Development and assessment of stressful life events subscales – A preliminary analysis

Author’s Accepted Manuscript Development and Assessment of Stressful Life Events Subscales - a Preliminary Analysis Teresa Buccheri, Salma Musaad, Kelly K. Bost, Barbara H. Fiese, The STRONG Kids Research Team www.elsevier.com/locate/jad

PII: DOI: Reference:

S0165-0327(17)30812-1 https://doi.org/10.1016/j.jad.2017.09.046 JAD9254

To appear in: Journal of Affective Disorders Received date: 21 April 2017 Revised date: 8 August 2017 Accepted date: 24 September 2017 Cite this article as: Teresa Buccheri, Salma Musaad, Kelly K. Bost, Barbara H. Fiese and The STRONG Kids Research Team, Development and Assessment of Stressful Life Events Subscales - a Preliminary Analysis, Journal of Affective Disorders, https://doi.org/10.1016/j.jad.2017.09.046 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Development and Assessment of Stressful Life Events Subscales - a Preliminary Analysis Teresa Buccheri1, Salma Musaad2, Kelly K. Bost2, Barbara H. Fiese2 and the STRONG Kids Research Team 3 1

2 3

University of Messina, Department of Humanities and Social Sciences - Psychology Unit, 1 Bivona st. 98122 Messina, Italy Family Resiliency Center, Department of Human Development and Family Studies, University of Illinois at Urbana Champaign, 904 W. Nevada, MC-081, Urbana, IL 61801, USA The STRONG Kids Team includes Kristen Harrison, Kelly Bost, Brent McBride, Sharon Donovan, Diana Grigsby-Toussaint, Juhee Kim, Janet Liechty, Angela Wiley, Margarita Teran-Garcia and Barbara Fiese

* Please address all correspondence to: Teresa Buccheri University of Messina, Department of Humanities and Social Sciences - Psychology Unit Permanent address: 10 Giacinto Carini st. 90144 Palermo (Italy) E-mail address:

[email protected]; [email protected]

Mobile phone: +393383626764; +12178190793

Abstract Background: Stress affects people of all ages, genders, and cultures and is associated with physical and psychological complications. Stressful life events are an important research focus and a psychometrically valid measure could provide useful clinical information. The purpose of the study was to develop a reliable and valid measurement of stressful life events and to assess its reliability and validity using established measures of social support, stress, depression, anxiety and maternal and child health. Methods: The authors used an adaptation from the Social Readjustment Rating Scale (SRRS) to describe the prevalence of life events; they developed a 4-factor stressful life events subscales and used Medical Outcomes Social Support Scale, Social Support Scale, Depression, Anxiety

and Stress Scale and 14 general health items for validity analysis. Analyses were performed with descriptive statistics, Cronbach’s alpha, Spearman’s rho, Chi-square test or Fisher’s exact test and Wilcoxon 2-sample test. Results: The 4-factor stressful life events subscales showed acceptable reliability. The resulting subscale scores were significantly associated with established measures of social support, depression, anxiety, stress, and caregiver health indicators. Limitations: The study presented a number of limitations in terms of design and recall bias. Conclusions: Despite the presence of a number of limitations, the study provided valuable insight and suggested that further investigation is needed in order to determine the effectiveness of the measures in revealing the family’s wellbeing and to develop and strengthen a more detailed analysis of the stressful life events/health association.

Keywords: Factor Analysis; Life events; Social Support; Depression; Anxiety; Stress.

1. Introduction Stress can be defined as “the non-specific response of the body to any demand for change” (Selye, 1936). Stress affects people of all ages, genders and cultures and is associated with physical and psychological complications such as heart disease, cancer, stroke (Cohen et al., 2007), depression, posttraumatic stress disorder (PTSD), pathologic aging (Green et al., 2010; Marin et al., 2011; Boardman, Alexander, 2011) and greater food consumption (Epel et al., 2001). Exposure to chronic stress in early life can significantly increase the risk of mental illness

and somatic disturbances in adolescence and adulthood (Anda et al., 2006; Anda et al., 2007; Chapman et al., 2004; Cutrona et al., 2005; Dong et al., 2004; Edwards et al., 2003; Pirkola et al., 2005; Coker et al., 2011). Parental and maternal stress are associated with child behavior problems such as attention-deficit hyperactivity disorder and sleep disturbances, alterations in child neuroendocrine-immune function, mood and anxiety disorders (Mash, Johnston, 1990; Lupien et al., 2009; Tosevski, Milovancevic, 2006; Riis et al. 2016) and may impact the healthfulness of the family food environment (Bauer et al., 2012). Stress in early childhood is associated with deficits in cognitive performance, memory, executive functioning and emotion regulation (Pechtel, Pizzagalli, 2011) and it is a risk factor for obesity in children (Wilson, Sato, 2014; De Vriendt et al., 2009; Dockray et al., 2009; Roemmich et al., 2007; van Jaarsveld et al., 2009) and women (Liu, Umberson, 2015). Some researchers have proposed that a measure of stress could be based on events that occur over the life course (DeLongis et al., 1988). Life events have been defined as significant experiences involving a relatively abrupt change that may produce serious and long lasting effects (Olafsson, Svensson, 1986; Settersten, Mayer, 1997) and can be classified as normative and non-normative or major and everyday life events. Normative life events are predictable happenings or transitions that families experience over the life course (e.g. parenthood, marriage, entering kindergarten) (Price et al., 2010). Non-normative life events are relatively unpredictable with potentially destabilizing effects on family functioning (e.g. death of a family member, change in residence) or positive situations that can create a deep imbalance within the family system (e.g. better changes in financial state). Major life events are stressful events that are typically experienced on an irregular basis such as the death of a relative and divorce, whereas everyday life events refer to stressful events experienced in the day-to-day lives of individuals

(e.g. increased child care duties, handling finances) (Wagner et al., 1988). Effects of life events on individual and family functioning depend on individual attitudes toward change, on the capacity to adjust, and on the support available (Berge et al., 2012). It could be the accumulation of several stressor events rather than one isolated event that determines a family’s level of stress (Holmes, Rahe, 1967; McCubbin et al., 1980; Sarason et al., 1978). The measurement of life events reveals the family’s well-being. Stressful life events and their impact on human life have attracted many researchers in recent years (Sali et al., 2013). Many studies have used the Social Readjustment Rating Scale (SRRS) (Holmes, Rahe, 1967). The Psychiatric Epidemiology Research Interview Life Events Scale (Dohrenwend et al., 1979), the Life Events and Difficulties Schedule (Brown, Harris, 1978) and the Standardized Event Rating System (Dohrenwend et al., 1993) are often used. Stressful experiences on the daily level are assessed with self-report daily diaries such as the Daily Life Experiences Checklist (Stone, Neal, 1982) and the Daily Stress Scale (Bolger, Schilling, 1991). Although the literature reveals that different scales and tools have been developed and used to assess life events, it seems that there is no single life events instrument that is appropriate for all populations or one that is generally accepted in the field (Cohen, Psychosocial Working Group, 2000). Therefore, a specific definitive life events scale with a reasonable and balanced representation of events (Cohen et al. 1995) should be created. Moreover, a psychometrically valid measure of life events could help screen and provide considerable information about family history, family life, caregivers and other family members. This type of measure is important not only for research in health-related fields to evaluate the impact of salient stressors on health outcomes but also for clinical applications in stress prevention.

Our study aims to develop a reliable and valid measurement of stressful life events by utilizing information about stressful family experiences in the past year. Using data from the Synergistic Theory and Research on Obesity and Nutrition Group (STRONG Kids) cohort, the objectives of the study were 1) to gain a preliminary understanding of the underlying structure of a stressful life events measure that was adapted from the Social Readjustment Rating Scale (SRRS) (Holmes, Rahe, 1967), and 2) to assess the reliability and validity of its subscales in determining parents’ and children’s psychological and physical wellbeing using established measures of social support, stress, depression, anxiety and maternal and child health.

