Journal of Affective Disorders 70 (2002) 229–240
To what extent do oral contraceptives influence mood and affect?
Kirsten A. Oinonen, Dwight Mazmanian
Department of Psychology, Lakehead University, Thunder Bay, Ont., Canada P7B 5E1
Received 4 October 1999; accepted 7 March 2001
Background: Studies examining the effects of oral contraceptives (OCs) on mood, affect, and affect variability are
reviewed. Methods: MEDLINE and PsycLIT data bases were examined to identify studies that compared OC users with
nonusers using daily ratings of mood, affect, or affect variability. Results: Compared to non-users, OC users experience less
variability in affect across the entire menstrual cycle, and less negative affect during menstruation (i.e. withdrawal bleeding).
In women with OC-related negative mood and affect change, potential mediators of the relation between OCs and mood or
affect were identified: a history of depression, psychiatric symptoms, dysmenorrhea, and premenstrual mood symptoms prior
to OC use; a history of pregnancy-related mood symptoms; a family history of OC-related mood complaints; being in the
postpartum period; and age. Furthermore, a lower ratio of progesterone to estrogen is associated with more negative mood
change in women with a history of premenstrual emotional symptoms, higher progesterone to estrogen ratios are associated
with increased negative mood effects in women without such a history, and monophasic OCs have a greater stabilizing effect
on mood than triphasic OCs. Limitations: The ‘survivor effect’, psychological factors, and indirect pharmacological effects
(e.g. weight gain) have not yet been systematically investigated. Furthermore, most studies have examined only negative
mood or affect, as opposed to both positive and negative affect and affect variability; and few affect studies have assessed
potential mediators of OC-related affect change. Conclusions: While the only consistent OC-related mood effects
experienced by most women are beneficial, a subgroup of women do experience negative mood change. Future research must
focus on expounding the individual difference and OC-related risk factors for negative mood change.
2002 Elsevier Science B.V. All rights reserved.
Keywords: Oral contraceptives; Affect; Hormones; Mood; Estrogen; Progesterone
Mood change remains one of the controversial
side effects of oral contraceptive (OC) use despite a
great deal of research. Change in mood, specifically
depression, is one of the most common reasons given
for discontinuing OC use (Goldzieher, 1994). How-
*Corresponding author. Tel.: 11-807-343-8257; fax: 11-807-
E-mail address: email@example.com (D. Mazma-
0165-0327/02/$ – see front matter
2002 Elsevier Science B.V. All rights reserved.
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240
ever, guidelines for the management of OC-related
mood change have not been based on comprehensive
reviews of prospective clinical studies (e.g. Dickey,
1998) . In order to provide a comprehensive review
of the OC literature, this paper will: (a) discuss the
biochemical mechanisms for OC-mediated mood
change; (b) clarify the distinctions between mood
and affect, and categorical versus dimensional mea-
surement approaches; (c) review the research on the
relationship between OCs and affect; (d) review the
research on individual difference variables and OC-
related variables that predispose women to OC-re-
lated affect/mood change; (e) discuss limitations of
previous research; and (f) suggest directions for
future research in the area.
also possible that estrogen potentiates any proges-
togenic effects on mood (Fink et al., 1998).
Although biochemical mechanisms for any posi-
tive mood effects of OCs are less frequently dis-
cussed in the literature, the most plausible mecha-
nism involves an increase in serotonin levels due to
an estrogen-mediated inhibition of the MAO path-
way (Chakravorty and Halbreich, 1997). In fact,
increases in estrogen are associated with site-specific
tryptamine(5-HT ) receptors, (b) the expression
of genes for the 5-HT receptor, and (c) serotonin
transporter mRNA (Fink and Sumner, 1996; Fink et
In terms of an OC-related effect on affect vari-
ability, Felthous and Robinson (1981) suggest that
monophasic OCs may stabilize affect across the
cycle through stabilization of cycle fluctuations in
estrogen and progesterone, resulting in a stabilization
of the functional levels of the relevant central
2. Biochemical theories: hormones and mood
Although no clear consensus exists as to how the
estrogen and progesterone in OCs could affect mood,
a number of mechanisms have been proposed
(Sheehan and Sheehan, 1976; Moller, 1981; Slap,
1981; Roy-Byrne et al., 1984; Deijen et al., 1992;
Patten and Love, 1993; Special Advisory Committee
on Reproductive Physiology, 1994; Tuiten et al.,
1995; Sherwin, 1996).
The most plausible explanations for hormone-me-
diated negative mood change include: (a) an es-
trogen-induced pyridoxine (one component of vita-
min B ) deficiency and subsequent decrease in
serotonin and GABA levels due to the low affinity of
the decarboxylases for pyridoxal phosphate (e.g.
Slap, 1981; Patten and Love, 1993; McCarty, 2000),
(b) a progesterone- and estrogen-mediated augmenta-
tion of GABA-induced inhibition and suppression of
glutamate excitation (e.g. Ghazal et al., 1976; Smith
et al., 1987), and (c) a progesterone-mediated in-
crease in monoamine oxidase (MAO) activity, re-
Sheehan and Sheehan, 1976; Sherwin, 1996). Since
estrogen induces brain progesterone receptors, it is
3. Mood versus affect
Mood and affect are closely related terms that are
often used interchangeably in the literature. There
are, however, important and useful distinctions be-
tween the two concepts. The American Psychiatric
Association (2000) compares mood and affect to
climate and weather, respectively. Mood is referred
to as ‘‘a more pervasive and sustained emotional
‘climate’’’ while affect involves ‘‘more fluctuating
changes in emotional ‘weather’’’ (p. 819). Early
research on the relationship between mood and OC
use took a categorical approach by focusing on the
presence or absence of diagnosable mood disorders
in women who use OCs. Most studies looked at the
incidence of clinical depression in OC users (e.g.
