Developmental Psychology
2007, Vol. 43, No. 2, 341–351
Copyright 2007 by the American Psychological Association
0012-1649/07/$12.00 DOI: 10.1037/0012-1649.43.2.341
Cumulative Risk, Maternal Responsiveness, and Allostatic Load Among
Young Adolescents
Gary W. Evans, Pilyoung Kim, Albert H. Ting, Harris B. Tesher, and Dana Shannis
Cornell University
The purpose of this study was to examine the impact of cumulative risk exposure in concert with maternal
responsiveness on physiological indicators of chronic stress in children and youth. Middle-school
children exposed to greater accumulated psychosocial (e.g., family turmoil, poverty) and physical (e.g.,
crowding, substandard housing) risk factors manifested higher levels of allostatic load, a physiological
marker of cumulative wear and tear on the body caused by the mobilization of multiple, physiological
response systems. This effect was longitudinal, residualizing allostatic load 3– 4 years earlier when the
youth were in elementary school. This effect, however, occurred only among adolescents with mothers
low in responsiveness. Cumulative risk was also associated with dynamic cardiovascular processes in
response to an acute stressor (mental arithmetic). Higher risk was associated with muted reactivity and
slower, less efficient recovery in blood pressure. These dynamic cardiovascular effects occurred irrespective of maternal responsiveness.
Keywords: cumulative risk, stress, maternal responsiveness, allostatic load
sure reactivity and recovery to an acute stressor) in relation to
cumulative risk exposure. This is the first study to examine this
topic and only the second study of cardiovascular recovery among
a sample of children or youth. All prior studies of recovery except
one used adult samples.
Overexposure to a combination of physiological mobilizations
across different systems in response to adaptive demands over time
alters the ability of the body to respond efficiently to environmental demands (Karlamangala, Singer, McEwen, Rowe, & Seeman,
2002; McEwen, 1998, 2002; Seeman, McEwen, Rowe, & Singer,
2001; Seeman, Singer, Rowe, Horwitz, & McEwen, 1997). In
contrast to homeostasis, allostasis posits a more dynamic and
interactive set of multiple physiological systems (i.e., neuronal,
endocrinological, cardiovascular, and immunological) of equilibrium maintenance. Operation of these dynamic interactive systems
causes the body to continuously adjust its operating range, producing downward regulation of metabolic activities in order to
maintain internal stability congruent with shifting environmental
demands (McEwen, 1998, 2000, 2002; McEwen & Seeman, 1999).
Allostasis has generated considerable interest in both the neuroscience and medical communities. Whereas alterations in singular
physiological risk factors (e.g., blood pressure) produced by environmental demands reveal modest consequences for morbidity and
eventually mortality, the combination of singular changes across
multiple physiological indicators captured by allostatic load portends marked elevations in physical and possibly psychological
morbidity (Karlamangala et al., 2002; McEwen, 1998, 2002; Seeman et al., 1997, 2001, 2004). Moreover, allostatic load also
appears to provide a cogent model of how physiological changes
accompanying adaptive demands from the environment can, in
turn, alter neurological activity responsible for mental and cognitive disorders (McEwen, 2000, 2002).
In addition to scant research on young adolescents and cumulative risk exposure, and no data on allostasis among adolescents,
this is an important age group to consider in its own right. Children
Elementary-school children exposed to accumulated psychosocial (e.g., family turmoil, violence) and physical (e.g., noise,
substandard housing) risk factors manifest higher allostatic load
relative to children with lower levels of cumulative risk (Evans,
2003). Allostatic load is a physiological marker of cumulative
wear and tear on the body caused by the mobilization of multiple
physiological systems in response to environmental demands
(McEwen, 2002; McEwen & Seeman, 1999; McEwen & Stellar,
1993; Sterling & Eyer, 1988). This study extends developmental
research on cumulative risk in several respects. We incorporate a
longitudinal design examining risk and allostatic load among a
young adolescent sample. Most cumulative risk research has examined preadolescent children, and only Evans (2003) examined
physiological sequelae of cumulative risk exposure. We also incorporate a protective factor—maternal responsiveness—that potentially could buffer the adverse effects of cumulative risk on
allostatic load. There is very little work on protective factors and
cumulative risk exposure in adolescents and no work on protective
factors and allostatic load among people of any age. We also
investigate dynamic cardiovascular functioning (i.e., blood pres-
Gary W. Evans, Departments of Design & Environmental Analysis and
Human Development, Cornell University; Pilyoung Kim, Harris B. Tesher,
and Dana Shannis, Department of Human Development, Cornell University; Albert H. Ting, Department of Design & Environmental Analysis,
Cornell University.
Partial support for this research came from the W. T. Grant Foundation
and the John D. and Catherine T. MacArthur Foundation Research Network on Socioeconomic Status and Health. We are grateful to the families
and children who have participated in our research. We thank Jana Cooperman, Kim English, Missy Globerman, Tina Merrilees, Chanelle Richardson, Adam Rokhsar, and Amy Schreier for assistance with data collection.
Correspondence concerning this article should be addressed to Gary W.
Evans, Departments of Design & Environmental Analysis and Human
Development, Cornell University, Ithaca, NY 14853-4401. E-mail:
[email protected]
341
342
EVANS, KIM, TING, TESHER, AND SHANNIS
between age 11 and 13 years undergo rapid dramatic changes in
physical growth and endocrinological activity. These physiological alterations coincide with marked social changes, particularly
enhanced autonomy, more intensive peer associations, and onset of
romantic relationships, that have obvious implications for cumulative risk exposure (Compas, Hinden, & Gerhardt, 1995; Grotevant, 1998; Steinberg & Morris, 2001). This age period also
marks a major transition point from the comforts of a relatively
small, often homogeneous elementary school, with most of the day
spent with one adult in one place along with the same group of
same-age peers, to a larger, often more heterogeneous middle
school with greater variability in where and with whom time is
spent during the day. Lengthening of the school day and, especially in rural areas, much longer bus rides can also dramatically
increase the amount of time one is away from home.
Nearly all studies of cumulative risk have relied on urban
samples, and most have focused on low-income children of color.
The major limitations of such sample restrictions include (a) most
high risk families in America are White (U.S. Census Bureau,
2005), (b) both the degree and persistence of material deprivation
is considerably worse in rural areas compared with urban centers
(Auchincloss & Hadden, 2002; Blank, 2005; Mathematica Policy
Research, 2005; Sherman, 1992), and (c) by focusing on lowincome samples, the degree of variance in cumulative risk exposure is truncated; thus, estimates of cumulative risk and developmental disarray may be downwardly biased.
Evans (2003) is the only published documentation of a link
between cumulative risk exposure and allostasis in children. One
study has uncovered similar results in older adults. B. H. Singer
and Ryff (1999) found that retrospective reports of cumulative risk
exposure over the life course among 59-year-old adults predicted
allostatic load. A handful of studies suggest that exposure to
singular rather than cumulative risk factors has the potential to
elevate allostatic load in children, although none of these studies
assessed allostatic load, instead monitoring a single physiological
marker of stress. Adolescents in more hostile, demanding, and
noisy environments had higher blood pressure throughout the day
(Southard et al., 1986). Experimental exposure to simulated interpersonal conflict among adults elevated acute blood pressure in
children (El-Sheik, Cummings, & Goetsch, 1989). Chronic exposure to high ambient noise levels as well as crowded homes
correlates with elevated blood pressure and catecholamines in
children (Evans, 2001).
Adverse environmental circumstances early in childhood, which
include neglect, abuse, or severely deprived physical and social
stimulation, influence the hypothalamic–pituitary axis (HPA) in
children. This has been documented among children growing up in
abusive families (Cicchetti & Rogosch, 2001; Hart, Gunnar, &
Cicchetti, 1996) or in severely deprived institutional settings such
as Romanian orphanages (Gunnar, 2000). Furthermore, children
chronically exposed to loud ambient noise manifest elevated, basal
cortisol levels (Evans, 2006; Ising & Braun, 2000).