2. Methods 2.1 Participants Parents (caregivers) and their 2-5-year-old child (n = 497) were recruited as part of the STRONG Kids cohort, a 3-wave study conducted over 5 years that explores childhood obesity within a developmental, ecological framework (Harrison et al., 2011). To ensure socio-economic and racial/ethnic diversity, an unequal probability sampling frame was used to identify licensed day care centers (n = 33) across five counties in [blinded for review]. Beginning in January 2009, 91% (n = 30) of the centers permitted recruitment of children and their parents. Written informed consent was obtained from the parents and assent was obtained from the children to collect height and weight. Parents completed surveys online or were mailed surveys if they did not have Internet access. Response rates among parents ranged from 60% to 95% across centers. This research was approved by the Institutional Review Board at [Institution blinded for review] and meets all requirements for ethical conduct for research with human subjects.

2.2 Measures Stressful life events scale This scale, adapted from the SRRS (Holmes, Rahe, 1967), asks a series of 43 life events including marriage, trouble with boss, death of a close friend, vacation, and change in residence. On the original scale, 394 participants (55% female, 92% White, 46% with less than College education) were asked to rate the 43 life events to reflect the degree of perceived readjustment (time and ability needed to accommodate to change). Marriage was rated as 50 and the participants were asked to assign a number for the remaining items based on whether they needed more or less readjustment than marriage (e.g., 100 or 40, respectively). In the current study, caregivers were asked if they or someone in their immediate family experienced the event the past year. The same life events were used in their original order. The scale was adapted by asking the caregivers to place a check mark next to the experienced event and a check mark if it was stressful. If the event was not experienced it was left blank. The purpose of this adaptation was to simplify the measures and increase its reliability when used in a survey. Moreover, it allowed for the examination of the accumulation of stressors. Life events that were marked as stressful were coded as 1, 0 otherwise and their sum was used to construct the life events total and subscales score.

Scales and items used for validity testing Three scales (Medical Outcomes Social Support Scale (MOSS), Social Support Scale, and Depression, Anxiety and Stress Scale (DASS-21) and 14 general health items (11 caregiverspecific items, 3 child-specific items) were used for validity analysis.

MOSS The abbreviated MOSS is a 12-item measure that assesses social support availability if needed (Gjesfjeld et al., 2007). It was derived from a 19-item version that was developed based on responses of 3,000 patients with chronic health conditions. It demonstrated acceptable model fit, internal consistency, and concurrent reliability in a sample of 330 mothers whose children were in treatment. Item responses range from 1 (none of the time) to 5 (all of the time), with higher scores indicating greater social support availability. It consists of 4 subscales: tangible (3 items, e.g. “Someone to take you to the doctor if you needed it”), affectionate (3 items, e.g. “Someone who shows you love and affection”), positive social interaction (3 items, e.g. “Someone to do something enjoyable with”) and emotional-informational (3 items, e.g. “Someone to give you good advice about a crisis”). The MOSS measure was selected for inclusion in the study because of the proposed inverse correlation between life events that are considered stressful and perceived support. Relevant items within each subscale were summed. Higher scores indicate more availability of social support. In this sample, mean (standard deviation), minimum, maximum and Cronbach’s alpha were 50.6 (10.2), 12, 60 and 0.951 for the total score; 12.0 (3.4), 3, 15 and 0.942 for tangible; 13.4 (2.5), 3, 15 and 0.921 for affectionate; 12.9 (2.8), 3, 15 and 0.934 for positive social interaction; 12.1 (3.2), 3, 15 and 0.945 for emotional-informational scores, respectively.

Social Support Scale The Social Support Scale asks about people in the environment who can provide help or support and the level of satisfaction (Carney-Crompton, Tan, 2002). It was developed in a sample of 63 female University-level students based on items selected from different sources. In

this study, 7 items were selected to indicate different types of perceived support: 1) ”Whom can you really count on to distract you from your worries and you feel under stress?”; 2) “Whom can you really count on to help you feel more relaxed when you are under pressure or tense?”; 3) “Who accepts you totally, including both your worst and your best points?”; 4) “Whom can you really count on to care about you, regardless of what is happening to you?”; 5) “Whom can you really count on to help you feel better when you are feeling generally down-in-the-dumps?”; 6) “Whom can you count on to console you when you are very upset?”, 7) “Overall, how satisfied are you with the help and support you receive from your network.” Participants were asked to rate how satisfied they were with each type of support using a 6-point scale with response options ranging from 1 (very dissatisfied) to 6 (very satisfied). Higher scores indicate more satisfaction with the specified type of social support. The Social Support Scale was selected for inclusion in the study because individuals with higher perceived satisfaction with social support can manage stress better and display more positive readjustment to stressful stimuli, potentially leading to better outcomes, irrespective of the actual level of support and stressors. In this sample, mean (standard deviation), minimum, and maximum was 5.3 (1.2), 1, 6 for all items except item 3 (5.4 (1.2), 1, 6) and item 4 (5.4 (1.1), 1, 6).

DASS-21 The DASS-21 is a 21-item measure, derived from a longer 42-item version, which covers a range of core symptoms of depression (e.g., “I felt downhearted and blue”), anxiety (e.g. “1 was aware of the action of my heart in the absence of physical exertion” and stress (e.g., “I found it difficult to relax”) with a high discriminating ability (Lovibond, Lovibond, 1995). Items assess the level of severity of symptoms over the past week on a 4-point severity/frequency scale

ranging from 0 (did not apply to me at all) to 3 (applied to me very much, or most of the time). It has been shown to have adequate psychometric properties in a sample of 717 first year psychology students. The DASS measure was selected for inclusion in the study because of the potential for stressful life events to precipitate depression, anxiety, and a stress response. Relevant items within each subscale were summed then converted to the full subscale score by multiplying by 2. Higher scores indicate greater levels of distress. No total score is calculated. In this sample, mean (standard deviation), minimum, maximum and Cronbach’s alpha were 4.6 (6.1), 0, 42 and 0.849 for depression; 3.3 (4.9), 0, 36 and 0.744 for anxiety and 9.1 (7.6), 0, 42 and 0.829 for stress, respectively.

General health items Items related to the health status of the caregiver or their child were obtained from the 36item short-form (SF-36) questionnaire1 (Ware, Sherbourne, 1992). There are 14 items, 1-11 are caregiver-specific, and 12-14 are child-specific. The items assessed role limitations due to physical problems (items 1 and 2) or emotional problems (3, 4), bodily pain (5), general mental health (6-8) and interference with social activities (9) in the past 4 weeks. The SF-36 questions were selected for inclusion in the study because individuals with limitations due to physical or emotional problems may potentially perceive life events as more stressful than usual, partially due to the impaired ability to adequately function in the presence of, manage, or readjust to stressful stimuli. Response choices for items 1-4 are yes, no. Response choices for item 5 ranges from 1 (not at all) to 5 (extremely) and were categorized into 2 groups by grouping the highest 3 responses to indicate substantial pain interference (yes) and the lowest 2 responses to indicate 1

The SF-36 is reproduced here (in part or in its entirety) with permission from the RAND Corporation. Copyright © the RAND Corporation. RAND's permission to reproduce the survey is not an endorsement of the products, services, or other uses in which the survey appears or is applied.

none or little pain (no). Response choices for items 6-9 range from 1 (all of the time) to 5 (none of the time); responses were categorized into 2 groups by combining the lowest 3 responses to indicate the presence of that construct (yes) and the highest 2 responses to indicate none or little (no). Other items ask about if the mother recently had a baby (items 10) and pregnancy complications related to the birth of study child (item 11). Child-specific items included the child’s physical health at the time of survey completion, diagnosis of the child with a chronic medical condition and child family history of specific medical conditions. Results using childspecific items are not presented in a table but are discussed in the text.

