Tài liệu tâm lý học sức khỏe: Stress và môi trường

20
Environmental Stress and Health
Gary W. Evans
Cmtell University
Although environmental conditions play a prominent role in health and
psychological processes, antecedent factors in these processes have largely
been neglected within health psychology. Instead, the focus has been on
various markers of health, with considerable attention to stress-related
mechanisms, interceding between the environment and health. Another
focus within health psychology that has directed attention away from
environmental factors has been coping resources, with the examination of
either social support, personality, or coping strategies that potentially alter
the impact of environmental demands on health. But what characteristics of
the environment itself are likely to impinge on health and psychological
processes? When this question has been addressed within health
psychology, environment has been operationalized primarily in social terms.
Family and work social climates, as well as sociocultural and economic
conditions, predominate in the few environmental studies in health
psychology. This chapter intends to draw greater attention to the potential
role of the physical environment in health and psychological processes.
Why might the physical environment be important to health psychology? For
one reason, the physical environment clearly impacts health. Adverse
physical conditions can cause toxicological reactions, challenge homeostatic
balance, produce physical trauma, or function as vectors bearing pathogens.
Physical factors can also be a source of environmental demands that
pressure coping resources.
A second reason the physical environment is worthy of scrutiny within health
psychology is because the environment can be modified and thus becomes a
potential intervention target to improve health and well-being. Third,
environmental conditions are objective and thus can be measured more
readily in reliable and valid ways. For example, researchers can systematically monitor density or noise levels in precise, accurate ways that can

then be examined as possible causal factors in health. Fourth, physical
environmental conditions tend to be stable. Increasingly, research suggests
that chronic environmental demands are most likely to have negative
impacts on health (Lepore, 1995). Finally, the concept of psychological stress
that is central to several formulations of health, behavior, and disease (see
chap. 17, this volume) has been utilized to broaden understanding of how
physical features of the environment can influence human health and wellbeing.
There are at least three major ways in which the physical environment might
operate as a psychological stressor, straining human adaptive capacities.
First, this can occur when a stressor directly loads, or pressures, the system.
Both crowding and noise, for example, create a surfeit of stimulation that
can directly overload the system, causing discomfort, negative affect, and
under some circumstances, the marshaling of adaptive resources. Both
negative affect and adaptive responses to challenge or threat in turn directly
affect neuroendocrine and cardiovascular functioning. Physical stressors can
also interact with psychosocial conditions to exacerbate negative affect
and/or psychophysiologic mobilization. For example, noise plus high
workload demands leads to more serious health outcomes than workload
levels alone. Noise and crowding frequently covary with other psychosocial


risk factors (e.g., poverty, inadequate working conditions), and thus have
the potential to contribute to multiple risk situations.
A second manner in which the physical environment can contribute to stress
is by damaging or ameliorating coping resources themselves. People rarely
respond to suboptimal physical or psychosocial conditions passively;
instead, they
-365invoke various coping strategies to reestablish some modicum of balance
between environmental demands and personal resources. Evidence is
presented herein, for example, that crowding interferes with the

development and maintenance of socially supportive relationships in the
residential environment. Both chronic noise and chronic crowding appear to
contribute to learned helplessness, adversely affecting self-efficacy and
related motivational processes.
The third way in which physical conditions can operate as stressors is to
elicit coping strategies that in turn lead to poor health. Studies of noise, for
example, reveal that increases in substance abuse occur under noisy
working conditions.
Another aspect of research on psychological stress and health relevant to
this chapter on environmental stressors are the concepts of vulnerability and
resilience (Cohen, Kessler, & Gordon, 1995; Rutter, 1983). Just as certain
personal or situational characteristics can render individuals more or less
vulnerable to social stressors, there is evidence of vulnerable subgroups
among the population who appear more adversely affected by noise and by
crowding, respectively. Thus throughout both direct, man effects and
associations between environmental stressors and health, as well as
occasions with vulnerable subgroups, are noted.
The field of environmental stress (Cohen, Evans, Stokols, & Krantz, 1986;
Evans, 1982; Evans & Cohen, 1987) is sufficiently developed such that
exhaustive coverage is impossible. A small amount of environmental stress
research has examined climatic conditions as potential psychological
stressors influencing human stress responses (Bell & Greene, 1982; Evans,
1994). Research on housing conditions as a possible stressor have also been
undertaken (Framan, 1984). The focus of this chapter is on the two most
studied environmental stressors, crowding and noise. Health outcomes
include physical health and psychological health. Moreover, the chapter
examines underlying psychosocial psychophysiological processes that may
help explain the linkages between noise and crowding and major physical
and psychological health outcomes. Psychophysiological mechanisms,
immune function, social resources, coping strategies, and motivational

processes are examined.

CROWDING
The element of crowding that relates most strongly to physical and
psychological health is people per room. Traffic congestion may also prove to
be a potent stressor. Area measures of crowding, such as people per acre,
generally have little or no relation to health. Although some studies of
crowding separate group size effects from density effects, the vast majority
of studies have confounded these two factors, manipulating or measuring


density as it covaries with group size. Therefore, some of the effects
attributed to crowding may be clue to group size rather than the amount of
space per person. At the same time, when attempts have been made to
distinguish between these related concepts, density and group size, the
impacts of density typically persist.

Physical Health
Early interest in crowding in the public health field emanated from concerns
about the spread of disease among crowded populations (Cox, Paulus,
McCain, & Karlovac, 1982). There is a large literature on this topic. Physical
health has been operationalized in this literature as rates of illness based on
archival data, visits to infirmary, physical development among children, and
self-reports of somatic symptoms. Archival evidence for positive associations
between crowding and ill health come from studies in prisons (McCain, Cox,
& Paulus, 1976; Paulus, 1988), refugee camps (Amow, Hierholzer, Higbee, &
Harris, 1977), and schools (Essen, Fogelman, & Head, 1978; Koopman,
1978). The Arnow et al. (1977) study is noteworthy because they
demonstrated over time between Vietnamese refugee camp population
fluctuations with changes in a highly contagious disease (acute

conjunctivitis). There is also evidence that crowded residential conditions are
linked to disease both among children (Booth & Johnson, 1975; Jacobson,
Chester, & Fraser, 1977) and among adults (Levy & Herzog, 1978;
McGlashen, 1977; Menton & Meyers, 1977; Sims, Downham, McQuillin, &
Gardner, 1976; Wyndham, Gonin, & Reid, 1978; Yarnell, 1979). Yodfat, Fidel,
Cohen, and Eliakim (1979) found that linkages among residential crowding
and asthma were due to number of children rather than density per se.
Booth (1976) found that male adults, but not women, had greater levels of
disease in crowded homes. Traffic congestion levels among commuters is
also associated with illness-related absenteeism from work (Novaco, Stokols,
& Milanesi, 1990).
Several studies of residential crowding find little or no correlates with
disease (Brett & Benjamin, 1957; Collette & Webb, 1974; Mackintosh, 1934;
McKinlay & Truelove, 1947; Quinn, Lowry, & Zwaag, 1978), and Winsborough
(1965) uncovered an inverse relation between area density and tuberculosis.
Schmitt and colleagues also found no relation between residential density
and disease rates, but found small, positive correlations with area density
(people per acre) measures (SGhrnitt, 1966; Schmic, Zane, & Nishi, 1978).
Similar trends have been uncovered by Levy and Herzog (1974). Kellett
(1984) made the important point that certain diseases should be expected a
priori to correlate with crowding more so than others. Kellett examined
morality patters for specific diseases in London for a S-year period. As in
prior work, persons per room rather than people per acre appeared more
useful in predicting mortality. Second, diseases wherein a major stress
component is believed to be operative (e.g., hypertension, myogardial
infarction, vascular disorders, asthma) were related to household crowding
whereas many other diseases (e.g., various forms of cancer) were not.
Fradman and colleagues challenged many of these studies of crowding and
disease, noting that poor or nonexisting controls for other variables such as
socioeconomic status are common in the crowding and epidemiological

literature. They found in a well-controlled residential crowding was not a
significant predictor of distiase (Freedman, Heshka, & Levy, 1975). However,
the prison
-366-


