Health Al

Location:

Tempe, AZ

Posted:

November 27, 2012

Contact this candidate

Resume:

YHBEH-*****; No. of pages: *; *C:

Hormones and Behavior xxx (2011) xxx xxx

Contents lists available at ScienceDirect

Hormones and Behavior

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y h b e h

Negative emotionality, depressive symptoms and cortisol diurnal rhythms:

Q4 1

Analysis of a community sample of middle-aged males

Leah D. Doane a,, Carol E. Franz b, Elizabeth Prom-Wormley c, Lindon J. Eaves c,

Sally P. Mendoza d, Dirk H. Hellhammer e, Sonia Lupien f, Hong Xian g, Michael J. Lyons h,

William Kremen b, i, j, Kristen C. Jacobson k

6 a

Department of Psychology, Arizona State University, PO Box 871104, Tempe, AZ 85287-1104, USA

7 b

Department of Psychiatry, University of California, San Diego, La Jolla, CA 92093, USA

Q2 8 c

Virginia Institute for Behavioral and Psychiatric Genetics, Virginia Commonwealth University, Richmond, VA 23298-0126, USA

9 d

Department of Psychology, University of California Davis, Davis, CA 95616, USA

10 e

Department of Psychobiology, University of Trier, Germany

11 f

Center for Studies on Human Stress, Centre de recherch Fernand-S guin, Universit de Montr al, Montreal, QC, Canada H1N 3M5

12 g

St. Louis VA Medical Center and Department of Internal Medicine, Washington University, St. Louis, MO 63106, USA

13 h

Department of Psychology, Boston University, Boston, MA 02215, USA

14 i

Center for Behavioral Genomics, University of California, San Diego, La Jolla, CA 92093, USA

15 j

VA San Diego Healthcare System, San Diego, CA 92161, USA

16 k

Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, USA

article info abstract

19 31

Article history: Prior research suggests that individuals with particular personality traits, like negative emotionality, are at

20 Received 21 October 2010 32

greater risk for adverse health outcomes. Despite bivariate associations between negative emotionality,

21 Revised 6 May 2011 33

depressive symptoms and the hypothalamic pituitary adrenal axis (HPA axis), few studies have sought to

22 Accepted 10 May 2011

34 Q3

understand the biological pathways through which negative emotionality, depressive symptomatology and

23 Available online xxxx

cortisol one of the primary hormonal products of the HPA axis are associated. The present study explored 35

whether negative emotionality in uenced cortisol dysregulation through current depressive symptomatol- 36

27 Keywords:

ogy and whether negative emotionality served as a moderator of the relationship between depressive

28 Cortisol

symptoms and cortisol. In the community-based Vietnam Era Twin Study of Aging, 783 male twins completed

29 Negative emotionality

30 two days of cortisol saliva sampling in their natural environments. Three measures of cortisol were analyzed:

Depressive symptomatology

waking levels, the cortisol awakening response, and the peak to bed slope. Depressive symptoms signi cantly 40

mediated the associations between negative emotionality and the peak to bed slope. A 2-way interaction

between depressive symptoms and negative emotionality was signi cant for the peak to bed slope and for 42

waking levels of cortisol. Exploration of the interactions illustrated that depressive symptoms only affected

cortisol slopes at average or high levels of negative emotionality and only affected waking levels at low levels

of negative emotionality. Negative emotionality and depressive symptoms were not related to the cortisol

awakening response. This is the rst study to nd indirect associations between negative emotionality and 46

peak to bed cortisol slopes through depressive symptoms. These ndings illustrate the complex interplay 47

between personality characteristics, depressive symptoms and different indices of the cortisol diurnal

rhythm.

2011 Published by Elsevier Inc.

54 52

55 61

Negative emotionality, depressive symptoms and cortisol diurnal negative emotionality were associated with risky health behaviors

(e.g. Caspi et al., 1997), in ammatory markers and subsequent disease

56 62

rhythms: mediating and moderating pathways

(Black, 2003; Sutin et al., 2010), and subjective well-being (e.g.