2.3 Data analysis The underlying factor structure of the stressful life events items was identified using exploratory factor analysis (EFA). Factor analysis was chosen instead of principal components because it is more conservative since it assumes that there is error variance (i.e., it acknowledges that the total variance cannot be accounted for by the resulting factor structure). Since the stressful life events items were scored on a binary scale (1, 0), a matrix of tetrachoric correlations was obtained using PROC FREQ in SAS (Han et al., 2001) and submitted to a principal factor analysis model using PROC FACTOR with orthogonal rotation. The orthogonal rotation was used to facilitate interpretation of the factors and to determine the extent at which items that were correlated could commit to a single factor rather than more than one. In order to accommodate the correlation matrix, the prior communality estimates (estimate of the variance of an item that is shared with other items) were specified as MAX (largest absolute correlation with any other item) prior to rotation. Small communality estimates (< 20%) indicated that the item needed to

be modified or dropped. Following rotation, communality estimates were also examined in order to determine the proportion variance explained by the factors. The extent of deviation between the observed and the predicted correlation matrix (residual correlation matrix) was examined. If the retained factors were sufficient to explain the correlations among the observed items, the residual correlation matrix was expected to approximate a null matrix with most values ≤ 0.10. The minimum number of items allowed in each factor was 3. If an item did not significantly load on any factor the EFA model was re-run with gradually fewer factors. The ultimate goal was to determine the solution that balanced factor parsimony with factor distinctiveness and interpretability. Initially, a 7-factor solution was obtained by examining the scree plot of eigenvalues. The 7-factor solution was reduced to a 4-factor solution due to the following reasons. The Cronbach’s alpha values in the 7-factor solution (ranging from 0.531 - 0.686) improved after combining 3 pairs of factors into 3 single factors leading to a 4-factor solution (0.568 to 0.742). The 4-factor solution provided similar EFA diagnostics as the 7-factor solution. The 7-factor solution explained 93% of the common variance (initial communality estimates) of the items. Since no items were deleted in the 4-factor solution, there was no change in the proportion of variance explained. Furthermore, if we had used the average eigenvalues as a criterion we would have selected 2 factors, which we felt would not capture enough diversity in the context of life events. Thus the 4-factor solution satisfied these concerns. Items with loadings of absolute value ≥ 4.0 were considered significant and assigned to the factor in which they had the highest loading. This loadings cutoff was selected due to the adequate sample size (n = 497) and of the presence of many items within each factor, thus providing an accurate solution for interpretation purposes (Guadagnoli, Velicer, 1988). No loading crossover of the same item across factors was observed. In order to assess the internal consistency (reliability) of the items

within each factor, the values of the Cronbach’s alpha and its change if an item was deleted were evaluated. All 43 items were retained because deleting items did not improve the model diagnostics or factor structures. A confirmatory factor analysis (CFA) was conducted on the tetrachoric correlation matrix to confirm the proposed 4-factor structure using PROC CALIS with the unweighted least squares method. Model fit was determined using the Standardized Root Mean Square Residual (SRMSR) (appropriate fit suggested for values of 0.08 or less), Goodness of Fit Index (GFI) and the Normed Fit Index (NFI) (appropriate fit suggested for values of ≥ 0.90 for both) (Hooper et al., 2008). The CFA model had an SRMSR of 0.1, GFI of 0.9 and NFI 0.9. Although the SRMSR is 0.1, we decided to utilize this model because the remainder of the fit indices followed the recommended cutoffs; the model was parsimonious and interpretable. Standardized parameter estimates (loadings) for the confirmatory model are reported. Table 1 shows the final 4 factors (subscales) and corresponding parameter estimates. A total score for the stressful life events scale was calculated by summing all the individual item scores that were marked as stressful. Subscale scores were created by summing the number of stressful life events within each subscale. The mean (standard deviation), minimum, maximum and Cronbach’s alpha were 3.3 (4.2), 0, 43 and 0.869 for the total score (43 items); 1.2 (1.8), 0, 14 and 0. 722 for social and financial readjustment (14 items); 0.9 (1.6), 0, 12 and 0.742 for Change in habits and family dynamics (12 items); 0.8 (1.2), 0, 11 and 0.617 for Social misconduct and work-time challenges (11 items); 0.3 (0.7), 0, 6 and 0.568 for Death and change in relationships (6 items) scores, respectively. The total score is highly correlated with all the subscale scores, with correlation coefficients ranging from 0.713 to 0.798, except for the death subscale score that has a lower correlation of 0.408 (not shown).

Descriptive statistics and Cronbach’s alpha were calculated for all scales. Construct validity of the stressful life events subscale scores was assessed by calculating the Spearman’s rank correlation coefficient (Spearman’s rho) between the stressful life events scores with the MOSS, Social Support, and DASS-21 continuous scales. Associations between the stressful life events scores with the continuous scales and general health items were examined as follows: comparisons of categorical data were performed using the Chi-square test or Fisher’s exact test; continuous data were compared across groups using the Wilcoxon 2-sample test with tapproximation. A 2-tailed p-value < 0.05 was considered statistically significant. No adjustment for multiple testing was performed in this preliminary analysis. All analyses were conducted using the Statistical Analysis Software (version 9.3; Institute Inc., Cary, NC).

3. Results 3.1 Description of the population Over half of the caregivers of the study sample were Non-Hispanic White (56%) (NonHispanic Black 26.4%, Other 10.9%, Hispanic 6.8%), college graduates or had a post-graduate degree (54%), employed (64%) with an annual household income $70,000 (35.4%). The mean age of the children was 39.0 ± 8.2 months with an almost balanced gender distribution (49% girls).

3.2 Validity of the stressful life events subscales Table 2 shows the correlation of the stressful life events total and subscale scores with the MOSS, DASS-21 and Social Support scales. MOSS total and subscale scores were negatively yet weakly correlated with the total stressful life events score and two of its subscales (social and financial readjustment, change in habits and family dynamics). DASS-21 subscales were

positively correlated with the stressful life events total and three subscale scores (social and financial readjustment, change in habits and family dynamics, social misconduct and work-time challenges), with the highest significant correlation coefficient being 0.351 between the DASS21 stress subscale and the total stressful life events score. The anxiety DASS-21 subscale was positively yet weakly correlated with the death and change in relationships stressful life events subscale (0.094). Social Support scale items were mostly negatively yet weakly correlated with the stressful life events subscales, with significant correlations ranging from -0.095 to -0.200. None of these items were significantly correlated with the death and change in relationships subscale of the stressful life events. To determine if the stressful life events were differentially associated with the health status of the caregiver or their child, the mean total and subscale scores were examined across caregiver responses to the general health items. In Table 3, the mean total score and all the subscale scores except social misconduct and work-time challenges were significantly higher among caregivers who reported role limitations due to emotional problems (accomplished less that they liked and didn’t do work or other activities as carefully as usual), interference with normal work activities due to bodily pain and interference with social activities due to physical health or emotional problems. In contrast, the mean total score and the score for the social and financial readjustment and change in habits and family dynamics subscales were significantly lower among those who reported feeling calm and peaceful and having a lot of energy in the past four weeks. The social misconduct and work-time challenges subscale was not associated with any of the general health items. The mean score among caregivers who reported no preexisting diabetes was 0.8 (standard deviation = 1.19) compared to 0 among caregivers who had