Herzberg et al., 1970; Kutner and Brown, 1972a;
Vessey et al., 1985). Another approach involves
viewing mood as a dimensional continuum and
measuring the positive or negative mood change
occurring during OC use. The dimensional approach
allows one to look at sub-clinical change and thus
investigate changes in all women as opposed to only
those who develop a mood disorder. It also allows
1Guidelines have been based primarily on laboratory measurement
of estrogenic, progestational, and androgenic activity of specific
OC formulations, and hypothetical relationships between relative
hormonal activity and specific mood symptoms.
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240 231
for the measurement of change in both positive affect
and negative affect (Watson and Tellegen, 1985).
Positive affect reflects the extent to which a person
feels a zest for life. Someone with high positive
affect is characterized as attentive, interested, alert,
and enthusiastic. Someone with low positive affect
feels sleepy and tired. Negative affect reflects the
extent to which a person conveys feeling distressed
or unpleasantly aroused. A person with high negative
affect feels downhearted, hostile, angry, fearful, and
guilty, while someone with low negative affect is
characterized as calm and content. The rationale for
measuring negative and positive affect separately
comes from studies, summarized by Watson and
Tellegen (1985), which have consistently found that
positive and negative affect are dominant and rela-
tively independent dimensions (e.g. Diener and Em-
mons, 1985; Kercher, 1992).
Thus, a categorical approach to investigating
mood change associated with OC use is concerned
with the development of a diagnosable mood disor-
der while the dimensional approach looks at shorter-
term changes in mood or levels of negative and/or
incidence of depression (e.g. Fleming and Seager,
1978; Vessey et al., 1985). While many reviews of
this early research are incomplete, Slap (1981),
Cullberg (1972), and Long and Kathol (1993) in-
clude most of the better studies.
5. Recent research: oral contraceptives and
affect (dimensional approach)
More recent studies have focused on differences in
affect by having subjects fill out daily self-rating
scales for 30–90 days (e.g. Walker and Bancroft,
1990; Almagor and Ben-Porath, 1991). Daily rating
scales can be used to examine cyclical changes in
affect across the menstrual cycle and general trends
or differences in mood between the OC users and
non-users. This methodology also permits mood
variability to be assessed, and allows researchers to
identify changes that might occur at specific phases
of the menstrual cycle. Daily mood ratings, or more
accurately, daily ratings of affect, are a more sensi-
tive indicator of differences between users and non-
users of oral contraceptives.
A review of the literature revealed 17 studies
comparing daily ratings of affect in both OC users
and non-users . Given the limitations of retrospective
designs, four studies were not included in the present
review (Moos, 1968; Rouse, 1978; Warner and
Bancroft, 1988; Bancroft and Rennie, 1993). Exami-
nation of the results of the 13 controlled prospective
daily-rating studies revealed that all but one study
(Marriott and Faragher, 1986) found group differ-
ences in affect between OC users and non-users.
However, the direction of the differences and the
menstrual cycle phases in which the differences
occurred were not consistent across studies.
Since most researchers divided the menstrual cycle
4. Early research: oral contraceptives and mood
The early research
with OC use has produced inconsistent findings.
Most studies employed designs that assessed mood
in never-users, past-users, and present users of OCs,
one to three times over the course of several months.
The early studies usually either compared rates of
depressive disorders between the two or three groups
(e.g. Royal College of General Practitioners, 1974)
or compared scores on mood rating scales designed
to measure general or average mood (e.g. Herzberg
et al., 1970). Some researchers found increased rates
of depression in OC users (e.g. Nilsson and Almgren,
1968; Herzberg et al., 1970; Cullberg, 1972), others
found a decrease in depression rates (e.g. Herzberg et
al., 1971; Deijen et al., 1992), and some studies did
not find any relationship between OC use and
on mood change associated
3Studies were excluded if there were no published statistical
analyses or raw data to examine differences in affect between OC
users and non-users (e.g. Gallant et al., 1992), or if a control group
of non-OC users was not included in the study (e.g. Forrest, 1979;
Alexander and Sherwin, 1993; Tuiten et al., 1995). [Bancroft and
Sartorius (1990) review a number of the studies covered in the
present review and include an excellent broad review of methodo-
logical considerations and the effects of OCs on well-being and
2It is important to keep in mind that a major reduction of the
estrogen dosage in oral contraceptives occurred in 1985.
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240
into different phases, the four most frequently used
phases were used for this review: all days of
menstrual flow (menstrual phase), the 7 days after
menstruation (postmenstrual phase), the remaining
days leading up to the 7 days before menstruation
(intermenstrual phase), and the 7 days before men-
struation (premenstrual phase). (Although, technical-
ly speaking, women taking OCs experience with-
drawal bleeding as opposed to menstruation, the
bleeding phase of the cycle will be referred to as the
menstrual phase for all women, as is common in this
literature.) Three types of findings are summarized
below: group differences in affect, group differences
in affect variability, and group differences during
each of the four phases of the menstrual cycle.While
most studies examined negative affect, a few studies
also examined positive affect.