One of the important objectives of the present article is to build
upon and extend prior studies of risk exposure and physiological
activity among children by examining cumulative rather than
singular risk exposure. Most if not all of the childhood risk factors
identified above as correlates of HPA or sympathetic–adrenal–
medullary (SAM) impacts do not occur independently of one
another. A unique and potentially key aspect of childhood poverty
is the confluence of psychosocial and physical environmental risk
factors that low-income children must contend with in their daily
lives (Evans, 2004; Repetti, Taylor, & Seeman, 2002; Schell,
1997; Taylor, Repetti, & Seeman, 1997). Risks rarely occur in
isolation, instead clustering among families living in poverty.
Moreover, singular risk exposure pales in comparison with cumulative risk exposure in accounting for adverse socioemotional and
cognitive outcomes in children (Ackerman, Kogos, Youngstrom,
Schoff, & Izard, 1999; Evans, 2003; Gutman, Sameroff, & Cole,
2003; Lengua, 2002; Liaw & Brooks-Gunn, 1994; Rutter, 1983;
Sameroff, 1998).
Another reason why it is valuable to examine cumulative risk
exposure and allostatic load during earlier periods in the life course is
because of the possibility that cumulative risks may accumulate to
affect changes in cardiovascular dynamics as the body mobilizes in
response to specific acute demands. Under normal conditions when
confronted with an acute environmental demand, the body mobilizes
metabolically to meet the requirements imposed by the immediate
circumstances. Such energy mobilization is manifested by a sharp,
rapid rise in cardiovascular activity, termed reactivity. As soon as the
demand is over, the body rapidly returns it its prior resting state
(Haynes, 1991; Krantz & Falconer, 1995; Linden, Earle, Gerin, &
Christenfeld, 1997; McEwen, 1998, 2000). This latter process is
called recovery. However, if the demands from the environment are
excessive and/or occur over a prolonged period of time, allostatic load
increases, and the system becomes overloaded. Typical healthy patterns of cardiovascular reactivity and recovery in response to an acute
stressor no longer occur. Instead, one witnesses cardiovascular hyporeactivity to acute environmental demands coupled with prolonged
recovery to baseline levels (Dienstbier, 1989; McEwen, 1998, 2000).
Animal research (Dienstbier, 1989) in addition to case reports of
traumatized children suggest that very high levels of stressor exposure
depress cardiovascular reactivity (Perry & Pollard, 1998). Exposure to
violence and harsh parenting in public housing projects has also been
associated with depressed cardiovascular reactivity among children
(Krenichyn, Saegert, & Evans, 2001). Murali and Chen (2005) found
similar relations between the experience of violence and muted cardiovascular reactivity among adolescents. Adolescents with a history
of a high number of life events manifested muted cardiovascular
reactivity to several acute stressors (Boyce & Chesterman, 1990).
Analogous results were uncovered by Mengel et al. (1992) in a study
of diabetic adolescents. In a fascinating series of studies of children
residing in rural Caribbean villages, Flinn (1999) has shown that
chronically stressed children (e.g., persistent family conflict) evidence
blunted cortisol responses to physical demands that evoke increased
cortisol in children without chronic stress.
Potential linkages between cardiovascular recovery and chronic
stressor exposure have not been examined among children or youth.
We were able to uncover only one study of cardiovascular recovery in
children. Jackson, Treiber, Turner, Davis, and Strong (1999) reported
that African American, 11–15-year-old males took longer to recover
from exposure to acute stressors compared with White children and
females, respectively. These authors did not examine chronic stress as
a factor in their study. Among adults, slower cardiovascular recovery
to an acute stressor is associated with higher levels of chronic stress
(Fleming, Baum, Davidson, Rectanus, & McArdle, 1987; Lepore,
Miles, & Levy, 1997; Pardine & Napoli, 1983; Schaubroeck &
Ganster, 1993). Cardiovascular recovery to an acute stressor has also
been linked to socioeconomic status (SES) in the Whitehall II study of
British civil servants. Lower SES was correlated with delayed recovery (Steptoe et al., 2002).
RISK AND ALLOSTATIC LOAD IN YOUNG ADOLESCENTS
We were also interested in scrutinizing maternal responsiveness
as a potential protective factor for the impact of cumulative risk
exposure on allostatic load. Although there is much literature on
the protective effects of maternal responsiveness on socioemotional development (Bornstein, 1989; Bradley, Corwyn, Burchinal,
McAdoo, & Garcia-Coll, 2001; Demo & Cox, 2000; Shonkoff &
Phillips, 2000), no research has examined maternal responsiveness
and allostatic load. There are, however, a few studies that have
explored the potential buffering effects of maternal responsiveness
on SAM and HPA activity levels in relation to stress. Children in
families with lower levels of social support (Woodall & Matthews,
1989) and those who receive harsher, more punitive parenting
(Gump, Matthews, & Raikkonen, 1999) evidence higher cardiovascular reactivity to an acute stressor. Children in families with
more punitive parenting also evidence higher levels of resting
blood pressure (Wright, Treiber, Davis, Bunch, & Strong, 1998).
Children with less responsive parents and less secure maternal
attachment show stronger HPA reactivity to acute stressors (Gunnar, 2000; Gunnar & Donzella, 2002). Children with more negative parent– child interactions also manifest elevated basal cortisol
levels (Flinn & England, 1997; Luecken, 1998). Thus, although
there is evidence that responsive parenting can influence physiological stress responses in relation to childhood exposure to singular stressors, none of this work has investigated the potential of
parental responsiveness to buffer cumulative risk exposure among
children or examined allostatic load outcome measures. Furthermore, the parenting research on physiological stress responses in
children has examined basal or reactivity patterns only, neglecting
the investigation of recovery from acute stressor exposure.
In sum, the primary objective of the present study was to
replicate and extend the one existing study on cumulative risk and
allostatic load in children with a longitudinal assessment of young
adolescents. We also wanted to explore whether dynamic cardiovascular activity (i.e., reactivity and recovery to an acute stressor)
would be influenced by cumulative risk. Finally, we were interested in the potential ameliorative effects of a well-documented
resource for adolescents—maternal responsiveness— on cumulative risk and allostatic load.
We hypothesized that cumulative risk would elevate allostatic
load over time. We expected that cumulative risk would also
influence dynamic SAM activity, dampening cardiovascular reactivity and inhibiting recovery to an acute stressor. Finally, both of
the above sets of physiological outcomes to cumulative risk were
expected to be moderated by maternal responsiveness. We predicted that youth with more responsive mothers would be less
adversely impacted by cumulative risk exposure.
Method
Participants
Participants for this second wave of a longitudinal study of rural
poverty and child development were 207 seventh and eighth grade
children (mean age 13.37 years, 52% male, 48% female) who had
physiological data. In Wave 1 there were 339 elementary-school
children (mean age 9.20 years).1 All of these youth lived in rural
areas in upstate New York. They were recruited from public
schools, New York State Cooperative Extension programs, and
various antipoverty programs (⬍ 5% Wave 1 refusal rate). Lowincome families (income-to-needs ratio ⱕ 1) were over sampled
343
(53%) because this research program is focused on poverty and
human development. Only one child per household participated in
the study. The sample was predominantly White (94%), reflecting
upstate, rural New York demographics. Attrition analyses revealed
that children with higher levels of cumulative risk at Wave 1
(mean cumulative risk index ⫽ 3.57) were less likely to remain in
the sample at Wave 2 (mean cumulative risk index ⫽ 1.79),
t(336) ⫽ 8.82, p ⬍ .001. The individual components of cumulative
risk that significantly reflected this pattern of selective attrition
included poverty, housing quality, family turmoil, child separation
from family, exposure to violence, and maternal high school drop
out. There was no selective attrition, however, as a function of
allostatic load.
Procedure
A standard protocol was used in data collection in each youth’s
home at Wave 1 when the participants were in third through fifth
grade (Evans, 2003) and then again when the youth was in seventh
or eighth grade. The target youth and his or her mother were
interviewed independently by two interviewers. The gender of the
interviewer was matched to the youth’s gender in Wave 2.