preexisting diabetes. However, these results should be interpreted with caution due to the limited sample size of those who had pre-existing diabetes (n = 5). Out of all the stressful life event subscales, only the change in habits and family dynamics subscale score was significantly associated with recently having a baby. Mean scores were significantly higher among those who recently had a baby (1.6 (1.87)) compared to those who did not (0.8 (1.52) (p = 0.0004). Furthermore, one of the items that make up this subscale, gaining a new family member (e.g. through birth, adoption, elder moving in), also followed the same trend, with a mean of 0.3 (0.44) and 0.05 (0.21), respectively (p < 0.0001) (not shown). Mean change in habits and family dynamics subscale score was also significantly higher among those who had gestational diabetes when pregnant with the study child (2.0 (2.74)) compared to those who did not (0.8 (1.38)) (p = 0.004). Both the total and the social and financial readjustment subscale scores were significantly higher among those who had a history of pregnancy-induced hypertension when pregnant with the study child than those who did not. Mean total score for the stressful life events was 4.9 (4.54) compared to 3.3 (4.15) (p = 0.011), respectively. Mean social and financial readjustment subscale score was 2.2 (2.09) compared to 1.2 (1.78) (p = 0.001), respectively. Table 4 presents comparisons across the child-specific general health items. No differences were noted across categories of child physical health whether the child was diagnosed with a chronic medical condition. Significant differences were observed among caregivers who reported presence of a child family history of depression compared to none for the mean total stressful life events score (5.3 (5.39) and 2.8 (3.70), respectively) (p < 0.0001) and all subscales (p < 0.05) as well as other mental illnesses for the mean total stressful life events score (5.5 (5.02) and 3.2 (4.17), respectively) (p = 0.006), the change in habits and family

dynamics subscale score (1.8 (1.88) and 0.9 (1.54), respectively) (p = 0.006) and the social misconduct and work-time challenges subscale score (1.5 (1.53) and 0.8 (1.18), respectively) (p = 0.013). Significant differences were also noted among caregivers who reported presence of a child family history of diabetes compared to none for the mean total stressful life events score (4.3 (5.61) and 2.9 (3.32), respectively) (p = 0.003), the social and financial readjustment subscale score (1.7 (2.35) and 1.0 (1.41), respectively) (p = 0.001) and the change in habits and family dynamics subscale score (1.3 (1.97) and 0.8 (1.32), respectively) (p = 0.003). Similar differences were observed for high total cholesterol in the mean total stressful life events score (3.8 (4.51) and 3.2 (4.11), respectively) (p = 0.034) and the social and financial readjustment subscale score (1.5 (2.07) and 1.1 (1.66), respectively) (p = 0.042). Significant differences were also observed among caregivers who reported a child family history of overweight/obesity compared to none for the mean total stressful life events score (3.8 (3.96) and 3.1 (4.32, respectively) (p = 0.013) and the social and financial readjustment subscale score (1.5 (1.82) and 1.1 (1.78), respectively) (p = 0.021). Other significant differences were observed for family history of cancer, stroke and eating disorders for the mean total stressful life events score and one or more of the subscale scores. In some cases, these results should be interpreted with caution due to the small sample size of individuals with the reported family history.

4. Discussion We built 4 valid subscales measuring different aspects of stressful life events (social and financial readjustment, change in habits and family dynamics, social misconduct and work-time challenges, death and change in relationships), developed as a result of an EFA with a reasonable model fit in the CFA. They demonstrate acceptable reliability with Cronbach’s alphas ranging

from 0.57 to 0.74. The alpha values are reasonable given the observed variability of the responses within each subscale. For example, the proportion of life events that happened and we found to be stressful for items that make up the social and financial readjustment subscale ranged from 2.4% (foreclosure on a mortgage or loan) to 25.2% (major changes in financial state). Moreover, since perfect tau-equivalence (linear related and differing by a constant) is impractical and rarely achieved, the raw Cronbach’s alpha is considered a lower bound estimate of reliability (Cortina, 1993). It was the authors’ intention to develop stressful life events subscales that differed from the traditional normative and non-normative structure. This enabled the authors to create more diversity in the meaning of the subscales and to reflect their different characteristics in this population. Some items that relate to the same physical environment were assigned to different subscales (e.g., being fired from work lies in a different subscale from retirement from work). This may be a reflection of the varied capacity of the study caregivers to perceive, adjust and adapt to these changes.

The resulting total and subscale scores were significantly associated with established measures of social support, depression, anxiety, stress, and caregiver health indicators in the expected directions. The only subscale that was associated with feeling downhearted and blue, recently having a baby and history of gestational diabetes was the change in habits and family dynamics subscale. The social and financial readjustment subscale was instead the only subscale found to be associated with pregnancy-induced hypertension. Social and financial readjustment and change in habits and family dynamics subscale scores were negatively correlated with social support availability (MOSS) and social support items. Our results are consistent with previous studies reporting that accumulation of stressful life events and lack of social support are related to psychological maladjustment and social

support may serve as a protective factor to help cope with stressful life events (Lila et al., 2013; Lanier, Maume, 2009; Silver, Teasdale, 2005). All DASS-21 subscales were positively correlated with the total score and the scores of 3 of the stressful life events subscales (social and financial readjustment, change in habits and family dynamics, social misconduct and work-time challenges). This finding confirms past studies whereby stressful life events consistently predict the onset or increase the symptoms of depression and anxiety (Michl et al., 2013; Hammen, 2005; Kendler et al., 2003). The only significant correlations observed for the social misconduct and work-time challenges subscale score was with DASS-21 subscales and one of the social support items, suggesting discrimination of the subscales domains. The death and change in relationships subscale score was weakly yet significantly correlated with the anxiety subscale of DASS-21 and this may be related to how this subscale contains the least number of items thus affecting its reliability (Cronbach’s alpha = 0.57). However, it also suggests that it characterizes individuals who respond to death and major changes in social relationships by becoming more anxious than others. The stressful life events subscales showed association with the general health items. The social and financial readjustment and the change in habits and family dynamics subscale scores were both higher among parents who reported role limitations due to emotional problems, interference with normal work activities due to bodily pain and physical health or emotional problems interfering with social activities. Studies revealed the association of stressful life events with immune alterations, endocrine and cardiovascular problems, neurological symptoms and mental health diseases (Herbert, Cohen, 1993; Schwarzer, Schulz, 2003). However, a relationship between stress and health problems is not always exclusive because other factors

operate at the same time (environment, social support, personal coping strategies). As events differ in their nature and impact, so do individuals differ in their immediate responses to an event. In regard to these aspects, the present study shows that, the mean total score as well as scores for the social and financial readjustment and change in habits and family dynamics subscales decreased among individuals who reported feeling calm and peaceful and having a lot of energy, confirming that the experience of a critical life event could be related to coping and social support, whereby these two factors may moderate the stress-illness connection (Cobb, 1976; Schwarzer, Schulz, 2003).

5. Limitations The authors acknowledge a number of limitations of this study. First, the design of this study was cross‐sectional and thus could not make causal inferences. Future intervention and longitudinal studies should examine causal relationships between the variables. Second, the STRONG Kids sample is not representative of the general population. Thus, the resulting stressful life events subscales should be examined in other samples to confirm their validity. Third, limitations can result from recall bias. Parents may not have recalled all life events or did not want to report them all. Finally, the authors did not use the original structure of the life events questionnaire.

