Journal of Adolescent Health 44 (2009) 452–457

Original article

Bummed Out Now, Feeling Sick Later: Weekday versus Weekend

Negative Affect and Physical Symptom Reports in High School Freshmen

Jorie M. Butler, Ph.D.a,*, Carol K. Whalen, Ph.D.b, and Larry D. Jamner, Ph.D.b

aUniversity of Utah, Salt Lake City, UtahbUniversity of California at Irvine, Irvine, California

Manuscript received April 25, 2008; manuscript accepted September 2, 2008

Abstract Purpose: This study examined adolescent negative affect (NA) in daily life on school days and week-

*Address correspo

East, Room 502, Salt

E-mail address: jo

1054-139X/09/$ – see

doi:10.1016/j.jadoheal

end days during the spring and associations with physical symptoms during the following summer.

Methods: Using experience sampling methodology (ESM), participants provided electronic diary

(eDiary) reports of NA on weekdays (Thursday and Friday) and weekend days during their 9th grade

year. In telephone interviews during the winter and summer months they reported physical symptoms.

Multiple regression analyses were conducted to examine associations between weekday NA, weekend

NA, and their interaction and four constellations of physical symptoms reported in summer (pain,

respiratory, gastrointestinal, and immune symptoms).

Results: Findings indicated that weekend NA was associated with later reports of pain, respiratory,

and immune symptoms. For gastrointestinal symptoms only adolescents who reported low NA on both

weekend and school days reported fewer gastric symptoms than other adolescents.

Conclusions: Mapping the predictors and correlates of weekend NA may be important not only for

understanding teenage mood patterns but also for enhancing the interpretation of physical symptom

reporting by adolescents � 2009 Society for Adolescent Medicine. All rights reserved.

Keywords: Negative affect; Daily diary; Physical symptoms; School days; Weekend days; High school

Negative affect (NA), characterized by negative emotions

such as sadness and anxiety, is associated with poor health

and somatic complaints [1,2]. The majority of studies of

NA and symptoms focus on adults or clinically depressed ad-

olescents [3]. Minimal attention has been paid to NA and its

association with physical symptoms in healthy adolescents

from the community. The current study focuses on NA mea-

sured frequently during a slice of adolescent daily life and the

associations between both school day and weekend NA and

later physical symptom reports.

Prior work investigating affect and physical symptoms in

adults has measured NA at state levels (e.g., momentary

experience), trait levels (e.g., the ‘‘typical’’ amount of NA ex-

perienced), or in combination. There are indications, at least

for adults, that both trait and state NA are associated with

vulnerability to illness. In a well-known prospective study,

healthy adults completed measures of trait and state NA

ndence to: Jorie M. Butler, Ph.D., 380 South 1530

Lake City, UT 84112.

rie.butler@psych.utah.edu

front matter � 2009 Society for Adolescent Medicine. All

th.2008.09.005

before exposure to a respiratory virus [1]. Mood descriptors

were rated both for how accurately the adjective described

the participant over the preceding 24 hours (state NA) and

how accurately the adjective described him or her usually

(trait NA). Both trait and state NA were associated with

increased symptom reports after exposure. The symptom

reports for the participants who were high in state NA were

associated with objective indices of illness severity, whereas

the reports of participants who were high in trait NA were un-

related to illness severity [1]. This finding suggests the possi-

bility that state NA may be more closely tied to objective

indicators of disease than trait NA.

NA is also associated with physical symptoms during ad-

olescence. Williams et al. [4] found that depressive symp-

toms were associated concurrently and longitudinally with

physical symptom reports in a sample of healthy adolescents.

In another study a specific source of NA, test anxiety, was

associated with an increase in reports of physical symptoms

during the school year [5]. Depressed mood was correlated

with the number of physical symptoms reported in a sample

of incarcerated female adolescents [6]. None of these studies

rights reserved.