The majority of the studies found no significant
group differences in negative affect across the entire
menstrual cycle (Paige, 1971; Wilcoxon et al., 1976;
Marriott and Faragher, 1986; Almagor and Ben-
Porath, 1991). While one study found that OC users
feel less negative affect across the cycle (Boyle and
Grant, 1992), one study found that monophasic OC
users experienced higher negative affect throughout
the cycle (Walker and Bancroft, 1990). In terms of
positive affect, Almagor and Ben-Porath (1991)
found that OC users experienced higher positive
affect. Two other studies, however, did not find
group differences in positive affect (Silbergeld et al.,
1971; Boyle and Grant, 1992).
The ingestion of the hormones in OCs does
provide a stabilizing effect on mood. Four studies
found that OC users showed less day-to-day vari-
ability in their affect ratings than non-users (Paige,
1971; Sutker et al., 1983; Walker and Bancroft,
1990; Graham and Sherwin, 1993). Only one study
(McFarlane et al., 1988) did not find any significant
differences in affect variability between the two
groups. It is noteworthy that researchers have not
examined group differences in positive affect vari-
ability across the menstrual cycle, or group differ-
ences in circadian affect variability.
When group differences in negative affect within
specific menstrual cycle phases were examined
(Paige, 1971; Morris and Udry, 1972; Wilcoxon et
al., 1976; Sutker et al., 1983; Marriott and Faragher,
1986; Alexander et al., 1990; Walker and Bancroft,
1990; Almagor and Ben-Porath, 1991; Boyle and
Grant, 1992; Graham and Sherwin, 1993), only the
findings from the menstrual phase were relatively
consistent.While three studies did not find any group
differences in negative affect during the menstrual
phase (Alexander et al., 1990; Almagor and Ben-
Porath, 1991; Graham and Sherwin, 1993) and one
study found higher negative affect for monophasic,
but not triphasic, users (Walker and Bancroft, 1990),
four studies suggest that OC users experience less
negative affect than non-users during the menstrual
phase (Paige, 1971; Wilcoxon et al., 1976; Sutker et
al., 1983; Boyle and Grant, 1992). Although three of
these studies did not provide an interpretable statisti-
cal analysis of the comparison (Paige, 1971; Sutker
et al., 1983; Boyle and Grant, 1992), graphs or data
suggested consistent trends in the direction of OC
users experiencing less negative affect than non-
users during menstruation. This reduction in negative
affect could, of course, be the consequence of an
OC-mediated reduction in somatic symptoms (i.e. an
indirect pharmacological effect). Three of these
studies that examined somatic symptoms did, in fact,
find lower levels of physical symptomatology in OC
users than non-users during menstruation, but it was
not possible to establish the extent to which this
OC-mediated reduction in physical symptoms could
account for the lower negative affect in OC users.
This remains an interesting and viable explanation,
Although four studies (McFarlane et al., 1988;
Walker and Bancroft, 1990; Almagor and Ben-
Porath, 1991; Boyle and Grant, 1992) measured
positive affect over the menstrual cycle, there is no
consistent difference in positive affect between OC
users and non-users during any of the menstrual
6. Risk factors or predisposing factors to mood/
6.1. Individual difference variables
As reviewed above, the only common effects of
OCs on affect for all women are better described as
beneficial than adverse effects (i.e. less variability in
negative affect across the cycle and less negative
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240233
affect during menstruation). However, the fact that
some women do experience OC-related increases in
negative affect/mood (Canadian Pharmaceutical As-
sociation, 2000) suggests that individual difference
variables predispose certain women to specific OC-
mediated changes in mood or affect. The majority of
research on such variables has focused on OC-related
changes in mood or depression. Only a few studies
have examined specific individual difference vari-
ables that may increase the likelihood of OC-me-
diated changes in affect. Furthermore, very few of
the studies were conducted using the low-dose OCs
that are currently in use.
Previous studies suggest that the following vari-
ables increase a woman’s risk of becoming depressed
or exhibiting depressive symptoms when taking OCs:
(a) a history of depression (Nilsson et al., 1967;
Lewis and Hoghughi, 1969; Herzberg et al., 1971);
(b) a history of moderate to severe ‘premenstrual
depression’ (Herzberg and Coppen, 1970) or severe
premenstrual weepiness prior to OC use (Herzberg et
al., 1971); (c) a history of dysmenorrhea prior to OC
use (Herzberg and Coppen, 1970); (d) a history of
depression during or after pregnancy (Nilsson et al.,
1967; Herzberg et al., 1971; Kutner and Brown,
1972b); (e) a family history of OC-related depressive
symptoms (Kendler et al., 1988); (f) a predisposition
to a vitamin B deficiency while taking OCs (Adams
et al., 1974; Wynn et al., 1975), and (g) a high level
of psychological distress prior to OC use (Cullberg,
1972; Tuiten et al., 1995). While one study found
that a history of psychological distress was not
associated with an increase in symptoms of depres-
sion (Grounds et al., 1970), this study only examined
the first 2 months of OC use.