6. Conclusions In health psychology, the relationship between individuals and environment have been analyzed along with mediating and moderating factors, such as stressful life events, coping and social support (Hobfoll et al., 1998; Schwarzer, Schulz, 2003). Family life cycle theory

underlined that stressful life events affect the individual and family functioning. The measurement of life events seems to be essential to provide information about family’s wellbeing. Despite a number of limitations of the study, it provides valuable insight into the different dimensions that underlie stressful life events in parents of preschoolers, and suggests that further investigation is needed in order to determine the effectiveness of the measures and to develop and strengthen a more detailed analysis of the stressful life events/health association, including mediators and moderators, such as coping strategies and social support. The authors believe that a psychometrically valid measure of stressful life events could help screen and characterize at-risk caregivers and other family members based on the pattern and amount of reported stressors. This type of measure is needed for research in health-related fields to determine the differential impact of salient stressors experienced by caregivers on health outcomes. Research outcomes using valid stress assessments, in turn, can greatly inform clinical applications in stress prevention.

Disclosure Statement The authors have declared that no conflict of interest exists.

Acknowledgements The authors would like to thank the participating families as well as the following funding sources: United States Department of Agriculture (Hatch 793-328) to Barbara Fiese (PI), and Kelly Bost; Illinois Council on Food and Agricultural Research Sentinel Grant to Kris Harrison, University of Illinois Health and Wellness grant to Sharon Donovan and Barbara Fiese, the Dairy Research Institute to Barbara Fiese and Sharon Donovan (PI’s) and the Christopher Family Foundation Food and Family Program.

Role of the Funding source

The present research has been supported by the following funding sources: United States Department of Agriculture (Hatch 793-328) to Barbara Fiese (PI), and Kelly Bost; Illinois Council on Food and Agricultural Research Sentinel Grant to Kris Harrison, University of Illinois Health and Wellness grant to Sharon Donovan and Barbara Fiese, the Dairy Research Institute to Barbara Fiese and Sharon Donovan (PI’s) and the Christopher Family Foundation Food and Family Program.

Ethical Approval The present research was approved by the Institutional Review Board at University of Illinois at UrbanaChampaign, and meets all requirements for ethical conduct for research with human subjects.

Contributors All authors were involved in the design, analysis and formulation of the manuscript. All authors have approved the final article.

Conflict of interest There were no conflicts of interest for any of the authors involved.

References Anda, R., Felitti, V., Bremner, J., Walker, J., Whitfield, C., Perry, B., et al., 2006. The enduring effects of abuse and related adverse experiences in childhood. Eur Arch Psychiatry Clin Neuroscience. 256, 174–186.

Anda, R.F., Brown, D.W., Felitti, V.J., Bremner, J.D., Dube, S.R., Giles, W.H., 2007. Adverse childhood experiences and prescribed psychotropic medications in adults. Am J Prev Med. 32, 389–394.

Bauer, K.W., Hearst, M.O., Escoto, K., Berge, J.M., Neumark- Sztainer, D., 2012. Parental employment and work-family stress: Association with family food environments. Social Science & Medicine. 75(3), 496-504.

Berge, J.M., Loth, K., Hanson, C., Croll‐ Lampert, J., Neumark‐ Sztainer, D., 2012. Family life cycle transitions and the onset of eating disorders: a retrospective grounded theory approach. Journal of Clinical Nursing. 21(9‐ 10), 1355-1363. Boardman, J.D., Alexander, K.B., 2011. Stress trajectories, health behaviors, and the mental health of black and white young adults. Social Science & Medicine. 72(10), 1659-1666. Bolger, N., Schilling, E.A., 1991. Personality and the problems of everyday life: The role of neuroticism in exposure and reactivity to daily stressors. Journal of Personality. 59, 355-386.

Brown, G.W., Harris, T.O., 1978. Social origins of depression. Free Press, New York.

Carney-Crompton, S., Tan, J., 2002. Support systems, psychological functioning, and academic performance of nontraditional female students. Adult Education Quarterly. 52(2),140-154.

Chapman, D.P., Whitfield, C.L., Felitti, V.J., Dube, S.R., Edwards, V.J., Anda, R.F., 2004. Adverse childhood experiences and the risk of depressive disorders in adulthood. J Affect Disord. 82(2), 217–225.

Cobb, S., 1976. Social support as a moderator of life stress. Psychosomatic Medicine. 38, 300314.

Cohen, S., Kessler, R., Underwood Gordon, L., 1995. Strategies for measuring stress in studies of psychiatric and physical disorders. In Cohen, S., Kessler, R., Underwood Gordon, L. (Ed.), Measuring Stress. Oxford University Press, New York, pp. 3-28.

Cohen, S., Psychosocial Working Group, John D. and Catherine T. MacArthur Research Network on Socioeconomic Status and Health. Measures of psychological stress. Available at: http://www.macses.ucsf.edu/research/psychosocial/stress.php#measurement. Accessed February, 2000.

Cohen, S., Janicki-Deverts, D., 2007. Stress and disease. Paper Commissioned by the Institute of Medicine Committee on Psychosocial Services to Cancer Patients and Families in a Community Setting.

Coker, T.R., Elliott, M.N., Wallander, J.L., Cuccaro, P., Grunbaum, J.A., Corona, R., et al., 2011. Association of family stressful life-change events and health-related quality of life in fifthgrade children. Archives of Pediatrics & Adolescent Medicine. 165(4), 354-359.

Cortina, J.M., 1993. What is Coefficient Alpha? An examination of theory and applications. Journal of Applied Psychology. 78(1), 98-104.

Cutrona, C.E., Russell, D.W., Brown, P.A., Clark, L.A., Hessling, R.M., Gardner, K.A., 2005. Neighborhood context, personality, and stressful life events as predictors of depression among African American women. J Abnorm Psychol. 114(1), 3–15.

DeLongis, A., Folkman, S., Lazarus, R.S., 1988. The impact of daily stress on health and mood: psychological and social resources as mediators. Journal of Personality and Social Psychology. 54(3), 486.

De Vriendt, T., Moreno, L.A., De Henauw, S., 2009. Chronic stress and obesity in adolescents: scientific evidence and methodological issues for epidemiological research. Nutrition, Metabolism and Cardiovascular Diseases. 19(7), 511-519.

Dockray, S., Susman, E. J., Dorn, L.D., 2009. Depression, cortisol reactivity, and obesity in childhood and adolescence. Journal of Adolescent Health. 45(4), 344-350.

Dohrenwend, B., Askenasy, A., Krasnoff, L., Dohrenwend, B., 1978. Exemplification of a Method for Scaling Life Events: The PERI Life Events Scale. Journal of Health and Social Behavior. 19(2), 205-229.

Dohrenwend, B.P., Raphael, K.G., Schwartz, S., Stueve, A., Skodol, A., 1993. The structured event probe and narrative rating method for measuring stressful life events, in: Goldberger, L., Breznitz, S. (Eds.), Handbook of Stress. Free Press, New York, pp. 174-199.

Dong, M., Giles, W.H., Felitti, V.J., Dube, S.R., Williams, J.E., Chapman, D.P., et al., 2004. Insights Into causal pathways for ischemic heart disease: Adverse Childhood Experiences Study. Circulation. 110(13), 1761–1766.

Edwards, V.J., Holden, G.W., Felitti, V.J., Anda, R.F., 2003. Relationship between multiple forms of childhood maltreatment and adult mental health in community respondents: Results from the Adverse Childhood Experiences Study. Am J Psychiatry. 160(8), 1453–1460.

Epel, E., Lapidus, R., McEwen, B., Brownell, K., 2001. Stress may add bite to appetite in women:

a

laboratory

study

of

stress-induced

cortisol

and

eating

behavior. Psychoneuroendocrinology. 26(1), 37-49.

Gjesfjeld, C.D., Greeno, C.G., Kim, K.H., 2007. A Confirmatory Factor Analysis of an Abbreviated Social Support Instrument-The MOSS-SSS. Research on Social Work Practice.

