Mental Health, Mental Illness,

and Psychological Distress:

Same Continuum or Distinct Phenomena?*
ANDREW R. PAYTON
University of North Carolina at Chapel Hill

Journal of Health and Social Behavior 2009, Vol 50 (June):213–227

In this article, I argue that the relationships among mental health, disorder, and
distress are a key source of conflict in the sociology of mental health and that
the features of the conflict have the potential to call into question much of the
accumulated scientific knowledge on mental health. To address this issue, I at-
tempt to empirically assess three competing frameworks regarding these rela-
tionships: (1) the “modal perspective,” (2) the “Mirowsky and Ross perspec-
tive,” and (3) the “positive psychology perspective.” Results, however, support a
“discontinuous perspective:” no underlying continuum among any of the three
concepts. These results suggest that researchers need to avoid the common prac-
tice of “lumping together” distress, disorder, and mental health and study each
in their own right. Subsequent tests attempt to further specify the relationships
among these concepts. Results indicate a strong positive directional association
from distress to disorder, a strong negative directional association from distress
to mental health, and no significant relationship between mental health and dis-
order. These results are used to generate a number of directions for future re-
search.

In a 2002 special issue of the Journal of
Health and Social Behavior entitled “Selecting
Outcomes for the Sociology of Mental Health:
Issues of Measurement and Dimensionality” a
number of researchers wrestled with the rela-
tionships among distress, disorder, and mental
health. Aneshensel (2002), in a commentary on
the articles, referred to these relationships as
“several recurrent and unresolved epistemo-
logical issues that are especially consequential
for sociologists” (p. 236). Indeed, without a
clear conceptualization of the relationships
among these concepts researchers in the soci-
ology of mental health must question the very
* I would like to express my most profound appreci-
ation to Peggy Thoits, Ken Bollen, and Andy Perrin.
I also wish to thank the Editor and the anonymous
reviewers at JHSB. All shortcomings are of my own
doing. Please address correspondence to Andrew
Payton, University of North Carolina at Chapel Hill,
Department of Sociology, CB#3210, Hamilton Hall,
Chapel Hill, NC 27599-3210 (e-mail: arpayton
@unc.edu).
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
213
nature of what it is that they are studying. It dri-
ves to the very heart of the enterprise for how
we can make claims to understand the sociolo-
gy of mental health and to be accumulating
knowledge regarding mental health if we oper-
ationalize our dependent variable on the basis
of multiple, opaque concepts that contain con-
tradictory assumptions about the nature of one
another.
In this issue, Mirowsky and Ross (2002) re-
iterated their long-standing framework that
links psychological distress with mental health
and that views disorder as a conceptually dis-
tinct and socially constructed phenomenon.
Additionally, Keyes (2002) offered evidence to
support the positive psychology movement,
which seeks to confront the tendency toward
treating mental health and mental disorder as a
continuous phenomenon. Aneshensel’s review
of the articles in this special issue suggested
that the important insights of Mirowsky and
Ross and Keyes offered compelling arguments
in favor of pulling apart mental disorder and
214
mental health, but in no way provided a defin-
itive assessment of the relationships among
distress, disorder, and mental health. This de-
bate resurfaced in a recent issue (late 2007) of
Health: An Interdisciplinary Journal for the
Social Study of Health, Illness, and Medicine,
with contributions from some of the same, and
other, researchers. While a large body of re-
search has accumulated on mental health, dis-
order, and distress, the relationships among
these phenomena remain contested and, be-
cause they carry assumptions about one anoth-
er, implicitly challenge the accumulated re-
search.
The goal of this article is to move beyond the
present impasse surrounding the dependent
variable(s) of interest to sociologists of health
and illness. It seeks to do so by turning the key
features of the debate into a series of testable
hypotheses subject to empirical analysis. First,
I review the literature on distress, disorder, and
mental health and distill alternative under-
standings of the relationships among these
concepts into three competing perspectives. I
label these perspectives (1) the “modal per-
spective,” which treats mental health and men-
tal disorder as a single continuum and either
ignores distress or relegates it to some sub-
sidiary status; (2) the “Mirowsky and Ross per-
spective,” which treats distress and mental
health as a single continuum, and disorder as
artificial thresholds imposed along the contin-
uum; and (3) the “positive psychology per-
spective,” which argues explicitly against the
modal perspective, instead viewing disorder
and mental health as distinct phenomena. The
positive psychology perspective almost wholly
ignores distress. I test these three perspectives
using confirmatory factor analysis and then
turn to structural equation modeling (SEM) in
order to further assess the relationships among
distress, disorder, and mental health. The fol-
lowing section offers a brief sketch of the de-
bate organized around these competing per-
spectives.
LITERATURE REVIEW
The modal perspective is less a well-articu-
lated position supported by a substantial body
of empirical research, but more a prevailing
norm representing the long-standing tendency
for researchers to conflate mental health and
mental illness. From this perspective, disorder
(or mental illness) is the “master” or “umbrel-
la” concept. Mental health is a state character-
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
ized only by the relative presence or absence of
disorder, as defined by the Diagnostic and
Statistical Manual of Mental Disorders (DSM)
of the American Psychiatric Association. This
perspective is perhaps best signified by the fact
that the American Sociological Association
section entitled the “Sociology of Mental
Health” has very little research explicitly fo-
cused on mental health. While labeled “mental
health” research, the actual substance is almost
exclusively focused on disorders: depression,
anxiety, alcohol and drug abuse, comorbidity,
etc. (Keyes 2002). To be fair, the terms mental
health and mental illness naturally lead to the
assumption that they exist on a single continu-
um. Within the modal perspective distress is ei-
ther ignored or relegated to some subsidiary
status.
The positive psychology perspective is an
explicit alternative to the modal perspective. It
is both not new and emergent (Ryff 1989a). It
argues that the absence of disorder is a neces-
sary but insufficient condition for mental
health and thus seeks to study mental health in
its own right, as a separate yet related continu-
um to disorder. As such, it maintains that re-
search stemming from the modal perspective
tells us very little about mental health. One of
the leading approaches within the positive psy-
chology perspective conceptualizes mental
health as consisting of six key dimensions:
self-acceptance, positive relations with others,
personal growth, purpose in life, environmen-
tal mastery, and autonomy (Ryff 1989a,
1989b). These dimensions represent liking
oneself, having warm and trusting relation-
ships with others, having goals of personal de-
velopment, having direction in life, having the
ability to control and satisfy one’s needs, and
having a sense of self-determination (Ryff
1989b).
Marie Jahoda (1958) was perhaps the first
scholar to articulate this perspective when she
argued that the idea of “normality” is “unspe-
cific and bare of psychological content”
(1958:22). I trace the empirical origins of the
modern perspective to the classic finding that
positive affect (e.g., good spirits) and negative
affect (e.g., hopelessness) are weakly correlat-
ed (Bradburn 1969). If we assume that positive
affect indicates mental health and negative af-
fect indicates disorder then this finding repre-
sents early evidence in favor of separating
mental health and disorder into two distinct
continua.
RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS
The bulk of the evidence is more recent.
Research suggests mental health, operational-
ized using the “quality of life inventory,” cor-
relates approximately –.40 to –.50 with several
disorder items, suggesting disorder and mental
health are distinct concepts (Frisch et al. 1992).
Other research has found that each of the six
dimensions of mental health listed above cor-
relate, on average, approximately –.51 with the
Zung depression inventory and approximately
–.55 with the Center for Epidemiologic Studies
Depression Scale (Ryff and Keyes 1995). From
the positive psychology perspective such cor-
relations suggest that the concepts are related