Cumulative risk assessment. Nine domains of risk factors for
young adolescents were assessed. Residential density was determined by dividing the number of people living in the household by
the number of rooms (including bathrooms). Anyone sleeping in
the home three or more nights per week was deemed a resident.
Only rooms regularly used by family members were counted.
Garages, hallways, attics, and basements were not counted unless
regularly occupied (e.g., basement converted into a family room).
Noise levels were estimated by a 2-hr monitoring of decibel levels
(Leq, dBA) in the primary social space (typically the living room)
in the home. Housing quality was assessed with a standardized,
observer rating scale consisting of subscales measuring structural
quality, maintenance, cleanliness and clutter, safety hazards, children’s resources, and climatic conditions (Evans, Wells, Chan, &
Saltzman, 2000). The housing quality instrument has undergone
extensive psychometric development. Exposure to family turmoil
and child–family separation were determined by combining maternal and youth reports. Maternal responses came from the Life
Events and Circumstances Checklist (Work, Cowen, Parker, &
Wyman, 1990; Wyman, Cowen, Work, & Parker, 1991), which
consists of multiple dichotomous (yes/no) item subscales for each
of these stressor domains. Sample items include “Our child has
been involved in serious family arguments” and “A close family
member was away from home a lot.” The time period was the time
between the initial wave of data collection and the present one,
typically 2–3 years later. A calendar that the mother and interviewer filled in together with important personal and family events
was used as a mnemonic device. Youth were also asked to indicate
the occurrence of stressors (again with the aid of personally
constructed calendar to facilitate recall) with a life event scale
based on a revised Adolescent Perceived Events Scale (Compas,
1997) supplied to us by Bruce Compas. Youth indicated “yes” or
“no” to whether each stressful event or circumstance had occurred.
Sample items in each of the domains of family turmoil and
child–family separation included “Pretty serious arguments or
1
See Evans (2003) for further details on Wave 1.
344
EVANS, KIM, TING, TESHER, AND SHANNIS
fights between parents” and “Parents getting divorced or separated.” From mothers, information about exposure to violence was
determined from a third subscale of the Life Events and Circumstances Checklist (e.g., “Our neighborhood has been unsafe”).
For each of these stressor domains, risk was defined dichotomously (as 0 or 1) based upon a statistical criterion with 1 equal to
a value greater than one standard deviation above the mean for the
distribution of the specific risk factor across the entire sample of
youth. A value of 0 was assigned for levels of exposure less than
the cutoff. In addition to these six continuous risk factors, three
additional categorical risks were included in the cumulative risk
index: maternal high school drop out, single parent, and household
income at or below the poverty line (income-to-needs ratio ⱕ 1.0).
Thus, the cumulative risk index could vary for each participant
from 0 to 9.
Maternal responsiveness. Maternal responsiveness was measured by combining the youth’s perception of maternal responsiveness with observations of mother– child interaction during a cooperative game. A scale consisting of 11 items tapping both instrumental
(e.g., help with homework) and emotional (e.g., willing to talk to me
when needed) responsiveness was constructed. Youth answered each
question on a 5-point scale (never, hardly ever, sometimes, fairly
often, very often). The scale has good internal consistency (␣ ⫽ .84)
and strong test–retest reliability (r ⫽ .92) over a 3-month period.
Evidence for validity includes confirmatory factor analysis (two moderately correlated subscales of instrumental and emotional responsiveness), a nomological network of associations with other constructs
(significant but modest negative correlations with income, household
crowding, and a positive higher correlation with Moos’ Family Cohesion scale (Moos & Moos, 1986). We also rated maternal sensitivity
to the child during a game of Jenga (Hasbro; Pawtucket, RI). Mothers
and the target child took turns working together to build a tower of
blocks as high as possible. The Jenga game was videotaped and then
coded for maternal responsiveness. Interrater reliability for the sensitivity ratings was good (Ebel r ⫽ .95; Ebel, 1951). Raters were blind
to the child’s cumulative risk and perceived maternal responsiveness
scores. Sensitivity was coded for each 2.5-min period over the 15-min
duration of Jenga. Sensitivity ratings ranged from 1 (Not at all
characteristic. There are almost no signs of parent sensitivity. The
parent rarely responds appropriately to the child’s cues.) to 5 (Highly
sensitive/responsive. The parent displays consistent sensitivity to the
child throughout the rating period.). Behavioral and affective indicators of sensitivity were operationalized in a coders’ manual summarized in Belsky, Crnic, and Woodsworth (1995). These six observational ratings (␣ ⫽ .97) were then averaged for the overall
observational rating. Perceived maternal responsiveness was significantly correlated (r ⫽ .68) with observer codings of maternal parental
responsiveness to the child during the game. The perceived maternal
responsiveness scale and the mean observational rating were each
converted to z scores and then added to develop our index of maternal
responsiveness.
Allostatic load. Resting blood pressure was monitored with a
Critikon Dinamap Pro 100 (Critikon; Tampa, FL) while the youth
sat and read quietly. Seven readings were taken every 2 min and
the average of the second through seventh reading was used as the
resting index. This procedure yields highly reliable indices of
chronic resting blood pressure (Kamarck et al., 1992). Overnight
(8 p.m. to 8 a.m.) urinary catecholamines and cortisol were assayed from the evening of the interview with the youth. All urine
voided during this time was stored on ice in a container with a
preservative (metabisulfite). Total volume was recorded, and four
10-ml samples were randomly extracted and then deep frozen at
⫺80° C until subsequent assays were completed. The pH of two of
the 10-ml samples was adjusted to 3 to further inhibit oxidation of
catecholamines. Total unbound cortisol was assayed with a radioimmune assay (Contreras, Hane, & Tyrrell, 1986). Epinephrine
and norepinephrine were assayed with high-pressure liquid chromatography with electrochemical detection (Riggin & Kissinger,
1977). Creatinine was assayed to control for differences in body
size and incomplete urine voidings (Tietz, 1976). The samples
were assayed by technicians blind to the respondent’s status.
Allostatic load (0 – 6) was calculated by summing the number of
singular physiological indictors on which each child scored in the
top quartile of risk. Allostatic load included overnight cortisol,
epinephrine, and norepinephrine, resting diastolic and systolic
blood pressure, and an index of fat deposition (body mass index ⫽
kg/m2). Prior studies of allostatic load in adults (Kubzansky,
Kawachi, & Sparrow, 1999; Seeman et al., 2002; B. H. Singer &
Ryff, 1999) and in children (Evans, 2003) have used similar
metrics, combining multiple physiological indicators of risk into
one total allostatic load index.
We did not incorporate cardiovascular reactivity and recovery
into the allostatic load index for several reasons. First, prior studies
of allostatic load have not included reactivity or recovery measures. Second, unlike the other measures included in prior allostatic load indices, reactivity and recovery have not yet been
shown to be independent risk factors for morbidity. Third, dynamic SAM or HPA processes in response to an acute stressor
likely reflect tertiary outcomes, which are caused by chronically
elevated SAM and HPA levels (McEwen, 2000; McEwen & Seeman, 1999). Fourth, because there is only one prior study of
cardiovascular recovery in children and no studies of cumulative
risk and reactivity, we felt it was important to examine these
impacts separately from allostatic load.
Cardiovascular reactivity and recovery. Immediately following the physiological resting period, the youth was given a “math
test” without warning to assess cardiovascular reactivity and recovery. Blood pressure monitoring continued during and immediately following the mental arithmetic task. The youth’s task was to
mentally subtract a two-digit number from a four-digit number
continuously and report aloud the correct answer. The youth did
this task for 12 min without writing down any calculations. Every
4 min, the number to be subtracted was changed. Thirty seconds
after the test was completed, the first of five blood pressure
readings was taken while the youth again sat quietly and read for
a period of 10 min. This constituted the cardiovascular recovery
procedure. At the end of the math test the participant was assured
that no more tests would be given and he or she could relax and
read while the blood pressure monitoring continued for a little
longer. Mental arithmetic is a common stress induction stimulus
used in reactivity protocols for both children and adults (Gump &
Matthews, 1999; Krantz & Manuck, 1984; Matthews, Gump,
Block, & Allen, 1997).