studies and a few of the residential studies (e.g., Menton & Meyers, 1977) do
have good controls for SES. Furthermore, there are trends in the data
indicating that when individual levels of exposure to density and individual
indices of health are compared rather than aggregated population statistics,
such as used by Freedman and colleagues, stronger results occur.
Nonetheless, Freedman and colleagues' cautious perspective on crowding
and disease is well taken. Overall findings are suggestive but not rigorously
or consistently supportive of a crowding-disease link. It would be useful to
include, in the same individual-based study, disease rates for disorders that
ought to vary with stress exposure plus inclusion of immunological
measures.
A handful of studies in institutional contexts have examined crowding and
infirmary visits. These studies converge on positive associations between
levels of crowding and infirmary visits among shipboard military personnel
(Dean, Pugh, & Gunderson, 1975, 1978), college campus residents (Baron,
Mandel, Adams, & Griffen, 1976; Stokols, Ohlig, & Resnik, 1978), and
prisoners (Paulus, 1988). The prison effects were most noticeable among
inmates forced to live under dormitory- like conditions rather than in single
cells. Trends also indicated that the associations in prisons were somewhat
stronger for men than women and for African American in comparison to
Anglo prisoners (Paulus, 1988). Wener and Keys (1988) found that increases
in density (doubling up cell mates) markedly elevated (nearly 50%) sick call
rates among prison inmates.
A few studies have examined physical development among crowded

children, uncovering evidence of negative associations between household
density and physical stature (Booth, 1976; Essen et al., 1978; Goduka, Poole,
& Aotaki- Phenice, 1992). Crowded children, particularly boys, are shorter.
Shapiro (1974) also found that boys, but not girls, motoric development
appeared to be inhibited in crowded homes. Moreover, this effect was
amplified among children of less educated mothers. More recently,
Widmayer and colleagues (1990) found delayed psychomotor development
among infants as a function of household density, controlling for
socioeconomic status (SES).
Self-reported levels of physical illness are positively associated with
crowding in prisons (Cox, Paulus, & McCain, 1984; McCain et al., 1976),
among college dormitory women but not men (Karlin, Epstein, & Aiello,
1978), and among crowded home settings (Gove & Hughes, 1983)-although
Booth (1976) found this association among men, but not women, in crowded
homes. Giel and Ormel (1977) and Baldassare (1979) failed to replicate the
association between home crowding levels and self-reported illness. The
validity of all the self-report data on illness and crowding is suspect given
retrospective self-report indices. On the other hand, Cox et al. (1984) found
a dose-response function between number of inmates per cell and selfreported illness levels among male prisoners. Of additional interest, Gove
and Hughes (1983) provided some evidence that heightened illness levels
associated with crowded residences are related to lack of sleep and lower
resistance when exposed to other sick family members (all self-reported).
There is evidence that some of the association between crowded living
conditions and self-reported health symptoms is mediated by loss of
perceived control over the living environment. Ruback and associates found
that both female and male prisoners' reports of ill health in association with
crowding were also negatively related to perceived control (Ruback & Carr,
1984; Ruback, Carr, & Hopper, 1986). Another way in which environmental
stressors like crowding can impinge on health is through injuries. Rhesus
monkeys when crowded, for example, show a 5-fold increase in incidents of

injuries (Boyce, O'Neill-Wagner, Price, Haines, & Suomi, 1998).

Psychophpysiologieal
Several studies have examined the relation between crowding and blood
pressure in people. Laboratory studies with random assignment to density
levels have found small but significant elevations among crowded versus
uncrowded participants (Epstein, Lehrer, Csz Woolfolk, 1978; Evans, 1979).
Field studies of prisoners (D'Atri, 1975; Paulus, McCain, & Cox, 1978) and
automobile commuters (Novaco, D. Stokols, Campbell, & J. Stokols, 1979;
Schaeffer, Street, Singer, & Baum, 1988; Stokols et al., 1978) have also
revealed correlational evidence for elevated blood pressure under more
crowded or congested living or commuting conditions. The commuting
studies have found that the effects are stronger for car poolers rather than
solo drivers, for Type B rather than Type A drivers, among external versus
internal locus of control drivers, and among drivers with less residential
choice. One field study found no relations between chronic residential
crowding and blood pressure or neuroendocrine indices among adults
(Booth, 1976), although small, statistically significant elevations in serum
cholesterol were noted among crowded men. No such correlation was noted
among women. Booth's sample did not vary much in density, which may
have weakened his findings. Evans, Lepore, Shej wal, and Palsane (1998)
found elevated blood pressure among crowded boys, but not girls, among
working- class families in India.
Another cardiovascular function, blood pressure reactivity, has been related
to chronic crowding in adults. Residents of more crowded neighborhoods had
higher reactivity (increase from baseline in blood pressure levels) and took
longer to return to resting baseline levels (Fleming, Baum, Davidson,
Rectanus, & McArdle, 1987). Both heightened reactivity and protracted

recovery to baseline are potentially important precursors to the
development of coronary heart disease.
Neuroendocrine markers of stress, typically urinary catecholamines and
cortisol, have been noted in several studies of crowded commuters
(Lundberg, 1976; Singer, Lundberg & Frankenhaeuser, 1978) and among bus
drivers operating under more congested driving conditions (Evans & Carrere,
1991).
Pedestrian exposure to more crowded urban areas elevates neuroendocrine
activity, at least for males (Heshka & Pylypuk, 1975), and residence in
neighborhoods perceived as more crowded because of commercial
establishments and more people on the street is associated with increased
urinary catecholamine levels (Fleming, Baum, & Weiss, 1987). Dormitory
-367crowding, however, had no apparent effects on neuroendocrine activity
among college students (Karlin et al., 1978). A small sample size may have
rendered low power. These authors did find, however, that uncrowded
residents' neuroendocrine indices dropped over the course of the semester,
whereas crowded residents' neuroendocrine levels increased over the same
time period. Schaeffer, Baum, Paulus, and Gaes (1988) found that prisoners
housed in more open, unpartitioned dormitories felt more crowded and
experienced elevated chronic catecholamine levels in comparison to
prisoners living in smaller groups.


The critical role of control has been implicated in some of these
psychophysiological crowding studies. Lundberg (1976) and his colleagues
found that passengers with greater choice over seating were less negatively
impacted by congested commuting. Evans and Carrere (1991) found that the
neuroendocrine effects of traffic congestion on bus drivers were largely
mediated by perceived control on the job. On the other hand, perceived
control did not mediate the positive relation between prison crowding and

neuroendocrine elevations (Schaeffer, Baum, Paulus,
A few laboratory studies have also utilized skin conductance as an index of
psychophysiologic stress, generally finding elevations among more crowded
participants (Aiello, Epstein, & Kalin, 1975; Aiello, Nicosia, & Thompson,
1979; Bergman, 1971; Nicosia, Hyman, Karlin, Epstein, & Aiello, 1979).
Studies of crowding and skin conductance are evenly split on gender
differences, with some studies finding more pronounced effects among
males than females and other studies finding no sex differences. There is
also evidence that skin conductance may be more strongly affected by
crowding when physical touching occurs. McCallum, Walden, and Schopler
(1979) found that acute crowding elevated palmar sweat but only when
experimental subjects were motivated to maintain high levels of group
performance. When performance was permitted to deteriorate under
crowding, no physiological elevations were noted. Finally, in a field study,
Cox, Paulus, McCain, and Schkade (1979) found a significant positive
correlation between the palmar sweat index and crowding among prison
inmates.
Although indirect, some findings by Hutt and Vaizey (1966) may shed some
light on psychophysiological mechanisms associated with crowding and
psychophysiologic responses. Chronically overaroused children responded to
high density laboratory conditions by extreme social and physical
withdrawal; whereas chronically underaroused children and children without
arousal disturbance reacted in the opposite direction, becoming more
engaged and aggressive with other children.
Many animal studies have examined endocrine activity among crowded
species both under laboratory and field conditions (see Evans, 1978, for a
review). Generally, this research indicates support for a population
regulation feedback mechanism whereby crowded animals' fertility declines.
This occurs more markedly among subordinate rather than dominant
animals and appears to be mediated by adrenal cortical activity. Attempts to

link crowding with population regulation among human beings have proven
futile.

Immune Function
Animal but not human work has examined immunological processes as a
function of crowding, generally finding evidence of compromised immune
functioning among more crowded animals (Christian, 1963; E. A. Edwards &
Dean, 1977; Thiessen & Rodgers, 1961). These effects appear to be stronger
among subordinate rather than dominant animals and among animals
without a history of crowded living conditions (Cassel, 1971). Cassel(1974)
pointed out, however, that compromised immune function alone cannot
account for changes in morbidity among crowded animals because both
infectious and noninfectious diseases are elevated among crowded animals.