yet distinct. Others (Dohrenwend et al. 1980)
argue that measurement error suppresses the
“true” correlation and interpret such evidence
as support for the modal perspective. Rather
than assume that measurement error accounts
for roughly half the “missing” correlation, I
take this as an empirical question to address in
my subsequent analyses.
The single most compelling piece of evi-
dence in support of the positive psychology
perspective is from a recent article by Keyes
(2005). Keyes uses CFA to examine the rela-
tionship between a higher-order factor for dis-
order (comprised of major depressive episode
[MDE], generalized anxiety disorder [GAD],
panic attacks, and substance abuse) with a
higher-order factor for mental health (com-
prised of Ryff’s dimensions of mental health).
Results found that the best-fitting model treats
disorder and mental health as two axes rather
than a single axis, suggesting the two phenom-
ena are conceptually distinct. However, Keyes’
research, and the positive psychology perspec-
tive more generally neglect the larger issue of
how distress fits into the equation.
The Mirowsky and Ross perspective (1989,
2002) treats distress as the umbrella concept. It
defines psychological distress as the emotion-
al suffering experienced by individuals
(Mirowsky 2007). Distress is typically thought
to consist of two major forms of symptoms, de-
pression and anxiety, with each of these forms
consisting of two types of symptoms, mood
and malaise (Mirowsky and Ross [1989]
2003). Mood refers to feelings such as sadness
or angst while malaise refers to bodily states
such as restlessness and upset stomach.
From this perspective, disorder is a “view
taken of [suffering] by professionals and cor-
porations with services and drugs to sell”
(Mirowosky 2007:301), and viewing disorder
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
215
as “unusual distress” undermines efforts to-
ward understanding the underlying causal
processes of human suffering (Mirowsky
2007). Recent research based on newly devel-
oped scales for the assessment of psychologi-
cal distress (Kessler et al. 2002) is able to dis-
tinguish DSM-IV cases of serious mental ill-
ness from noncases with excellent precision
(Kessler et al. 2003). If the presence or absence
of disorder can be detected with the use of a
distress scale, then this leads to the conclusion
that disorder is nothing more than an artificial-
ly constructed cut-point of severe distress.
Such evidence suggests that all the research on
disorders is irrelevant and can be discarded
with little cause for concern.
Others argue that distress and disorder are
conceptually distinct phenomena. Using his-
torical developments in studies of distress and
disorder as evidence, Horwitz (2007a; Horwitz
and Wakefield 2007) argues that distress and
disorder have been conflated as a result of the
inability to distinguish between “appropriate”
and “inappropriate” responses to stressful situ-
ations. Distress and disorder are qualitatively
distinct phenomena hinging on the issue of
“expectable response” in the DSM definition
of disorder (American Psychiatric Association
2000). Following this logic, if symptoms are
short-lived and expected given the stressful sit-
uation, the individual is distressed. If symp-
toms are excessive or disproportionate given
the situation, the individual likely suffers from
a disorder. In short, distress deserves attention,
and disorder is more than a reified social con-
struction.
The Mirowsky and Ross perspective views
distress as the opposite of mental health.
Similar to the modal perspective, mental health
is simply an absence of distress. Drawing on
correlations between distress and mental health
in the –.50 range, Mirowsky and Ross ([1989]
2003) argue that these correlations support
treating distress and mental health as a single
continuum because the “true” correlation is
suppressed by measurement error. They esti-
mate that controlling for sources of error rais-
es the correlation to the –.70 range, suggesting
distress and mental health exist on a single
continuum and we need only study distress to
“capture” what we are after. Again, this is an
empirical question that I return to in my analy-
ses.
Taken together, these perspectives offer a va-
riety of competing conceptualizations of the
216
relationships among mental health, disorder,
and distress, which have plagued researchers
for some time, and it is this competition of per-
spectives that is at the heart of the sociology of
mental health section in the ASA. The positive
psychology perspective makes this last point
explicit, for if it is right in its view that mental
health is more than the mere absence of disor-
der it calls into question the lion’s share of re-
search in the sociology of mental health, which
claims to provide insight into mental health
while studying disorders. The goal of the pre-
sent research is to distill these competing
frameworks into testable hypotheses and bring
new empirical evidence to the table with the
hope of providing some insight into how re-
searchers ought to best conceptualize the rela-
tionships among mental health, disorder, and
distress. I turn now to these hypotheses.
HYPOTHESES
Three underlying hypotheses emerge regard-
ing the competing conceptualizations of the re-
lationships among mental health, disorder, and
distress. These hypotheses are as follows:
H1: Distress and Disorder exist on a single con-
tinuum.
The Mirowsky and Ross perspective views
disorder as socially constructed and argues that
it should be discarded while others argue that
distress and disorder are important in their own
right. The available evidence reveals mixed
support for these claims, and it is not all di-
rectly comparable, creating the first impasse.
Confirmation of this hypothesis would lend
support to the Mirowsky and Ross perspective,
which seeks to discard the separate measure-
ment of disorder, while disconfirmation would
refute their argument, instead suggesting that
they are different constructs operating inde-
pendently of one another.
H2: Distress and Mental Health exist on a sin-
gle continuum.
The Mirowsky and Ross perspective further
argues that distress and mental health exist on
a single continuum while the positive psychol-
ogy perspective seeks to study mental health in
its own right. The evidence that the Mirowsky
and Ross perspective draws upon provides ten-
tative evidence in favor of a distress-mental
health continuum, while research from the pos-
itive psychology perspective has neglected to
consider this possible dimension, instead fo-
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
cusing on the relationship between mental
health and disorder. This has created a second
impasse in the literature. Confirmation of this
hypothesis would lend support to the
Mirowsky and Ross perspective, while discon-
firmation of this hypothesis would support the
positive psychology perspective, suggesting
mental health is a distinct entity.
H3: Disorder and Mental Health exist on a sin-
gle continuum.
The final hypothesis is a test of the modal
versus positive psychology perspectives. While
previous research provides very compelling
evidence in favor of the positive psychology
perspective, the modal perspective remains the
dominant paradigm, and thus a third impasse
exists. One reason for this may be because
Keyes’s analysis included alcohol abuse, which
may function differently than the other three
disorders (see, for instance, Horwitz, White,
and Howell-White 1996 and Kessler, Olfson,
and Berglund 1998). While this is a speculative
claim, if correct, it might bias the results in fa-
vor of the positive psychology view that disor-
der and mental health are independent con-
structs. Since the norm is the modal perspec-
tive, the burden of proof must be placed on the
positive psychology perspective. As such, my
analyses exclude substance abuse from the op-
erationalization of disorder. Confirmation of
this hypothesis would lend support to the
modal perspective, while disconfirmation
would support the positive psychology per-
spective.
The three hypotheses offer a variety of com-
peting outcomes. Confirmation of all three hy-
potheses would support the Mirowsky and
Ross perspective, suggesting that all three con-
cepts tap into a single underlying construct,
and that work stemming from the modal and
positive psychology perspectives is ultimately
redundant, since there is no value added to
measuring mental health and/or disorder in ad-
dition to distress. Disconfirmation of all three
hypotheses lends support to the positive psy-
chology perspective as well as a more general
“discontinuous perspective” suggesting that
disorder, distress, and mental health are dis-
tinct phenomena, each deserving substantive
attention in its own right. Confirmation of H2
alone would provide partial evidence in favor
of the Mirowsky and Ross perspective, show-
ing that distress and mental health exist on a
single continuum, but that would refute the
RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS
claim that disorder can be folded into distress.