Results
Cumulative Risk Exposure
Table 1 provides descriptive information on each of the singular
risk factors as well as the total cumulative risk index. Thirty
RISK AND ALLOSTATIC LOAD IN YOUNG ADOLESCENTS
345
Table 1
Descriptive Statistics on Cumulative Risk Factors
Measure
M
Proportion of sample
with risk factor
SD
Cumulative risk factors
Crowding (no. of people per room)
Noise (Leq, dBA)
Housing problems (0–2)
Family separation (0–12)
Family turmoil (0–9)
Violence (0–5)
Poverty line (0 or 1)
Single parent (0 or 1)
Maternal high school dropout (0 or 1)
Cumulative risk (0–9)
0.56
60.57
0.59
2.23
2.78
0.68
0.19
6.46
0.31
1.73
2.04
0.88
1.67
1.55
.02
.11
.07
.24
.14
.11
.18
.43
.07
with mothers who were low in responsiveness, t(169) ⫽ 1.85, p ⬍
.05. The simple slopes at the mean, t(169) ⬍ 1.00, and at one
standard deviation above the mean, t(169) ⬍ 1.00, were not
significant.
percent of the sample had zero risk factors, 24% had one risk, 17%
two risks, 14% had three risks, 9% had four risks, and 6% had five
or more risks. For the analyses of cumulative risk and allostatic
load, cumulative risks of five or more were combined into one
category, given small sample sizes at the upper end.
Cardiovascular Reactivity and Recovery
Allostatic Load
There were main effects of cumulative risk on both cardiovascular reactivity and recovery. Maternal responsiveness and its
interaction with cumulative risk had no effect on dynamic cardiovascular functioning. Reactivity was assessed by subtracting the
baseline level (i.e., the mean of Readings 2–7 [Reading 1 was
discarded] while sitting quietly and reading) from the mean of the
six blood pressure readings during the mental arithmetic task. This
is the standard procedure for the calculation of cardiovascular
reactivity (Krantz & Falconer, 1995; Matthews et al., 1987, 1997).
Gender was also included as a control. Gender did not interact with
cumulative risk to impact either reactivity or recovery. Because we
did not assess dynamic cardiovascular parameters in Wave 1 of
this study, the reactivity and recovery data are cross-sectional and
not longitudinal, as is the allostatic load analysis above. As cumulative risk increased, both systolic reactivity, b ⫽ ⫺.98 (SE ⫽ .33),
p ⬍ .01, f2 ⫽ .04, and diastolic reactivity, b ⫽ ⫺.48 (SE ⫽ .24),
p ⬍ .05, f2 ⫽ .03, were muted (see Table 3). As can be seen in
Figure 2, participants exposed to more cumulative risk exhibited
lower levels of reactivity in response to the acute stressor, mental
arithmetic.
Recovery, the efficiency with which physiological function approached basal levels, was analyzed with multilevel (time nested
within cumulative risk) hierarchical linear modeling (J. D. Singer
& Willett, 2003). This analytic approach allows one to calculate
interindividual heterogeneity in change over time: in the present
Longitudinal analyses with orthogonal least squares regression
examined the relations between cumulative risk, maternal responsiveness, and the interaction of cumulative risk and maternal
responsiveness on allostatic load. The analysis incorporated Wave
1 allostatic load as an initial control variable in the regression
model. Gender was also included as a control given its significant
relations with many of the physiological measures. We examined
potential interactive effects for gender and cumulative risk but
found none. The statistical controls, each of the main effects, and
their multiplicative term were also centered to reduce multicollinearity (Aiken & West, 1991).
As indicated in Table 2 and shown in Figure 1, there was an
interaction of maternal responsiveness and cumulative risk on
allostatic load at Wave 2, b ⫽ ⫺.08 (SE ⫽ .04), p ⬍ .05, f2 ⫽ .02,
statistically controlling for Wave 1 allostatic load and gender.
Neither of the main effects for cumulative risk or maternal responsiveness were significant (see Table 2). The regression plots depicted in Figure 1 were calculated at one standard deviation above
and below the mean of maternal responsiveness, as well as at the
mean. This was done for descriptive purposes only; the orthogonal
least squares analysis maintained the continuous nature of cumulative risk, maternal responsiveness, and their multiplicative interaction term. Analyses of the simple slopes revealed that as cumulative risk increased, allostatic load rose but only for those youth
Table 2
Regression Results for Allostatic Load, Cumulative Risk, and Maternal Responsiveness,
Controlling for Wave 1 Allostatic Load and Gender
Predictor
Total R2
⌬R2
F
df
b
SE
Cumulative risk
Maternal responsiveness
Cumulative Risk ⫻ Maternal Responsiveness
.12
.12
.14
.00
.00
.02
⬍1.00
⬍1.00
4.11*
1, 171
1, 170
1, 169
.06
⫺.11
⫺.08*
.07
.08
.04
*
p ⬍ .05.
EVANS, KIM, TING, TESHER, AND SHANNIS
346
2
1.8
1.6
Allostatic Load
1.4
Low Maternal
Responsiveness
1.2
Medium Maternal
Responsiveness
1
0.8
High Maternal
Responsiveness
0.6
0.4
0.2
0
-2.8
-1.4
0
1.4
2.8
Cumulative Risk
Figure 1. Regression plots showing the relationship between cumulative risk exposure, maternal responsiveness, and allostatic load (as measured by indices described in the text).
case, how intraindividual blood pressure scores recover as a function of cumulative risk exposure. Figure 3 shows millimeters of
mercury for diastolic blood pressure recovery and the timing of the
blood pressure measures taken every 2 min throughout the recovery period. The initial recovery reading occurred 30 s after cessation of the mental arithmetic task and every 2 min after that for a
total of five readings. Recovery was analyzed by setting the
intercept for each participant at their value for the last reading
during the reactivity period. The three line graphs in Figure 3
depict exposure to 0 –1, 2– 4, and 5 or more cumulative risks. This
is for descriptive purposes only. The actual analyses maintained
the continuous nature of the cumulative risk metric (i.e., 0 –5).2
As can be seen in Figure 3, diastolic blood pressure recovery is
less efficient (i.e. slower), with higher levels of cumulative risk
exposure, b ⫽ .13 (SE ⫽ .05), p ⬍ .01. This significant effect is for
the quadratic function, controlling for the linear function, which
was also significant. No effect size estimates are available for
hierarchical linear modeling parameters. The 95% confidence interval for the diastolic recovery quadratic b weight is .04 –.22 (see
Table 4). Although the pattern was similar for systolic blood
pressure recovery, suggesting less efficient recovery with greater
cumulative risk exposure, this effect was not significant for either
the linear or quadratic terms.
Discussion
Childhood exposure to singular social and physical environmental risk factors is associated with adverse developmental outcomes
including depression, behavioral conduct problems, poorer academic achievement, dysregulated HPA and SAM, and premature
morbidity and mortality in adulthood (Evans, 2001; Gallo & Matthews, 2003; Repetti et al., 2002; Taylor et al., 1997). One explanatory construct that may explain some of the linkage between early
childhood risk and adverse outcomes both in childhood and subsequently over the life course is allostatic load. Allostatic load is
indicative of changes in activity levels across multiple physiological systems. Each specific physiological change, although modest
when aggregated across different systems, appears to be an important precursor to long-term morbidity and possibly premature
mortality (Karlamangala et al., 2002; McEwen, 1998, 2002; Seeman et al., 1997, 2001, 2004).
The first objective of this study was to replicate and extend
earlier work in younger children, showing a cross-sectional relation between cumulative risk exposure and allostatic load. In
previous research, allostatic load was associated with higher levels
of cumulative risk exposure among 8 –10-year-olds (Evans, 2003).
Inspection of Figure 1 reveals that greater cumulative risk is also
related to elevated allostatic load among young adolescents. Furthermore, this relation holds longitudinally, controlling for levels
of allostatic load 3– 4 years prior, when these adolescents were in
elementary school. However, the main effect of cumulative risk on
allostatic load is qualified by a significant interaction with maternal responsiveness. Cumulative risk elevates allostatic load in
young adolescents only when maternal responsiveness is low.
