Cumulative Life Stress

Overview:

Measures that incorporate information about multiple stressors or risk factors, rather than looking at a single stressor or risk factor, are sometimes referred to as assessments of cumulative life stress (i.e., “cumulative life stress”).(1) Cumulative life stress may be assessed prospectively (e.g., with repeated measures across time) or cross-sectionally (e.g., with measures of current stress experiences and retrospective reports about the past). By assessing how stress accumulates, we can gain important insights into the role of stress on health and behavior.(2) The measurement of cumulative life stress builds upon self-report life event checklist measures that typically assess stressors occurring over the past year only. It does this by spanning a longer period of time (e.g., several years, the entire lifespan) and by incorporating information about both acute and chronic life stressors (e.g., acute events, work stress, relationship stress, discrimination, financial stress, neighborhood-related stress, etc.). Inherent in this approach is a temporal dimension of stress exposure that helps investigators appreciate how stressor exposure occurring across time affects health outcomes.

Background:

Psychosocial stressors, defined as social or contextual exposures that overwhelm the adaptive capacities of an individual (3) can negatively impact physical and mental health, and the accumulation of psychosocial stressors is thought to be especially harmful.(4) Across multiple disciplines, researchers have shown that the accumulation of stressors, or repeated exposure to the same stressor, is more strongly associated with poor health outcomes relative to that observed for single exposures.(4,5) Therefore, failure to assess stressors comprehensively can lead to biased estimates of the associations between stress and health. Accordingly, it is valuable to integrate information about multiple forms of stressors and across time(4), including both acute and chronic stressors. A number of approaches have been used to operationalize cumulative stress, including (1) the use of a broad selection of existing survey items that can be combined together, and (2) a single instrument designed to study this construct.(1,5-7) We discuss both types of measures, below.

Collection and measurement:

Formal measures of cumulative life stress

There are several measures of cumulative life stress, some of which are investigator-based life stress interviews (e.g., Life Events and Difficulty Schedule (8), UCLA Life Stress Interview (9)). These interviews are conducted by highly trained interviewers who focus on biographical details of the respondent and record objective characteristics of stressors that are later categorized and coded for severity according to a manual. Interviewer-based life stress interviews are considered to be a gold standard in terms of the detailed information solicited, with regard to frequency, duration, and impact. This approach was developed to address the conundrum of how to move beyond event frequency counts (which ignores impact and meaning to the individual) to attain a measure of severity while avoiding confounding of the respondents’ mental state with his/her answers. A primary reason and major strength of this approach are to contextualize the reporting of an event without reliance on the respondent’s own severity rating. While these approaches reflect a gold-standard, they often only cover a limited time frame, and they are costly given the need for highly trained interviewers and raters and the time required from participants, and for collection and coding by the investigators.

To address the high cost and practical challenges associated with using investigator-based systems for assessing life stress, Slavich developed an automated life stress interview that is easier to use and much more affordable than traditional investigator-based systems. This system called the Stress and Adversity Inventory (i.e., STRAIN(6)), is an online instrument specifically designed to assess cumulative lifetime stress exposure (for more information about the measure and applications, see https://www.STRAINsetup.com). There are two basic versions of the STRAIN: the Adolescent STRAIN(9), which focuses on 75 developmentally relevant stressors, and the Adult STRAIN(6), which focuses on 55 developmentally relevant stressors. Both systems employ extensive branching logic to administer follow-up questions similar to how a highly-qualified life stress interviewer would administer follow-up questions to capture information about when various stressors occurred across the lifespan, in addition to their severity, frequency, and duration. Based on the information collected, the STRAIN can produce 445 raw variables that can be combined to create more than 115 stress exposure scores and life charts that comprehensively summarize a person’s exposure to acute and chronic stress across the entire life course. The Adolescent STRAIN and Adult STRAIN both take approximately 18-25 minutes to completed (depending on the population being interviewed), are available in several languages, and have been extensively validated in relation to numerous different psychological, cognitive, biological, and clinical outcomes.(10-20)

Construction of Composite Scores using Existing Measures

In approaches that make use of a broad range of existing survey items (i.e., as opposed to a formal, existing life stress instrument), researchers have operationalized cumulative life stress by creating a composite index that includes multiple domains of stress (10-14). This approach is informed by empirical research documenting that the quantity of exposure may be the most important feature for identifying associations with health outcomes (in contrast to specific constellations of stressors) (4,5,12,15).

In this approach, each domain (or, type) of stressor tends to be weighted equally, thus avoiding assumptions about which may be more or less consequential for health. Although this is sometimes noted as a criticism of composite scores for assessing cumulative stress and health, there currently exists limited information for weighting the relative importance of component stressors, which could vary based on specific health outcomes, or for determining their effects on specific subgroups of the population (e.g., by age, sex, race/ethnicity, etc.). Furthermore, empirical studies of this issue suggest that weighting schemes do not provide any meaningful improvement (relative to equal weighting) for the prediction of health outcomes (10-12, 14, 17-19).