Confirmation of H3 alone would support the
modal perspective by suggesting, first, that the
positive psychology perspective is redundant
and, second, that the Mirowsky and Ross per-
spective’s exclusive focus on distress is less
than comprehensive.
The implications of these hypotheses tests
vary considerably as well. On one end of the
spectrum, we need only study one of the three
constructs because each serves as a proxy for
the others. For instance, research on psycho-
logical distress is equally research on mental
health and disorder. On the other end of the
spectrum, any attempt to synthesize these con-
structs is reductionistic and neglects important
variation. Research on disorders, for instance,
should not be taken to reveal important insights
into mental health and thus the overwhelming
majority of research in the sociology of mental
health in fact tells us very little about mental
health. In the following sections, I present sys-
tematic tests of these competing hypotheses.
METHODS
I test these hypotheses using confirmatory
factor analysis. The model I develop creates
three key latent variables: mental health, disor-
der, and distress. Mental health is constructed
following Ryff’s conceptualization of psycho-
logical well-being; disorder is constructed
based on three measures of disorder, MDE,
GAD, and panic attacks; and distress is a six-
item scale that consists of items generally
found in indices of distress.
Sample
I draw upon the Midlife Development in the
United States (MIDUS; Brim et al. 2003) sur-
vey. The data collection was funded by the
MacArthur Foundation. In 1994–1995 respon-
dents were drawn from a nationally representa-
tive random-digit-dial sample of noninstitu-
tionalized, English-speaking adults, aged
25–74, with oversampling for older persons
and men. Respondents participated in an initial
telephone interview and then a follow-up mail
questionnaire. The data set I use contains re-
sponses from the main survey of 4,242 respon-
dents without use of sample weights, as they do
not substantially impact the results. The re-
sponse rate was 70.0 percent for the telephone
interview, 86.8 percent for the completion of
the mail questionnaire among the telephone re-
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
217
spondents, and 60.8 percent for the overall re-
sponse rate (.700 ⫻ .868).
Measures
Disorder. In the MIDUS survey, the authors
provide a continuous indicator for MDE, GAD,
and panic attacks, which I use to create a sin-
gle second-order factor for disorder. These
measures are constructed based on screening
mechanisms meant to mirror DSM-IV assess-
ment criteria. In order to conserve space, I have
included the details of the screening mecha-
nisms in Appendix A.
For MDE, 576 respondents screened posi-
tively on feeling “sad, blue, or depressed” and
were asked seven questions about whether they
lost interest, felt tired, lost appetite, had trouble
falling asleep “every night” or “nearly every
night,” had trouble concentrating, felt worth-
less, and thought a lot about death. One-hun-
dred twenty-two respondents screened posi-
tively on “lost interest in most things” and were
asked six questions concerning fatigue, ap-
petite increase or decrease, trouble concentrat-
ing, feelings of worthlessness, and thinking
about death a lot. The questions are dichoto-
mous, where a yes response indicates a depres-
sive symptom; “yes” responses were summed
for each group. “Sad, blue, or depressed” feel-
ings range from 0 to 7, where 0 represents no
symptoms and 7 represents the most severe
symptoms. Feelings of having “lost interest in
most things” range from 0 to 6, where 0 repre-
sents no symptoms and 6 represents the most
severe symptoms. Major depressive episode is
then constructed as a continuous variable
based on the sum of yes responses to each of
the two dimensions and ranges from 0 to 7,
where 0 represents no symptoms and 7 repre-
sents the most severe symptoms.
Both dimensions of major depressive
episode have mean values that are quite low, as
does the mean MDE score (see Table 1). This
is due, in part, to the fact that a large number
of the respondents were screened out of the in-
strument and thus are scored as zero on the
symptom items. This may bias the symptom
count downward since some portion of the re-
spondents may have experienced some of the
remaining symptoms and thus not be “true” ze-
ros. While this may raise a slight cause for con-
cern, this approach is meant to correspond to
the DSM-IV assessment criteria for disorder.
For instance, following DSM-IV assessment
criteria (American Psychiatric Association
218
TABLE 1. Descriptive Statistics for Two Dimensions of Major Depressive Episode, Generalized
Anxiety Disorder, Panic Attacks, and Distress
Variable N Mean
Major Depressive Episode 4,242
sad, blue,depressed 4,242
lost interest 4,242
Generalized Anxiety Disorder 4,242
Panic Attacks 4,242
Distress 3,675
a
Created by combining the yes scores from the two dimensions in italics.
2000) a respondent is not diagnosed as having
experienced a major depressive episode unless
they experienced either (1) depressed mood or
(2) loss of interest or pleasure most of the day
and nearly every day during a two-week peri-
od, which mirrors the screening mechanism
employed in the MIDUS. In short, there may
be a “true” zeros problem, but the instrument
does not undercount “true” disorders following
the DSM-IV assessment criteria. This holds
true for all measures of disorder.
Respondents screening positively for GAD
were administered a series of 10 questions.
These items are restlessness, nervousness, irri-
tability, trouble falling asleep, trouble staying
asleep, inability to concentrate, trouble re-
membering things, low energy, tiring easily,
and sore or aching muscles. Two measures—
“trouble remembering things” and “sore or
aching muscles” due to worry—do not fit eas-
ily into the DSM criteria. Additionally, the
DSM definition of GAD has a time frame of
six months, while these questions cover the
past 12 months. While these two points are
noteworthy, the overall measurement of GAD
corresponds quite closely to the DSM-IV as-
sessment criteria. Generalized anxiety disorder
was then constructed as a continuous variable
summing the yes responses for the 10 ques-
tions and coded so 10 represents the most se-
vere symptoms. The mean value of GAD is
quite low as well (see Table 1).
Respondents passing the screen for panic at-
tacks were administered six items concerning
feelings and sensations of heart pounding;
tightness, pain, or discomfort in the chest or
stomach; sweating; trembling or shaking; hot
flashes or chills; and whether things around
them seem unreal. “Yes” responses were
summed, and the descriptive statistics are
shown in Table 1.
Conceptually, disorder is the underlying
construct to test. In order to do so, MDE, GAD,
and panic attacks are used to create a single
second-order factor for disorder. The internal
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
SD Min. Max. Reliability
.84 2.00 0 7 N/Aa
.73 1.91 0 7 N/A
.11 .70 0 6 N/A
.19 1.05 0 10 .84
.36 1.09 0 6 .60
1.57 .64 1 6 .84
reliability for the latent construct disorder is
.84. Internal reliability is measured using the
equation
(⌺␭i)2 Var (␩) ,
(⌺␭i)2 Var (␩) + ⌺ Var (⑀i) + 2⌺⌺Cov (εiεj)
i<j
following Bollen (1980), where i,j = 1, 2, .|.|. ,
total number of indicators. I depart from the
standard use of Cronbach’s alpha because it
makes two assumptions that do not hold for my
analyses. First, Cronbach’s alpha assumes that
there are no correlated errors, but CFA models
correlated error terms. In the above equation
the last term of the denominator controls for
this assumption. Second, Cronbach’s alpha as-
sumes that all of the indicators of the underly-
ing construct equally contribute to it. While I
have no clear theoretical basis to suspect that
this assumption is violated, in practice it is
common. The factor loadings in the present
analysis demonstrate that this assumption is vi-
olated (see Figure 1).
Distress. Distress is constructed differently
from the three measures of disorder. The ques-
tions for distress were framed within the past
30 days and asked how much of the time re-
spondents felt “so sad nothing could cheer you
up,” “hopeless,” “worthless,” “nervous,” “rest-
less or fidgety,” and “that everything was an ef-
fort.” Responses range from “all of the time”
(coded 1) to “none of the time” (5). Variables
were reverse-coded so that 5 represents nega-
tive symptoms “all of the time” and 1 “none of
the time.” Based on Mirowsky and Ross’s view
that distress contains two types of symptoms,
mood and malaise, I intercorrelated the first
three items as questions concerning mood, and
the second three items as questions concerning
malaise to model hypothesized covariation.
The final sample size is 3,675 cases, with a
mean score of 1.57 (see Table 1). The smaller
sample size reflects the fact that questions con-
cerning disorder were asked in the initial tele-
 FIGURE 1. Path Diagram of the Relationships among Mental Health, Disorder, and Distress RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS

Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015

Note: SBD = “sad, blue, or depressed,” Int = “lost interest,” MDE = Major Depressive Episode, GAD = Generalized Anxiety Disorder, PAN = Panic Attacks, MH = Mental Health, MF =
219

Method Factor, AU = Autonomy, EM = Environmental Mastery, PG = Personal Growth, PR = Positive Relations, PU = Purpose in Life, and SA = Self-Acceptance.
220
phone survey, while the questions concerning
distress and mental health were asked in the
follow-up mail questionnaire.
It is important to note that these questions
are quite similar to many of the questions used
for disorder, which may raise concerns of op-
erational confounding. However, this concern
is attenuated by the fact that the questions are
not identical, that they involve very different
time frames, and that they do not overlap en-
tirely with any one dimension of disorder. In
the results I report on a test that created a high-
er-order factor composed of distress and disor-
der in order to explicitly model this possibility.
The results provide little evidence to support
such a model.
Mental health. Mental health consists of six
latent variables corresponding to Ryff’s six di-
mensions of psychological well-being, and one
second-order factor that synthesizes these six
dimensions. Each of the six dimensions is mea-
sured with three items giving a total of 18 in-
dicators (see Appendix A for the wording of all
18 items). Response categories range from
“strongly disagree” (coded 1) to “strongly
agree” (7). Each dimension ranges from 3 to 21
(the sum of three items that range from 1 to 7
each). These six dimensions are then used to
create a single second-order factor of mental
health (see Figure 1).
In the descriptive statistics presented below
(Table 2), all negatively worded items have
been reverse-coded so that 1 corresponds to
poor mental health while 7 corresponds to ex-
cellent mental health. Positively worded items
receive higher scores, all with a mean above
5.0, and most above 6.0, while negatively
worded items are slightly lower, with the low-
est mean (“demands get me down”) equal to
4.23. The means for all six dimensions are
slightly above 16.0, except for personal
growth, which has a mean of almost 18.0 (see
Table 2). The internal reliability for mental
health is .895.
Analysis
Based on previous research (Springer and
Hauser 2006), I introduce two effects for the
questionnaire design; however, I model the ef-
fects in a slightly different fashion. First, I
model a method factor to control for the lower
correlations among negatively-worded items
(labeled “MF” in the path diagram, Figure 1).
This approach is much more common in the
CFA tradition than intercorrelating negatively-
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
worded items, which was the approach of the
previous research.
Although the items for a given dimension
are generally distributed throughout the psy-
chological well-being instrument, there are
some inconsistencies. Thus, the second adjust-
ment is for a question ordering effect, which is
represented by the straight arrow from one
item to the next in the 18 items used to con-
struct psychological well-being. The previous
approach (Springer and Hauser 2006) correlat-
ed a given item with the next item in the ques-
tionnaire; however, I believe my approach to be
more intuitively appealing because it can be in-
terpreted, quite literally, as saying a response to
a given question has an effect on the response
to the subsequent question (Bollen and
Medrano 1998). Measures of overall model fit
suggested that the inclusion of these adjust-
ments improved model fit (data available upon
request).
One final note is in order before I proceed.
The various latent measures are based on ob-
served indicators that employ different time
frames. The measures of disorder are based on
a 12-month time frame, the measures of dis-
tress are based on a 30-day time frame, and the
measures of mental health are based on a sub-
jective time frame. This is an artifact of the pre-
vailing measurement norms for each of the
three concepts. At this time, no theoretical or
empirical basis leads me to expect bias in a par-
ticular direction due to this measurement arti-
fact. If there are time-based measurement ef-
fects, then it would be most suitably addressed
using longitudinal analyses, which is beyond
the scope of the present work. On the positive
side, these differential time frames may help
further avoid issues of operational confound-
ing.
I use the data analysis program Mplus
(Muthén and Muthén 1998) to estimate maxi-
mum likelihood for confirmatory factor analy-
ses. I rely on CFA for a number of reasons.
First, I intend to test specific hypotheses re-
garding the relationships among mental health,
disorder, and distress, which lends itself to
CFA rather than a “brute force” exploratory
factor-analytic approach. Further, these con-
cepts are measured with a number of different
observed variables and modeled as latent fac-
tors. Using CFA it is possible to construct sev-
eral latent factors and to test the relationships
among multiple factors simultaneously. Last,
CFA controls for measurement error by corre-
RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS 221
TABLE 2. Descriptive Statistics for 18 Indicators and Six Dimensions of Psychological Well-being
Variable N Mean SD Min. Max. Reliability
Self-Acceptance 3,663 16.47 3.5 3 21 .782
—Like personality 3,647 6.19 0.94 1 7 N/A
—Pleased with life 3,646 5.71 1.45 1 7 N/A
—Disappointed with lifea 3,637 4.57 2.1 1 7 N/A