Psychological Health


Ever since 1962 when Calhoun published his famous Scientific American
study of pathology among overpopulated rats, researchers and policymakers
alike have wondered about the potential role of crowded living conditions on
mental health. The chapter first reviews research on linkages between
density and psychological distress and then turns its attention to recent
work examining possible underlying mechanisms for this linkage.
Many studies have uncovered positive relations between residential density
and self-reported psychological distress (Edwards, Fuller, Sermsri, &
Vorakitphokatorn, 1990; Evans, Palsane, Lepore, & Martin, 1989; Gabe &
Williams, 1987; Gove & Hughes, 1983; Hassen, 1977; Jain, 1987; Lakey,
1989; Marsella, Escudero, & Gordon, 1970). Mitchell (1971) found greater
worrying among crowded families but only if they were also poor. Crowding
in Mitchell's study was unrelated, however, to more serious indices of

psychiatric illness. Lepore, Evans, and Schneider (1991) found evidence. that
residential crowding causes psychological distress in a prospective,
longitudinal study of crowding and mental health. Controlling for educational
levels and income, they found that crowded residents did not differ from
uncrowded residents in psychological distress symptoms during initial
occupancy (I- = .12), but after 2 months and 8 months the associations
became significant (r = .21; r = .27). This is the only prospective study of
crowding and health. Webb and Collette (1975) found an association
between residential density and use of prescription hypnotics.
Booth (1976), Baldassare (1979), and Giel and Ormel (1977) failed to find a
positive association between residential crowding and psychological distress.
These studies had little variance in density. Moreover, Baldassare relied on
mental health indices of questionable sensitivity (one dichotomous item in
one case, and three dichotomous items in a second case). Two studies of
neighborhood crowding levels have also found linkages to psychological
distress (Collette & Webb, 1974; Fleming, Baum, $I Weiss, 1987).
Studies utilizing archival indices such as psychiatric admissions or suicide
rates generally find very weak or insignificant associations between
crowding and pathology when measured in the aggregate (Freedman et al.,
1975; Gove &
-368Hughes, 1980; Schmitt, 1966; Schmitt et al., 1978). In some studies
negative associations between density and psychiatric admissions have
been uncovered, probably created by the association of living alone and
mental disorder (Galle, Gove, & McPherson, 1972; Levy & Herzog,
1974/1978). One exception to these generally negative trends in archival
indices of mental health and crowding is notable. Several prison studies
have found clear, strong associations between the total population size of
prison populations and indices of psychiatric illness (Paulus, 1988).
Quite a number of studies have examined psychological symptoms among
children living in crowded homes. Plant (1937) described several case

studies noting a pattern of low self-sufficiency and little idealism among
children from crowded homes. He attributed these patterns to mental strain
associated from always having to get along with others and to exposure to
adults under close quarters that made it difficult to look up to or idealize
grownups. Crowded children have increased levels of various symptoms of
psychological distress (Booth, 1976; Gasparini, 1973; Murray, 1974; Saegert,
1982; Wachs, 1987). Parents in more crowded homes report relief when their
children are outside (Gove & Hughes, 1983), have more difficulty supervising


their children (Mitchell, 197 l), and are generally less responsive and
involved with their children (Bradley & Caldwell, 1984; Evans, Maxwell, &
Hart, 1999; Wachs & Camli, 1991) in comparison to uncrowded parents of
comparable social class. These trends appear to be exacerbated in the
presence of other risk factors, particularly poverty (Baldassare, 1981;
Bradley et al., 1994).

Psychosocial Resources
Some of the relation between high residential density and psychological
distress in children may be linked to family interactions, which have been
found to be more contentious under crowded living conditions (Booth, 1976;
Gasparini, 1973; Saegert, 1982). There may also be greater incidence of
physical punishment and open expression of anger between parents and
children in crowded homes (Booth & Edwards, 1976; Light, 1973), although
Gove and Hughes' (1983) study did not support this finding.
Another factor that may help explain the link between high residential
crowding and symptoms of psychological distress in children is withdrawal.
Aiello, Thompson, and Baum (1985) reviewed several field and laboratory
studies documenting increased social withdrawal under crowded conditions
among young children. Similar trends exist in the adult literature, indicating

that crowded adults interact with housemates less (Baum & Valins, 1977,
1979; Proshansky, Ittelson, & Rivlin, 1970); are less friendly with their
neighbors (McCarthy & Saegert, 1978), and have impaired social support
with those they live with (Evans et al., 1989; Lakey, 1989; Lepore et al.,
1991). Baldassare (1979) did not replicate linkages between residential
crowding and neighboring. People under crowded conditions also tend to be
less affiliative in their behaviors toward others (R. L. Munroe & R. H. Munroe,
1972) and view others in more negative or suspicious terms (Bickman et al.,
1973; Griffit & Veitch, 1971; McCarthy & Saegert, 1978). There is also
evidence that crowded working conditions lead to greater social withdrawal
from coworkers (Oldham & Fried, 1987). Finally, as already noted, parents in
crowded homes are less responsive to their children (Bradley & Caldwell,
1984; Wachs Bt Camli, 1991). Furthermore, this relative unresponsiveness
partially accounts for less complex parent to child verbalizations to infants
and toddlers (Evans, et al., 1999).
Evidence that social withdrawal and impaired social relationships are a
primary mechanism accounting for the relation between crowding and
psychological distress has been documented in some detail by two research
programs. Baum and colleagues found that more crowded dorm residents
report more unwanted social interaction in their dorms. These same crowded
residents also evidence greater behavioral indices of withdrawal outside of
the dorm. They sit farther away from other research participants and
withdraw more in group interaction games (Baum & Valins, 1977, 1979).
Residential exposure to high levels of street traffic is also associated with
less neighboring (Appleyard & Lintell, 1972; Halpern, 1995). Evans and
Lepore showed direct evidence for a similar pattern. They found both crosssectionally (Evans et al., 1989) and in a prospective, longitudinal design
(Lepore et al., 1991) that the negative effects of residential crowding on
psychological distress (with controls for social class) are mediated by social
support. Similar patterns also appear to occur among children in crowded
residences (Evans, et al., 1998).

Evans and Lepore (1993a) also found that crowded relative to uncrowded
residents were less likely to offer support to a confederate under stress in an


uncrowded laboratory setting. Of additional interest, crowded residents in
comparison to uncrowded residents were also less responsive to offers of
social support during a stressful situation (see Fig. 20.1 ). Ignored meant
that the subject did not look at or made no verbal acknowledgment of the
confederate; acknowledgment meant some brief comment or a head nod
was given in response to offers of support; and accepted meant the subject
was very responsive

FIG. 20.1
-369to the confederate's offers of support, elaborating or embellishing on their
offers of support.

Motivation
Many theorists have postulated that a prime reason why crowding can have
negative impacts on psychological health is because of reduced behavioral
options and greater difficulty in regulating social interaction (Altman, 1975;
Baron 8z Rodin, 1978; Schmidt & Keating, 1979). An important psychological
consequence of prolonged exposure to an aversive, uncontrollable stressor,
such as crowding, may be learned helplessness. Persons chronically exposed
to crowding report feeling a greater sense of powerlessness over their living
environments than their less crowded counterparts (Baum & Valins, 1977,
1979; Baron et al., 1976; Carr, Hopper & Ruback, 1986; Saegert, 1978).
Sherrod (1976), Aiello, DeRisi, Epstein, and Karlin (1977), Evans (1979),
Nicosia et al. (1979), and Doofey (1978) all found negative aftereffects
immediately following laboratory exposure to crowded conditions. Sherrod,
Evans, and Nicosia and colleagues each utilized the Glass and Singer (1972)

aftereffects paradigm that measures persistence on challenging puzzles.
Giving up sooner in the face of challenge in an achievement context may be
indicative of greater helplessness (Cohen, 1980; Glass & Singer, 1972).
Dooley (1978) incorporated proofreading performance as her aftereffects
measure. Saegert, Mackintosh, and West (1975) reported that crowded train
stations produced negative aftereffects in women only. Nicosia's data also
indicated more severe aftereffects of crowding among women.
Parallel trends to the laboratory work have been noted in studies of more
chronic, crowded living conditions, finding less persistence on difficult
puzzles among persons living in more crowde.d neighborhoods (Fleming,
Baum, & Weiss, 1987). Moreover, perception of control over social
interactions largely accounted for the main effect of neighborhood crowding
on the helplessness indicator. Residents of crowded dorms feel less control
over social interaction than their uncrowded counterparts (Baum & Valins,
1977, 1979) exhibit behavioral strategies in a group' prisoners' dilemma
game consistent with helplessness (Baum, Aiello, & Calesnick, 1978; Baum,
Gatchel, Aiello, & Thompson, 1981). Interestingly, the development of
helplessness strategies in the game over the course of the initial semester
under crowded conditions was mirrored by residents growing external
attributions for problems in the dormitory over this same time period (Baum
et al., 1981). Uncrowded residents generally felt self-efficacy over problems
in their dormitory over the course of the semester and these internal
attributions remained stable over time. Crowded dormitory residents are
also less likely to seek clarification when given ambiguous instructions about
an impending laboratory procedure than were uncrowded dormitory
residents.