57 64

Prior research suggests that individuals with particular personality Friedman et al., 2010). Negative emotionality has been shown to be

58 65

traits, like negative emotionality, are at greater risk for adverse health stable over adulthood and consistent across situations; thus it has

been conceptualized as trait marker of vulnerability toward general

59 66

outcomes (for reviews see Friedman, 2000; Kern and Friedman, 2011;

60 67

Smith and Gallo, 2001). Studies have found that higher levels of distress.

An indicator of health that has not been closely examined in

relation to negative emotionality is the hypothalamic pituitary

adrenal axis (HPA axis). The HPA axis is one of the body's major

Corresponding author. Fax: +1-480-***-****.

stress responding systems. This association is particularly important

E-mail address: ****.*****@***.*** (L.D. Doane).

0018-506X/$ see front matter 2011 Published by Elsevier Inc.

doi:10.1016/j.yhbeh.2011.05.003

Please cite this article as: Doane, L.D., et al., Negative emotionality, depressive symptoms and cortisol diurnal rhythms: Analysis of a

community sample of middle-aged males, Horm. Behav. (2011), doi:10.1016/j.yhbeh.2011.05.003

L.D. Doane et al. / Hormones and Behavior xxx (2011) xxx xxx

as negative emotionality may in uence how an individual perceives in uences cortisol dysregulation through current depressive symp-

72 134

tomatology. In addition, previous studies have not speci cally

73 135

the stressors in their environment. The HPA axis reacts to stressors in

74 136

the environment by activating a complex cascading of events, examined whether the relationship between depressive symptoms

and HPA axis dysregulation is modi ed by personality characteristics

75 137

eventually resulting in the release of cortisol. Under normal

circumstances (i.e., in the absence of speci c stressors), cortisol

76 138

such as negative emotionality. To the extent that negative emotion-

follows a daily rhythm. Speci cally, cortisol follows a diurnal pattern

77 139

ality can be considered a marker of a liability to general distress, it

78 140

by which cortisol levels are high upon awakening in the morning, may exacerbate the effects of depressive symptomatology on cortisol.

79 141

peak about thirty minutes after waking (commonly called the cortisol Thus, experiences of depressive symptomatology may be more

80 142

awakening response; CAR), and decline across the waking day strongly associated with HPA axis dysregulation among individuals

81 143

(Kirschbaum and Hellhammer, 1989; Pruessner et al., 1997). Prior with high levels of negative emotionality.

82 research has found elevations of cortisol in relation to negative or

83 depressed mood in non-clinical populations (Knight et al., 2010;

144

The present study

84 Pruessner et al., 2003; Sj gren et al., 2006) and more persistent,

Q6 85 chronic life stressors (Miller et al., 2007), but its associations with the

145

In this study we examined the potential pathways among negative

86 personality trait negative emotionality has yet to be examined.

146

emotionality, depressive symptoms and indices of the cortisol diurnal

147

rhythm. The present study extends previous work by considering the

87 Negative emotionality and cortisol

148

joint pathways through which negative emotionality and depressive

149

symptoms are associated with diurnal rhythms of cortisol and by

88 To our knowledge, there are no studies of adults that have

150

testing whether negative emotionality moderates the association

89 investigated associations between negative emotionality and cortisol.

151

between depressive symptoms and diurnal rhythms of cortisol in a

90 However, negative emotionality is closely associated with neuroti-

152

large sample of community-dwelling men. We focused on three

91 cism (Tellegen, 1985), and is linked with broad negative affectivity

measures of the diurnal pattern of cortisol waking levels, slope of 153

92 (Patrick et al., 2002), which both have been linked with HPA axis

the diurnal rhythm across the day and the CAR because of the ndings 154

93 activity. Although not all studies have found differences in diurnal

155

between these parameters and major depression or depressive

94 cortisol rhythms in subjects high in neuroticism as compared to those

156

symptoms in prior research (as outlined above). We hypothesized

Q7 95 with low neuroticism (Kirschbaum et al., 1992a, 1992b; Schommer et

157

the following: 1) negative emotionality and depressive symptoms

96 al., 1999; van Santen et al., 2010), neuroticism has been associated

158

would be independently associated with indicators of the cortisol

97 with differences in HPA axis regulation in everyday life (e.g. diurnal

159

diurnal rhythm; 2) depressive symptoms would partially mediate the

98 measures; Hauner et al., 2008; Schlotz et al., 2006). Some studies have

160

associations between negative emotionality and cortisol; and 3) levels

99 found increased levels of cortisol at waking in individuals high in

161

of negative emotionality would moderate the associations between

100 neuroticism (Portella et al., 2005; Wetherell et al., 2006), while others

162

depressive symptoms and cortisol such that the relationship between

101 reported positive associations between neuroticism and higher

163

depressive symptoms and cortisol dysregulation is higher among

102 evening levels of cortisol (Gerritsen et al., 2009). A recent study

164

people with higher levels of negative emotionality.