Purpose in Life 3,665 16.38 3.67 3 21 .39

—Not wander aimlessly 3,624 5.83 1.56 1 7 N/A
—Don’t look aheada 3,642 4.88 2.08 1 7 N/A
—Nothing left to d a 3,634 5.69 1.8 1 7 N/A

Environmental Mastery 3,666 16.04 3.44 3 21 .766

—Demands get me downa 3,627 4.23 1.98 1 7 N/A
—In charge 3,636 5.72 1.51 1 7 N/A
—Good at managing life 3,650 6.09 1.18 1 7 N/A

Positive Relations 3,666 16.05 4.12 3 21 .57

—No close relationshipsa 3,645 4.78 2.17 1 7 N/A
—Giving person 3,654 6.08 1.13 1 7 N/A
—No warm relationshipsa 3,647 5.19 2.08 1 7 N/A

Personal Growth 3,663 17.85 3.17 3 21 .627

—Life about growth and learning 3,658 6.24 1.15 1 7 N/A
—New experiences good 3,659 6.19 1.12 1 7 N/A
—Gave up trying tochangea 3,650 5.42 1.93 1 7 N/A

Autonomy 3,655 16.51 3.32 3 21 .655

—Easily influenceda 3,644 4.46 1.98 1 7 N/A
—Confidence in own opinion 3,645 6.02 1.3 1 7 N/A
—Judge self by own values 3,644 6.03 1.32 1 7 N/A
—a Negatively-worded item

lating error terms, modeling the question or-
dering effect, and modeling the question word-
ing effect. This is perhaps the most important
reason to use CFA because measurement error
is a key aspect of the evidence used to support
the Mirowsky and Ross perspective and to re-
fute the positive psychology perspective.
A related yet distinct argument concerns
whether disorder should be treated as continu-
ous or discrete in nature (Mirowsky and Ross
[1989] 2003; Horwitz 2002a; Horwitz 2002b).
The one downside of CFA and structural equa-
tion approaches is that they treat all latent vari-
ables as continuous, thus making an assump-
tion about the nature of disorder. From the
SEM perspective, this is a softer assumption
than it may first appear because even the most
categorical measures often obscure important
underlying variation. Gender is one example.
Quantitative researchers almost uniformly treat
gender as a categorical measure, which ignores
the underlying heterogeneity that is one of the
important insights of Gender and Queer
Theories stemming from the work of Foucault
(1990), Judith Butler (2006), and many others.
In relation to disorder, treating it as continuous
may be a rather soft assumption that simply
suggests some people have more severe disor-
ders or more severe symptoms of a disorder
than others. Expressions such as “borderline
depression” suggest there is a continuum with
a threshold. Finally, if there is bias in treating
disorder as continuous then it is in favor of the
Mirowsky and Ross perspective, which oppos-
es treating disorder as a discrete state.
Figure 1 presents the path diagram for the
test of the relationships among distress, disor-
der, and mental health, as well as the factor
loadings for the model and the correlations
among the three key constructs. Each of the
squares represents an observed variable, which
is then used to construct the latent variables un-
der question. Latent variables are signified by
circles. The present analysis concentrates on
the three circles labeled “MH,” “Disorder,” and
“Distress,” which correspond to the concepts
under question. The three curved two-headed
arrows between each of these circles represent
the three hypotheses presented above. For ex-
ample, the curved two-headed arrow between
MH and Disorder represents the hypothesis
that mental health and disorder exist on a sin-
gle continuum.

Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015

222
In CFA and SEM-based approaches a given
hypothesis is confirmed if the two variables be-
ing tested correlate to a near-perfect degree
(~1.00). At first glance this may strike the read-
er as an impossibly high standard, and this is
the logic behind the aforementioned measure-
ment error arguments (Dohrenwend et al.
1980; Mirowsky and Ross [1989] 2003).
However, the standards for a CFA correlation
matrix differ from the standards for a “normal”
correlation matrix because CFA models the re-
lationships among variables while controlling
for error. Thus, controlling for error, we should
expect what is left in the equation to be almost
perfectly correlated if two concepts exist on a
single continuum. Even if we assume “almost
perfectly correlated” to mean .90 or .80, well
below the threshold in CFA-based research
(Bollen and Grandjean 1981), the results pre-
sented below still hold.
RESULTS
In order to assess the goodness-of-fit for the
model, I rely on a number of fit measures,
since some offer desirable properties that oth-
ers do not (Bollen 1989; Bollen and Long
1993). The Comparative Fit Index (CFI) and
the Tucker-Lewis Index (TLI; Tucker and
Lewis 1973) range from 0 to 1.00, where 1.00
is considered ideal fit. Values above .900 are
considered acceptable, while values below this
threshold would suggest that the model poorly
specifies the relationships among the variables
of interest and would make all results not sig-
nificant. The Root Mean-Squared Error of
Approximation (RMSEA) has a lower bound
of .00 and no upper bound. Lower values sug-
gest better model fit and .00 is perfect model
fit. The rule of thumb varies, but the most con-
servative threshold for the RMSEA is .05. The
final fit measure I use is the Bayesian
Information Criterion (BIC), which I calculate
following Raftery (1995) with the equation Tm
– df * ln(n), where Tm is the chi-square test sta-
tistic, df is degrees of freedom, and n is the
sample size. For this measure, acceptable mod-
el fit is BIC < .00. Values above .00 would
mean the model is poorly specified.
All fit indices give a consistent picture of the
relative fit of the model. The value for the CFI
is .957, and the value for the TLI is .947, both
above the .900 threshold. The RMSEA test sta-
tistic for my model is .033, below the .05
threshold. Finally, the BIC for my model is
–754.87, which is well below the .00 threshold.
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
In sum, the hypothesized model performs well
for all measures of fit, which suggests that the
model is well specified and thus allows me to
assess the relationships among distress, disor-
der, and mental health. The curved two-headed
arrows in Figure 1 present the correlations
among these three variables, which are dis-
cussed below.
Given the goal of testing the competing con-
ceptualizations of the relationships among
mental health, disorder, and distress, I focus
my attention exclusively on these results. I do
not take the time to discuss the details of the
rest of the model because it tells us very little
about the hypothesis tests and would take con-
siderable space to do so. For the purposes of
the present analysis, the measures of overall
model fit suggest that the modeling is accept-
able, and thus further attention to the minutiae
of the model, while potentially interesting, are
unwarranted.
Hypothesis H1—that distress and disorder
exist on a single continuum—is not supported.
The correlation between distress and disorder
is quite high, .771, for a “normal” correlation
matrix; however, this model controls for mea-
surement error, and thus the ~1.00 criterion is
not met. Due to the high correlation between
disorder and distress, I tested an alternative
model with disorder and distress as a single
higher-order factor. This model assumed that
disorder and distress were on a single continu-
um and assessed their relationship as a single
continuum with mental health as a separate en-
tity. If the alternative model performed better
than the main model it would have provided ev-
idence to support hypothesis H1; however, all
measures of model fit were worse than the
main model (CFI = .940, TLI = .922, RMSEA
= .040; BIC = –118.97). In short, conceptual-
izing disorder and distress as a single continu-
um is inappropriate.
Hypothesis H2—that distress and mental
health exist on a continuum—is not supported.
There is also a strong correlation, –.646, be-
tween distress and mental health, yet the corre-
lation is still well below the threshold. The re-
sults thus refute the view of the Mirowsky and
Ross perspective, which treats mental health
and distress as opposite ends of a single con-
tinuum. While the positive psychology per-
spective does not address the issue of distress,
it lends support to the general view that mental
health is more than the absence of negative
symptoms or suffering.
RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS
Hypothesis H3—that disorder and mental
health exist on a continuum—is also not sup-
ported. There is a moderate-to-strong negative
correlation of –.510 between mental health and
disorder. Thus, the modal perspective fails to
find support in its conceptualization of mental
health and disorder as opposite poles of a sin-
gle continuum. Instead, the positive psycholo-
gy perspective seems to be accurate in its view
of mental health as more than the absence of
disorder. Interestingly, this is the weakest cor-
relation among any of the three hypotheses.
This suggests that the norm of assuming disor-
der and mental health as a single continuum in
the sociology of mental health is actually based
on the weakest grounds of any of the concep-
tualizations offered in the literature review.
Given that the results fail to support any of
the three hypotheses distilled from the litera-
ture, the analysis simultaneously supports two
perspectives. First, it provides support for the
positive psychology perspective because the
results indicate that mental health and disorder
do not exist on a single continuum. However,
the positive psychology perspective neglects
consideration of how mental health relates to
distress. Since the other two hypotheses were
not supported, a second, more general perspec-
tive emerges. I label this the “discontinuous
perspective”: There is no underlying continu-
um among any of the three concepts. From this
perspective, mental health, disorder, and dis-
tress may be correlated, yet each is a conceptu-
ally distinct phenomenon, and thus it would be
a mistake to assume an underlying continuum
among any two or all three of the concepts. In
sum, distress, disorder, and mental health
should not be conflated.
Nonetheless, evidence suggesting that men-
tal health, disorder, and distress are conceptu-
ally distinct phenomena does not suggest that
they operate entirely independent of one an-
other. In fact, the moderate-to-strong correla-
tions among the concepts suggest that the sub-
stantive relationships among the concepts are
likely more important than the mere fact that
they are conceptually distinct phenomena. It is
for this reason that I briefly turn to additional
analyses in order to provide some preliminary
insights and future directions regarding the
substance of the relationships among mental
health, disorder, and distress.
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
223
Moving Beyond
The analysis above suggests that mental
health, disorder, and distress are related yet dis-
tinct concepts but reveals nothing about the ac-
tual substance of how they relate to one anoth-
er. In order to begin to understand how they
operate in relation to one another, I use struc-
tural equation models to test several competing
possible relationships among these concepts.
SEMs move a step beyond CFA because they
replace the free association of mental health,
disorder, and distress with specific directional
associations. Rather than testing, for instance,
whether distress and disorder are correlated
with one another, I am able to test how they
tend to be correlated with one another. Put suc-
cinctly, it is a question of direction, not just of
correlation. While it is no replacement for
good longitudinal data, it is an excellent means
of exploratory analysis to identify future direc-
tions for research. Figure 2 presents the possi-
ble relationships among these concepts.
To conserve space and highlight the most
important aspects of each model, I have elimi-
nated all the measures used to construct men-
tal health, disorder, and distress. The structure
of each concept is identical to that presented in
Figure 1, with one difference: Figure 2 replaces
the curved, two-headed arrows of Figure 1,
which represented a test of the various contin-
ua hypotheses, with straight, single-headed ar-
rows, which represent tests of directional asso-
ciation. At first glance, Figure 2 seems to sug-
gest I am testing four hypotheses with three da-
ta points; however, this is an artifact of my at-
tempt to make the information more easily di-
gestible.
In order to better understand the information
such as that presented in Figure 2, Bollen
(1989) suggests a “mental experiment” for the
first row: These two models suggest that as one
varies mental health one should expect distress
and disorder to systematically vary with it, i.e.,
a one unit increase in mental health tends to be
associated with a one unit decrease in both dis-
tress and disorder. This claim is quite strong,
and I find little reason to support it, particular-
ly given the previous analysis which suggested
each construct must be relatively independent
of the others. The difference between the two
models on the first row is the relationship be-
tween distress and disorder. The model in the
first row and first column can be interpreted as
suggesting that changes in disorder tend to be
224