J.M. Butler et al. / Journal of Adolescent Health 44 (2009) 452–457 453

addressed daily levels of NA in relation to physical symp-

toms in adolescents, leaving this area in need of further

exploration.

Studies with adult populations suggest that particular con-

stellations of physical symptoms may relate to NA in unique

ways. Charles and Almeida [7] investigated adult reports of

NA in association with three constellations of physical symp-

toms: pain, gastrointestinal illness, and respiratory/flu symp-

toms. The authors defined trait NA as the average of state NA

scores within-person across an 8-day daily diary. The authors’

separation of physical symptoms into distinct constellations

allowed the relationship with affect to be examined separately

for respiratory/flu symptoms, which are directly observable,

in comparison to pain and gastrointestinal symptoms, which

cannot be measured objectively. Results indicated that

trait NA and state NA were concurrently associated with

symptoms of pain, gastrointestinal illness, and respiratory/

flu infection and state NA was also associated with pain and

gastrointestinal symptoms one day later. The lagged relation-

ship between state NA and gastrointestinal symptoms was

stronger for individuals high in trait NA. Although it is impor-

tant to acknowledge that the distinction between trait NA and

state NA may be overly simplified, the pattern of findings sug-

gest that state NA is an important predictor of concomitant

and at least some short-term future somatic complaints (in

this case, the more covert constellations of gastrointestinal

and pain symptoms), and particularly in concert with trait NA.

The relationship between affect states and physical symp-

toms may vary by context, such as the different context expe-

rienced by adolescents on weekdays and weekend days. In

adults, nonwork hassles on the weekends were associated

with psychosomatic complaints on subsequent workdays,

whereas positive social experiences were associated with

fewer complaints [8]. This suggests that weekend recovery

provided a subsequent reduction in health complaints,

whereas difficult weekend experiences were detrimental.

Very little is known about adolescents’ weekday versus

weekend experiences and their subsequent health reports.

Weekday activities are more structured than weekend activi-

ties, and the ability to manage NA during less structured

times may be important for avoiding illness or overly vigilant

attendance to physical symptoms.

Affect states are presumed to be transient, yet it is likely

that the variability in an individual’s mood hovers about

a mean influenced by trait-like characteristics [9,10]. That

is, on a given day the sadness a person feels is influenced

by that person’s dispositional tendency to experience and

to report sadness, as well as by transient events of the day.

In the unique approach used in the current study, frequent

mood reports logged over many days likely reflect trait-like

characteristics in addition to more transient and specific af-

fect states in adolescent daily life. To investigate trait-influ-

enced state NA across contexts that have been found to

influence physical health complaints in adults, this study cen-

tered on differential associations between physical symptoms

and school day versus weekend NA.

The aims of the present study were to examine: (1)

whether experience sampling reports of NA are associated

with later physical symptoms in adolescents; and (2)

whether this relationship varies by when NA occurs (school

days versus weekend days or an interactive combination)

and by constellation of symptoms (pain, respiratory, gastro-

intestinal, and immune). Based on previous research, we

hypothesized that NA on school days and weekend days

would be associated interactively, as adolescents reporting

higher levels of NA during both time periods are likely

higher in trait NA. In addition, given the relationship in

adults between state NA and covert (but not overt) symp-

toms documented by Charles and Almeida [7] we predicted

that the association between diary NA and physical symp-

toms would be stronger for covert physical symptoms

(e.g., pain and gastrointestinal symptoms) compared to

overt, easily observable symptoms (e.g., respiratory and im-

mune problems).

Methods

The data for this study are from a study of adolescent

health behaviors and susceptibility to tobacco use called

Project MASH (Monitoring Adolescent Stress and Health).

For a full description of methods and procedures see Whalen

et al. [11]. All procedures were approved by the Institutional

Review Board of the University of California at Irvine.

Participants and procedures

Incoming 9th-grade students from five high schools were

recruited for participation. The students completed a health

screening questionnaire at their high school about their

everyday health behaviors and smoking history. A random

sample of nonsmoking students, and all students with a his-

tory of smoking (oversampled because the larger study was

focused on adolescent smoking vulnerability) were invited

to attend an orientation session with their parents. Of those

who attended, 90% (474) agreed to participate.