Two studies suggest that the following variables
are unrelated to the development of depression while
taking OCs: age; age at menarche; length of menstru-
al cycle; parental separation, divorce, or death; birth
weight; feeding problems or serious illness during
the first year of life (Herzberg et al., 1971); educa-
tion; religion; and family size (Murawski et al.,
Only four studies have investigated individual
difference variables that predispose women to OC-
related changes in affect. Bancroft and colleagues
(1987) found that women younger than 20 ex-
perienced more negative affect, or perhaps less
positive affect, during days 13–18 of OC use than
older women. This is in line with the finding that
women older than 30 experience less tension than
non-users during menstruation while taking OCs
(Rouse, 1978). Therefore, while women under age
20 are more likely to experience an increase in
negative affect, women over 30 experience a de-
crease in negative affect. Another study found that
women who are premenstrually ‘depressed’ at
baseline show greater improvement in behavioural
positive affect measures premenstrually than ‘non-
depressed’ women while taking a triphasic OC
(Graham and Sherwin, 1992). A final study ex-
amined the following variables and found that they
could not account for the observed differences in
negative affect variability between monophasic OC
users and non-users: age at menarche, reaction to
menarche, mother’s attitude to menstruation, amount
of preparation for menstruation, degree of menstrual
discomfort during teens, and prior expectations about
the effects of OCs on mood (Paige, 1971).
6.2. Oral contraceptive-related variables
As well as the individual difference variables
noted above, there are a number of OC-related
variables that may affect women directly or interact
with other variables to influence mood or affect. For
example, OCs can differ in terms of constituents
(types of estrogen and progesterone), dosage, ratio of
progesterone to estrogen, and temporal pattern of
dosage (monophasic, biphasic, or triphasic). A num-
ber of these variables have been investigated in terms
of their effects on mood and a few in terms of their
effects on affect.
While there is a consensus that high doses of both
estrogen and progesterone in OCs are associated with
an increase in negative mood side effects (e.g. Deijen
et al., 1992), there is no consensus of findings
indicating whether higher doses of progesterone,
higher doses of estrogen, or specific ratios of these
hormones are likely to increase the risk of mood or
affect change for all women. Four studies suggest
that OCs with either higher dosages of progesterone
or higher ratios of progesterone to estrogen produce
more depression or negative mood effects (Peterson,
1969, cited in Cullberg, 1972; Grant and Pryse-
Davies, 1968; Lewis and Hoghughi, 1969; Akerlund
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240
et al., 1993). Interestingly, in the study by Akerlund
and colleagues, when the progesterone dosage was
held constant, women taking the OC with the lower
estrogen dosage reported more negative mood
change. In contrast, a large-scale study indicated that
rates of depression increase as estrogen dose in-
creases (Kay, 1984) and two studies found that lower
dosages of progesterone or lower ratios of progester-
one to estrogen were associated with greater symp-
toms and severity of depression (Cullberg, 1972;
Kutner and Brown, 1972a). Another study found that
the dosage of progesterone was unrelated to rates of
depression (Kutner and Brown, 1972b). The incon-
sistency in the findings suggests the possibility of a
third variable that moderates the relationship be-
tween progesterone or estrogen dosage and sub-
sequent mood or affect change.
Two excellent studies indicate that premenstrual
mood symptoms moderate the relationship between
type of hormone, dosage, and negative mood change
(Bancroft et al., 1987; Cullberg, 1972). Women with
a history of premenstrual irritability/emotionality
prior to OC use had an increased risk of negative
mood effects if they took an OC with lower proges-
terone levels or lower progesterone to estrogen
ratios. Furthermore, women without a history of
premenstrual irritability/emotionality had an in-
creased risk of negative mood effects if they were
taking an OC with higher progesterone levels or
higher progesterone to estrogen ratios. The inter-
action between progesterone dosage and premenstru-
al irritability history on mood suggests that history of
premenstrual emotional symptoms should be consid-
ered when selecting an OC.
The relative effects of monophasic, biphasic, and
triphasic OCs on mood and affect have also been
examined. Although two studies did not find any
difference in rates of severity of depression between
monophasic, triphasic, and non-users (Herzberg et
al., 1970; Bancroft and Rennie, 1993), three studies
suggest that monophasic OC use is associated with
greater mood stability than triphasic OC use (Moos,
1968; Warner and Bancroft, 1988), biphasic OC use
(Paige, 1971), and no use (Paige, 1971; Warner and
Bancroft, 1988). Kutner and Brown (1972a) also
found that monophasics are associated with less
‘premenstrual depression’ than biphasics.
While one study examining affect found that
monophasic users experience more negative affect
compared to triphasic users and non-users (Walker
and Bancroft, 1990), the subjects were all experienc-
ing premenstrual symptoms while taking OCs and
had been taking OCs for a mean of 3.7 years. Thus,
the research indicates that monophasic OCs stabilize
mood, and as illustrated in the study by Bancroft and
colleagues (1987), are much less likely than triphasic
OCs to cause negative mood changes in women with
high baseline premenstrual emotional symptoms.
An examination of research on OC discontinuance
and duration of use indicates that the severity of
premenstrual irritability symptoms increases over
time after OC discontinuance (Kutner and Brown,
1972a), and that there are contradictory findings
regarding whether duration of pill use is related to
depression (Herzberg et al., 1970; Kutner and
Brown, 1972a). Given that the ‘survivor effect’
(discussed below) may account for some of the
inconsistency in studies on the relationship between
affect and OCs, duration of use should be investi-
The review of OC-related variables that affect
mood/affect suggests three conclusions. First, for
women with a history of premenstrual emotional
symptoms prior to OC use, OC formulations with
higher progesterone to estrogen dosage ratios are
associated with less negative mood changes, as are
the use of monophasic versus triphasic OCs. Second,
for women without a history of premenstrual ir-
ritability, formulations with lower ratios of proges-
terone to estrogen decrease the risk of negative mood
change. Finally, adding to the earlier discussion
about the stabilizing effects of OCs on mood,
monophasic OCs have a greater stabilizing effect
than triphasic OCs.