Traffic congestion also is related to motivational deficits. Greater traffic
congestion levels have been related to decreased task motivation on

challenging puzzles and proofreading (Novaco et al., 1979; Schaeffer et al.,
1988; D. Stokols, et al., 1978).
The most direct evidence for helplessness induced by crowding comes from
a pair of studies on residential crowding and children by Rodin (1976).
Matched on socioeconomic indicators, elementary-aged schoolchildren living
in more crowded public housing were less likely to control the administration
of outcomes in an operant conditioning paradigm in comparison to their less
crowded counterparts. In a second study, helplessness was induced in
adolescents by a classic helplessness paradigm, pretreatment with an
insoluble versus a soluble puzzle. Helplessness was monitored on a second
challenging but solvable puzzle. The main effect of pretreatment solvability
(the helplessness induction) was significantly moderated by residential
crowding with heightened vulnerability to the induction of helplessness
among the more crowded children. In their study of children in India, Evans,
et al., (1998) replicated Rodin's effects (but for girls only). Saegert (1982),
however, did not replicate these findings examining a sample of children
from public housing projects in New York City.

Summary
Residential crowding has little impact on physical morbidity among the
general population. Residential crowding may be linked, however, to ill
health among vulnerable subgroups of the populations, particularly young
children and extremely crowded, captive populations (e.g., prisons, refugee
camps). Evidence linking high density exposure either under controlled
conditions or in the field to elevated cardiovascular functioning is quite
strong. Neuroendocrine functioning also appears elevated, although less
data are available. The potential clinical implications of these two data
patterns has not been explored in the crowding literature.
Psychological distress is increased by residential crowding. Individual but not
aggregate level analyses continue to uncover a positive association between

crowded living conditions and poorer psychological health. Several studies
have excellent controls for sociodemographic factors and one is a
prospective, longitudinal analysis. Psychological distress associated with
residential crowding may be caused by a typical coping strategy for dealing
with chronic high density living conditions-social of this social withdrawal
may be a breakdown in socially supportive relationships. There is not strong
evidence, however, that human crowding is associated with more extreme
forms of psychopathology characterized in some animal studies as a
behavioral sink.
There is also evidence that crowding may lead to the development of
motivational deficits, particularly among children in achievement-related
contexts. There is indirect evidence suggesting that these motivational
deficits are related to learned helplessness from diminished perceived
control over the environment.

NOISE
Noise, which is defined as unwanted sound, is typically measured in


decibels. Decibels is a logarithmic scale with a
-370change in 10 decibels perceived as approximately twice as loud. There is
considerably more research on noise and health in comparison to research
on crowding and health. The bulk of the noise and health research has
occurred in industrial settings. More recently, studies of noise and health
have also focused on people living in airport impact zones or near to road
traffic noise. Prolonged exposure to high levels of noise is clearly linked to
hearing damage (Kryter, 1994). Because the thrust of this chapter is on
environmental stress, the noise- related hearing damage literature is not
discussed.

Physical Health
Studies have examined exposure to either occupational noise or community
noise and disease. Outside of cardiovascular problems, there appears to be
little relation between noise exposure and physical disease. In industrial
settings, noise has been associated with increased risk for myocardial
infarction (Ising, Babisch, & Giinther, 1999), reductions in cardiorespiratory
efficiency (Semczuk & Gorny, 197 l), difficulties in peripheral circulation and
cardiac problems generally (Jansen, 1961), electrocardiogram abnormalities
suggestive of coronary heart disease (Cuesdan general sickness- related
absenteeism (Cohen, 1973), and self-reported fatigue (Carlestam, Karlsson &
Levi, 1973; Melamed & Bruhis, 1996). Several industrial studies have found
no associations between occupational noise exposure and rates of coronary
heart disease (Lees, Romeril, & Wetherall, 1980) or rates of total illness
(Lees et al., 1980).
Community airport noise studies have shown that higher levels of noise
exposure are associated with greater contact with physicians for coronaryrelated problems (Knipschild, 1977a) and, for women only, use of drugs to
treat hypertension (Knipschild & Oudshoorn, 1977; Koszarny, Maziarka, &
Szata, 1981). These studies also show an association with greater physician
contact in general (Knipschild, 1977b), rates of colds (Ising et al., 1990), as
well as total health symptoms (Pulles & Stewart, 1990), and higher levels of
coronary heart disease symptoms among women but not men (Koszarny et
al., 1981). Graeven (1974) and Hiramatsu, Tamamoto, Taira, Ito, and
Nakasone (1993), however, found no differences in self-reported health
symptoms between persons living in airport impact zones versus citizens in
quiet neighborhoods.
Turning to road traffic noise, Cameron, Robertson, and Zaks (1972) found
little relation between community noise levels (self-reported) and illness
rates. Babisch, Elwood, Ising, and Kruppa (1993) found slight elevated risk
(1.2 odds ratio) in noisier traffic areas in three different sites for men
residing in areas above 65 dBA Leq. However, when comparing across

different noise levels varying from > 50 dBA Leq to 70, they uncovered no
linear relation.
Another area of physical health worthy of note in the noise literature is birth
defects and other abnormalities during pregnancy. Not surprisingly, findings
in this area are highly controversial and not at all definitive. Jones and
Tauscher (1978) found higher rates of birth defects in high airport noise
impact zones relative to quieter areas, but Edmonds, Layde, and Erickson
(1979) could not replicate the findings. Several rodent studies have found
abnormal fetal development following noise exposure (Welch, 1973). There


is evidence that women working under very noisy conditions, particularly if
subjected to additional stressors like shiftwork, have more pregnancy
complications such as vaginal bleeding and pregnancy- induced
hypertension (Nurminen & Kurppa, 1989). Babies born in areas with high
noise impact have lower birth weights (Ando & Hattori, 1977; Knipschild et
al. (1981) with controls for socioeconomic status. Ando (1987) also found an
increase in low birth weight babies following the opening of a new airport.
Schell (1981) also noted that female infants, but not males, had significantly
shorter gestation periods in high airport noise impact zones. Moreover, Ando
and Hattori (1977) showed diminished levels of human placental lactogen
levels in mothers living in high noise airport impact zones. Finally, Schell and
Ando (1991) found a dose- response function relating airport noise levels
and 3-year-olds' physical stature (but not weight) in a large epidemiological
study. The data on possible linkages between noise and early development
are sobering to consider in light of environmental surveys of neonatal,
intensive care units that are often populated by premature babies. Levels of
noise match or exceed recommended standards for ambient traffic exposure
and health (Lawson, Daum, & Turkewitz, 1977).

Psychophysiogical
Although previous reviews of noise indicate that cardiovascular responses
(typically blood pressure or pulse) to noise under acute exposures rapidly
habituate (Glass & Singer, 1972; Kryter, 1994), more careful scrutiny of this
literature indicates important exceptions. Persons who are noise sensitive do
not easily habituate (Conrad, 1973; Stansfeld & Shine, 1993), nor do
individuals who are hypertensive (von Eiff, Friedrich, & Neus, 1982). Shortterm habituation is blocked when people perform demanding cognitive tasks
under noise (Carter & Beh, 1989; Conrad, 1973; Mosskov & Ettema, 1977).
Evans et al. (1996) also showed that noise significantly increases blood
pressure over a 20-minute period without habituation, if it follows exposure
to a psychological stressor (i.e., giving a speech, taking a final examination).
Other psychophysiological indices examined under acute noise have
included electrodermal activity, ECG, EEG, and neuroendocrine activity.
Results parallel the cardiovascular data, indicating rapid habituation (Finkle
& Poppen, 1948; Fruhstorfer & Hensel, 1980). Recent findings suggest,
however, that when short-term exposure to loud noise is accompanied by
demanding tasks, habituation may be blocked (Frankenhaeuser & Lundberg,
1977; Ising, Rebentisch, Poustka, & Curio, 1990; Lundberg &
Frankenhaeuser, 1978). Work by Tafalla and Evans (1997) indicated a central
role of effort in the performance/physiological activation tradeoff.
Performance can be maintained, at least under many circumstances (e.g.,
short-term tasks that do not demand large amounts of attention or
memory), by additional cognitive effort. Such maintenance of performance,
however, appears to exact a cost of greater psychophysiological
-371activation. It is noteworthy that McCallum et al. (1979) found a very similar
pattern for performance under crowded laboratory conditions.
There is also evidence indicating that habituation is interfered with by calling
attention to the potential negative impacts of noise on the person (Vera,
Vila, & Godoy, 1992). This latter finding might explain why noise sensitive
persons apparently do not readily habituate to repeated exposures of acute