The inclusion criteria for the current study were comple-

tion of spring eDiary recordings of NA and physical symp-

tom telephone interviews in the winter and summer. These

criteria were met by 330 of the participants who had enrolled

in the larger study (51% female; average age at enrollment

14.46, SD ¼ .48). Half reported some history of smoking,

with 20% reporting more than one or two smoking experi-

ences. The ethnic breakdown was 56% European American,

19% Asian American, 2% African American, and 23%

Latino or other. Most parents, 75%, had attended college,

with 63% having earned a B.A. or advanced degree. There

were no differences in age, reports of a chronic health condi-

tion, aggregated levels of weekend or weekday NA, or num-

ber of physical symptoms reported between participants who

completed all phases of the current study and those who did

not. Noncompleters were slightly more likely to be male

(56%).

Table 1

Physical symptom constellations

Pain Respiratory Gastrointestinal Immune

Earache Sore throat Nausea Swollen lymph nodes

Headache Runny nose Stomachache Urinary tract infection

symptoms

Toothache Wheezing Diarrhea Cold sore

Coughing

Coughing up

phlegm

Fever

J.M. Butler et al. / Journal of Adolescent Health 44 (2009) 452–457454

Diary recordingsParticipants completed a total of 4 consecutive days of

ambulatory diary recordings in the spring of their 9th-grade

year. The multiple choice responses were entered into hand-

held computers from the Palm III family (Sunnyvale, CA).

The 4-day recording periods comprised 2 school days (Thurs-

day and Friday) and the weekend (Saturday and Sunday).

Auditory prompts directing participants to complete a diary

entry were emitted on a variable schedule approximately ev-

ery 25 minutes (6 10 minutes) during the participant’s self-

reported waking hours across the recording period. This

high-frequency schedule was used to maximize opportunities

for capturing rapid state changes in mood typical of adoles-

cents [12]. In recognition of the intensive nature of the study,

participants were paid up to $100 for their participation.

Health interviewsParticipants were contacted during the winter and summer

of their 9th grade year to complete a telephone interview about

their health and physical symptoms. Precise information

about the time interval between the health interviews and

diary recordings is available for a subsample of participants

(n¼ 225). For this subsample, the winter health interview pre-

ceded the first day of the diary data collection by an average of

6 weeks (mean ¼ 54 days, SD ¼ 24) and the summer health

interview followed the first day of diary data collection an

average of 3 months later (mean ¼ 104 days, SD ¼ 25).

Table 2

Descriptors for negative affect diary prompts and their corresponding

numeric values

Anxiety

descriptors

Sadness

descriptors

Anger

descriptors

Stress

descriptors

Numeric

values for

descriptors

Overwhelmed Devastated Furious Intense 5

Pressured Depressed Pissed (irritated) Strong 4

Worried Sad Mad Moderate 3

Tense Down/low Annoyed Mild 2

Uneasy Unhappy Sore Just Noticeable 1

None None None None 0

Constructs and measures

Chronic health conditionsParents of 18% of the participants reported that the adoles-

cent had a chronic physical health condition. More than 29

chronic conditions were reported and dummy coded as pres-

ent (1) or not present (0). Most of these were relatively minor

(e.g., knee inflammation, acne). The most common chronic

health conditions reported were asthma or exercise-induced

asthma (6%).

Smoking statusParticipants’ smoking status was dummy coded as (1) re-

ported smoking on one or more occasions or (0) never

smoked.