Significant longitudinal relations between cumulative risk exposure and allostatic load may help illuminate why early childhood
risk exposure is associated with increased morbidity and premature
mortality later in life (Felitti et al., 1998; Power, Manor, & Fox,
1991; Power & Matthews, 1998; Repetti et al., 2002; Taylor et al.,
1997). B. H. Singer and Ryff (1999) documented that similar, early
childhood elevated risk profiles, as shown herein, were associated
with allostatic load in older adults. Elevated allostatic load early in
life could be a viable process to explain how childhood exposure
to excess risk causes morbidity later in life. In the MacArthur
studies of successful aging, allostatic load explained 35% of the
SES-related variance in mortality (Seeman et al., 2004). Clearly,
longer term data collection over the life course is required to
directly test the role of elevated allostatic load in linking early
2
See J. D. Singer and Willett (2003) for more details on the application
of hierarchical linear modeling to modeling intraindividual change over
time in relation to an exogenous variable.
RISK AND ALLOSTATIC LOAD IN YOUNG ADOLESCENTS
347
Table 3
Regression Results for Cardiovascular Reactivity and Cumulative Risk, Controlling for Gender
Outcome
Diastolic reactivity
Systolic reactivity
*
p ⬍ .05.
**
Total R2
Predictor
Cumulative risk
Cumulative risk
.03
.04
⌬R2
.02
.04
F
*
4.08
8.63**
df
b
SE
1,203
1,203
⫺.48
⫺.98**
*
.24
.33
p ⬍ .01.
cumulative risk exposure to morbidity. Our results are assessed too
early in the life course to uncover disease endpoints.
A second objective of this study was to examine the role of
cumulative risk exposure in dynamic cardiovascular activity.
Youth exposed to greater cumulative risk had muted cardiovascular reactivity to an acute stressor (see Figure 2). These effects
match several adult studies showing that higher chronic stressor
exposure leads to muted cardiovascular reactivity (Gump & Matthews, 1999). A much smaller body of work shows similar trends
among children (Boyce & Chesterman, 1990; Krenichyn et al.,
2001; Mengel et al., 1992; Perry & Pollard, 1998). None of these
studies, however, examined reactivity to an acute stressor in relation to cumulative risk exposure. Moreover, there is only one prior
study of cardiovascular recovery to an acute stressor in children
(Jackson, Treiber, Turner, Davis, & Strong, 1999), and these
investigators did not examine chronic stressor exposure. Adult
studies (Gump & Matthews, 1999) reveal slower, less efficient
recovery in conjunction with greater chronic stress. To our knowledge, this is the first study with children or youth to reveal the
same pattern. As shown in Figure 3, exposure to higher levels of
cumulative risk impedes cardiovascular recovery to an acute stressor in young adolescents. The nonlinearity of the data was unexpected and is difficult to interpret. One speculative explanation
might be the toughening hypothesis: Some modest amount of
exposure to challenges over time, somewhat analogous to physical
exercise, enables the body to improve its efficiency in dealing with
an acute demand (Dienstbier, 1989). What is clear is that at higher
levels of cumulative risk exposure, cardiovascular recovery is
inhibited.
A less efficient stress response (i.e., muted reactivity, slower
recovery) to an acute stressor in relation to cumulative risk exposure is potentially important for at least two reasons. First, there is
growing evidence that inefficient physiological mobilization and
recovery rather than resting indices of physiological stress are
particularly important precursors of morbidity (Gunnar &
Vasquez, 2001; Haynes, 1991; Heim, Newport, Heit, & Graham,
2000; Linden et al., 1997; McEwen, 1998, 2002; Seeman et al.,
2001). Changes in the patterns of responsiveness to demands rather
than resting, basal levels may be more indicative of developing
pathology. Second, the pattern of hyporeactivity and delayed recovery uncovered in relation to cumulative risk exposure is in
accord with allostatic load theory. One of the costs of accumulated
wear and tear on the body is loss of a fast acting, robust stress
response to acute demands (McEwen, 1998, 2002; Seeman et al.,
1997, 2001). One limitation of our study was the inability to also
examine HPA reactivity and recovery to an acute stressor. There is
very little work on this in children, and none has examined
cumulative risk exposure. Research in adults (Cohen et al., 2006;
Heim et al., 2000) and in children (Flinn, 1999; Gunnar, 2000;
Gunnar & Donzella, 2002; Gunnar & Vasquez, 2001; Watamura,
Donzella, Alwin, & Gunnar, 2003) suggests that experiences of
chronic stress lead to a flattening of the daytime diurnal cortisol
rhythm, particularly the absence of a sharp afternoon decline in
cortisol levels. Flinn (1999) has also found evidence for muted
cortisol reactivity to physical demands that typically elevate cortisol among chronically stressed children. These downwardregulated, daytime diurnal rhythms could be a byproduct of the
sustained overnight activity of the HPA as indexed in the allostatic
7
Cardiovascular Reactivity
6
5
4
—— Diastolic Blood Pressure
3
- - - - Systolic Blood Pressure
2
1
0
0
1
2
3
4
5
Cumulative Risk
Figure 2. Cumulative risk exposure (as measured by the cumulative risk index) and cardiovascular reactivity
level.
EVANS, KIM, TING, TESHER, AND SHANNIS
348
69
Cum ulative Ris k
Low (0-1)
Medium (2-4)
68
Diastolic Blood Pressure
High ( 5)
67
66
65
64
63
62
0
1
2
3
4
5
Time
Figure 3. Cumulative risk exposure (as measured by the cumulative risk
index) and diastolic blood pressure recovery (in millimeters of mercury).
The x-axis depicts the timing of the blood pressure measures taken every
2 min throughout the recovery period.
load metric (e.g., higher levels of overnight cortisol; McEwen,
2000). Increasingly, evidence points toward the importance of not
only examining mean levels of SAM and HPA activity and other
markers of chronic stress but scrutinizing the temporal patterns of
activity throughout the day.
The null results for maternal-responsiveness buffering of cardiovascular reactivity and recovery could reflect temporal aspects
of allostatic load development. Changes in resting, basal levels of
multiple physiological systems are theorized to reflect secondary
outcomes caused by primary allostatic load mechanisms at the
cellular level, including downward regulation of neurotransmitters,
trafficking of immune cells, and excitatory amino acids. Reactivity
and recovery as tertiary processes may take longer to occur,
resulting from wear and tear on the body produced over time by
elevated blood pressure levels, higher levels of circulating cat-
echolamines, and elevated cortisol levels (McEwen, 2000; McEwen & Seeman, 1999).
The buffering of cumulative risk impacts on allostatic load by
maternal responsiveness uncovered herein is in accord with a few
studies that revealed protective effects of parenting practices on
cumulative risk and socioemotional and cognitive development.
Socioemotional difficulties associated with elevated cumulative
risk exposure were attenuated by better social relationships between adolescents and other adults, including parents (Jessor, Van
Den Bos, Vanderryn, Costa, & Turbin, 1995), and for adolescents
whose mothers expressed less negative affect toward them (Seifer,
Sameroff, Baldwin, & Baldwin, 1992). Similar protective effects
were shown for parental monitoring among elementary-school-age
children exposed to more cumulative risk (Klein, Forehand, &
Group, 2000). The adverse consequences of cumulative risk exposure on adolescent academic achievement have also been shown
to be moderated by high parental expectations for success (Reynolds, 1998), consistent discipline (Gutman, Sameroff, & Eccles,
2002), and granting of lower autonomy (Gutman et al., 2002). We
show a similar buffering pattern for allostatic load, an index of
chronic physiological stress. Some caution is warranted, however,
in interpreting these findings solely as a reflection of parental
responsiveness. It is difficult to disentangle maternal responsiveness from at least two other constructs that might also affect
reactions to cumulative risk exposure: (a) Maternal responsiveness
might be a reflection of a more general positive youth disposition.