noise in the laboratory. Perhaps they are more threatened or concerned


about potential harmful effects of the noise.
Field research on noise and psychophysiologic outcomes has occurred
primarily in industrial settings. The occupational noise and
psychophysiologic literature is too large to review exhaustively herein.
Several reviews of this literature (Kryter, 1994; Thompson, 1981, 1993;
Welch; 1979) have characterized the findings similarly. Unfortunately,
nonexperimental designs have frequently been employed in the
occupational noise and health literature with poor or nonexisting controls,
and many studies have relied on poor estimates of noise exposure.
Furthermore, blood pressure is often poorly measured. Many of the industrial
studies have relied on one or two measures of blood pressure taken during a
physical at work by medical personnel. Moreover, annual medical
examinations or other medical screenings may seriously bias estimates
since some workers become excluded.
Thompson concluded from her two reviews that workers with adequate
hearing protection are unlikely to show much effect of noise on the
cardiovascular system. Kryter (1994) reflected greater concern but also
remained skeptical, noting the paucity of well-designed research studies;
Welch (1979) sounded a considerably greater sense of alarm about
cardiovascular health risks from chronic, occupational exposure to noise.
Interestingly, Welch's review is based primarily on Eastern European
literature that includes worksites with generally very high levels of
occupational noise exposure, often coupled with a paucity of hearing
protection programs. The bulk of the literature in the other major reviews is
based on North American and Western European studies where occupational
noise levels tend to be lower and hearing protection programs more
common.

Difficulties in exposure estimation in industrial studies of noise and
cardiovascular functioning are illustrated by one of the most thorough
investigations (Talbott et al., 1985). Although these investigators found no
significant differences in blood pressure readings that were carefully
administered to men from noisy and from quiet manufacturing plants, they
also uncovered a clear, consistent positive link between elevated diastolic
blood pressure and severe hearing loss in the noisy plant. Moreover, looking
at the subset of men who had worked for at least 15 years in the two
respective manufacturing plants, occupational noise exposure did
significantly relate to both systolic and diastolic blood pressure (Talbott et
al., 1990). See Lercher (1996) for an in-depth discussion of noise exposure
estimation and health outcomes.
It is also conceivable that subsets of workers may be particularly vulnerable
to the chronic effects of noise exposure on their cardiovascular systems. For
example, Tarter and Robins (1990) found that male, African American
automobile plant workers suffered increased blood pressure, whereas their
Anglo counterparts, who were exposed to comparable levels of high noise at
work, did not show this relation. Tarter and Robins speculated that perhaps
racial differences in propensity for hypertension might explain these
findings. Given the fact that individual differences in noise sensitivity
interfere with habituation to acute noise exposure as reviewed earlier, it
might be hypothesized that noise sensitivity creates vulnerable subgroups
within occupationally noise-exposed groups. This idea has not been tested,
although mixed support of such a pattern has been uncovered in community
studies of aircraft noise (Neus, Ruddel, & Schulte, 1983; Stansfeld, 1993).
A few longitudinal studies of noise and cardiovascular functioning in


industrial settings have been conducted. By comparing the same worker in
quiet and noisy periods, some of the weaknesses most endemic to crosssectional studies (e.g., selection bias) are reduced. The U.S. Raytheon (1975)

study, for example, found a significant reduction in medical problems after
the implementation of a hearing conservation program in the plant. No
changes in similar health indices occurred over the same time period among
workers in quiet plant areas. Moreover, the greater the level of compliance
observed (e.g., wearing hearing protection), the greater the apparent health
benefit. Hypertension and cardiovascular disease were included in the
overall health records monitored but could not be singled out because of
insufficient sample size. Antonova (1971) compared miners before and after
their workshifts in either noisy or quiet areas of the mine. Noise significantly
elevated mean arterial pressure with no changes in the quiet group pre and
post work. Systolic blood pressure was significantly elevated among brewery
workers when they did not wear ear plugs in comparison to days in which
they did (Ising & Melchert, 1980). Cortisol fluctuations were also shown to be
dependent on the use of earplugs in a similar design (Melamed & Bruhis,
1996).
Another more rigorous approach to studying industrial noise exposure and
psychophysiological responses is to simulate occupational noise exposure
under experimental conditions with random assigment to noise conditions.
Three-hour exposure to jet turbines significantly elevated blood pressure
over resting levels among workers in a jet assembly plant (Ortiz, Arguelles,
Crespin, Sposari, & Villafane, 1974). Mosskov and Ettema (1977) and
Rovekamp (1983) found elevations in blood pressure in 2- to 3-hour noise
exposures but at much lower intensities of noise than employed by Ortiz and
colleagues. Cartwright and Thompson (1975) found no effects, however, of a
l-hour exposure to loud noise, but Carter and Beh (1989) were able to
significantly elevate cardiovascular parameters from 1 hour of exposure, as
long as participants simultaneously worked at a difficult task.
This latter finding, along with other experimental findings reviewed earlier
on the multiplicative effects of noise and task demands on cardiovascular
and neuroendocrine functioning, is interesting to consider in light of a small

number of occupational noise studies that have also incorporated additional
measures of working conditions. A Russian industrial study reviewed by
Welch (1979) found elevated cardiovascular functioning in a noisy
manufacturing plant among workers
-372with higher levels of workload demands. Workers with low workloads did not
reveal any cardiovascular correlates of occupational noise exposure. Parallel
results were recently uncovered in a longitudinal study (Melamed, BonehKristal, & Froom, 1999). Cottington, Matthews, Talbott, and Kuller (1983) also
reported a significant interaction of job stress and noise on diastolic blood
pressure. Job stress was associated with higher blood pressure in a noisy
manufacturing plant but not a quiet one with good controls for SES and
cardiovascular risk. Similarly, Lercher, Hortnagl, and Kofler (1993) found that
annoyance with noise at work had a small positive association with diastolic
blood pressure. This relation was significantly amplified, however, among
workers who also reported job dissatisfaction and low levels of social support
on the job. Occupational exposure to noise levels may also interact with shift
work. Ottmann, Rutenfranz, Neidhart, and Boucsein (1987) and Cesana et al.
(1982) both found elevated catecholamine levels related to noise levels at
work but only among workers on rotating shifts. Nonshift workers in noisy
work areas did not reveal these associations. Lercher et al. (1993) also found
higher levels of blood pressure among workers annoyed by noise who also


engaged in shiftwork relative to nonshiftwork employees.
There has been a small number of industrial studies or simulation studies
with prolonged noise exposure that have examined neuroendocrine and
other biochemical markers of stress rather than cardiovascular functioning.
Mixed results have been uncovered with no relation between noise exposure
and cholesterol (Brown, Thompson, & Folk, 1975), cortisol (Brandenberger,
Follenius, & Tremolieres, 1977; Cavatorta et al., 1987; Slob, Wink, & Radder,
1973), and with one or more catecholamines (Carlestam et al., 1973; J.

Osguthorpe, Mills, & N. Osguthorpe, 1983; Paulocci, 1975; Slob et al., 1973).
Other studies have uncovered significant, although typically small,
associations between noise levels on the job or from simulated exposures
and various psychophysiologic indicators, such as reduced urine volume and
17-ketosteroid levels (Gibbons, Lewis, & Lord, 1975), elevated fatty acids
(Ortiz et al., 1974; Proniewska et al., 1972), higher levels of cholesterol
(Cantrell, 1974; Ortiz et al., 1974; Rai, Singh, Upadkyay, Patil, & Nayer, 198
l), increased epinephrine levels (Cavatorta et al., 1987; Ortiz et al., 1974;
Slobet al., 1973), elevated cortisol (Cantrell, 1974; J. Osguthorpe et al.,
1983; Rai et al., 1981), and increased levels of ACTH and oxytocin
(Fruhstorfer & Hensel, 1980). Although there are more published positive
findings, it is important to keep in mind that most of these noise and
biochemical studies find small changes, and null results are more difficult to
get published. On the other hand, there is also a large animal literature
generally consistent with significant biochemical outcomes from acute noise
exposure under controlled conditions (B. Welch & A. Welch, 1970).
Increasingly, researchers have turned their attention to community studies
of noise and psychophysiologic parameters, particularly blood pressure.
Traffic noise levels appear to have no relation to blood pressure in
community samples (Elwood, Ising, & Babisch, 1993; Lercher & Kofler, 1993;
Knipschild & Salle, 1979) or show a small positive associa tion (von Eiff,
Friedrich, & Neus, 1982; Neus, Ruddel, Schulte, & von Eiff, 1983, Wu, Chiang,
Huang, & Chang, 1993). Regecova and Kellcrova (1995) found that traffic
noise both at home and at school was associated with elevated blood
pressure among 3- to 7-year-olds. were multiplicative effects as well of
school and home noise. The Neus study is noteworthy because it is
longitudinal. The Wu study bears mention as well since they found that
traffic noise elevated young children's blood pressure as a function of
hearing status. Congenitally deaf children were unaffected by road noise,
whereas their able-hearing counterparts suffered small elevations. Herbold,

