, it is acknowledged that the current study is unable to draw conclusions about causality, and that the authors rely heavily on other studies to conclude that stressful life events cause depression (line 376). The arguments in the conclusion then appear to depend on the causality evidenced by other studies, rather than on the data presented in the current study. Given that this study is apparently the first to measure the moderating effects of sex and optimism, how can we be confident that this assumption of causality is justified?

�  In our study, we cannot show causal relationships, only associations, since our study is a cross-sectional observational study. As explained in our revised introduction, previous studies with more complex study designs like longitudinal analyses of population-based samples including dizygotic and monozygotic twins have made it possible to conclude that life events actually trigger, that is causally influence depressive reactions instead of just being symptoms of depression. However, these studies did not include continuous scores of life event stress and did not analyze moderators of the associations between life events and depression. We now clarified that our study is not suited to draw causal conclusions in our revised limitations section. To prevent confusion, we deleted the section about reliance on previous studies.

12. In the discussion section, there is a list (beginning on line 315) of several previous studies that have explored similar research questions, albeit using different scales, sampling methods, etc. Some of the findings of these previous studies were consistent with the current findings, while others were inconsistent. The list format is quite repetitive, and it is difficult to follow the argument throughout – what is the overall message? The argument would be clearer if the findings from the previous studies could be aggregated in a meaningful way, such as by the methods used, the time period sampled, etc.

�  We now ordered the previous population-based studies by the methods used beginning with:

Dalgard et al., who used a list of 12 negative life events and assessed depression with BDI, followed by

Honkalampi., et al who used a similar list of 12 negative life events and also BDI,

followed by the two studies assessing depression with structured diagnostic interview instead of screening, which are

Assari & Lankarani who used a list 8 stressful life events and

Maciejewski et al who used a list of 11 negative life events.

These studies were also inserted in the revised introduction section in the same order.

13. It is not clear why a less comprehensive assessment of life events using a checklist of 8 events vs. free response would potentially reverse the sex effect (lines 334-335). This argument needs to be made more explicit. The same applies to the potential difference between clinical and subclinical depression (lines 336-338).

�  Men tend to report different life events than women, which we now also show in our new Supplemental Table S3 and as it is operationalized in the SRRS, different life events have different stress-inducing properties so that for example the experience of the death of a spouse has a high probability of inducing a clinical depression whereas minor violations of laws rather lead to short-term negative emotions. Moreover, specific life events are more harmful for one sex than another, i.e., job loss is usually more severe for men than for women while loss of a family member is usually more severe for women than for men. With short life event checklists, there is a higher probability of having an overrepresentation of life events more often experienced by one sex or overrepresentation of life events leading to clinical vs subclinical depression. We now added these arguments to our discussion section.

14. On lines 356-363, there is a section on previous uses of SRRS (refs 11, 12, 13) showing that depressives had higher stress scores (i.e., more/more serious life events) than controls in 2/3 studies, and that stress score associated with depression severity in one study. This information would be better in the introduction (see point 2, above).

�  We now present this information in the revised introduction section as suggested.

15. On lines 378-380, the authors argue that even if the causality is reversed, such that depression leads to the experience of stressful life events, “this still has important clinical implications in a way that the experience of life events must be considered more in the prevention and treatment of depression”. This argument was difficult to follow; if depression causes stressful life events, how could consideration of the experience of life events prevent or treat depression?

�  It is important because it is a vicious cycle: depressive patients experience more negative life events, which prevents remission and leads to recurrent depressive episodes. In depression treatment, it is important to convey this knowledge in psychoeducation, identify life events leading to recurrent depressive episodes and work on ways to interrupt this vicious cycle. We now added this information to our discussion section.

16. A new study is introduced in the conclusion on lines 389-396. It would be better to include this argument in the main body of the discussion so that the conclusion remains focused on the current investigation.

�  We now included this argument in the new “clinical implications” section of our revised discussion.

17. Minor points:

Line 84-85 reads “From the clinical point of view, especially subclinical depression is important because it is associated with…”; it would read better as “From a clinical point of view, subclinical depression is especially important because….”

�  Changed as suggested.

Lines 152-155 – this sentence is quite long, and would read better with commas.

�  Changed as suggested.

Lines 310-311 – “In case important life events were experienced” would read better as “In participants who experienced important life events”

�  Changed as suggested.

Line 311-312 – “higher stress… was again significantly associated…” – the “again” appears to be redundant here

�  Changed as suggested.

Line 313 – “Women and individuals with pessimistic personality were more susceptible to the negative influence of life events on depression”. The wording is ambiguous, as it seems to suggest there is a negative association between life events and depression, rather than the positive association that was observed.