7. Problems with previous studies
The above review reveals that future research
must: (a) control for the ‘survivor effect’ (discussed
below) and duration of OC use, (b) independently
measure both negative and positive affect, (c) ex-
pound mediating or moderating individual difference
or OC-related variables for specific types of affect
change, (d) assess group differences in both negative
and positive affect variability, and (e) control for
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240235
three potentially important psychological or indirect
The ‘survivor effect’ (Kutner and Brown, 1972a)
refers to the fact that women who experience prob-
lems while taking OCs stop taking them, which, by
definition, leaves a group of women who are not
experiencing adverse effects (mood or otherwise).
Roughly 25% of women discontinue OC use within
1 year of use (Trussell and Kost, 1987) and depres-
sion is in fact one of the most common reasons given
for discontinuation (Kay, 1984). Therefore, the
incidence rates of negative mood change associated
with being on OCs are likely underestimations in
studies using groups of women taking OCs for many
years, since any women who experience moderate to
severe negative mood effects would likely have
Three factors can contribute to the survivor effect:
(a) the selection criteria for the OC user and non-user
groups, (b) the length of time which the OC users
have been taking the OC pill, and (c) the study’s
attrition rate. In order to minimize survivor effects,
researchers should distinguish between first-time
users and brand ‘switchers’ in the OC user group,
and between never-users and previous-users in the
non-user group. The inclusion of ‘switchers’ in the
user group and of previous users in the non-user
group introduces a bias, since it may have been
negative mood effects which led to switching brands
or discontinuing use (i.e. Deijen et al., 1992). With
regards to duration of pill use, it is important to
distinguish between ‘early’ and ‘late’ use of OCs
(e.g. Bancroft and Sartorius, 1990). While an early-
use group would still contain women who may
discontinue OCs in the future due to negative effects,
a late-use OC group would have already lost the
women experiencing negative mood side effects.
Finally, in prospective studies and controlled experi-
ments comparing OC users and non-users, large
attrition rates increase the size of the survivor effect.
It seems logical that a woman just starting on the pill
who drops out of an experiment would be more
likely to do so as a result of negative than positive
mood side effects. Losing this type of data makes the
OC users groups appear to be experiencing less
negative affect than they actually are. The ‘survivor
group’ would include women who are taking OCs
and potentially experiencing no mood effects, posi-
tive mood effects, or only mild negative effects on
A second suggestion for future research is to use
independent measures of change in positive affect,
negative affect, positive affect variability, and nega-
tive affect variability. If women’s subjective com-
plaints of OC-related increased depression or nega-
tive affect, or increased moodiness or lability, actual-
ly reflect decreased positive affect and increased
positive affect variability, respectively, traditional
measurement of negative affect and negative affect
variability will not identify such differences. Given
the independence of positive and negative affect,
separate measures of these dimensions are required
to detect any effects of OCs on affect or affect
A third suggestion is to evaluate individual differ-
ence variables and OC-related variables associated
with specific OC-related affect changes. While many
of the studies reviewed above examined predisposing
variables for OC-related depression, very few studies
have focused on examining such risk factors for
changes in positive and negative affect and affect
variability. If OC-related affect change varies as a
function of OC type, a failure to separately analyse
users of monophasic, biphasic, and triphasic OC
preparations may make it difficult to detect mood
change if women taking different OC types are
grouped together. Morris and Udry (1972) found that
of the OC users who felt ‘differently’, one-half felt
‘better than usual’ and the other half felt ‘worse than
usual’. This suggests that if some women experience
increased negative affect and others experience
decreased negative affect with OC use, a failure to
separately examine these groups means that any
negative affect changes may cancel each other out if
the overall mean of OC users is compared with
non-users. Obviously, a failure to separately examine
groups of women experiencing different types of
mood change can result in an obfuscation of actual
mood change as well as an increased likelihood of
overlooking potentially important predisposing in-
dividual difference and OC-related variables.
As reviewed above, women taking OCs ex-
perience less variability in negative affect than non-
users. It may be that affect variability, rather than
average affect, is primarily altered by OCs. This
hypothesis suggests the importance of assessing
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240
group differences in positive as well as negative
affect variability. Furthermore, diurnal and ultradian
variation in mood has long been observed in affec-
tive disorders and there are known diurnal and
circadian rhythms in neuroendocrine activity (e.g.
adrenocorticotropic hormone, cortisol, growth hor-
mone, thyroid stimulating hormone; Goodwin and
Jamison, 1990; Checkley, 1992; Thase and Howland,
1995). Unfortunately, studies comparing OC users
and non-users have failed to examine within-day
variability in affect.
A fifth suggestion is to control for three possible
psychological and indirect pharmacological causes of
mood change due to OC use (e.g. Cullberg, 1972).