Happier and better adjusted youths likely see their parents as more
responsive. (b) It is difficult to determine whether parental responsiveness is the key process or simply a marker for better parent–
child relationships in general.
The principal limitation of the present study is its research
design. Although we have longitudinal data for one of our outcome
measures, allostatic load, our design is still correlational, and thus,
causal conclusions are not warranted. We also lack measures of
cardiovascular dynamics in Wave 1, so the reactivity and recovery
data are cross-sectional. As noted above, there is emerging research suggesting that chronic stressor exposure can influence
HPA patterns of response to acute stressors as well. A valuable
adjunct to the present physiological protocol of basal cardiovascular and neuroendocrine assessments and cardiovascular dynamic
assessment would have been the inclusion of HPA reactivity and
recovery to an acute stressor. It is also possible that children
experiencing more risk did not perceive or react to the task, mental
arithmetic, in the same manner as those exposed to less risk. Evans
(2003), for example, found that cumulative risk exposure was
Table 4
Hierarchical Linear Modeling Results for Diastolic Blood Pressure Recovery and Cumulative
Risk, Controlling for Gender
Parameter
Intercept
Gender
Cumulative
Cumulative
Trials
Cumulative
Cumulative
**
p ⬍ .01.
Coefficient
**
risk
risk quadratic
Risk ⫻ Trials
Risk Quadratic ⫻ Trials
66.07
⫺0.24
0.04
⫺0.02
⫺0.10
⫺0.58**
0.13**
SE
t ratio
df
1.03
0.92
1.05
0.23
0.19
0.22
0.05
64.44
⫺0.26
0.04
⫺0.06
⫺0.55
⫺2.68
2.74
246
212
222
223
211
212
215
RISK AND ALLOSTATIC LOAD IN YOUNG ADOLESCENTS
associated with greater learned helplessness in an achievement
task (solving puzzles) among 8 –10-year-olds. Thus, a valuable
addition to the reactivity and recovery physiological protocol
would be assessments of perceived challenge, stress, engagement,
and so on with the task itself.
The sample of the present study is not representative of young
adolescents. We over sampled low-income children given our
interest in risk and poverty, and all of the participants were from
rural areas in upstate New York and predominantly Caucasian,
reflecting that geographic area. At the same time, it is worth
reiterating the marked paucity of research on rural poverty and
low-income, high-risk White children. Given that the majority of
children in America who are poor are White (U.S. Census Bureau,
2005) and that rural poverty constitutes greater and more persistent
material deprivation than urban poverty (Auchincloss & Hadden,
2002; Blank, 2005; Mathematica Policy Research, 2005; Sherman,
1992), more work on populations like the present one are needed.
Research by Conger and Elder (Conger & Elder, 1994; Elder &
Conger, 2000) on Midwestern families struggling with the farm
crisis of the 1980s and research by Brody, Flor, and Gibson (1999)
on low-income, African-American families in the Southeast stand
out as excellent exceptions to the dominance of research on urban
high-risk populations conducted to date. Another limitation of the
present study is selective attrition. Children with higher levels of
cumulative risk exposure in Wave 1 were less likely to remain in
the sample. The net effect of this selective attrition is probably to
underestimate the impacts of cumulative risk on allostatic load and
dynamic cardiovascular processes.
Early childhood risk exposure has consistently been associated
with concurrent, and in a smaller number of studies, subsequent,
socioemotional and cognitive difficulties (Ackerman et al., 1999;
Evans, 2003; Gutman et al., 2003; Lengua, 2002; Liaw & BrooksGunn, 1994; Rutter, 1993; Sameroff, 1998). We know much less
about cumulative risk exposure and adolescent development and
almost nothing about physiological processes such as allostatic
load that may be associated with early exposure to accumulated
psychosocial and physical environmental risk factors. We have
shown that cumulative risk exposure is positively associated with
allostatic load in middle-school children, after controlling for prior
levels of allostatic load in elementary school. This is especially
true for those with mothers who are less responsive. Middle-school
children exposed to higher levels of cumulative risk also evidence
muted cardiovascular reactivity to a cognitive task and are slower
to recover physiologically thereafter. These effects, however, are
not moderated by maternal responsiveness. There is a growing
body of evidence documenting links between early childhood risk
exposure and adult morbidity and premature mortality. Most of
this work has examined specific risk factors, largely ignoring the
fact that many childhood risk factors tend to covary. Allostatic
load may play a critical role in understanding how early childhood
exposures to cumulative risk factors translate into poorer adjustment and elevated morbidity across the lifespan.
References
Ackerman, B. P., Kogos, J., Youngstrom, E., Schoff, K., & Izard, C.
(1999). Family instability and the problem behaviors of children from
economically disadvantaged families. Developmental Psychology, 35,
258 –268.
Aiken, L., & West, S. (1991). Multiple regression: Testing and interpreting
interactions. Los Angeles: Sage.
349
Auchincloss, A. H., & Hadden, W. (2002). The health effects of ruralurban residence and concentrated poverty. Journal of Rural Health, 18,
319 –336.
Belsky, J., Crnic, K., & Woodsworth, S. (1995). Personality and parenting:
Exploring the mediating role of transient mood and daily hassles. Journal of Personality, 63, 905–929.
Blank, R. M. (2005). Poverty, policy, and place: How poverty and policies
to alleviate poverty are shaped by local characteristics. International
Regional Science Review, 28, 441– 464.
Bornstein, M. H. (Ed.). (1989). Maternal responsiveness: Characteristics
and consequences. San Francisco: Jossey-Bass.
Boyce, W. T., & Chesterman, E. (1990). Life events, social support, and
cardiovascular reactivity in adolescence. Developmental and Behavioral
Pediatrics, 11, 105–111.
Bradley, R. H., Corwyn, R. F., Burchinal, M., McAdoo, H. P., & GarciaColl, C. (2001). The home environment of children in the United States,
part II: Relations with behavioral development through age thirteen.
Child Development, 72, 1868 –1886.
Brody, G. H., Flor, D., & Gibson, N. M. (1999). Linking maternal efficacy
beliefs, developmental goals, parenting practices, and child competence
in rural single-parent African American families. Child Development,
70, 1197–1208.
Cicchetti, D., & Rogosch, F. A. (2001). The impact of child maltreatment
and psychopathology on neuroendocrine functioning. Development and
Psychopathology, 13, 783– 804.
Cohen, S., Schwartz, J. E., Epel, E., Kirschbaum, C., Sidney, S., &
Seeman, T. E. (2006). Socioeconomic status, race, and diurnal cortisol
decline in the coronary artery risk development in young adults (CARDIA) study. Psychosomatic Medicine, 68, 41–50.
Compas, B. E. (1997). Responses to Stress Questionnaire. Burlington, VT:
University of Vermont.
Compas, B. E., Hinden, B. R., & Gerhardt, C. A. (1995). Adolescent
development: Pathways and processes of risk and resilience. Annual
Review of Psychology, 46, 265–294.
Conger, R. D., & Elder, G. H., Jr. (1994). Families in troubled times.
Hawthorne, NY: Aldine de Gruyter.
Contreras, L. N., Hane, S., & Tyrrell, J. B. (1986). Urinary cortisol in the
assessment of pituitary-adrenal function: Utility of 24-hour and spot
determinations. Journal of Clinical Endocrinological Metabolism, 62,
965–969.
Demo, D. H., & Cox, M. J. (2000). Families and young children: A review
of research in the 1990s. Journal of Marriage and the Family, 62,
876 – 895.
Dienstbier, R. (1989). Arousal and physiological toughness: Implications
for mental and physical health. Psychological Review, 96, 84 –100.
Ebel, R. L. (1951). Estimation of the reliability of ratings. Psychometrika,
16, 407– 424.
Elder, G. H., Jr., & Conger, R. D. (2000). Children of the land. Chicago:
University of Chicago Press.
El-Sheik, M., Cummings, E., & Goetsch, V. (1989). Coping with adults’
angry behavior: Behavioral, physiological, and verbal responses in preschoolers. Developmental Psychology, 25, 490 – 498.
Evans, G. W. (2001). Environmental stress and health. In A. Baum, T.