Illness symptomsParticipants completed a telephone interview, adapted

from the Monthly Health Review, validated as a measure

of infectious disease [13]. During the telephone health inter-

views, participants were asked to think back to 1 month ago

and to report whether they experienced each of 15 physical

symptoms at that time. A reference date was provided, e.g.,

‘‘Memorial Day weekend,’’ to assist the adolescents in spe-

cific recall. Then participants were asked to think back to 3

weeks ago, 2 weeks ago, and 1 week ago, and to report the

occurrence of the symptoms at each time. Total symptoms

in each of four categories were generated: pain, respiratory,

gastrointestinal, and immune (Table 1). These constellations

were chosen to represent overt symptoms, easily observable

as illness-related (immune and respiratory symptoms), and

symptoms that are less observable (pain and gastrointestinal)

[7].

State NADuring the diary recording periods, participants provided

ratings of four NA states (sadness, anger, anxiety, and stress)

using six-point adjective scales [11] (Table 2). Adjective

scale points were converted to numeric values (0–5), and in-

dividual means for the weekdays (Thursday/Friday) and

weekend (Saturday/Sunday) were calculated for each partic-

ipant for the four NA states; to create a more robust variable

representing NA, the individual means of the discrete NA

states were combined into grand NA means for weekday

and weekend. Internal consistency for the combined NA

means was excellent (a ¼ .93; .94, respectively).

Results

Participants completed a mean of 106 diary reports and

missed a mean of 23 reports (occasions on which they were

prompted but did not complete an electronic diary entry) dur-

ing the 4-day recording period. More physical symptoms

were reported during winter than summer months for all con-

stellations except gastrointestinal (Table 3).

In a series of hierarchical regression analyses, gender,

smoking status, presence of a chronic health condition, and

winter physical symptom reports were entered as control var-

iables in Step 1, along with the NA variables centered at the

sample mean as predictors (Table 4). In Step 2, interactions

(or cross products) between weekend and weekday NA

Table 3

Correlations, means, and standard deviations for study variables

S I S GI S R S P WI W GI W R W P NA WN NAWD CC SS Mean (SD)

Age .02 .03 .04 .03 .08 .03 �.03 .01 �.01 �.03 .01 .15** 14.48 (.82)

Smoking Status (SS) .15** .14** .14** .17* .19** .21** .16* .16** .12* .12* �.03 — .52 (.50)

Chronic condition (CC) .08 .19* .09 .16* .04 .05 .08 .10* .09 .05 — .18 (.39)

NA Weekday (NA WD) �.00 .17** .07 .23** .12* .20** .18** .14** .84** — .79 (.63)

NA weekend (NA WN) .07 .16** .15** .29** .12** .19** .19** .13* — .70 (.71)

Winter pain symptoms (WP) .16** .26** .22** .47** .25** .29** .34** — .12 (.12)

Winter respiratory symptoms (WR) .14** .20* .27** .22** .37** .36** — .17 (.15)

Winter gastrointestinal symptoms (W GI) .14** .29** .20** .20** .29** — .10 (.12)

Winter immune symptoms (W I) .32** .16** .19** .26** — .05 (.09)

Summer pain symptoms (SP) .30** .38** .30** — .10 (.12)

Summer respiratory symptoms (SR) .23** .36** — .10 (.12)

Summer gastrointestinal symptoms (SGI) .17** — .10 (.13)

Summer immune symptoms (SI) — .03 (.07)

** p < .01.

* p < .05.

J.M. Butler et al. / Journal of Adolescent Health 44 (2009) 452–457 455

were entered as predictors [14]. This sequential entering pro-

cedure was used to allow interpretation of main effects if in-

teractive effects were not found.

In contrast to our hypotheses, only weekend NA was asso-

ciated with summer pain, respiratory, and immune symptoms

when control variables were included in the models (Table

4). For gastrointestinal symptoms, there were no main effects

of weekend or weekday NA, but a significant interaction was

present. Examination of the residuals for each regression anal-

ysis revealed some evidence of heteroscedasticity, consistent

with the use of a count variable as a dependent variable [15].