FIGURE 2. Hypothesized Relationships among Distress, Disorder, and Mental Health
associated with changes in distress, while the
alternative model hypothesizes the opposite.
The two models on the second row switch
the direction of the arrows between distress
and mental health and disorder and mental
health. These models suggest that distress and
disorder tend to be associated with changes in
mental health, but the changes need not vary
systematically. These models also test the di-
rection of the association between distress and
disorder.
The BIC for each of these models suggests
good overall fit; however, using the BIC allows
for adjudication between competing models in
a straightforward manner: the lower the BIC,
the better the model (Raftery 1995). Of the
four models tested, the model in the second
row and second column performs the best. The
BIC difference between it and the next best
model is 53.37. A difference greater than 10 is
very strong evidence in favor of a particular
model (Raftery 1995). Other goodness-of-fit
measures give a similar portrait (data available
upon request).
The parameter estimates for the best-fitting
model (see Figure 2) reveal interesting rela-
tionships among mental health, disorder, and
distress that are more specific than those found
in the CFA-based analysis above. There is a
strong directional association of .67 from dis-
tress to disorder. This suggests that changes in
distress tend to be positively associated with
changes in disorder. There is also a strong di-
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
rectional association of –.63 from distress to
mental health. This suggests changes in dis-
tress tend to be negatively associated with
changes in mental health. Finally, and perhaps
most interestingly, the parameter from disorder
to mental health is quite small, –.03, and this
effect is not significant (p = .537). This sug-
gests disorder has no clear directional associa-
tion with mental health. Counter to the positive
psychology perspective, which views the ab-
sence of disorder as a necessary but insuffi-
cient condition for mental health, this result in-
dicates that the absence of disorder may be nei-
ther a necessary nor sufficient condition for
mental health.
DISCUSSION
Using the MIDUS, these analyses focused
on a perennial debate in the sociology of men-
tal health: the relationships among mental
health, illness, and distress. The goal was to
distill the debate into testable hypotheses and
assess these with new empirical evidence in or-
der to move beyond several impasses in the ex-
isting literature. The results yielded several im-
portant findings.
First, results of the confirmatory factor
analysis supported a “discontinuous perspec-
tive” that views each concept as distinct rather
than treating any two or all of them as points
on a single, underlying continuum. These re-
sults demonstrate the necessity of studying
each of these phenomena in their own right
RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS
rather than assuming, for instance, that what
holds true for mental disorders must also hold
true for mental health or distress, as is present-
ly common practice. Furthermore, the fact that
there is no umbrella concept suggests that each
of these three concepts might be caused or pre-
dicted by different factors and may also cause
or predict different outcomes. Along this vein,
recent research (Keyes 2002, 2005) suggests
that those who are “languishing,” that is re-
spondents who have both a disorder and poor
mental health, are considerably more likely to
suffer serious health consequences than those
who report only one or the other.
Second, the present research attempted to
begin to understand the substance of the rela-
tionships among distress, disorder, and mental
health and move beyond the finding that there
is no unifying concept. While only preliminary,
results of the structural equations models sug-
gest that distress tends to be positively associ-
ated with disorder and poorer mental health,
and that disorder and mental health do not dis-
play any clear directional associations. The ab-
sence of a clear directional association between
mental health and disorder is particularly note-
worthy since it calls into question much of the
available research on disorders that purports to
be concerned with mental health. In order to
better grasp these implications, imagine a sce-
nario in which an individual that suffers from
occasional panic attacks is asked about his or
her mental health in general. The respondent
might say, “All things considered, my mental
health is quite good except when I experience
the occasional panic attack.” An even more
telling example might be that of an individual
who suffers from a mild case of chronic tic dis-
order but is able to cope quite well with it, and
thus the disorder has no appreciable effect on
her or his mental health. Given such possibili-
ties, it makes sense that the two are correlated
and, at the same time, that there is not a clear
directional association between disorder and
mental health because the two would only be
associated in isolated incidents.
The major task of future research will be to
move beyond the SEM analysis and begin to
better understand the dynamics lying “under-
neath” the fact that mental health, distress, and
disorder appear to be distinct entities. As
Horwitz (2007b) has suggested for the rela-
tionship between disorder and distress, the
SEM analysis presented here suggests that
contextual factors are likely to play a signifi-
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
225
cant role in the relationships among disorder,
distress, and mental health. Sociological
processes might intervene or, perhaps, specify
the relationships among these concepts. For in-
stance, exposure to different forms of stress
might differentially impact mental health, dis-
tress, and disorder. Equally plausible, classic
sociological concerns such as race, class, and
gender might be important in determining how
and when mental health, distress, and disorder
are related to one another. Researchers might
also explore differential vulnerability within
and among such groups.
A number of other possibilities exist as well.
One important component might be the ways
in which different individuals and groups make
sense of these processes. It is likely that how
the individual, her or his significant others,
professionals, and corporations view these
phenomena plays a part in how these phenom-
ena relate to one another in the actual experi-
ence of them. Additionally, the present work
neglected to consider physical health, which is
likely to play a role. Future research should al-
so take a longitudinal approach in order to bet-
ter understand how mental health, distress, and
disorder influence one another over time and in
different ways. Finally, it is worth taking seri-
ously the possibility that much of our current
knowledge will need to be reformulated, since
so much of the existing literature in the sociol-
ogy of mental health has used distress or dis-
order as a proxy for the remaining concepts.
CONCLUSION
While none of these questions can be an-
swered within the scope of the present work,
the findings within this research signal a
wealth of potential directions for future re-
search. The present research simply attempts to
move beyond several key impasses in the pre-
sent literature. Results suggest that the tenden-
cy to conflate distress, disorder, and mental
health likely obscures important underlying
variation. It is clear from the above analyses
that the positive psychology perspective is cor-
rect in viewing mental health as not simply the
absence of disorder. It is equally clear that dis-
tress deserves attention in its own right.
Minimally, for now, any study claiming to fo-
cus on mental health and illness ought to in-
corporate measures of distress, disorder, and
mental health, since they appear to be associat-
ed yet independent constructs.
226 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
APPENDIX A.—Screening Mechanisms for Measures of Disorder
Screening Process for Major Depressive Episode:
First, respondents were asked whether they felt “sad, blue, or depressed” or “lost interest in most things” in the past 12
months for two weeks or more. If so, respondents were then asked whether these feelings lasted “all day long” or
“most of the day.” If so, respondents were then asked whether these feelings were “every day” or “almost every day.” If
so, respondents were then administered the corresponding items listed in the methods section.
Screening Process for Generalized Anxiety Disorder and Panic Attacks:
First, respondents were screened for worry “about half the days” to “every day” in the past 12 months. Next, respon-
dents were asked if they usually worry about “one particular thing” or “more than one thing,” and if they ever have dif-
ferent worries on their mind at the same time. If respondents answered “one particular thing” they were administered a
series of 10 questions related to GAD, which are listed in the methods section.
Respondents were screened for panic attacks if, during the above screening, they answered that they usually worry
about more than one thing and ever have different worries at the same time. They were then administered two ques-
tions: (1) “Did you ever have a spell or an attack when ALL OF A SUDDEN you felt frightened, anxious, or very un-
easy, in a situation when most people would not be afraid or anxious?” and (2) “Did you ever have a spell or attack
when for no reason your heart suddenly began to race, you felt faint, or you couldn’t catch your breath? When we say,
“for no reason,” we mean that it was NOT due to any physical cause, like a heart problem.” If respondents answered af-
firmatively to either question they were administered six questions, which are listed in the methods section.