Hense, and Keil (1974) noted a small positive relation between traffic noise
levels and hypertension prevalence among adults and Babisch, Fromme,
Beyer, and Ising (1996) found elevated overnight neuroendocrine stress
hormones. Simulated exposure to traffic noise under controlled conditions
elevates both cardiovascular and neuroendocrine activity as a function of
sound intensity (Ising, Dienel, & Markert, 1980; Osada, Ogawa, Hirokawa, &
Haruta, 1973). Ising's study is particularly interesting because, as in several
of the acute noise exposure studies noted earlier, he found that exposure to
simulated traffic noise while working had significant effects on both
cardiovascular and neuroendocrine levels, especially when mental loads
were higher. In one of the more rigorous tests of ambient noise exposure and
cardiovascular health, Peterson, J. S. Augenstein, Tanis, and D. G. Augenstein
(1981) were able to produce sustained, elevated arterial blood pressure in
monkeys exposed for long periods of time to simulated recordings of aircraft
and traffic noise played at typical ambient levels (Leq = 78). Their work also
showed that these monkeys sustained no hearing damage.
Studies of airport noise, which is typically louder and less predictable than
road traffic noise, generally find stronger associations between noise
exposure and elevated cardiovascular functioning in comparison to the road


traffic noise studies. Most studies have focused on children rather than
adults, which might also explain the generally more consistent, positive
results than those uncovered in the road traffic noise literature.
Knipschild (1977a) found a dose-response relation between community
airport noise exposure and hypertension among adult residents in
Amsterdam. Two studies of simulated, military aircraft flights at low altitude
have shown significant increases in blood pressure among elderly residents
(Michalak, Ising & Rebentisch, 1990) and in catecholamines among middleaged adults (Maschke, Breinl, Grimm, & Ising, 1992). Several studies have
found significant relations between exposure to aircraft noise and elevated

blood pressure in children (Cohen, Evans, Krantz, & Stokols, 1980; Cohen et
al., 1986; Evans, Hygge & Bullinger, 1995; Ising et al., 1990; Karagodina,
Soldatkina, Vinokur, & Klimukhin, 1969; Karsdorf & Klappach, 1968;
Schmeck & Poustka, 1993). Several of these studies have very thorough
statistical controls for socioeconomic status. One study has found no relation
between airport noise levels and blood pressure (Cohen, Evans, Krantz,
Stokols, & Kelly, 1981), but these data were explained by selective attrition
(persons in noisy areas with
-373the highest levels of blood pressure left the area). Troche, Chumlea, ambient
noise exposure in suburban communities with no nearby airports or major
highways. This study is flawed because of unreliable blood pressure
measurement procedures and use of self-reports for noise exposure
estimation.
Evans and colleagues (1995) also investigated reactivity to a noise source,
as well as chronic neuroendocrine activity levels as a function of community
airport noise exposure. As shown in Table 20.1, they found evidence of
elevated catecholamine activity, but no shifts in cortisol among elementary
schoolchildren living in the flight path of a major international airport. Of
further interest, children chronically exposed to noise appeared less reactive
to an acute noise source while reading.
Ising and his colleagues found parallel trends for epinephrine, but not
norepinephrine, and also found elevated cortisol in two sets of studies with
adults that simulated exposure to night-time aircraft operations (Maschke,
Ising, & Arndt, 1995). Of additional interest, in one study they generated a
dose-response function between elevated overnight hormonal levels and
sound intensity levels. Finally, Evans, Bullinger, and Hygge (1998) replicated
their cross-sectional aircraft noise and young children's health findings in a
prospective, longitudinal study of children living in the vicinity of the new,
Munich international airport.

Immune Function
A large number of animal studies have utilized noise as a stressor to
investigate altered immune function. The results, like those of the few
human studies are quite mixed (Bly, Goodard, & McLean, 1993). Sieber et al.
(1992), for example, found that uncontrollable but not controllable noise
significantly decreased natural killer cells among healthy male subjects;
Weisse et al. (1990) found the opposite pattern with controllable noise
causing lymphocyte resistance to mitogens to drop.


Coping Behaviors
An alternative pathway by which noise and other environmental stressors
may impact physical health is the exacerbation of substance abuse.
Cigarette smoking and alcohol consumption both increase under stress
(Cohen et al., 1986). In the presence of loud noise, nicotine ingestion
reduces muscle tension (Hutchinson & Emley, 1973) and accelerates
habituation (Friedman, Horvath, & Meares, 1974). In a particularly
interesting study, Cherek (1985) demonstrated a dose-response function
between cigarette smoking (objective, experimental measures) and
controlled exposures to varying noise levels (60–90 dBA).

Psychological Health
Several different types of studies have examined chronic noise exposure and
mental health. The first set of studies explored possible relations between
psychiatric admissions and aircraft noise exposure with decidedly mixed
results. Several studies have found positive correlations between admission
rates and high noise exposure (Abey-Wickrama, A'Brook, Gattioni, &
Herridge, 1969; Herridge & Chin, 1972; Jenkins, Tarnopolsky, & Hand, 1981;
Meecham & Smith, 1977). Nonsignificant relations have been found by
Gattoni and Tarnopolsky (1973), and Jenkins, Tarnopolsky, Hand, and Barker

(1979) found an inverse relation between noise levels and psychiatric
admissions in the same region (Heathrow, to the West of London) utilized by
Abey-Wickrama and by Jenkins et al. (1981). Kryter (1990), in a further
analysis of some of Jenkins' data, discovered large ethnic differences that
might have explained Jenkins' puzzling findings. Many of these studies have
poor controls for social class and all are cross-sectional.
Self-reports of psychological distress were unrelated to road traffic noise
levels in two cross-sectional studies (Tarnopolsky & Morton-Williams, 1980;
Tarnopolsky, Watkins, & Hand, 1980) and in a prospective, longitudinal study
(Stansfeld, 1993). The absence of support for a link between road traffic
noise exposure and psychological health could be due, in part, to noise
measurement. Halpem (1995) found that peak noise levels predicted several
indices of psychological health, controlling for socioeconomic status of
residents. Mean levels of traffic noise had no mental health correlates.
Physician treatment for psychological problems, as well as use of hypnotic
drugs, was associated with aircraft noise around Amsterdam (Knipschild,
1977b). Koszarny et al. (1981) demonstrated a similar relation, but only
among women. Knipschild and Oudshoorn (1977) also found a clear relation
among prescription rates for tranquilizers and aircraft noise over a 7-year
period. Moreover, these authors found longitudinal trends in use of hypnotic
pharmaceuticals that tracked changes in noise levels in airport impact
zones. At the same time, they noted lower and consistently similar utilization
rates among quiet neighborhoods of comparable socioeconomic
composition. Grandjean, Graf, Lauber, Meier, and Muller (1976) found a
dose-response function linking airport noise exposure to self-reported use of
sleeping pills and tranquilizers. Watkins, Ttamopolsky, and Jenkins (1981),
however, could not replicate the linkages between drug usage and aircraft
noise exposure. One study has also uncovered a coarse dose-response
function between occupational noise exposure and psychological symptoms
among blue-collar workers (Mc- Donald, 1989). Interestingly, in light of

earlier work on crowding, social support and psychological health, MC
Donald also noted that impaired interpersonal relationships at work