First, expectations of a positive or negative mood
change side effect can lead to mood change (or the
placebo effect; e.g. Cullberg et al., 1969). Second,
what Cullberg (1972) refers to as the ‘‘symbolic
effect of the ‘anti-baby pill’’’ may also lead to
changes in mood (e.g. increased negative affect due
to feelings of guilt over preventing the conception of
a child, or mood improvements due to increased
reassurance about pregnancy prevention). Third,
mood changes may also occur secondary to somatic
side effects (an indirect pharmacological effect) (e.g.
increased negative affect following weight gain or
increased positive affect following a desired breast
size increase). Since these factors could hypothetical-
ly lead to an improvement or deterioration in mood,
pre-test measures of each should be included when
attempting to determine the cause of a mood change
associated with OC use. However, it is important to
realize that the correlation of any of the three factors
with the predicted mood change does not rule out the
possibility of a pharmacological effect.
affect across the menstrual cycle and less negative
affect during the menstrual phase. Research also
suggests that the following factors predispose certain
women to OC-related negative changes in mood/
affect: a history of depression, other symptoms of
psychological distress (e.g. anxiety, stress, neuro-
symptoms prior to OC use; a history of pregnancy-
related mood symptoms; a family history of OC-
related mood complaints; being in the postpartum
period; and age. A number of OC-related variables
also mediate mood/affect change for certain in-
dividuals: types of estrogen and progesterone in the
OC; dosage; ratio of progesterone to estrogen in the
OC; temporal pattern of dosage; and duration of OC
use. More specifically, women with premenstrual
mood symptoms prior to OC use experience more
negative mood symptoms when taking OC formula-
tions with lower progesterone to estrogen dosage
ratios or when taking triphasic (versus monophasic)
OCs. Second, for women who do not complain of
premenstrual irritability, formulations with higher
ratios of progesterone to estrogen increase the risk of
negative mood change. Finally, monophasic OCs
have a greater stabilizing effect than triphasic OCs
In the OC literature it is commonly suggested that,
due to the 1985 estrogen reduction in OCs, there is
no need for concern about the possibility of OC-
related negative affect change. However, four factors
suggest that the above individual difference and
OC-related variables cannot be discounted as pre-
disposing factors or mediators for negative mood
change. First, women continue to experience nega-
tive mood effects while taking OCs (Canadian
Pharmaceutical Association, 2000). Second, some
studies have indicated that OCs with lower estrogen
dosages are associated with more negative mood
complaints (e.g. Akerlund et al., 1993). Third, while
the earlier studies included women taking higher
dose OCs than are commonly taken today, most of
the studies also included a substantial number of
women taking OCs that have equivalent estrogen
dosages to formulations prescribed today (e.g. Her-
zberg et al., 1971; Kutner and Brown, 1972a).
Fourth, while only a few recent studies have ex-
amined individual difference and OC-related mediat-
ing factors in mood/affect change (e.g. Bancroft et
A number of plausible biochemical explanations
have been put forward to explain how the estrogen
and progesterone in OCs could have effects on mood
or affect. As the dimensional approach of studying
affect is more sensitive than the categorical approach
of examining rates of mood disorders, studies that
compared daily ratings of affect in OC users and
non-users were reviewed. Compared to non-users,
women taking OCs experience less variability in
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240237
al., 1987; Warner and Bancroft, 1988; Akerlund et
al., 1993), there are no studies that contradict the
Other individual difference variables deserving
exploration include physiological measures such as
baseline levels of endogenous hormones (e.g. es-
trogen, progesterone, cortisol), personal and family
medical history indicators (e.g. thyroid disorders,
breast cancer, and diabetes), and variables indicative
of differential hormone sensitivity (e.g. body mass
index, waist-to-hip ratio or body fat distribution,
acne history, and menses duration).
Another important area of future research involves
the examination of the interaction between OC type
and the lipid, lipoprotein, and hormone parameters
known both to be affected by OCs and to be
associated with changes in mood or affect. For
example, while the use of some OCs is associated
with a decrease in total cholesterol, LDL cholesterol,
and triglyceride levels (e.g. Wiegratz et al., 1998;
Mostafavi et al., 1999; Stefanick, 1999), the use of
other OC types is associated with increases in total
cholesterol and triglyceride levels (e.g. Schiele et al.,
1998; Wiegratz et al., 1998; Cheung et al., 1999;
Mostafavi et al., 1999). These changes may play an
etiological role in OC-related mood change as some
studies have indicated an association between higher
levels of depression and anxiety and low levels of
total cholesterol, triglycerides, and LDL cholesterol
(Olusi and Fido, 1996; Suarez, 1999). Other studies
have found associations between high triglyceride
and cholesterol levels and Generalized Anxiety
Disorder (Kuczmiercyk et al., 1996), and high
triglyceride levels and depressive symptoms (Glueck
et al., 1993). Some women may be more sensitive to
OC-related changes in lipid parameters and/or any
effects of lipid parameters on mood. With further
research it may be possible for physicians to choose
an OC least likely to cause mood change, based on a
woman’s baseline lipid levels, the known effects of
each OC type on lipid levels, family and personal
physical/psychiatric history, as well as information
about behavioural factors known to affect lipid levels
(e.g. diet, exercise, and drug use). While the effects
of OCs on lipid levels have been investigated and
monitored in order to prevent cardiovascular disease,
these same variables deserve further attention as
possible mediators in OC-related affect change.
In addition, a number of methodological guide-
lines could improve future research. While these
guidelines may seem obvious, they should be re-
stated given the fact that, despite over 40 years of
research, many questions remain unanswered about
the relationship between OCs and mood or affect
changes. Moreover, recent studies (e.g. Haugen et
al., 1996; Sivin et al., 1998) investigating potential
side-effects of subdermal contraceptive implants are
limited by many of the same methodological prob-
lems as the early OC studies. First, subjects taking
OCs should be defined as long-time users or first-
time users, and non-users should be defined as
previous-users or never-users. Any long-time user
study may underestimate negative mood side effects
due to the survivor effect. Second, all studies should
use prospective, as opposed to retrospective, ratings
of affect. Third, independent measures of positive
and negative affect should be included as both types
of mood change have been reported. Fourth, group
differences in both positive and negative affect
variability should be examined as OCs provide a
stabilizing effect on mood. Fifth, the effects of OCs
on the variability of circadian and ultradian rhythms
as well as daily ratings of affect should be examined.