�  We changed “negative” into “adverse”.

Line 317 – “The so far largest population-based study” would read better as “So far, the largest population-based study”

�  Changed as suggested.

Line 350 – “this study could show….” presumably means “this study showed that….”?

�  Changed as suggested.

Submitted filename: ResponseToReviewers_PlosOne.docx

Click here for additional data file.

9 May 2022

7 Jun 2022

Gronewold J, et al.

‘Association between life events and later depression in the population-based Heinz Nixdorf Recall study – the role of sex and optimism’ (MS ID: PONE-D-21-02934R1)

Responses to reviewers

Please note that all new/revised text is highlighted in yellow in the manuscript.

Reviewer #1

The authors have addressed the previous points that were raised, and the re-working of the introduction and discussion makes the message much clearer. The inclusion of the additional information in the supplementary tables is also very useful. I have one main question arising from the inclusion of these new tables:

Table S4 shows the frequency of different life events stratified by optimism, and shows that the cut off for determining whether a participant is an optimist or a pessimist results in many more pessimists than optimists (i.e., 0.8% of the optimist group is 6 participants, but 0.8% of the pessimist group is 32 people). Is it possible that the stronger association between life event stress and depression in the pessimist group is due to increased statistical power because the sample size is larger? To put it another way, does the association look smaller in the optimist group because of lower power? If this is a possibility, how would it affect the conclusions?

�  In Table S4, the number and percentage of optimists and pessimists reporting different life events is shown. 6 participants of the optimist group (0.8%) reported that they experienced the life event “death of a spouse”, which was reported by the same percentage of participants from the pessimist group (0.8%, n=32). Since we defined optimists as subjects with Life Orientation Test-Revised score above the upper quartile (>Q3) and pessimists as subjects with Life Orientation Test-Revised score below the upper quartile (≤Q3), we have a smaller sample size in the optimist group (n=781) than in the pessimist group (n=3820). Even though we thus have higher statistical power in the pessimist group, resulting in narrower confidence intervals and lower p-values, the effect estimates for the association of life event stress and depression clearly differ between optimists and pessimist with the unstandardized regression weight from linear regression being 0.9 in optimists and 2.3 in pessimists (Table 3). According to up-to-date statistical recommendations, relevant effect modification is assumed in case of a difference in effect estimates of more than 10%. In our study, we have a difference of 55%, which clearly indicates that the effect of life events on depression is modified by the character trait optimism vs pessimism. This is also supported by the observed statistically highly significant interaction term (p<0.001) between life event stress and the character trait optimism vs pessimism (line 325). Consequently, the different sample size of optimists and pessimists does not affect our conclusion of a stronger association between life event stress and depression pessimists.

Aside from that question, I have just a few minor comments after my second reading of the manuscript. In the following comments, line numbers refer to the line numbers in the revised manuscript.

1. line 70: "self-report screenings which are often not feasible in large population based studies" - I think the authors are arguing that self-report measures are usually used in larger studies because structured interviews are usually not feasible, but the wording makes it sound like it's the self-report measures that are not feasible.

�  The reviewer is completely right, that sentence was misleading, we changed it to “The minority of population-based studies assessed depression with structured diagnostic interviews because – compared with self-report screenings – structured diagnostic interviews require a lot more time and psychiatrically trained staff and thus are often not feasible in large population-based studies designed to address multiple research questions.

2. lines 151-152: I think it would be good to include the details of the second-coding and reliability analysis for the life event scoring in here, so that readers can see that the scoring of life events was carried out more rigorously.

�  We now included details of the second-coding and reliability analysis as suggested (line 152/153, line 172).

3. line 175: "optimism is regarded AS a stable trait" - the word as is missing

�  Changed as suggested.

4. line 327: stronger influence of stress on optimists compared to pessimists - this appears to be the opposite of what is claimed elsewhere

�  That was a mistake, we changed it to “This interaction revealed the stronger influence of life event stress on pessimists compared to optimists.”

5. line 392-393: an argument is made about some life events having more impact on men vs. women, and vice versa. A reference is needed here

�  Reference inserted as suggested.

Submitted filename: ResponseToReviewers_Revision2.docx

Click here for additional data file.

7 Jul 2022

Association between life events and later depression in the population-based Heinz Nixdorf Recall study – the role of sex and optimism

PONE-D-21-02934R2

Dear Dr. Gronewold,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Barbara Dritschel, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

11 Jul 2022

PONE-D-21-02934R2

Association between life events and later depression in the population-based Heinz Nixdorf Recall study – the role of sex and optimism

Dear Dr. Gronewold:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Barbara Dritschel

Academic Editor

PLOS ONE