The hierarchical regression analytic approach used here pro-

vides a more conservative test of effects in this circumstance

than alternative approaches. Predicted values for the interac-

tion were calculated from the regression equation by

substituting scores one standard deviation above and below

the mean [14]. Participants who reported low (one SD below

the mean or more) NA both on weekdays and weekend days

in their eDiaries also reported lower levels of gastrointestinal

symptoms in summer than did other groups (Figure 1).

Discussion

Consistent with other studies suggesting that indices of

global NA are associated with physical symptoms in adoles-

cents, we found that daily experiences of NA as measured by

experience sampling are associated with later physical symp-

toms. We also found that there is a particular role for NA on

weekends in this association for pain and respiratory symp-

toms. To verify a unique association between physical symp-

toms and weekend NA, parallel analyses were run with

weekend NA and weekday NA as single predictors. The pat-

tern of results was consistent with the interactive models for

respiratory and gastrointestinal symptoms. For immune

symptoms, neither weekday nor weekend NA were signifi-

cant when considered individually, whereas for pain symp-

toms, both weekend and weekday NA were associated with

pain symptoms. For immune symptoms, a somewhat different

pattern emerged. Weekend NA was associated with higher

levels of immune symptoms, whereas weekday NA was asso-

ciated with lower levels of immune symptoms. NA may serve

a functional role to generate needed attention to symptoms

[16]. Perhaps increased vigilance to immune symptoms

produced a long-term reduction in immune symptoms. Note

that there were no interactive findings in predicting immune

symptoms, suggesting that individuals experiencing weekday

NA were independent of those experiencing weekend NA.

Thus weekend NA could have a protective association with

immune symptoms, whereas weekday NA might enhance

health care behaviors. Taken together, the pattern of findings

predicting pain, respiratory, and immune symptoms suggests

a unique and consistent role of weekend NA that differs from

the role of weekday NA.

Gastrointestinal symptoms were associated with a differ-

ent pattern of NA. Low weekend NA served a protective

effect only in concert with low weekday NA, suggesting

that gastrointestinal symptoms were reported at a consistent

rate among adolescents in our sample except for those whose

NA was below average on school days and weekend days.

These individuals who experience consistently low state

NA are likely to be low in trait NA as well.

Why is weekend NA so important, relative to weekday

NA, in explaining physical symptom reports? Prior research

has found that weekends are particularly important to adoles-

cents as the focus of relaxation and fun for the week [17].

During weekends, adolescents’ social environments may be

more variable than they are on school days. Differences in

adolescents’ social and family experiences, perhaps includ-

ing increased family conflict on weekends, may be pivotal

in contributing to vulnerabilities to physical symptoms, or

to proneness to experience physical symptoms for reasons

that may be unrelated to illness.

Methodological advantages and limitations

Participants were instructed to rate their current moods

each time they heard an auditory signal prompt from the

PDA. These procedures greatly reduce the possibility of

Table 4

Multiple regression analyses predicting summer physical symptom reports from diary reports of negative affect (NA) 2 months earlier controlling for gender,

smoking status, chronic condition, and winter physical symptom reports

DV Summer pain symptoms

DV Summer

respiratory symptoms

DV Summer

gastro-intestinal symptoms

DV Summer

immune symptoms

Predictor variable b SE b SE b SE b SE

Step 1 Gender �.04 .01 �.01 .01 �.04** .01 �.01 .01

Chronic condition .01 .01 .02 .02 .05** .02 �.00 .01

Smoking status .02* .02 .02 .01 .01 .01 .01 .01

Winter symptoms

(pain, respiratory,

gastrointestinal,

and immune)

.34** .01 .18** .04 .26** .05 .12** .04

Weekday NA �.15 .02 �.04* .02 .02 .02 �.02* .01

Weekend NA .05** .01 .04* .02 �.00 .02 .02* .01

DR2 ¼ .25 F(6, 319) ¼ 17.53**

DR2 ¼ .09 F(6, 318) ¼ 5.45**

DR2 ¼ .19 F(6, 321) ¼ 12.16**

DR2 ¼ .06 F(6, 320) ¼ 3.47**

Step 2 Gender �.00 .01 �.01 .01 �.04** .01 �.04 .01

Chronic condition .02 .01 .02 .02 .05** .02 �.00 .01

Smoking status .02* .01 .01 .01 .01 .02 .01 .01

Winter symptoms

(pain, respiratory,

gastrointestinal,

and immune)