Finally, psychological (e.g. expectations) and indirect
pharmacological factors (e.g. somatic changes such
as weight gain) that could affect mood should be
assessed to determine what role, if any, they may
In response to the original question posed in the
title of this article, the only OC-related effects on
mood that are experienced by most women are
positive: mood stability and decreased negative
affect during menstruation. However, a subgroup of
women also experience negative mood effects, and
specific individual difference and OC-related vari-
ables increase one’s risk of experiencing these
Adams, P.W., Wynn, V., Seed, M., 1974. Vitamin B6, depression
and oral contraception. Lancet 2, 516–517.
Akerlund, M., Rode, A., Westergaard, J., 1993. Comparative
profiles of reliability, cycle control, and side effects of two oral
contraceptive formulations containing 150 micrograms de-
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240
sogestrel and either 30 micrograms or 20 micrograms ethinyl
oestradiol. Br. J. Obstet. Gynecol. 100, 832–838.
Alexander, G., Sherwin, B., 1993. Sex steroids, sexual behavior,
and selection attention for erotic stimuli in women using oral
contraceptives. Psychoneuroendocrinology 18, 91–102.
Alexander, G., Sherwin, B., Bancroft, J., Davidson, D., 1990.
Testosterone and sexual behavior in oral contraceptive users
and nonusers: a prospective study. Horm. Behav. 24, 388–402.
Almagor, M., Ben-Porath, Y., 1991. Mood changes during the
menstrual cycle and their relation to the use of oral contracep-
tives. J. Psychosom. Res. 35, 721–728.
American Psychiatric Association, 2000. Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. Author, Washington,
Bancroft, J., Rennie, D., 1993. The impact of oral contraceptives
on the experience of perimenstrual mood, clumsiness, food
craving and other symptoms. J. Psychosom. Res. 37, 195–202.
Bancroft, J., Sartorius, N., 1990. The effects of oral contraceptives
on well-being and sexuality. Oxford Rev. Reprod. Biol. 12,
Bancroft, J., Sanders, D., Warner, P., Loudon, N., 1987. The
effects of oral contraceptives on mood and sexuality: a
comparison of triphasic and combined preparations. J. Psycho-
som. Obstet. Gynecol. 7, 1–8.
Boyle, G., Grant, A., 1992. Prospective versus retrospective
assessment of menstrual cycle symptoms and moods: role of
attitudes and beliefs. J. Psychopathol. Behav. Assess. 14, 307–
Canadian Pharmaceutical Association, 2000. CPS: Compendium
of Pharmaceuticals and Specialties. 35th Edition. Ottawa, Ont.
Chakravorty, S.G., Halbreich, U., 1997. The influence of estrogen
on monoamine oxidase activity. Psychopharmacol. Bull. 33,
Checkley, S., 1992. Neuroendocrinology. In: Paykel, E.S. (Ed.),
Handbook of Affective Disorders, 2nd Edition. Guilford Press,
New York, pp. 255–266.
Cheung, M.C., Walden, C.E., Knopp, R.H., 1999. Comparison of
the effects of triphasic oral contraceptives with desogestrel or
levonorgestrel on apolipoprotein A-I-containing high density
lipoprotein particles. Metab. Clin. Exp. 48, 658–664.
Cullberg, J., 1972. Mood changes and menstrual symptoms with
different gestagen/estrogen combinations. Acta Psychiatr.
Scand. 236, 1–86.
Cullberg, J., Gelli, M., Jonsson, C., 1969. Mental and sexual
adjustment before and after six months use of an oral contra-
ceptive. Acta Psychiatr. Scand. 45, 259–276.
Deijen, D.B., Duyn, K., Jansen, W., Klitsie, J., 1992. Use of
monophasic, low-dose oral contraceptives in relation to mental
functioning. Contraception 46, 539–567.
Dickey, R.P., 1998. Managing Contraceptive Pill Patients, 9th
Edition. EMIS, New Orleans, LA.
Diener, E., Emmons, R.A., 1985. The independence of positive
and negative affect. J. Pers. Soc. Psychol. 47, 1105–1117.
Felthous, A.R., Robinson, D.B., 1981. Oral contraceptive medica-
tion in prevention of psychotic exacerbations associated with
phases of the menstrual cycle. J. Prev. Psychiatr. 1, 5–15.
Fink, G., Sumner, B.E.H., 1996. Oestrogen and mental state.
Nature 383, 306.
Fink, B., Sumner, B.E.H., McQueen, J.K., Wilson, H., Rosie, R.,
1998. Sex steroid control of mood, mental state, and memory.
Clin. Exp. Pharmacol. Physiol. 25, 764–775.
Fleming, O., Seager, C., 1978. Incidence of depressive symptoms
in users of the oral contraceptive. Br. J. Psychiatry 132,
Forrest, A., 1979. Cyclical variations in mood in normal women
taking oral contraceptives. Br. Med. J. 1, 1403.
Gallant, S., Popiel, D., Hoffman, D., Chakraborty, P., Hamilton, J.,
1992. Using daily ratings to confirm premenstrual syndrome/
late luteal phase dysphoric disorder. Part II. What makes a
‘real’ difference? Psychosom. Med. 54, 167–181.