Revenson, & J. E. Singer (Eds.), Handbook of health psychology (pp.
365–385). Hillsdale, NJ: Erlbaum.
Evans, G. W. (2003). A multimethodological analysis of cumulative risk
and allostatic load among rural children. Developmental Psychology, 39,
924 –933.
Evans, G. W. (2004). The environment of childhood poverty. American
Psychologist, 59, 77–92.
Evans, G. W. (2006). Child development and the physical environment.
Annual Review of Psychology, 57, 423– 451.
Evans, G. W., Wells, N. M., Chan, H. E., & Saltzman, H. (2000). Housing
and mental health. Journal of Consulting and Clinical Psychology, 68,
526 –530.
350
EVANS, KIM, TING, TESHER, AND SHANNIS
Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M.,
Edwards, W., et al. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults.
American Journal of Preventative Medicine, 14, 245–258.
Fleming, I., Baum, A., Davidson, L. M., Rectanus, E., & McArdle, S.
(1987). Chronic stress as a factor in physiological reactivity to challenge.
Health Psychology, 6, 221–237.
Flinn, M. V. (1999). Family environment, stress, and health during childhood. In C. Panter-Brick & C. M. Worthman (Eds.), Hormones, health,
and behavior (pp. 105–138). New York: Cambridge University Press.
Flinn, M. V., & England, B. (1997). Social economics of childhood
glucocorticoid stress response and health. American Journal of Physical
Anthropology, 102, 33–53.
Gallo, L. C., & Matthews, K. A. (2003). Understanding the association
between socioeconomic status and physical health: Do negative emotions play a role? Psychological Bulletin, 129, 10 –31.
Grotevant, H. (1998). Adolescent development in family contexts. In W.
Damon & N. Eisenberg (Eds.), Handbook of child psychology: Vol. 3.
Social, emotional, and personality development (pp. 1097–1149). New
York: Wiley.
Gump, B. B., & Matthews, K. A. (1999). Do background stressors influence reactivity to and recovery from acute stressors? Journal of Applied
Social Psychology, 29, 469 – 494.
Gump, B. B., Matthews, K. A., & Raikkonen, K. (1999). Modeling
relationships among socioeconomic status, hostility, cardiovascular reactivity, and left ventricular mass in African American and White
children. Health Psychology, 18, 140 –150.
Gunnar, M. R. (2000). Early adversity and the development of stress
reactivity and regulation. In C. A. Nelson (Ed.), The effects of early
adversity on neurobehavioral development (pp. 163–200). Mahwah, NJ:
Erlbaum.
Gunnar, M. R., & Donzella, B. (2002). Social regulation of the cortisol
levels in early human development. Psychoneuroendocrinology, 27,
199 –220.
Gunnar, M. R., & Vasquez, D. M. (2001). Low control and flattening of
expected daytime rhythm: Potential indices of risk in human development. Development and Psychopathology, 13, 515–538.
Gutman, L. M., Sameroff, A. J., & Cole, R. (2003). Academic growth
curve trajectories from 1st grade to 12th grade: Effects of multiple social
risk factors and preschool child factors. Developmental Psychology, 39,
777–790.
Gutman, L. M., Sameroff, A. J., & Eccles, J. S. (2002). The academic
achievement of African American students during early adolescence: An
examination of multiple risk, promotive, and protective factors. American Journal of Community Psychology, 30, 367–399.
Hart, J., Gunnar, M., & Cicchetti, D. (1996). Altered neuroendocrine
activity in maltreated children related to symptoms of depression. Development and Psychopathology, 8, 201–214.
Haynes, S. N. (1991). Psychophysiological assessment of poststress recovery. Psychological Assessment, 3, 356 –365.
Heim, C., Newport, J. D., Heit, S., & Graham, Y. P. (2000). Pituitaryadrenal and autonomic response to stress in women after sexual and
physical abuse in childhood. Journal of the American Medical Association, 284, 592–597.
Ising, H., & Braun, C. (2000). Acute and chronic endocrine effects of
noise: Review of the research conducted at the institute for water, soil
and air hygiene. Noise and Health, 7, 7–24.
Jackson, R. W., Treiber, F. A., Turner, J. R., Davis, H., & Strong, W. B.
(1999). Effects of race, sex, and socioeconomic status upon cardiovascular stress responsivity and recovery in youth. International Journal of
Psychophysiology, 31, 111–119.
Jessor, R., Van Den Bos, J., Vanderryn, J., Costa, F., & Turbin, M. (1995).
Protective factors in adolescent problem behavior: Moderator effects and
developmental change. Developmental Psychology, 31, 923–933.
Kamarck, T., Jennings, R., Debski, T., Glicksman-Weis, E., Johnson, P.,
Eddy, M., & Manuck, S. B. (1992). Reliable measures of behaviorally
evoked cardiovascular reactivity from a PC-based test battery. Psychophysiology, 29, 17–28.
Karlamangala, A. S., Singer, B. H., McEwen, B. S., Rowe, J. W., &
Seeman, T. E. (2002). Allostatic load as a predictor of functional
decline: MacArthur studies of successful aging. Journal of Clinical
Epidemiology, 55, 696 –710.
Klein, K., Forehand, R., & Group, F. H. R. (2000). Family processes as
resources for African American children exposed to a constellation of
sociodemographic risk factors. Journal of Clinical Child Psychology,
29, 53– 65.
Krantz, D. S., & Falconer, J. (1995). Measurement of cardiovascular
responses. In S. Cohen, R. C. Kessler, & L. Gordon (Eds.), Measuring
stress (pp. 193–212). New York: Oxford University Press.
Krantz, D. S., & Manuck, S. B. (1984). Acute psychophysiologic reactivity
and risk of cardiovascular disease: A review and methodologic critique.
Psychological Bulletin, 96, 435– 464.
Krenichyn, K., Saegert, S., & Evans, G. W. (2001). Parents as moderators
of psychological and physiological correlates of inner-city children’s
exposure to violence. Journal of Applied Developmental Psychology, 22,
581– 602.
Kubzansky, L. D., Kawachi, I., & Sparrow, D. (1999). Socioeconomic
status, hostility, and risk factor clustering in the normative aging study:
Any help from the concept of allostatic load? Annals of Behavioral
Medicine, 21, 330 –338.
Lengua, L. J. (2002). The contribution of emotionality and self-regulation
to the understanding of children’s response to multiple risk. Child
Development, 73, 144 –161.
Lepore, S. J., Miles, H. J., & Levy, J. S. (1997). Relations of chronic and
episodic stressors to psychological distress, reactivity, and health problems. International Journal of Behavioral Medicine, 4, 39 –59.
Liaw, F., & Brooks-Gunn, J. (1994). Cumulative familial risks and low
birth weight children’s cognitive and behavioral development. Journal
of Clinical Child Psychology, 23, 360 –372.
Linden, W., Earle, T. L., Gerin, W., & Christenfeld, N. (1997). Psychological stress reactivity and recovery: Conceptual siblings separated at
birth? Journal of Psychosomatic Research, 42, 117–135.
Luecken, L. J. (1998). Childhood attachment and loss experiences affect
adult cardiovascular and cortisol function. Psychosomatic Medicine, 60,
765–772.
Mathematica Policy Research. (2005). Rural research needs and data
sources for selected human service topics. Retrieved February 14, 2005,
from http://aspe.hhs.gov/hsp/05/rural-data/index.htm
Matthews, K. A., Gump, B. B., Block, D. R., & Allen, M. T. (1997). Does
background stress heighten or dampen children’s cardiovascular responses to acute stress? Psychosomatic Medicine, 59, 488 – 496.
Matthews, K. A., Weiss, S. M., Detre, T., Dembroski, T. M., Falkner, B.,
Manuck, S. B., & Williams, R. B. (1987). Handbook of stress, reactivity,
and cardiovascular disease. New York: Wiley.
McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 338, 171–179.
McEwen, B. S. (2000). The neurobiology of stress: From serendipity to
clinical relevance. Brain Research, 886, 172–189.
