Electronic Supplementary Material
associated with
“Does Exposure to Stressors Predict Changes in Physiological Dysregulation?”
Table of Contents
Additional Information Regarding Measures 2
Cohort Selection and Attrition 7
Predictors of having valid data for stress indicators 7
Predictors of remaining in the analysis (given valid data for stress indicators) 7
Sensitivity Analyses 9
Inclusion of controls for baseline health status and changes in medication use 9
Alternative formulations for measuring physiological dysregulation 9
References 10
List of Tables and Figures
Table S1. Various formulations for measuring physiological dsyregulation 2
Table S2. Indicators of stressful events, trauma, and chronic strain, weighted estimates 4
Table S3. Items used to construct index of social integration 5
Table S4. Descriptive statistics for outcomes and potential confounders, weighted estimates 6
Figure S1: Selection process for the SEBAS cohort examined in 2000 and 2006 7
1
Additional Information Regarding Measures
Table S1. Various formulations for measuring physiological dsyregulation
/ (1) / (2) / (3) / (4) / (5) / (6) // Formulation
used in this
study
(17 markers) / Alternative
(20 markers) / Alternative
(quartile
cutoffs for all
20 markers) / Z-score
Formulation
(20 markers) / Formulation used
by Glei et al.
(2007)
(16 markers) / Original
(10 markers) /
Definition of high risk / Clinical cutoffs
(where available);
Otherwise, high
risk quartileb
(bottom 25% for
DHEAS, IGF1,
CrCl, & Albumin;
Top 25% for all
others) / Clinical cutoffs
(where available);
Otherwise, high
risk quartile
(bottom 25%
for DHEAS, IGF1,
CrCl, & Albumin;
Top 25% for all
others) / Bottom 25% for
HDL, DHEAS,
IGF1, CrCl, &
Albumin;
Top 25% for
all others / Each marker is
transformed as
indicated and
standardized / Top & Bottom 10%
for all but TC/HDL
(Top 10%) &
DHEAS
(Bottom 10%) / Bottom 25% for
HDL & DHEAS;
Top 25% for
all others
Metabolic
SBP (mmHg) / SBP 130 OR
DBP 85a / >140c / 148 / ln(SBP) / 113 OR 163 / 148
DBP (mmHg) / >90c / 90 / ln(DBP) / 70 OR 97 / 90
TC (mg/dL) / 160 OR 254
HDL (mg/dL) / <40(M)/<50(F)a / <40d / 39 / -[ln(HDL)] / 39
Ratio TC/HDL / 5e / 5.1 / ln(TC/HDL) / 6.1 / 5.1
Triglycerides (mg/dL) / 150a / 200d / 147 / ln(TG) / 54 OR 202
Fasting glucose (mg/dL) / 110a / 84 OR 126
HbA1c (%) / >6.5f / 5.8 / / 4.8 OR 6.7 / 5.8
BMI / <18.5 or 27g / 26.8 / ln(BMI) / 20.4 OR >29.0
Waist circumference (cm) / >102(M)/>88(F)a / >102(M)/>88(F)a / 91.5 / untransformed
Waist-hip ratio / 0.79 OR 0.96 / 0.92
Inflammation
IL-6 (pg/mL) / >3.69 / >3.69 / >3.69 / ln(IL-6) / 0.70 OR 5.80
CRP (mg/L) / >3 / >3 / 2.75 / ln(CRP)
sICAM-1 (ng/mL) / >290.9 / >290.9 / >290.9 /
sE-selectin (ng/mL) / >56.5 / >56.5 / >56.5 / ln(sE-selectin)
Neuroendocrine
DHEAS (μg/dL) / 42.2 / 42.2 / 42.2 / / 21.6 / 42.2
Cortisol (μg/g creatinine) / 28.2 / 28.2 / 28.2 / ln(cortisol) / 8.3 OR 45.3 / 28.2
Epinephrine (μg/g creatinine) / 4.44 / 4.44 / 4.44 / ln(epinephrine) / 1.51 OR 6.31 / 4.44
Norepinephrine (μg/g creatinine) / 26.4 / 26.4 / 26.4 / ln(norepinphrine) / 11.1 OR 33.2 / 26.4
Dopamine (μg/g creatinine) / 91.9 OR 230.2
Other markers
IGF-1 (ng/mL) / 73 / 73 / 73 / / 53.9 OR 169.6
Creatinine Clearance (ml/min) / 52.6 / 52.6 / 52.6 /
Albumin (g/dL) / 4.4 / 4.4 / 4.4 /
Homocysteine (μmol/L) / 16.7 / 16.7 / 16.7 / ln(homocysteine)
Abbreviations: BMI = body mass index; CrCl = creatinine clearance; CRP = C-reactive protein; DBP = diastolic blood pressure; DHEAS = dehydroepiandrosterone sulfate; F = Females; HbA1c = glycoslyated hemoglobin; HDL = high-density lipoprotein cholesterol; IGF-1 = insulin-like growth factor 1; IL-6 = interleukin-6; M = Males; SBP = systolic blood pressure; sE-selectin = soluble E-selectin; sICAM-1 = soluble intercellular adhesion molecule 1; TC = total cholesterol.