.34** .05 .17** .04 .26** .05 .12** .04

Weekday NA �.01 .02 �.03 .01 .03 .02 �.02* .01

Weekend NA .02** .06 .05* .02 .01 .02 .02* .01

Weekday NA

3 Weekend NA

�.02y .01 �.01 .01 �.03** .01 �.00 .01

DR2 ¼ .01 F

(7, 318) ¼ 15.54**

DR2 ¼ .01 F

(7, 317) ¼ 4.91**

DR2 ¼ .02F

(7, 320) ¼ 11.70**

DR2 ¼ .00 F

(7, 319) ¼ 2.98**

** p < .01.

* p < .05.y p � .10.

J.M. Butler et al. / Journal of Adolescent Health 44 (2009) 452–457456

retrospective bias or memory error in the measurement of

state NA and provide unique social ecological information

about variability of affect in adolescent daily life. However,

several methodological limitations should be noted. Adoles-

cents with a history of smoking behavior were oversampled,

potentially limiting generalizability of the findings. Measures

of state NA and physical symptoms were obtained from a sin-

gle source, the adolescent participants. NA reports were col-

lected on only 2 weekday school days, Thursday and Friday,

which may not be representative of the full week, nor do

we know how typical the monitored week was for each

adolescent. Moreover, it is not clear to what extent the state

NA reports reflect trait NA.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

Low weekend NA High Weekend NA

Sum

mer

Gas

troi

ntes

tina

lSy

mpt

oms

LowWeekdayNAHighWeekdayNA

Figure 1. Predicted means for the weekend negative affect (NA) 3 weekday

NA interaction predicting summer gastrointestinal symptoms.

Another limitation is that, given the complexity of physi-

cal symptoms, some of the symptoms could have been

assigned to either of two constellations, for example, ‘‘stom-

achache’’ could qualify as pain but here is categorized as

gastrointestinal. Because it was not possible to obtain

independent or objective verification of health status, the as-

sociations between weekend NA and physical illness cannot

be considered causal. Relationships may also be due, in part,

to a third factor such as a negative reporting bias. Finally, it

should be kept in mind that the effect sizes were relatively

small.

Implications

Given the links between NA and physical symptom

reports, health care providers and parents would be well ad-

vised to attend to adolescents’ NA, particularly on weekends,

when they discuss and try to understand adolescent health

complaints. Helping adolescents manage their affect states,

especially during weekends when loneliness may be espe-

cially salient or family conflict may be at a peak, could

produce positive health consequences later in the year.

Whether the physical symptoms reported here were

related to actual physical illness is beyond the scope of this

investigation. Some evidence in support of the accuracy of

these reports stems from the finding that higher mean levels

J.M. Butler et al. / Journal of Adolescent Health 44 (2009) 452–457 457

of physical symptom reports in all constellations were corre-

lated with seeking medical care in the subsample for whom

this information was available (n ¼ 302; rs ¼ .16–.31;

p < .01). It is also important to note that symptom reports

are important for study in their own right [18], particularly

during adolescence when conceptions of, and reactions to

such symptoms may have long-term consequences for

health-care–seeking behaviors and health outcomes.

An important next step is to replicate and extend the find-

ings using affect data from multiple weekends as well as

physical symptom data on more occasions. Further research

should also examine the contextual factors associated with

weekend NA in adolescents (e.g., family conflict) that may

relate to the development of illness or to the perceptions of

and/or need to report physical symptoms.

Acknowledgments

This research was supported by grants from the National

Cancer Institute (RO1 CA 80301), the California Tobacco-

Related Disease Research Program (6PT-3003), and the Na-

tional Institute of Drug Abuse (TTURC; P50 DA 13332).