appeared to play a role in the mental health-noise links.
-374-

Motivation
Interestingly, the initial study of helplessness and human beings utilized
inescapable noise as the induction stimulus. Hiroto (1974) demonstrated
that short-term exposure to inescapable noise induces helplessness. Adults
were exposed to noise or quiet during an initial phase of an experiment. Half
of the noise subjects could avoid the noise by learning an avoidance
response. For the other half of the noise subjects, the noise was inescapable.
The groups were then tested in a similar situation where noise could easily
be avoided by a simple manual response. A second series of experiments
replicated Hiroto's findings and also demonstrated that the helplessness
induced by inescapable noise generalized to persistence on subsequent task
performance (Hiroto & Seligman, 1975). Subjects exposed to inescapable
noise exhibited significantly greater helplessness in the second testing
phase, regardless of the similarity of the helplessness induction and testing
phase (Hiroto & Seligman, 1975). Furthermore, the helplessness effects of
inescapable noise were greater for external locus of control individuals
(Hiroto, 1974). Krantz, Glkass, and Snyder (1974) found similar results in two
studies of inescapable versus escapable noise. One final detail of Hiroto and
Seligman's work worthy of note is that the learned helplessness effects of
inescapable noise were quite similar to the induction of helplessness
produced by exposing subjects to insoluble concept formation problems.
A large number of studies, initiated by Glass and Singer's pioneering work on
perceived control and stress (1972) have examined performance

aftereffects, immediately following exposure to uncontrollable noise. The
basic paradigm includes exposing participants to noise while working on a
cognitive task for a period of about 30 minutes. The participant then leaves
the room and is asked to do another, apparently unrelated task where noise
is no longer present (see Cohen, 1980, for an overview of this paradigm).
Uncontrollable noise causes deficits in task persistence on puzzles (Gardner,
1978; Glass & Singer, 1972; Glass, Singer, and Friedman, 1969; Percival &
Loeb, 1980; Sherrod, Hage, Halpern, & Moore, 1977; Wohlwill, Nasar, DeJoy,
& Foruzani, 1976). Work by Glass and Singer (1972) also showed that the
controllability, and to a lesser extent the predictability, of the noise is a
critical component of these aftereffects. In a test of the external validity of
the initial Glass and Singer findings, Moran and Loeb (1977) utilized taperecorded aircraft noise and found, unexpectedly, that such noise did not
appear to induce aftereffects in the laboratory. Percival and Loeb (1980)
reasoned that perhaps airport noise, because of its temporal qualities, is
rather predictable. Thus, they replicated the original Moran and Loeb finding
utilizing the same stimuli, but of particular interest, found that when the
aircraft noise bursts were sudden rather than the typical slow onset pattern
of an approaching aircraft, negative aftereffects could be reliably produced.
Rotton, Olszewski, Charleton, and Soler (1978) also showed that meaningful
speech rather than noise could induce the same negative aftereffect. Evans
et al. (1996) indicated that these negative aftereffects are amplified if
exposure to uncontrollable noise occurs among subjects already under
psychological stress. Finally, Glass and Singer (1972) found that
uncontrollable noise interferes with subsequent proofreading accuracy.
A small number of studies has also examined possible relations between
chronic noise exposure and susceptibility to helplessness. Evans et al.


(1995) adapted the Glass and Singer aftereffects puzzle for young children.
They found that children living in high airport noise zones were less likely to

persist at solving line tracing puzzles than their quiet community
counterparts. Cohen and colleagues (Cohen et al., 1980, 1981) found that
aircraft noise-exposed children were significantly less likely to solve a
difficult, challenging puzzle than quiet comparison groups. Of particular
interest, noise-impacted children were also more likely to simply give up on
the puzzle before the allotted 4 minutes had passed. Fifteen percent of
children from noisy schools failed the puzzle by giving up in comparison to
only 2% of children from quiet schools. It is worth noting that the puzzles
were designed and pretested to be fun and engaging to elementary-aged
schoolchildren. These effects were replicated by Cohen and colleagues and
similar trends were also found for home noise levels (Cohen et al., 1986).
Both the Evans and Cohen studies had well-matched SES comparison
groups. Moth-Sibony (1984) found very similar results in kindergarten
children exposed to higher levels of aircraft noise in Paris. Wachs (1987) also
showed that infants exposed to more noise at home manifest less masteryoriented play as indexed by a standardized observation instrument. Of
additional interest, teachers in noisy schools frequently report more
difficulties motivating students than do teachers from quiet schools (see
Evans & Lepore, 1973b, for a review). Finally, Cohen et al. (1986) uncovered
a relation between children's willingness to relinquish choice and chronic
noise exposure. Children from noisy schools relative to quiet schools were
significantly more likely to allow an experimenter to choose a reward at the
conclusion of their experiments rather than make their own choice.

Summary
Both industrial and community studies find no clear, consistent pattern of
data on noise and morbidity. Similarly, data on acute noise exposure and
altered immune functions are mixed. Although not plentiful, there is a
confluence of findings suggestive of noise impacts on in utero development
that warrant followup. Several studies point to noise as a factor in elevated
smoking.

Acute noise produces short-lived elevations in cardiovascular and
neuroendocrine functioning. Recent research suggests, however, that
individuals sensitive to noise as well as situations with high workload
demands can diminish and perhaps even block such habituation. A plethora
of methodologically weak, occupational noise and health studies reveal
decidedly mixed findings on noise and blood pressure. Some longitudinal
studies indicate small, positive associations between occupational noise
exposure and blood pressure elevations. Road traffic noise appears to have
no significant impact on blood pressure of community residents, but persons
living in the proximity of airports, particularly children, are at risk
-375for elevated blood pressure. The clinical significance of these elevations is
unknown at this time.
Data on noise and psychological health are unclear. The preponderance of
poorly designed studies links community noise levels to rates of psychiatric
illness. There are better studies indicating some link between community
noise exposure and utilization of pharmaceutical hypnotics. Both laboratory
and field studies reveal that noise, particularly uncontrollable noise, can
contribute to diminished motivation related to learned helplessness. Children


chronically exposed to noise may be particularly susceptible to this
phenomenon.

DISCUSSION
Application of the construct of psychological stress to examine the role of
the physical environment in human health has proven useful in the case of
crowding and noise. The primary contributions to date have been the
identification of stress- related outcome measures likely to be related to
environmental stressors and the preliminary development of a conceptual
model for thinking about how and under what conditions noise, crowding,

and other environmental stressors might adversely impact human wellbeing.

Conceptual Issues
A central deficiency has been an inattention to the role of underlying
psychophysiological processes or social resources in the environmental
stressor-disease link. In searching for answers to the question, why does
crowding or noise cause disease?, there are very little data that has tested
mechanisms like elevated cardiovascular functioning or diminished selfefficacy. What the data generally show, as depicted in Fig. 20.2, is a broad
set of outcome measures independently assessed.
More studies are needed that simultaneously investigate physical or
psychological health outcomes and one or more underlying processes in the
same sample of individuals. For example, Evans and Lepore (Evans et al.,
1989; Evans & Lepore, 1993a; Lepore et al., 1991) showed evidence for the
model shown in Fig. 20.3 -namely, that high residential density causes
deterioration in social support resources, which in turn accounts for the
linkage between density and psychological ill health.
There are an unbelievably large number of studies of noise and
cardiovascular functioning (principally blood pressure) that have not also
looked at some disease endpoint. Similarly, no studies have examined
crowding, immune function, and physical morbidity.
Several psychophysiologic mechanisms are prime candidates for more indepth scrutiny as intervening processes that could link environmental
conditions to ill health. Alterations in neuroendocrine functioning affect
cardiovascular activity, primarily via adrenomedullary action as well as alter
immune functioning via adrenocortical pathways (Baum & Grunberg, 1995).
Cardiovascular reactivity is another process warranting analysis. Two viable,
competing hypotheses exist. Sustained, chronic exposure to uncontrollable,
environmental stressors like crowding or noise may deplete the organism's
ability to respond adequately to challenge with cardiovascular mobilization
(Dienstbier, 1989). Alternatively, heightened sensitivity and vigilance from
chronic stressor exposure might exacerbate reactivity (Krantz & Manuck,