a Based on the cutoffs used to define the ATPIII clinical criteria for metabolic syndrome (National Cholesterol Education Program, National Heart, Lung and Blood Institute, 2001, Table II.6-1) and the cutoff for CRP is based on Pearson et al. (2003).
b Quartile cutoffs are based on the weighted distribution in 2000 among the longitudinal cohort (n=639).
c Based on JNC 7 cutoffs for stage I hypertension (Chobanian et al., 2003).
d Based on ATP III cutoffs for low HDL and high triglycerides (National Cholesterol Education Program, National Heart, Lung and Blood Institute, 2001).
e Based on the American Heart Association recommendation (http://www.heart.org/HEARTORG/Conditions/Cholesterol/AboutCholesterol/What-Your-Cholesterol-Levels-Mean_UCM_305562_Article.jsp).
f Based on the AACE guidelines (Rodbard et al., 2007).
g Based on the BMI cutoffs used by the Taiwan Department of Health (Department of Health, 2002).
1
Table S2. Indicators of stressful events, trauma, and chronic strain, weighted estimates
Items by Survey Wave / SummaryMeasure
Variable (observed range) / 1996 / 1999 / 2000
A) Stressful life events
Marital disruption, % / 16.3a / 2.6b / 1.8b
Death of child, % / 15.1a / 0.9b / —
Moved to a new residence, % / 7.4c / 7.6b / 1.9b
R/spouse/children victimized by crime/fraud in past year, % / — / — / 13.1
Total number of events (0-5), mean (SD) / 0.7 (0.9)
B) Trauma
Substantially affected by the 1999 earthquaked, % / — / — / 26.4
C) Chronic strain/perceived stress
1) Financial
Difficulty meeting living expenses (0-3)e, mean (SD) / 1.2 (0.6) / 1.2 (0.7) / 1.2 (0.7)
Financial decline (0-4)f, mean (SD) / 2.1 (0.8) / 2.4 (0.8) / —
Perceived stress, own finances (0-2)g, mean (SD) / — / 0.4 (0.6) / 0.4 (0.6)
Summary index (-1.4 to 2.3)h, mean (SD) / 0.0 (0.7)
2) Family-related
Spouse in “not so good” or “poor” health, % / 18.0 / 13.1 / 19.6
Perceived stress (0-2)g regarding:
Getting along with family members, mean (SD) / 0.1 (0.3) / 0.1 (0.3)
Family members’ health, financial situation, job, marriage, etc., mean (SD) / 0.5 (0.7)
Family members’ health, mean (SD) / 0.3 (0.6)
Family members’ financial situation, mean (SD) / 0.3 (0.6)
Family members’ job, mean (SD) / 0.3 (0.6)
Family members’ marital situation, mean (SD) / 0.3 (0.6)
Summary index (-0.5 to 2.2)h, mean (SD) / 0.0 (0.5)
Overall chronic strain (-1.9 to 3.5)i, mean (SD) / 0.0 (1.0)
Abbreviations: R = respondent
a Percentage who ever experienced the event (by the 1996 survey).