103 found higher average level of cortisol across the whole day in subjects

104 high in neuroticism (Nater et al., 2010).

165

Method

105 Relationships among depression, negative emotionality and cortisol

166

Participants

106 There is evidence that personality traits are associated with

107 167

depression (Fanous et al., 2007; Kahn et al., 2005, Kendler et al., The sample for this study included a subset of male monozygotic

108 168

1993). Because negative emotionality is a propensity to experience and dizygotic twins who took part in the Vietnam Era Twin Study of

109 169

psychological distress that has been shown to be stable over Aging (VETSA). The VETSA has been described in detail elsewhere

110 170

adulthood and consistent across situations, it has been conceptualized (Kremen et al., 2006). VETSA twins were randomly selected for

as a trait marker of vulnerability toward general distress and may

111 171

recruitment from a pool of 3322 twin pairs in the Vietnam Era Twin

112 172

predispose an individual to experiencing higher levels of depressive Registry who served in the US military at some time during the

Vietnam era (1965 1975), were aged 51 60 during the VETSA data

113 173

symptoms. Therefore, negative emotionality may impact HPA axis

114 174

regulation through higher levels of depressive symptomatology. In collection, and who took part in a prior study of psychological health

115 175

addition, there is both cross-sectional and prospective evidence that in 1992 (Tsuang et al., 2001). These individuals were not selected as

116 176

cortisol is associated with depressive symptoms, although relation- Veteran Affairs patients and most were never in combat or in Vietnam

117 177

ships with salivary cortisol may be weaker (for review see Stetler and (Kremen et al., 2006). There were 1237 participants in the VETSA

118 178

Miller, 2011). van den Bergh et al. (2008) argued that by looking at (mean age = 55.9, SD = 2.58). Eighty-eight percent of the sample was

119 179

symptom levels rather than simply the presence of a disorder we may Caucasian, 4.3% African-American, 2.9% Hispanic, .9% Native American,

and .4% Paci c Islander. Compared to national data for men in their

120 180

be able to identify pre-disease pathways in all parts of the depressive

symptoms continuum by using biological markers as vulnerability 181 Q8

121 50s (National Health and Nutrition Examination Survey, NHANES III,

markers. Indeed, Pruessner et al. (2003) found that higher levels of 182 Q9

122 2007; National Center for Disease Statistics. Health and United States,

123 183

depressive symptoms were associated with greater cortisol awaken- 2003), the VETSA participants are similar to the larger US population

124 184

ing responses, and others report associations between depressed of men in this age range based on demographic (age, education,

mood and atter diurnal cortisol rhythms (Knight et al., 2010; Sj gren

125 185

income, marital status, employment) and health data (prevalence of

126 186

et al., 2006). A recent meta-analysis investigating chronic stress and chronic health problems, diabetes, and hypertension). Two years after

127 187

alterations in HPA-axis regulation argued that even when a person the VETSA study began, an additional study was initiated to

128 188

does not develop a psychiatric condition, greater emotional distress is understand the role of cortisol and other hormonal dysregulation as

associated with atter diurnal cortisol (Miller et al., 2007).

129 189

a risk factor for cognitive aging. Approximately one-third (N = 442) of

130 190

Despite bivariate associations between negative emotionality, VETSA participants had already been studied and were ineligible for

131 191

depressive symptoms and cortisol, no studies have examined these the additional cortisol study. Of the remaining 795 subjects, 786

132 192

three constructs simultaneously. In the present study we examine (98.9%) participated in the hormone data collection study between

133 193

whether a stable personality characteristic, negative emotionality, 2005 and 2007. The analyses reported herein were non-twin analyses.