Green, J.G., McLaughlin, K.A., Berglund, P.A., Gruber, M.J., Sampson, N.A., Zaslavsky, A.M., et al., 2010. Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication I: associations with first onset of DSM-IV disorders. Archives of General Psychiatry. 67(2), 113-123.

Guadagnoli, E., Velicer, W.F., 1988. Relation of sample size to the stability of component patterns. Psychol Bull. 103(2), 265-275.

Hammen, C., 2005. Stress and depression. Annual Review of Clinical Psychology. 1, 293–319.

Han, L., Neilands, T.B., Dolcini, M.M., 2001. Factor analysis of categorical data in SAS. Oral presentation at the Ninth Annual Conference of the Western Users of the SAS System, San Francisco, California.

Harrison, K., Bost, K.K., McBride, B.A., Donovan, S.M., Grigsby-Toussaint, D.S., Kim, J. et al., 2011. Toward a developmental conceptualization of contributors to overweight and obesity in childhood: The Six-Cs Model. Child Development Perspectives. 5(1), 50–8.

Herbert, T.B., Cohen, S., 1993. Stress and immunity in humans: A meta-analytic review. Psychosomatic Medicine. 55, 364-379.

Hobfoll, S.E., Schwarzer, R., Chon, K.K., 1998. Disentangling the stress labyrinth: Interpreting the meaning of the term stress as it is studied in health context. Anxiety, Stress, and Coping. 11(3), 181-212.

Holmes, T.H., Rahe, R.H., 1967. The social readjustment rating scale. Journal of psychosomatic research. 11(2), 213-218.

Hooper, D., Coughlan, J., Mullen, M.R., 2008. Structural Equation Modelling: Guidelines for Determining Model Fit. The Electronic Journal of Business Research Methods. 6(1), 53 – 60.

Kendler, K.S., Hettema, J.M., Butera, F., Gardner, C.O., Prescott, C.A., 2003. Life event dimensions of loss, humiliation, entrapment, and danger in the prediction of onset of major depression and generalized anxiety. Archives of General Psychiatry. 60, 789–796.

Lanier, C., Maume, M.O., 2009. Intimate partner violence and social isolation across the rural/urban divide. Violence Against Women. 15(11), 1311-1330.

Lila, M., Gracia, E., Murgui, S., 2013. Psychological adjustment and victim-blaming among intimate partner violence offenders: The role of social support and stressful life events. The European Journal of Psychology Applied to Legal Context. 5(2), 147-153.

Liu, H., Umberson, D., 2015. Gender, Stress in Childhood and Adulthood, and Trajectories of Change in Body Mass. Social Science & Medicine. 139, 61–69.

Lovibond, P.F., Lovibond, S.H., 1995. The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behaviour Research and Therapy. 33(3), 335-343.

Lupien, S.J., McEwen, B.S., Gunnar, M.R., Heim, C., 2009. Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nature Reviews Neuroscience. 10(6), 434-445.

Marin, M.F., Lord, C., Andrews, J., Juster, R.P., Sindi, S., Arsenault-Lapierre, G., et al., 2011. Chronic stress, cognitive functioning and mental health. Neurobiology of Learning and Memory. 96(4), 583-595.

Mash, E.J., Johnston, C., 1990. Determinants of parenting stress: Illustrations from families of hyperactive children and families of physically abused children. Journal of Clinical Child Psychology. 19, 313–328.

McCubbin, H., Patterson, J., Wilson, L., 1980. Family Inventory of Life Events and Changes (FILE) Form A. Family Social Science, St Paul.

Meltzer, L.J., Mongtomery-Downs, H.E., 2011. Sleep in the family. Pediatr Clin North Am. 58(3), 765-74.

Michl, L.C., McLaughlin, K.A., Shepherd, K., Nolen-Hoeksema, S., 2013. Rumination as a mechanism linking stressful life events to symptoms of depression and anxiety: Longitudinal evidence in early adolescents and adults. Journal of Abnormal Psychology. 122(2), 339–352.

Olafsson, O., Svensson, P., 1986. Unemployment-related lifestyle changes and health disturbances in adolescents and children in the western countries. Social Science & Medicine. 22(11), 1105-1113.

Pechtel, P., Pizzagalli, D.A., 2011. Effects of early life stress on cognitive and affective function: an integrated review of human literature. Psychopharmacology. 214(1), 55-70.

Pirkola, S., Isometsä, E., Aro, H., Kestilä, L., Hämäläinen, J., Veijola, J., et al., 2005. Childhood adversities as risk factors for adult mental disorders. Soc Psychiatry Psychiatr Epidemiol. 40, 769–777.

Price, S.J., Price, C.A., McKenry, P.C., 2010. Families & change: Coping with stressful events and transitions, ed. Sage, Los Angeles.

Riis, J.L., Granger, D.A., Minkovitz, C.S., Bandeen-Roche, K., DiPietro, J.A., Johnson, S.B., 2016. Maternal distress and child neuroendocrine and immune regulation. Social Science & Medicine. 151, 206-214.

Roemmich, J.N., Smith, J.R., Epstein, L.H., Lambiase, M., 2007. Stress reactivity and adiposity of youth. Obesity. 15(9), 2303-2310.

Sali, R., Roohafza, H., Sadeghi, M., Andalib, E., Shavandi, H., Sarrafzadegan, N., 2013. Validation of the Revised Stressful Life Event Questionnaire Using a Hybrid Model of Genetic Algorithm and Artificial Neural Networks. Computational and Mathematical Methods in Medicine. Vol. 2013, Article ID 601640.

Sarason, I.G., Johnson, J.H., Siegel, J.M., 1978. Assessing the impact of life changes: development

of

the

Psychology. 46(5), 932.

Life

Experiences

Survey. Journal

of

Consulting and

Clinical

Schwarzer, R., Schulz, U., 2003. The Role of Stressful Life Events, in: Nezu, A.M., Nezu, C.M., Geller, P.A. (Ed.), Health Psychology. Wiley, New York, Vol. 9, pp. 27-49.

Selye, H., 1936. A syndrome produced by diverse nocuous agents. Nature. 138, 32.

Settersten Jr., R.A., Mayer, K.U., 1997. The measurement of age, age structuring, and the life course. Annual Review of Sociology. 23, 233-261.

Silver, E., Teasdale, B., 2005. Mental disorder and violence: An examination of stressful life events and impaired social support. Social Problems. 52(1), 62-78.

Stone, A.A., Neale, J.M., 1982. Development of a methodology for assessing daily experiences, in: Baum, A., Singer, J.E. (Eds.), Advances in Environmental Psychology: Environment and health. Lawrence Erlbaum, Hillsdale, Vol. 4, pp. 49-83.

Tosevski, D.L., Milovancevic, M.P., 2006. Stressful life events and physical health. Current Opinion in Psychiatry. 19(2), 184-189.

van Jaarsveld, C.H.M., Fidler, J.A., Steptoe, A., Boniface, D., Wardle, J., 2009. Perceived Stress and Weight Gain in Adolescence: A Longitudinal Analysis. Obesity, 17, 2155–2161.

Ware, Jr., J.E., Sherbourne, C.D., 1992. The MOS 36-item short-form health survey (SF-36): I. Conceptual framework and item selection. Medical care. 30(6), 473-483.

Wagner, B.M., Compas, B.E., Howell, D.C., 1988. Daily and major life events: A test of an integrative model of psychosocial stress. American Journal of Community Psychology. 16(2), 189-205.

Wilson, S.M., Sato, A.F., 2014. Stress and paediatric obesity: what we know and where to go. Stress and Health. 30(2), 91-102.