b Percentage who experienced the event since the previous survey wave.
c Percentage who moved in the past three years.
d Earthquake resulted in injury of the respondent, injury or death of a family member or close friend, temporary displacement of the respondent, or damage to the respondent’s home or other property.
e Response categories were: enough money, with some left over (0); just enough money, no difficulty (1); some difficulty (2); much difficulty (3).
f Respondent’s perception of their current economic situation compared with three years ago (0=much better; 1=better; 2=about the same; 3=worse; 4=much worse).
g Respondents were asked to report their level of stress or anxiety (0=none, 1=some, 2=a lot) related to each specified situation.
h Each item was standardized (mean=0, SD=1) and then we calculated the mean across valid items if at least half of the items had valid data (α=0.78 for 7 financial items; α=0.68 for 10 family-related items).
i Calculated by summing the two indices for financial and family-related chronic strain.
1
Table S3. Items used to construct index of social integration
Indicator / Definition / CodingNetwork size / Number of friends and relatives with whom the respondent lives or has regular contact / Recoded <5, 5-7, 8-10, 11-14, 15-19, 20-29, 30+.
Network range / Number of types of relationships in social network / One point each for spouse/partner, kids, other relatives, non-relatives; range=0-4.
Married/partner / Dummy indicating that the respondent is married or lives with a companion.
Household size / Top-coded at 10+.
Does not live alone / Dummy indicating that the respondent does not live alone.
Number of friends / Number of close friends and neighbors with whom the respondent has weekly contact / Recoded 0, 1-2, 3-4, 5-9, 10-19, 20+.
Number of children / Top-coded at 6+.
Number of grandchildren / Recoded 0, 1-2, 3-5, 6-9, 10-14, 15+.
Religious attendance / How often the respondent attends church or temple / Response categories: never, rarely, sometimes, often.
Socializing / How often the respondent socializes with friends, neighbors, or relatives. / Response categories: never, less than once a month, two to three times a month, once or twice a week, nearly daily.
Participation in social organizations / Whether respondent participates in the following activities/organizations:
1) Group activities (e.g., singing, dancing, tai chi, or karaoke)
2) Neighborhood association (e.g., women’s association or arts & crafts classes)
3) Religious organization (e.g., church or temple committee)
4) Occupational associations for farmers, fishermen, or other professional group, civic group, Lion’s Club, etc.
5) Political association (e.g., political party)
6) Social service groups (e.g., Lifeline, relief association, benevolent societies, charities, etc.)
7) Village or lineage association
8) Elderly club (e.g., Elderly Association, Evergreen Recreation Club, etc.) / One point for each type of activity/organization in which the respondent participates; range = 0-8.
1
Table S4. Descriptive statistics for outcomes and potential confounders, weighted estimates
Variable (observed range) / Total(n=539) / Men
(n=315) / Women
(n=224)
Demographic Characteristics
Age in years (54-86), mean (SD) / 64.9 (7.2) / 64.9 (7.2) / 64.8 (7.2)
Male, % / 58.1 / -- / --
Urban resident, % / 44.2 / 40.9 / 48.8
Psychosocial Resources
Years of completed education (0-17), mean (SD) / 5.6 (4.5) / 6.9 (4.5) / 3.9 (4.1)
Social integration in 1999 (-1.6 to 1.3), mean (SD) / 0.1 (0.5) / 0.1 (0.5) / 0.1 (0.4)
Perceived support in 1999 (0.5-4.0), mean (SD) / 3.2 (0.7) / 3.2 (0.8) / 3.2 (0.7)
Personal mastery in 2000 (0-3)a, mean (SD) / 1.6 (0.5) / 1.7 (0.4) / 1.5 (0.5)
Physiological dysregulation
Overall dysregulation in 2000 (0-13), mean (SD) / 4.7 (2.4) / 4.1 (2.1) / 5.4 (2.5)
Overall dysregulation in 2006 (0-14), mean (SD) / 5.1 (2.4) / 4.7 (2.3) / 5.8 (2.4)