APPENDIX B.—Ryff’s Psychological Well-Being Instrument

Environmental Mastery
The demands of everyday life often get me down.
In general, I feel I am in charge of the situation in which I live.
I am good at managing the responsibilities of daily life.
Autonomy
I tend to be influenced by people with strong opinions.
I have confidence in my own opinions, even if they are different from the way most other people think.
I judge myself by what I think is important, not by the values of what others think is important.
Positive Relations with Others
Maintaining close relationships has been difficult and frustrating for me.
People would describe me as a giving person, willing to share my time with others.
I have not experienced many warm and trusting relationships with others.
Self-Acceptance
I like most parts of my personality.
When I look at the story of my life, I am pleased with how things have turned out so far.
In many ways I feel disappointed about my achievements in life.
Personal Growth
For me, life has been a continuous process of learning, changing and growth.
I think it is important to have new experiences that challenge how I think about myself and the world.
I gave up trying to make big improvements in my life a long time ago.
Purpose in Life
Some people wander aimlessly through life, but I am not one of them.
I live life one day at a time and don’t really think about the future.
I sometimes feel as if I’ve done all there is to do in life.

REFERENCES cal Democracy.” American Sociological Review

American Psychiatric Association. 2000. Diagnos-
tic and Statistical Manual of Mental Disorders
IV—Text Revision. Washington, DC: American
Psychiatric Association.
Aneshensel, Carol D. 2002. “Answers and Ques-
tions in the Sociology of Mental Health.” Jour-
nal of Health and Social Behavior 43:236–46.
Bollen, Kenneth A. 1980. “Issues in the Compara-
tive Measurement of Political Democracy.”
American Sociological Review 45:370–90.
———. 1989. Structural Equations with Latent
Variables. New York: Wiley.
Bollen, Kenneth A. and Burke D. Grandjean. 1981.
“The Dimension(s) of Democracy: Further
Issues in the Measurement and Effects of Politi-
46:651–59.
Bollen, Kenneth A. and J. Scott Long. 1993. Testing
Structural Equation Models. Thousand Oaks,
CA: Sage Publications, Inc.
Bollen, Kenneth A. and Juan Díez Medrano. 1998.
“Who are the Spaniards? Nationalism and Iden-
tification in Spain.” Social Forces 77:587–621.
Bradburn, Norman M. 1969. The Structure of Psy-
chological Well-Being. Chicago, IL:Aldine.
Brim, Orville G., Paul B. Baltes, Larry L. Bumpass,
Paul D. Cleary, David L. Featherman,William R.
Hazzard, Ronald C. Kessler, Margie E. Lach-
man, Hazel Rose Markus, Michael G. Marmot,
Alice S. Rossi, Carol D. Ryff, and Richard A.
Shweder. 2003. National Survey of Midlife

Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015

RELATIONSHIPS AMONG HEALTH, DISORDER, AND DISTRESS
Development in the United States (MIDUS),
1995–1996 [MRDF]. 2nd ICPSR version. Ann
Arbor, MI: DataStat, Inc. Boston, MA: Harvard
Medical School, Dept. of Health Care Policy
[producers], 1996. Ann Arbor, MI: Inter-univer-
sity Consortium for Political and Social
Research [distributor].
Butler, Judith. 2006. Gender Trouble: Feminism and
the Subversion of Identity. New York: Routledge.
Dohrenwend, Bruce P., Patrick E. Shrout, Gladys
Egri, and Frederick S. Mendelsohn. 1980.”Non-
specific Psychological Distress and Other
Dimensions of Psychopathology.” Archives of
General Psychiatry 37:1229–36.
Foucault, Michel. 1990. The History of Sexuality:
An Introduction. New York: Vintage.
Frisch, Michael B., J. Cornell, M. Villanueva, and P.
J. Retzlaff. 1992. “Clinical Validation of the
Quality of Life Inventory: A Measure of Life
Satisfaction for Use in Treatment Planning and
Outcome Assessment.” Psychological Assess-
ment 4:92–101.
Horwitz, Allan V. 2002a. Creating Mental Illness.
Chicago, IL: University of Chicago Press.
———. 2002b. “Outcomes in the Sociology of
Mental Health and Illness: Where Have We Been
and Where Are We Going?” Journal of Health
and Social Behavior 43:143–51.
———. 2007a. “Distinguishing Distress from
Disorder as Psychological Outcomes of Stressful
Social Arrangements.” Health: An Interdiscipli-
nary Journal for the Social Study of Health, Ill-
ness, and Medicine 11:273–89.
———. 2007b. “Transforming normality into
pathology: the DSM and the outcomes of stress-
ful social arrangements.” Journal of Health and
Social Behavior 48: 211–22.
Horwitz, Allan V. and Jerome C. Wakefield. 2007.
The Loss of Sadness: How Psychiatry Trans-
formed Normal Sadness into Depressive Disor-
der. New York: Oxford University Press.
Horwitz, Allan V., Helene White, and Sandra How-
ell-White. 1996. “The Use of Multiple Out-
comes in Stress Research: A Case Study of Gen-
der Differences in Responses to Marital Dissolu-
tion.” Journal of Health and Social Behavior
37:278–91.
Jahoda, Marie. 1958. Current Concepts of Positive
Mental Health. New York: Basic Books.
Kessler, Ronald C., G. Andrews, L. J. Colpe, E.
Hiripi, D. K. Mroczek, S. L. T. Normand, E. E.
Walters,and A. M. Zaslavsky. 2002. “Short
Screening Scales to Monitor Population Preva-
lences and Trends in Non-Specific Psychologi-
cal Distress.” Psychological Medicine 32:
959–76.
Andrew Payton is a doctoral candidate in the Department of Sociology at the University of North Carolina at
Chapel Hill. His primary research interests are the sociology of culture, medical sociology, and the sociology of
health and illness. His dissertation seeks to synthesize his interests in culture and health/illness by examining how
members of Alcoholics Anonymous engage in “cultural work” in order to maintain sobriety.
Downloaded from hsb.sagepub.com at NORTHERN KENTUCKY UNIV on June 19, 2015
227
Kessler, Ronald C., P. R. Barker, L. J. Colpe, J. F.
Epstein, J. C. Gfroerer, E. Hiripi, M. J. Howes, S.
L. T. Normand, R. W. Manderscheid, E. E. Wal-
ters, A. M. Zaslavsky. 2003. “Screening for Seri-
ous Mental Illness in the General Population.”
Archives of General Psychiatry 60(2):184–89.
Kessler, Ronald C., M. Olfson, and P. A. Berglund.
1998. “Patterns and Predictors of Treatment
Contact after First Onset of Psychiatric Disor-
ders.” American Journal of Psychiatry
155:62–69.
Keyes, Corey L. M. 2002. “The Mental Health Con-
tinuum: From Languishing to Flourishing in
Life.” Journal of Health and Social Behavior
43:207–22.
———. 2005. “Mental Illness and/or Mental
Health? Investing Axioms of the Complete State
Model of Health.” Journal of Consulting and
Clinical Psychology 73:539–48.
Mirowsky, John. 2007. “The Distribution’s Tail: A
Comment.” Health: An Interdisciplinary Journal
for the Social Study of Health, Illness, and Med-
icine 11:301–02.
Mirowsky, John and Catherine E. Ross. [1989]
2003. Social Causes of Psychological Distress.
New York: Aldine de Gruyter.
———. 2002. “Measurement for a Human Sci-
ence.” Journal of Health and Social Behavior
43:152–70.
Muthén, Linda K. and Bengt O. Muthén. 1998.
Mplus User’s Guide. Los Angeles, CA: Muthén
and Muthén.
Raftery, Adrian E. 1995. “Bayesian Model Selection
in Social Research.” Sociological Methodology
25:111–63.
Ryff, Carol D. 1989a. “Beyond Ponce de Leon and
Life-Satisfaction: New Directions in the Quest
of Successful Ageing.” International Journal of
Behavioral Development 12:35–55.
———. 1989b. “Happiness Is Everything, or Is It?
Explorations on the Meaning of Psychological
Well-Being.” Journal of Personality and Social
Psychology 57:1069–81.
Ryff, Carol D. and Corey L. M. Keyes. 1995. “The
Structure of Psychological Well-Being Revisit-
ed.” Journal of Personality and Social Psycholo-
gy 69:719–27.
Springer, Kristen W. and Robert M. Hauser. 2006.
“An Assessment of the Construct Validity of
Ryff ’s Scale of Psychological Well-Being:
Method, Mode, and Measurement Effects.”
Social Science Research 35:1080–1102.
Tucker, L.R. and C. Lewis. 1973. “A reliability coeffi-
cient for maximum likelihood factor analysis.”
Psychometrika 38:1–10.
376 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
ERRATUM
The following figure in the article, Andrew R. Payton, “Mental Health, Mental Illness, and
Psychological Distress: Same Continuum or Distinct Phenomena?” published in the June 2009
issue of JHSB had some errors. The correct figure appears below.

FIGURE 2. Hypothesized Relationships among Distress, Disorder, and Mental Health