Additional financial support was provided to the first author

by Dean Ron Huff of the School of Social Ecology by way of

the Dean’s Dissertation Fellowship. A second valuable

source of fellowship support was the Transdisciplinary

Tobacco Research Center of the University of California,

Irvine. In addition, we would like to thank the participants

in Project MASH and our dedicated research assistants,

particularly Dennis Dierck and Rod Jiminez. The first author

would like to thank members of her dissertation committee

for their generous assistance, Susan Charles and Barbara

Henker.

References

[1] Cohen S, Doyle WJ, Skoner DP, et al. State and trait negative affect as

predictors of objective and subjective symptoms of respiratory viral

infections. J Pers Soc Psychol 1995;68:159–69.

[2] Neeleman J, Ormel J, Bijl R. The distribution of psychiatric and

somatic ill health: Associations with personality and socioeconomic

status. Psychosom Med 2001;63:239–47.

[3] Bardone AM, Moffitt TE, Caspi A, et al. Adult physical health

outcomes of adolescent girls with conduct disorder, depression, and

anxiety. J Am Acad Child Adolesc Psychiatry 1998;37:594–601.

[4] Williams PG, Colder CR, Richards MH, et al. The role of self-assessed

health in the relationship between gender and depressive symptoms

among adolescents. J Pediatr Psychol 2002;27:509–17.

[5] Marschall P. Self-report and the stability of physical symptoms by

adolescents. Adolescence 1989;24:209–16.

[6] Williams RA, Hollis HM. Health beliefs and reported symptoms among

a sample of incarcerated adolescent females. J Adolesc Health 1999;

24:21–7.

[7] Charles ST, Almeida DM. Daily reports of symptoms and negative

affect. Not all symptoms are the same. Psychol Health 2006;21:

1–17.

[8] Fritz C, Sonnentag S. Recovery, health, and job performance: Effects

of weekend experiences. J Occup Health Psychol 2005;10:187–99.

[9] Boker SM, Nesselroade JR. A method for modeling the intrinsic dy-

namics of intraindividual variability: Recovering the parameters of sim-

ulated oscillators in multi-wave panel data. Multivar Behav Res 2002;

37:127–60.

[10] Fleeson W. Toward a structure- and process-integrated view of person-

ality traits as density distributions of states. J Pers Soc Psychol 2001;

80:1011–27.

[11] Whalen CK, Jamner LD, Henker B, et al. Smoking and moods in ado-

lescents with depressive and aggressive dispositions: Evidence from

surveys and electronic diaries. Health Psychology 2001;20:99–111.

[12] Larson R, Csikszentmihalyi M, Graef R. Mood variability and the

psychosocial adjustment of adolescents. J Youth Adolesc 1980;

9:469–90.

[13] Jenkins CD, Kreger BE, Rose RM, et al. Use of a monthly health review

to ascertain illness and injuries. Am J Public Health 1980;70:82–4.

[14] Aiken LS, West SG. Multiple Regression: Testing and Interpreting

Interactions. Newbury Park, CA: Sage Publications, 1991.

[15] Cohen J, Cohen P, West SG, Aiken LS. Applied Multiple Regression/

Correlation Analysis for the Behavioral Sciences. Mahweh, NJ:

Lawrence Earlbaum 2003.

[16] Mayne TJ. Negative affect and health: The importance of being earnest.

Cognit Emot 1999;13:601–35.

[17] Larson R, Richards M. Waiting for the weekend: Friday and Saturday

nights as the emotional climax of the week. In: Crouter AC, Larson R,

eds. Temporal Rhythms in Adolescence: Clocks, Calendars, and the

Coordination of Daily Life: New Directions for Child and Adolescent

Development. San Francisco, CA: Jossey Bass, 1998:37–51.

[18] Petrie KJ, Weinman J. More focus needed on symptom appraisal.

J Psychosom Res 2003;54:401–3.