Please cite this article as: Doane, L.D., et al., Negative emotionality, depressive symptoms and cortisol diurnal rhythms: Analysis of a

community sample of middle-aged males, Horm. Behav. (2011), doi:10.1016/j.yhbeh.2011.05.003

L.D. Doane et al. / Hormones and Behavior xxx (2011) xxx xxx 3

194 256

Procedures (Tellegen, 1985). The stress reaction scale has 14 items and assesses

257

whether someone is prone to negative emotions or is easily worried

or anxious ( = .86). The alienation scale has 17 items and detects

195 258

Participants completed salivary collection in their naturalistic

settings on two non-consecutive typical working days (for full

196 259

whether someone often feels victimized, mistreated or pushed around

( = .83). Lastly, the aggression scale has 18 items and assesses

197 260

description of saliva collection procedures see Franz et al., 2010).

198 261

Cortisol kits were mailed via courier and participants received a whether someone enjoys aggression, is physically aggressive, or is

vindictive ( = .75). These three subscales were summed to create a

199 262

reminder call the day before they were to begin sampling to ensure

200 263

that materials were received, to go over procedures and to answer any total NEM score with a possible range of 0 to 49.

201 questions. Materials included: 4.5 mL Cryotube vials, original Trident

202 sugarless gum, straws, tissues, detailed instructions, a daily log, pen, a

264

Depressive symptoms

203 reminder watch set for all of the sampling times, and a storage

265

Depressive symptoms were assessed using the Center for

container with a MEMS 6 (Aardax) track cap for detecting

204

266

Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977). The

205 compliance with protocol. The materials were tested to ensure that

267

scale is made up of 20 items designed to measure levels of depressive

206 they did not alter the assays. On each day participants were asked to

268

symptoms experienced in the past week. Participants circled re-

207 provide samples at waking, 30 min post waking, 10:00 am, 3:00 pm,

269

sponses ranging from 0 to 3, with 0 indicating that they had

208 and 9:00 pm or bedtime for a total of 10 samples per participant.

experienced that symptom rarely or b 1 day of the week and 3 270

209 Participants with non-standard schedules (i.e., night shift workers)

271

indicating that they had experienced that symptoms most of that

210 provided samples at equivalent time periods based on their wake

week or 5 7 days. The CES-D has been checked for reliability across 272

211 time. At each sampling point, participants provided a saliva sample,

273

several age groups, has good internal consistency within the VETSA

212 placed it in the storage container (with the track cap so the entry was

sample ( = .90), and has been found to be highly correlated with 274

logged) and lled out a written log reporting on their mood, food and

213

275 Q10

indicators of major depression based on accepted cut-off scores (Rush

214 drink consumption, medication use, alcohol use and whether or not

276 Q11

et al., 2000; Shafer, 2006). Items were summed to create a composite

215 they smoked or exercised in the last hour. Lastly, participants also

277

scale with a possible range of 0 to 60. This measure was positively

lled out an extensive psychosocial questionnaire at home and

216

278

skewed (skewness = 1.72) and therefore was log transformed prior to

217 brought the questionnaire to their study site.

279

analysis.

218 Measures

280

Covariates

219 Cortisol 281

As part of the cortisol collection procedure, participants were

Participants provided ve samples a day for two days. Cortisol was

220

282

asked to record whether they had smoked, eaten, drunk alcohol, taken

221 collected by passive drool until the participant had provided at least 283

medication, or exercised before each sample. Responses at each time

222 2.25 mL of saliva. If necessary, they chewed Trident gum to stimulate 284

were coded as 0 (no) or 1 (yes). In accordance with prior research

223 saliva and removed the gum prior to providing the sample. 285 Q12

(Kirschbaum et al., 1992a, 1992b), smoking prior to the sample

224 Participants stored and refrigerated their saliva samples in an 286

collection was the only variable consistently related to cortisol level.

225 insulated lunch bag included with the cortisol kits. At the end of the 287

In addition, data were collected on several other variables known to

226 two days, they sent the saliva samples via overnight mail to the 288

be associated with cortisol including wake time and number of hours

227 University of California, Davis to be assayed. Salivary assays were 289

slept prior to the days of sampling (see Doane et al., 2010 for details).