In the study of cumulative stress and health, individual researchers make decisions about which domains of stressors to include, how to combine information from multiple scales or contextual “levels” (e.g., home, school, neighborhood), and how they will handle issues related to accumulation across time. If researchers are creating scores by combining information from dissimilar scales, a variety of approaches have been used. We discuss two common approaches to achieve a single metric, below: (10-12,14,17-19).

Composite Index of Continuously Scored Stressors: in this method, researchers form a single composite metric with numerous continuous scores (each reflecting a single construct, measured by correlated items) by standardizing the component scores, e.g., to a z-score. Then, the z-scores can be combined and standardized once more for purposes of interpretation. This method has an advantage of preserving information about the severity of the stressor (as the components enter the composite metric on a continuous scale), but it is not able to incorporate risk factors that exist in a dichotomous format (e.g., history of school expulsion, below the federal poverty level, marital dissolution). Relative to the approach described below, another limitation relates to the issue of covariance. If the items do not show at least moderate correlations, there will be low reliability because the internal consistency will not be high. Please see Evans et. al (5). for a thorough discussion of the statistical issues related to this approach, and the problems created by combining variables that are not highly correlated.
Cumulative Risk Exposure Index: in this method, researchers can combine both continuous measures of stressors and dichotomous events or conditions to form a single composite metric measure of cumulative risk. To do this, researchers first transform the continuous measures into dichotomous indicators by following the method described above to standardize individual scores, and then establishing indicator variables based on some threshold (e.g., top quartile is one example). These indicators can then be combined with risk factors that naturally exist as dichotomous variables to create a count score for the number of risk factors. (12,15). A major strength of this approach (in contrast to the Composite Index of Continuously Scored Stressors describe, above) is that it does not require any assumptions about the degree of covariation between stressors.5 Numerous studies considering a range of health outcomes have shown that associations tend not to be influenced by the threshold selected to create indicators of high risk for each domain (e.g., top tertile, quartile, or quintile) (10,12,15).

Limitations and Future Directions:

There are a number of important future directions for the refinement of measurements of cumulative life stress, based on limitations of the existing measures. Some of the main points are summarized below. For detailed discussions of statistical and conceptual issues, see reviews by Evans et al. (5), Epel et al. (1), and Slavich (20).

A limitation of cumulative life stress and health literature is that the implications for prevention and intervention programs can become vague, obscured, or seemingly insurmountable. Proposed methods to address this issue are by (a) continuing to test the individual stressors in relation to health, which can identify domains with the strongest associations, (b) using latent class models to identify common profiles (or, patterns) of stressors that are most important for the study outcome, or (c) exploring recursive partitioning methods (21) which can designate risk across stressors based relationships with to the outcome (see Evans et al (5) for discussion).
Studies vary in terms of the particular stressors that are included in a score. Further research will be helpful to examine whether it is advantageous to develop cumulative life stress scores that are tailored to particular subgroups of the population, with regard to domains of stressors that are selected for inclusion in the assessment (e.g., parents, older adults, lower income samples, by race/ethnicity, etc.)
Future research should explore methods to incorporate variability in stressor exposure over time (i.e., duration/chronicity), and age at exposure (see Evans for detailed discussion(5)). At present, most scales of cumulative life stress are not designed to account for dynamic changes in stressors over time, and do not acknowledge interdependent relationships between risk factors that may lead to a cascade of exposures.
Additional methodological work is needed to extend our understanding of using additive versus multiplicative models to examine cumulative stress. A major unresolved issue in the operationalization of cumulative life stress relates to assumptions of linearity (versus non-linearity) in the relationship between number of risk exposures and outcomes. If it is nonlinear, this indirectly suggests there may be interactions (multiplicative effects); however, in studies that include more than 3 stressors, the sample size requirements become too large, and it is extremely challenging to make sense of the higher-level interactions.
Additional research is needed to identify whether cumulative life stress measures that rely on retrospective reports of childhood perform similarly relative to measures based on prospective measures (e.g., it is known that this is not the case for psychiatric outcomes (22,23), but this has not been explored across all types of health outcomes). If prospective and retrospective measures do not function similarly, it will be important to understand how and why these measurement approaches differ.
Finally, at the analysis phase, research is needed to explore the ideal way to integrate childhood and adulthood information (e.g., included together as a total cumulative life score, considered separately in a single model, modeled as an interaction, etc.) and how this may vary across health outcomes.