Change in overall dysregulation (-6 to 7), mean (SD) / 0.5 (2.3) / 0.5 (2.2) / 0.4 (2.3)
Decreased by two or more points, % / 17.6 / 15.2 / 20.8
Little (+/- one point) or no change, % / 51.4 / 53.7 / 48.3
Increased by two or more points, % / 31.0 / 31.1 / 30.9
Metabolic in 2000 (0-5), mean (SD) / 1.6 (1.2) / 1.4 (1.1) / 2.0 (1.3)
Metabolic in 2006 (0-5), mean (SD) / 1.6 (1.2) / 1.4 (1.0) / 1.9 (1.3)
Change metabolic (-3 to 3), mean (SD) / 0.0 (1.0) / 0.0 (1.0) / -0.1 (1.1)
Decreased by one or more points, % / 28.7 / 25.5 / 33.2
No change, % / 43.4 / 45.8 / 40.1
Increased by one or more points, % / 27.9 / 28.8 / 26.8
Inflammation in 2000 (0-4), mean (SD) / 1.0 (1.1) / 0.9 (1.1) / 1.0 (1.1)
Inflammation in 2006 (0-4), mean (SD) / 1.1 (1.1) / 1.1 (1.2) / 1.0 (1.1)
Change in inflammation (-4 to 4), mean (SD) / 0.1 (1.2) / 0.2 (1.2) / 0.0 (1.1)
Decreased by one or more points, % / 23.3 / 21.5 / 25.9
No change, % / 45.7 / 45.2 / 46.4
Increased by one or more points, % / 31.0 / 33.3 / 27.8
Neuroendocrine in 2000 (0-4), mean (SD) / 1.0 (1.0) / 0.6 (0.8) / 1.5 (1.1)
Neuroendocrine in 2006 (0-4), mean (SD) / 1.2 (1.0) / 0.9 (0.9) / 1.6 (0.9)
Change in neuroendocrine (-3 to 4), mean (SD) / 0.2 (1.1) / 0.3 (1.1) / 0.1 (1.3)
Decreased by one or more points, % / 23.8 / 18.9 / 30.7
No change, % / 37.3 / 43.1 / 29.3
Increased by one or more points, % / 38.9 / 38.0 / 40.0
a Higher values indicate greater personal mastery (i.e., internal locus of control).
Cohort Selection and Attrition
Figure S1 shows detailed information about the process of selection and attrition for the SEBAS longitudinal cohort. Attrition includes losses resulting from mortality, loss to follow-up, ineligibility for the physical examination, refusal to participate in the exam, and non-response.
Predictors of having valid data for stress indicators
Among those selected for SEBAS and living within the region served by contract hospitals (n=1698), 1497 respondents participated in the interview and 1382 (81% of the targeted sample) provided valid data on the stress indicators. A logit model comparing those with valid data (n=1382) with the remainder (n=316) indicates that attrition was higher for respondents aged 75 and older, those with limitations in activities of daily living (ADL), and those who had a history of cancer (data not shown). On the other hand, attrition was lower for respondents with more social connections and those with better cognitive function (in 1999). Sex, education, and income did not predict attrition.
A multinomial logit model comparing those with valid data to those who died (n=76), those lost to follow-up (n=125), and those missing data for one of the stress indicators (n=115) revealed that a history of cancer was associated with higher mortality, but not loss to follow-up or non-response. Poor cognitive function predicted non-response (in part because subjective questions were not asked if the interview was completed by proxy), but not mortality or loss to follow-up.
Predictors of remaining in the analysis (given valid data for stress indicators)
Among those with valid data on the stress indicators (n=1382), 406 did not participate in the 2000 exam, another 167 died before the 2006 SEBAS, 86 were lost to follow-up in 2006, 107 did not participate in the 2006 exam, and 77 were missing data for physiological dysregulation in 2000 or 2006 or one of the confounders. A logit model comparing those remaining in the analysis (n=539) with the remainder (n=843) indicates that attrition was higher for older and less educated respondents. Attrition was also more common for those interviewed by proxy, people who had ever had a stroke, those with an ADL limitation, individuals with diabetes, those with fewer social connections, and those with cognitive impairments. Nonetheless, the stress indicators (life events, earthquake exposure, chronic strain) were not significantly associated with attrition.