Development and Assessment of Stressful Life Events Subscales - a Preliminary Analysis

Table 1 Components and standardized estimates (loadings) of the final stressful life events subscales. Social

and

financial

Change in habits and

Social misconduct and

Death and change

family dynamics

work-time challenges

in relationships

readjustment Major personal injury or illness

0.7

Foreclosure on a mortgage or

0.7

loan Major

changes

in

social

0.7

Major change in number of

0.7

activities

family get-togethers Major changes in financial state

0.7

Being fired from work or laid off

0.6

Changing to a different line of

0.6

work Taking on a mortgage greater

0.6

than $150,000 Taking on a loan or mortgage

0.6

less than $150,000 Major changes in working hours

0.6

or conditions Major changes in the health or

0.6

behavior of a family member Major

changes

in

living

0.6

conditions Change in residence

0.5

Beginning or ceasing formal

0.5

schooling Major change in usual type and/or amount of

0.9

recreation Reconciliation with partner

0.9

Major changes in eating habits

0.7

Major changes in number of arguments with spouse

0.7

Sexual difficulties

0.7

Social

and

financial

Change in habits and

Social misconduct and

Death and change

family dynamics

work-time challenges

in relationships

readjustment Changing to a new school

0.6

Marriage

0.6

Divorce

0.6

Separation from partner

0.6

Gaining a new family member

0.6

Major changes in sleeping habits

0.6

Pregnancy

0.5

Retirement from work

0.9

Major business readjustment

0.8

Detention

0.8

in

jail

or

other

institution Partner beginning or ceasing work outside the home

0.7

Major changes in responsibilities at work

0.6

Outstanding

0.6

personal

achievement Trouble with boss

0.5

Minor violation of the law

0.5

Vacation

0.5

Major holiday

0.4

In-law troubles

0.4

Changes in personal habits

0.9

Major change in church or other group activities

0.8

Death of a partner

0.7

Son or daughter leaving home

0.7

Death of a close family member

0.6

Death of a close friend

0.5

Table 2 Spearman correlation coefficient (r) between stressful life events total and subscale scores and scores for MOSS, DASS-21, and the Social Support Scale items. Stressful life events Total

Social and financial

Change

score

readjustment

and

in

habits

Social misconduct and

Death and change in

family

work-time challenges

relationships

dynamics MOSS Total score

-0.200*

-0.217*

-0.190*

-0.064

-0.001

Emotional/

*

*

*

-0.024

0.029

-0.137

-0.166

-0.123

informational Affectionate Positive

-0.118*

-0.110*

-0.150*

-0.011

-0.031

*

*

*

-0.083

-0.027

-0.185

-0.175

-0.204

interaction Tangible

-0.181*

-0.198*

-0.165*

-0.049

-0.024

0.331*

0.276*

0.317*

0.194*

0.065

*

*

*

*

DASS-21 Depression

0.211

Stress

0.351*

0.259*

0.321*

0.284*

0.027

1

-0.123*

-0.102*

-0.118*

-0.059

0.017

2

-0.200

*

*

-0.191

*

3

-0.121*

-0.126*

-0.058

4

-0.118

*

-0.129

*

-0.064

-0.127

*

-0.127

*

6

-0.169

*

-0.165

*

7

-0.173*

Social

0.138

0.245

0.198

0.094*

Anxiety

Support

Scale

5

*

-0.155

-0.194* p

-0.095

*

0.025

-0.042

0.026

-0.041

0.039

-0.098

*

-0.071

-0.007

-0.101

*

-0.082

0.017

-0.120*

-0.069

0.019

<

0.05

Table 3 Mean (standard deviation) of the stressful life events subscale scores across the general health items (SF-36 and Other Items) Role limitations due to physical problems

Role limitations due to emotional problems

1. Accomplished less

2. Were limited in the

3. Accomplished less than you

4. Didn’t do work or

than you would like?

kind of work or other

would like?

other activities as

activities? Yes

No

(n = 55)

p

*

Yes

No

(n =

(n =

(n = 451)

436)

41)

Total

5.7

3.1

0.07

5.3

score

(7.49)

(3.55)

8

(8.23)

2.0

1.1

0.17

2.0

ale 1

(2.87)

(1.58)

5

(3.02)

1.2 (1.63)

Subsc

1.9

0.8

0.00

1.7

ale 2

(2.82)

(1.29)

7

(2.95)

0.9 (1.37)

Subsc

0.9

0.8

0.52

0.9

ale 3

(1.69)

(1.14)

2

(1.85)

0.8 (1.14)

Subsc

0.7

0.3

0.00

0.6

ale 4

(1.19)

(0.62)

3

(1.28)

5. Bodily pain

p

3.2 (3.65)

Subsc

carefully as usual? *

0.3 (0.64)

Yes

No

(n = 68)

(n = 420)

0.9

5.9

2.9 (3.58)

87

(6.67)

0.5

2.2

1.1 (1.58)

16

(2.67)

0.4

1.9

0.78 (1.29)

51

(2.57)

0.6

1.1

0.75 (1.11)

89

(1.68)

0.1

0.6

0.3 (0.63)

54

(1.06)

6. Calm and peaceful

p

*

< 0.0001

0.001

< 0.0001

0.078

< 0.0001

7. A lot of energy

Yes

No

(n =

(n =

48)

441)

6.5

3.0

(7.61

(3.57

)

)

2.5

1.1

(2.95

(1.58

)

)

2.1

0.8

0.000

(2.91

(1.29

4

)

)

1.2

0.8

(1.92

(1.10

)

)

0.6

0.3

(1.18

(0.64

)

)

Yes

No

(n = 41)

No

(n =

(n =

450

359

)

)

5.7

3.1

score

(5.96

No

(n =

(n =

(n = 133)

131)

359)

4.5

(4.0

(3.7

(5.19

)

0)

8)

)

Subscale

2.3

1.2

1.1

1.6

1

(2.64

(1.6

(1.6

(2.08

0.011

p

*

Yes

2.9

0.008

0.001

0.196

0.003

blue Yes

Total

p

0.001

8. Downhearted and

interferes with work *

p*

< 0.0001

0.002

2.9

Yes

No

(n =

(n =

36)

456)

p*

4.4

3.3

0.00

(3.68

(3.94

(4.26

9

)

)

)

1.6

1.2

0.17

(1.92

(1.79

9

1.2 (1.68

4.4 (5.38)

p

*

1.5 (2.08)

0.001

0.041

32

)

8)

7)

)

Subscale

1.8

0.9

0.8

1.4

2

(2.59

(1.4

(1.3

(2.06

(1.29

)

3)

2)

)

)

Subscale

0.8

0.8

0.7

0.9

3

(1.05

(1.2

(1.1

(1.45

)

2)

1)

)

Subscale

0.7

0.3

0.3

0.4

4

(1.06

(0.6

(0.6

(0.88

)

7)

5)

)

0.007

0.676

0.002

) < 0.0001

0.109

)

)

0.000

1.6

0.9

0.01

2

(2.34

(1.49

5

)

)

0.7

0.8

0.99

(1.04

(0.76

(1.24

5

)

)

)

0.5

0.3

0.06

(0.67

(0.94

(0.69

2

)

)

)

0.8

1.4 (2.10)

0.7

0.072

1.0 (1.55)

0.3

9. Interference with

10. Have you recently had a

social activities due to

baby?