1984).
Learned helplessness and other motivational processes related to chronic
environmental stressor exposure have not been adequately developed. It
seems clear that one of the potentially most injurious aspects of chronic
environmental stressors is their intractability. Several aspects of motivation


and chronic environmental stress warrant additional research. The role of
attributional processes, which is well documented in the helplessness
literature, has not been applied to environmental stress research. It is clear
that attributional processes are salient to environmental stressors like noise
and crowding. Noise annoyance is strongly affected by 'attributions about
the origins of noise stimuli, as well as their perceived health impacts
(Koelega, 1987). Feelings of arousal induced by personal space invasions
(Worchel & Teddlie, 1976), expectancies (Schmidt & Keating, 1979), or
informational cues (Langer & Saegert, 1977; Paulus & Matthews, 1980) can
all be attributed to crowding or other environmental conditions with varying
consequences. The potential interplay among environmental stressors and
uncontrollability, helplessness, and negative health outcomes (such as
depression) is an area ripe for further study. Motivation or effort to maintain
task performance or productivity under suboptimal conditions may be a
salient factor, as well, in determining the long-term health consequences of
chronic exposure to adverse environmental conditions. Several noise studies
both in the laboratory and the field, as well as one crowding study indicate
that task performance can be sustained under adverse conditions but at a
“cost” of psychophysiological activation. The longterm health consequences
of people expending additional effort to do their job when the environment is
not optimal is an important and unresearched topic.
Studies of underlying psychosocial processes, such as social support or
control, also raise provocative conceptual issues about environment, stress,

and coping. Social support and control have each traditionally been
conceptualized as exogenous factors that moderate stressor-outcome
relations. As can be seen herein, however, chronic exposure to crowding or
to noise directly effects social support and control processes, respectively.
These psychosocial processes mediate rather than moderate the impacts of
these chronic environmental demands. Other chronic stressors may have
similar effects on coping resources.
In considering hypothetical mechanisms, it is also prudent to carefully
scrutinize the traditional practice of statistically controlling risk factors in
environmental epidemiology. For example, several noise and coronary heart
disease investigations control for smoking levels. However, what if noise
exposure increases smoking as a coping device, as suggested by some
studies already reviewed? By statistically partialling out a “risk” factor, a
psychologically relevant process that may underlie the noise-health link has
been eliminated.
At a more abstract level, the construct validity implications of statistical
controls or the practice of random assignment
-376FIG. 20.2
FIG. 20.3
in experimental studies of stress and health should be carefully considered.
By removing environmental stressors from their natural context (i.e.,
poverty, other suboptimal environmental factors) for the purposes of study,
ecological validity of the stressor-health relation may be distorted. Perhaps
crowding and poverty together or noise and certain job requirements
together, respectively, lead to pathology. By isolating one independent


variable either through statistical or experimental design means for
purposes of causal modeling, the actual incidence of adverse outcomes from
suboptimal environmental conditions may be dramatically underestimated

(Lepore & Evans, 1996).
It might also be valuable to conceptualize the physical environment not only
as a source of stress but also as a source of coping resources (Becker, 1990).
Research on coping, like stress, tends to overly focus on intrapsychic
mechanisms, missing the potential role of the social and physical
environment to promote or interfere with health. For example, research on
crowding suggests that floorplan layouts that incorporate greater
intervening, hierarchically arranged
-377spaces, buffer the negative effects of residential crowding on psychological
distress (Evans, Lepore, & Schroeder, 1996). Similarly, children in crowded
homes who have a place where they can spend some time by themselves
appear to suffer fewer negative outcomes (Wachs & Gruen, 1982).
Although children exposed to multiple risk factors are more likely to suffer
adverse physical and mental health outcomes, some children are more
resilient than others (Rutter, 1987). Bradley and colleagues (1994) found
that low residential density was a significant, independent, protective factor
among poor, low birth weight babies tested 1 and 3 years later on a wide
array of physical and psychological health measures.

Methodological Issues
In addition to some of the conceptual issues associated with statistically
controlling for risk factors, statistical approaches that partial out variables in
order to “control” for possible confounding effects are also fraught with
analytic problems. Statistical models that partial or covary out variables are
based on the assumption of no interaction between the independent
variable and the potential control variable on the outcome. Controlling for
social class, for example, presupposes that noise or crowding do not interact
with social class to affect health or well-being. The same statistical problem
may occur with controls for certain risk factors, such as family history of
coronary heart disease, hypertension status, or age, to name some common

examples. Utilization of analysis of covariance or its regression equivalents
assumes that the slopes of the respective regression lines between the
outcome variable and the independent variable and the covariate (partial b)
are parallel (i.e., no statistical interaction). Researchers should not employ
covariance or analogous regression Iprocedures to control for risk factors or
contextual factors, such as socioeconomic status, without first assessing this
basic statistical assumption.
Another analytic issue concerns effect size considerations. When the
correlation coefficient between noise exposure and blood pressure, for
example, is squared, not a lot of variance is explained. But this is also true if
the same is done for cigarette smoking and lung cancer. It is also true that
the variance explained in mental health by crowding is on the same order of
magnitude as the variance explained by income (Evans et al., 1998; Gove &
Hughes, 1983). Critics of the apparently small role of the physical
environment in health need to grapple with this issue more in comparative,


rather than in absolute, terms.
There is critical need for prospective, longitudinal designs in the field. There
is only one crowding study incorporating such a design (Lepore et al., 1991)
and just a handful of industrial studies of noise and health that incorporate a
longitudinal component. Self-selection into noisy or crowded environments,
as well as possible spuriousness, loom as major threats to internal validity in
most of the field studies reviewed herein. Too many cross-sectional field
studies exist. Furthermore, not enough integrated research programs have
examined the same environmental stressor and health in the lab and in the
field. The value of integrating lab and field work is illustrated by Cohen and
colleagues' work on aircraft noise and children where laboratory-based
concepts and measures were brought to bear on the study of chronic,
community noise exposure (Cohen et al., 1986).

Caution is needed in generalizing from aggregate level, epidemiological
studies to individual health responses to the physical environment. Several
examples of this ecological fallacy were previously reviewed, particularly in
crowding field studies, where people per room as indexed by census tract
did not yield the same pattern of results as when individual health measures
were assessed. Parallel trends were noted in the noise and health literature
(Lercher, 1996). One reason aggregate-level comparisons can be misleading
is related to exposure estimation. Large degrees of variance in exposure are
truncated into a single estimate of exposure when aggregate level data are
examined. Furthermore, the actual environment as experienced is even
further removed from the exposure metric in comparison to individual
residential or work environment assessments.
The problems of exposure estimation and adequate representation of
physical stressors in studies are common in the environmental stress
literature. Many studies have only gross estimates of actual exposure to the
physical stressor. Crowding and noise are typically estimated indirectly and
do not account for individual movement throughout the day across settings.
One indication of the importance of this issue is found in the noise literature
where several studies on industrial noise show that duration of exposure is a
critical variable. Similarly, utilization of hearing protection affects noises and
health findings in industrial settings. Residential room location can also
impact noise exposure (Lercher, 1996). An interesting example of the
importance of exposure estimation comes from a recent study by Maxwell
(1996) of crowding in preschoolchildren. Children in more crowded day-care
centers had greater behavioral and emotional problems only if they also
lived in crowded homes.
Moreover, the range of environmental variables in many studies is often
truncated and/or the distribution of environmental exposures is badly
skewed. Both of these problems strain the general linear model that forms
the underlying statistical basis employed in'most studies of the physical

environment, stress, and health. For example, many studies of crowding
have hardly any people in homes with more than 1.5 persons per room. Most
laboratory studies expose people to quiet or noise and several community
noise studies transform continuous data into a noise/quiet dichotomy. Badly
skewed data, as well as use of dichotomous categories, reduce statistical
power.
Furthermore, there is some indication of threshold effects for noise and
crowding health effects. Recall for example some recent evidence that traffic
noise above 65 dBA Leq appears necessary before cardiovascular risk
elevates. This nonlinearity also adversely affects statistical power. Studies of


traffic congestion and health outcomes indicate that log transformations
(Halpern, 1995), or use of indices such as percentage of time at high
congestion levels (Evans & Carrere, 1991), predict outcomes significantly
better than do mean levels of exposure. This nonlinearity can also appear at
the opposite end of the environmental exposure spectrum. Living alone is
associated
-378with psychological impairment as well as low social support (Gabe &
Williams, 1987; Galle et al., 1972). Crowding studies that calculate persons
per room as the density' metric that include people living alone distory the
“stirna” of association between crowding and health outcomes.
Outcome measures are also wanting. Several studies of psychological health
employed one item or scales of unknown psychometric standardized,
sensitive indicators. Immune function would be a particularly valuable
adjunct to environmental morbidity studies. Several studies of blood
pressure incorporated one or two readings, often taken in a medical setting.
Such data are unreliable and of questionable validity.
Health psychology has demonstrated that individual, biological, and
personological characteristics are central to understanding health and

disease. A smaller body of work within health psychology has also examined
the potential role of sociocultural factors in human health. Hopefully, this
chapter has directed attention to the potential direct, indirect, and
interactive roles the physical environment can play in health and human
behavior.

ACKNOWLEDGMENTS

Preparation of this chapter was partially supported by grants from the
National Institutes of Health (1 ROl I-IL4732501) the National Science
Foundation (BNS-8920483) and the U.S. Department of Agriculture (HatchNo 327407). I am grateful to Jana Cooperman and Tamir Ebbin for
bibliographic assistance. I thank Steve Lepore, Paul Paulus, and Shirley
Thompson for critical feedback on earlier drafts.