McEwen, B. S. (2002). The end of stress as we know it. Washington, DC:
John Henry Press.
McEwen, B. S., & Seeman, T. E. (1999). Protective and damaging effects
of mediators of stress: Elaborating and testing the concept of allostasis
and allostatic load. In N. E. Adler, M. Marmot, B. S. McEwen, & J.
Stewart (Eds.), Socioeconomic status and health in industrial nations
(pp. 30 – 47). New York: New York Academy of Sciences.
McEwen, B. S., & Stellar, E. (1993). Stress and the individual: Mechanisms leading to disease. Annals of Internal Medicine, 153, 2093–2101.
Mengel, M. B., Blackett, P. R., Lawler, M. K., Volk, R. J., Viviani, N. J.,
Stamps, G. S., et al. (1992). Cardiovascular and neuroendocrine responsiveness in diabetic adolescents within a family context: Association of
RISK AND ALLOSTATIC LOAD IN YOUNG ADOLESCENTS
poor diabetic control and dysfunctional family dynamics. Family Systems Medicine, 10, 5–33.
Moos, R., & Moos, B. (1986). Family Environment Scale manual (2nd
ed.). Palo Alto, CA: Consulting Psychologists Press.
Murali, R., & Chen, E. (2005). Exposure to violence and cardiovascular
and neuroendocrine measures in adolescents. Annals of Behavioral Medicine, 30, 155–163.
Pardine, P., & Napoli, A. (1983). Physiological reactivity and recent
life-stress experience. Journal of Consulting and Clinical Psychology,
51, 467– 469.
Perry, B. D., & Pollard, R. (1998). Homeostasis, trauma, and adaptation: A
neurodevelopmental view of childhood trauma. Child and Adolescent
Psychiatric Clinics of North America, 7, 121–135.
Power, C., Manor, O., & Fox, J. (1991). Health and class. London:
Chapman & Hall.
Power, C., & Matthews, S. (1998). Accumulation of health risks across
social groups in a national longitudinal study. In S. S. Strickland & P. S.
Shetty (Eds.), Human biology and social inequality (pp. 36 –57). New
York: Cambridge University Press.
Repetti, R. L., Taylor, S. E., & Seeman, T. E. (2002). Risky families:
Family social environments and the mental and physical health of
offspring. Psychological Bulletin, 128, 330 –366.
Reynolds, A. J. (1998). Resilience among black urban youth. American
Journal of Orthopsychiatry, 68, 84 –100.
Riggin, R. M., & Kissinger, P. T. (1977). Determination of catecholamines
in urine by reverse phase liquid chromatography with electrochemical
detection. Analytic Chemistry, 49, 2109 –2111.
Rutter, M. (1983). Statistical and personal interactions: Facets and perspectives. In D. Magnusson & V. Allen (Eds.), Human development: An
interactional perspective (pp. 295–319). New York: Academic Press.
Rutter, M. (1993). Stress, coping, and development. In N. Garmezy & M.
Rutter (Eds.), Stress, coping, and development (pp. 1– 41). New York:
McGraw-Hill.
Sameroff, A. J. (1998). Environmental risk factors in infancy. Pediatrics,
102, 1287–1292.
Schaubroeck, J., & Ganster, D. C. (1993). Chronic demands and responsivity to challenge. Journal of Applied Psychology, 78, 73– 85.
Schell, L. M. (1997). Culture as a stressor: A revised model of bicultural
interaction. American Journal of Physical Anthropology, 102, 67–77.
Seeman, T. E., Crimmins, E., Huang, M. H., Singer, B., Bucur, A.,
Gruenewald, T., et al. (2004). Cumulative biological risk and socioeconomic differences in mortality: MacArthur studies of successful
aging. Social Science and Medicine, 58, 1985–1997.
Seeman, T. E., McEwen, B. S., Rowe, J. W., & Singer, B. H. (2001).
Allostatic load as a marker of cumulative biological risk: MacArthur
studies of successful aging. Proceedings of the National Academy of
Sciences, 98, 4770 – 4775.
Seeman, T. E., Singer, B. H., Rowe, J. W., Horwitz, R., & McEwen, B. S.
(1997). Price of adaptation—Allostatic load and its health consequences:
MacArthur studies of successful aging. Archives of Internal Medicine,
157, 2259 –2268.
Seeman, T. E., Singer, B. H., Ryff, C. D., Love, G., & Levy-Storms, L.
(2002). Social relationships, gender, and allostatic load across two age
cohorts. Psychosomatic Medicine, 64, 395– 406.
Seifer, R., Sameroff, A. J., Baldwin, C. P., & Baldwin, A. (1992). Child
and family factors that ameliorate risk between 4 and 13 years of age.
351
Journal of the American Academy of Child and Adolescent Psychiatry,
31, 893–903.
Sherman, A. (1992). Falling by the wayside. Washington, DC: Children’s
Defense Fund.
Shonkoff, J. P., & Phillips, D. A. (Eds.). (2000). From neurons to neighborhoods: The science of early childhood development. Washington,
DC: National Academy Press.
Singer, B. H., & Ryff, C. D. (1999). Hierarchies of life histories and
associated health risks. In N. E. Adler, M. Marmot, B. S. McEwen, & J.
Stewart (Eds.), Socioeconomic status and health in industrial nations
(pp. 96 –115). New York: New York Academy of Sciences.
Singer, J. D., & Willett, J. B. (2003). Applied longitudinal data analysis.
New York: Oxford University Press.
Southard, D., Coates, T., Kolodner, K., Parker, F., Padgett, N., & Kennedy,
H. (1986). Relationship between mood and blood pressure in the natural
environment: An adolescent population. Health Psychology, 5, 469 –
480.
Steinberg, L., & Morris, A. (2001). Adolescent development. Annual
Review of Psychology, 52, 83–110.
Steptoe, A., Feldman, P. J., Kunz, S., Owen, N., Willemsen, G., & Marmot,
M. (2002). Stress responsivity and socioeconomic status: A mechanism
for increased cardiovascular disease risk? European Heart Journal, 23,
1757–1763.
Sterling, P., & Eyer, J. (1988). Allostasis: A new paradigm to explain
arousal pathology. In S. Fisher & J. Reason (Eds.), Handbook of life
stress, cognition, and health (pp. 629 – 649). New York: Wiley.
Taylor, S. E., Repetti, R. L., & Seeman, T. E. (1997). Health psychology:
What is an unhealthy environment and how does it get under the skin?
Annual Review of Psychology, 48, 411– 447.
Tietz, N. (1976). Fundamentals of clinical chemistry (2nd ed.). Philadelphia: W. B. Saunders.
U.S. Census Bureau. (2005). 2005 American Community Survey: S1702.
Poverty status in the past 12 months of families [Data]. Retrieved
January 2, 2007, from http://factfinder.census.gov/servlet/
STTable?ds_name⫽ACS_2005_EST_G00_&qr_name⫽ACS_2005_EST
_G00_S1702&_lang⫽en
Watamura, S. E., Donzella, B., Alwin, J., & Gunnar, M. R. (2003).
Morning to afternoon increases in cortisol concentrations for infants and
toddlers at child care: Age differences and behavioral correlates. Child
Development, 74, 1006 –1020.
Woodall, K. L., & Matthews, K. A. (1989). Familial environment associated with Type A behaviors and psychophysiological responses to stress
in children. Health Psychology, 8, 403– 426.
Work, W., Cowen, E., Parker, G., & Wyman, P. (1990). Stress resilient
children in an urban setting. Journal of Primary Prevention, 11, 3–17.
Wright, L. B., Treiber, F., Davis, H., Bunch, C., & Strong, W. (1998). The
role of maternal hostility and family environment upon cardiovascular
functioning among youth two years later: Socioeconomic and ethnic
differences. Ethnicity and Disease, 8, 367–376.
Wyman, P., Cowen, E., Work, W., & Parker, G. (1991). Developmental
and family milieu correlates of resilience in urban children who have
experienced major life stress. American Journal of Community Psychology, 19, 405– 426.
Received February 21, 2005
Revision received September 24, 2006
Accepted October 3, 2006 䡲