0.4(0.83)

0.103

0.091

physical health or emotional problems Yes

No

(n = 30)

p*

Yes

No

(n =

(n =

(n =

460

70)

381)

3.8

3.3

p*

) Total

6.0

3.2

score

(5.79

(4.0

(3.9

(4.33

)

8)

6)

)

Subscale

2.2

1.2

1.2

1.3

1

(2.12

(1.7

(1.6

(1.85

)

7)

7)

)

Subscale

2.3

0.9

0.000

1.6

0.8

2

(2.88

(1.4

2

(1.8

(1.52

)

1)

7)

)

Subscale

0.8

0.8

0.7

0.8

3

(1.09

(1.2

(1.0

(1.22

)

2)

2)

)

Subscale

0.7

0.3

0.2

0.3

4

(1.09

(0.6

(0.4

(0.76

)

8)

8)

)

0.005

0.002

0.909

0.021

0.217

0.849

0.0004

0.771

0.379

11. Pregnancy complications related to birth of study child Pregnancy-induced

Pre-eclampsia or eclampsia

Pre-existing diabetes

Gestational diabetes

hypertension

33

Yes

No

(n = 30)

p*

Yes

No

(n =

(n =

422

p*

Yes

No

(n =

(n =

(n = 455)

14)

431)

5)

3.2

3.4

p*

Yes

No

(n =

(n =

33)

433)

p*

) Total

4.9

3.3

5.4

3.2

0.24

score

(4.54

(4.1

(4.1

(4.23

(1.34

(7.45

(3.77

5

)

5)

9)

)

)

)

)

Subscale

2.2

1.2

1.4

1.3

1.8

1.2

0.70

1

(2.09

(1.7

(1.9

(1.83

(0.45

(2.81

(1.70

2

)

8)

1)

)

)

)

)

Subscale

1.2

0.9

0.8

0.9

2.0

0.8

0.00

2

(1.87

(1.5

(1.6

(1.55

(0.89

(2.74

(1.38

4

)

3)

3)

)

)

)

)

Subscale

1.1

0.8

0.8

0.8

1.0

0.8

0.46

3

(1.28

(1.1

(1.3

(1.21

(1.59

(1.15

8

)

9)

7)

)

)

)

Subscale

0.3

0.3

0.3

0.3

0.6

0.3

0.32

4

(0.70

(0.7

(0.6

(0.73

(1.28

(0.65

7

)

2)

1)

)

)

)

0.011

0.001

0.611

0.079

0.822

0.721

0.922

0.524

0.659

0.947

0.6

0.2

0.4

0

0

3.3 (4.12)

1.2 (1.79)

0.9 (1.46)

0.8 (1.19)

0.3 (0.72)

0.025

0.105

0.370

0.044

0.244

*

For difference between the yes and no responses using the Wilcoxon 2-sample test with t-approximation. Stressful life events subscales: 1 = Social and financial readjustment; 2 = Change in habits and family dynamics; 3 = Social misconduct and work-time challenges; 4 = Death and change in relationships. Participants were asked using the past 4 weeks’ time frame. Response choices for item 5 ranges from 1 (not at all) to 5 (extremely) and were categorized into 2 groups by grouping the last 3 responses to indicate substantial pain interference (yes) and the first 2 responses to indicate none or little pain (no). Items 6-8 ask about general mental health. Response choices for items 6-8 range from 1 (all of the time) to 5 (none of the time); responses were categorized into 2 groups by combining the first 3 responses to indicate the presence of that construct (yes) and the last 2 responses to indicate none or little (no). For item 9 response choices range from 1 (all of the time) to 5 (none of the time); responses were categorized into 2 groups by combining the first 3 responses to indicate the presence of that construct (yes) and the last 2 responses to indicate none or little (no). Cells with zero values indicate no reporting of stressful life events.

Table 4 Mean (standard deviation) of the stressful life events subscale scores across the child’s family history of specific medical conditions. Child family history Diabetes Yes

No

High total cholesterol p

Yes

No

Overweight/obesity p

Yes

No

p

34

(n = 158) Total score

(5.61)

Subsc ale 2 Subsc ale 3

(1.41) 0.8 (1.32)

Subsc

0.8 (1.08)

0.4 (0.87)

0 .399

0.3 (0.62)

0

(0.69)

49) Total score

(4.48)

Subsc ale 2 Subsc ale 3

(1.72)

Subsc

(1.20)

039

0.3 (0.70)

Depression

0 .209

1.6 (1.82) 1.6

0

(1.23)

(0.57)

.004

(2.42)

.015

0.7

0.3

0

2.1

0

(1.67)

(1.13)

(4.87)

.005

0.9

0.9

5.8

0

(1.87)

(1.29) 0.

Yes (n = 18)

1.2

0.9

.611

0.8

0.5 (0.79)

0

0.7

0 .052

0.3

0 .257

No (n =

p *

479)

(4.52)

(1.59)

.0003

p *

3.2

1.4

0

0 .606

(0.75)

355)

(3.41)

0.9

Eating disorders

(n =

3.7

.002

0.9 (1.57)

(0.61)

No

(n =

0

1.1

1.1 (1.25)

ale 4

(4.18)

1.0 (1.54)

Yes

142)

3.2

2.1 (2.19)

p *

448)

4.9

Subsc ale 1

(n =

.734

0 .021

(1.19)

0.3

Cancer

No

(n =

(1.25)

0

1.1

(1.51)

0.9

.280

0.3

Stroke Yes

0

(1.18)

(0.78)

(1.72)

0 .013

(1.78)

1.0

.909

0.8

0.3

.063

0

(1.49)

(1.26)

(1.82)

.042

0.9

0.9

*

3.1 (4.32)

1.5

0

(1.66)

(1.74)

.003

(3.97)

.034

1.1

1.0

0

3.8

0

(4.11)

(2.07)

.001

(n = 362)

3.1

1.5

0

(n = 135)

148)

(4.51)

.003

*

(n =

3.8

0

1.0

0.9 (1.45)

ale 4

(3.32)

1.3 (1.96)

(n = 249)

2.9

1.7 (2.35)

*

339)

4.3

Subsc ale 1

(n =

(1.42)

.004

0.3

0

(0.76)

.977

0.4 (0.70)

3.2 (4.19)

0 .009

1.2 (1.76)

0 .063

0.9 (1.56)

0 .064

0.8 (1.19)

0 .002

0.3 (0.71)

0 .206

Other mental illness (excluding anxiety)

Yes

No

(n =

(n =

106) Total score Subsc ale 1

p

*

Yes

No

(n = 24)

(n

391) 5.3

(5.39) 1.9 (2.34)

p* =

473) 2.8

(3.70) 1.0 (1.56)

< .0001 < .0001

5.5 (5.02) 1.7 (2.12)

3.2 (4.17)

0. 006

1.2 (1.78)

0. 150

35

Subsc ale 2

(2.09) Subsc

ale 3

*

0.7 (1.34)

1.2 (1.39)

Subsc ale 4

1.6

.0001

0.7 (1.13)

0.4 (0.87)

<

< .0001

0.3 (0.66)

0. 049

1.8 (1.88) 1.5 (1.53) 0.5 (0.88)

0.9 (1.54)

0. 006

0.8 (1.18)

0. 013

0.3 (0.70)

0. 249

For difference between the yes and no responses using the Wilcoxon 2-sample test with t-approximation.

Only family history conditions with one or more significant comparisons are presented. Stressful life events subscales: 1 = Social and financial readjustment; 2 = Change in habits and family dynamics; 3 = Social misconduct and work-time challenges; 4 = Death and change in relationships

1

Highlights 1)Development and Assessment of Stressful Life Events Subscales 2)Prevalence of Life Events measured using an adaptation from SRRS 3)Development of Stressful Life Events subscales using Exploratory Factor Analysis 4)Validation of Stressful Life Events subscales with MOSS, DASS21,Social Support scale 5)Validation of Stressful Life Events subscales with general health items

36