Author(s) and Reviewer(s): Prepared by Natalie Slopen, PhD, and Adolfo Cuevas, PhD. Reviewed by George Slavich, PhD, and Gary Evans, PhD. Please direct suggestions and feedback to Dr. Slopen (nslopen@umd.edu).

References:

Epel ES, Crosswell AD, Mayer SE, et al. More Than a Feeling: A Unified View of Stress Measurement for Population Science. Frontiers in neuroendocrinology. 2018;49:146-169; PMC6345505.
Cohen S, Kessler RC, Gordon LU. Measuring Stress: A Guide for Health and Social Scientists. Oxford University Press on Demand; 1997.
Cohen S, Janicki-Deverts D, Miller GE. Psychological Stress and Disease. Jama-Journal of the American Medical Association. 2007;298(14):1685-1687.
Turner RJ, Lloyd DA. Lifetime Traumas and Mental Health: The Significance of Cumulative Adversity. J Health Soc Behav. 1995;36(4):360-376.
Evans GW, Li D, Sepanski Whipple S. Cumulative Risk and Child Development. Psychological Bulletin. 2013;139(6):1342–1396.
Slavich GM, Shields GS. Assessing Lifetime Stress Exposure Using the Stress and Adversity Inventory for Adults (Adult Strain): An Overview and Initial Validation. Psychosomatic medicine. 2018;80(1):17-27.
Slopen N, Meyer C, Williams DR. Cumulative Stress and Health. The Oxford Handbook of Integrative Health Science. 2018:75.
Brown GW, Harris TO. Social Origins of Depression: A Study of Psychiatric Disorder in Women. New York: Free Press; 1978.
Hammen C, Rudolph K. Ucla Life Stress Interview for Children: Chronic Stress and Episodic Life Events. Manual University of Illinois. 1999.
Slopen N, Dutra LM, Williams DR, et al. Psychosocial Stressors and Cigarette Smoking among African American Adults in Midlife. Nicotine & Tobacco Research. 2012;14(10):1161-1169; PMC3457711.
Slopen N, Kontos E, Ryff C, Ayanian J, Albert M, Williams D. Psychosocial Stress and Cigarette Smoking Persistence, Cessation, and Relapse over 9–10 years: A Prospective Study of Middle-Aged Adults in the United States. Cancer Causes Control. 2013;24(10):1849-1863; PMC3776130.
Sternthal MJ, Slopen N, Williams DR. Racial Disparities in Health: How Much Does Stress Really Matter? Du Bois Review: Social Science Research on Race. 2011;8(01):95-113.
Slopen N, Williams DR. Discrimination, Other Psychosocial Stressors, and Self-Reported Sleep Duration and Difficulties. Sleep. 2014;37(1):147-156.
Lampert R, Tuit K, Hong KI, Donovan T, Lee F, Sinha R. Cumulative Stress and Autonomic Dysregulation in a Community Sample. Stress. 2016;19(3):269-279; PMC5072401.
Farrington DP, Loeber R. Some Benefits of Dichotomization in Psychiatric and Criminological Research. Criminal Behaviour & Mental Health. 2000;10(2):100.
Wainer H. Estimating Coefficients in Linear Models: It Don't Make No Nevermind. Psychological Bulletin. 1976;83(2):213-217.
Puterman E, Gemmill A, Karasek D, et al. Lifespan Adversity and Later Adulthood Telomere Length in the Nationally Representative Us Health and Retirement Study. Proceedings of the National Academy of Sciences. 2016;113(42):E6335-E6342.
Burroughs Pena MS, Mbassa RS, Slopen NB, Williams DR, Buring JE, Albert MA. Cumulative Psychosocial Stress and Ideal Cardiovascular Health in Older Women. Circulation. 2019;139(17):2012-2021; PMC6478505.
Brewer LC, Redmond N, Slusser JP, et al. Stress and Achievement of Cardiovascular Health Metrics: The American Heart Association Life's Simple 7 in Blacks of the Jackson Heart Study. Journal of the American Heart Association. 2018;7(11):e008855.
Slavich GM. Stressnology: The Primitive (and Problematic) Study of Life Stress Exposure and Pressing Need for Better Measurement. Brain Behav Immun. 2019;75:3-5; PMC6279572.
Kraemer HC, Lowe KK, Kupfer DJ. To Your Health: How to Understand What Research Tells Us About Risk. Oxford University Press; 2005.
Newbury JB, Arseneault L, Moffitt TE, et al. Measuring Childhood Maltreatment to Predict Early-Adult Psychopathology: Comparison of Prospective Informant-Reports and Retrospective Self-Reports. Journal of Psychiatric Research. 2018;96:57-64.
Baldwin JR, Reuben A, Newbury JB, Danese A. Agreement between Prospective and Retrospective Measures of Childhood Maltreatment: A Systematic Review and Meta-Analysis. JAMA psychiatry