228 estimated in duplicate using commercial radioimmunoassay kits

229 (Siemens Medical Solutions Diagnostics, Los Angeles, CA). The

230 290

sensitivity of the cortisol assay is 1.39 nmol/L and the intra- and Data preparation and manipulation

inter-assay coef cients of variation are 3.96 and 5.66 respectively. All

231

232 291

saliva samples from each individual participant were assayed together Cortisol follows a predictable diurnal rhythm with relatively high

levels at wake time, followed by an increase of 50 60% in the 20

233 292

in batches containing one to three individuals. Three individuals were

234 293

excluded due to lost (N = 1) or contaminated (N = 2) samples, 40 min after waking (known as the cortisol awakening response), a

resulting in a nal sample of N = 783. There was very little missing

235 294

rapid decline in the subsequent few hours, and a slower decrease

236 295

data due to participant lapses or technical problems (~ 1.0%) although throughout the remainder of the day to reach the lowest point near

237 296

0.1% of samples were re-coded as missing because they had values midnight (Pruessner et al., 1997). We simulated this growth curve

238 297

greater than 50 nanomoles per liter (nmol/L; Hellhammer et al., 2009; pattern using the statistical program WinBUGS (Spiegelhalter et al.,

239 298

Nicolson, 2008; see Franz et al., 2010 for more details). Individual 2004). WinBUGS uses Markov Chain Monte Carlo (MCMC) which

240 299

cortisol values were positively skewed (skewness range = 1.7 to 20.1) allowed us to control for the non-independence of samples intro-

241 300

and were log transformed to approximate normality. Analyses duced when individuals were assayed in batches, and further allowed

242 301

focused on three indices of cortisol patterns throughout the day, as us to include nicotine use as a time-varying covariate in the analysis.

243 302

described in detail below. To account for the batch effects, we set the ith assayed value of the jth

303

twin by Yijk, where the subscript k further indicates the batch in which

244 304

Negative emotionality the value was assayed such that assays in the same batch have the

245 305

Negative emotionality was assessed using Tellegen's Multidimen- same value of k. We used each participant's log-transformed cortisol

246 306

sional Personality Questionnaire (MPQ) factor-form NZ; the NZ values at each time point to model the waking value of cortisol

247 307

version is considered to be very similar to the Brief Form (Caspi, (WAKE), the cortisol slope from wake up to the peak (cortisol

248 308

2000; Caspi et al., 1997; Krueger et al., 2000; Patrick et al., 2002). awakening response, CAR), and the exponential decay after the peak

249 309

Validity of the psychometric properties of the MPQ is well documen- until bedtime or peak-to-bed slope for each day, while also accounting

250 310

ted (Krueger et al., 2000; Patrick et al., 2002; Tellegen, 1985). The for batch and nicotine use. A larger CAR indicates a greater difference

251 311

MPQ is a self-report questionnaire and consists of 11 scales. The three between waking and peak levels of cortisol at the wake and wake + 30

samples, while a lower PTB slope indicates atter cortisol slopes

252 312

subscales of interest in this study are Stress Reaction, Alienation, and

253 313

Aggression, which when summed together make up the broad trait of across the day. Parameter estimates for WAKE, CAR, and PTB for each

254 314

Negative Emotionality (NEM). This trait was our primary interest for participant were then exported for use as the main dependent

255 315

these analyses because of overlap with measures of trait neuroticism variables in all analyses.

Please cite this article as: Doane, L.D., et al., Negative emotionality, depressive symptoms and cortisol diurnal rhythms: Analysis of a

community sample of middle-aged males, Horm. Behav. (2011), doi:10.1016/j.yhbeh.2011.05.003

L.D. Doane et al. / Hormones and Behavior xxx (2011) xxx xxx

339

used to account for the correlated observations of participants from

340

the same family (i.e., twins within pairs).

Mediation models were t using MPlus 3.0 (Muth n and Muth n, 341 Q13

1998 2004). Maximum likelihood estimation was used and absolute 342

model t was assessed with the chi-squared statistic, the comparative 343

t index (CFI) and the root mean square error of approximation 344

(RMSEA). The criterion of acceptable t was N.95 for the CFI and b.06 345

346

RMSEA (Hu and Bentler, 1999).

347

Moderation

The role of NEM as a signi cant moderator of the depressive 348

symptoms cortisol associations was tested using three-level hierar- 349

350

chical linear models to account for the nesting of days within persons

Contact this candidate