Ghazal, A.E.M., Makar, A.B., Daabees, T.T., 1976. Effect of oral
contraceptives (Lyndiol) on rat brain gamma aminobutyric acid
system. Biochem. Pharmacol. 25, 115–118.
Glueck, C.J., Tieger, M., Kunkel, R., Tracy, T., 1993. Improve-
ment in symptoms of depression and in an index of life
stressors accompany treatment of severe hypertriglyceridemia.
Biol. Psychiatry 34, 240–252.
Goldzieher, J., 1994. Hormonal Contraception: Pills, Injections,
and Implants, 3rd Edition. Emis-Canada, London, Ont.
Goodwin, F.K., Jamison, K.R., 1990. Manic-Depressive Illness.
Oxford University Press, New York, NY.
Graham, C., Sherwin, B., 1992. A prospective treatment study of
premenstrual symptoms using a triphasic oral contraceptive. J.
Psychosom. Res. 36, 257–266.
Graham, C., Sherwin, B., 1993. The relationship between mood
and sexuality in women using an oral contraceptive as a
treatment for premenstrual symptoms. Psychoneuroendocrinol-
ogy 18, 273–281.
Grant, E.C.G., Pryse-Davies, J., 1968. Effect of oral contracep-
tives on depressive mood changes and on endometrial mono-
amine oxidase and phosphates. Br. Med. J. 3, 777–780.
Grounds, D., Davies, B., Mowbray, R., 1970. The contraceptive
pill, side effects and personality: report of a controlled double
blind trial. Br. J. Psychiatry 116, 169–172.
Haugen, M.M., Evans, C.B., Kim, M.H., 1996. Patient satisfaction
with a Levonorgestrel-releasing contraceptive implant: Reasons
for and patterns of removal. J. Reprod. Med. 41, 849–854.
Herzberg, B., Coppen, A., 1970. Changes in psychological
symptoms in women taking oral contraceptives. Br. J. Psychi-
atry 116, 161–164.
Herzberg, B.N., Johnson, A., Brown, S., 1970. Depressive symp-
toms and oral contraceptives. Br. Med. J. 4, 142–145.
Herzberg, B.N., Draper, K., Johnson, A., Nicol, G., 1971. Oral
contraceptives, depression, and libido. Br. Med. J. 6, 495–500.
Kay, C.R., 1984. The Royal College of General Practitioner’s oral
contraceptive study: some recent observations. Clin. Obstet.
Gynaecol. 11, 759–786.
Kendler, K.S., Martin, N.J., Heath, A.C., Handelsman, D., Eaves,
L., 1988. A twin study of the psychiatric side effects of oral
contraceptives. J. Nerv. Ment. Dis. 176, 153–160.
Kercher, K., 1992. Assessing subjective well-being in the old-old:
the PANAS as a measure of orthogonal dimensions of positive
and negative affect. Res. Aging 14, 131–168.
Kuczmiercyk, A.R., Barbee, J.G., Bologna, N.A., Townsend,
M.H., 1996. Serum cholesterol levels in patients with general-
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240 239
ized anxiety disorder (GAD) and with GAD and comorbid
major depression. Can. J. Psychiatry 41, 465–468.
Kutner, S., Brown, W., 1972a. Types of oral contraceptives,
depression, and premenstrual symptoms. J. Nerv. Ment. Dis.
Kutner, S.J., Brown, W.L., 1972b. History of depression as a risk
factor for depression with oral contraceptives and discon-
tinuance. J. Nerv. Ment. Dis. 155, 163–169.
Lewis, A., Hoghughi, M., 1969. An evaluation of depression as a
side effect of oral contraceptives. Br. J. Psychiatry 115, 697–
Long, T., Kathol, R.G., 1993. Critical review of data supporting
affective disorder caused by nonpsychotropic medication. Ann.
Clin. Psychiatry 5, 259–270.
Marriott, A., Faragher, E., 1986. An assessment of psychological
state associated with the menstrual cycle in users of oral
contraception. J. Psychosom. Res. 30, 41–47.
McCarty, M.F., 2000. High-dose pyridoxine as an ‘anti-stress’
strategy. Med. Hypotheses 54, 803–807.
McFarlane, J., Martin, C., Williams, T., 1988. Mood fluctuations:
women versus men and menstrual versus other cycles. Psychol.
Wom. Q. 12, 201–223.
Moller, E., 1981. Effect of oral contraceptives on tryptophan and
tyrosine availability: evidence for a possible contribution to
mental depression. Neuropsychobiology 7, 192–200.
Moos, R.H., 1968. Psychological aspects of oral contraceptives.
Arch. Gen. Psychiatry 19, 87–94.
Morris, N.M., Udry, J.R., 1972. Contraceptive pills and day-by-
day feelings of well-being. Am. J. Obstet. Gynecol. 113,
Mostafavi, H., Abdali, K., Zare, N., Rezaian, G.R., Ziyadlou, S.,
Parsanejad, M.E., 1999. A comparative analysis of three
methods of contraception: Effects of blood glucose and serum
lipid profiles. Ann. Saudi Med. 19, 8–11.
Murawski, B., Sapir, P., Shulman, N., Ryan, G., Sturgis, S., 1968.
An investigation of mood states in women taking oral contra-
ceptives. Fertil. Steril. 19, 50–63.
Nilsson, A., Almgren, P., 1968. Psychiatric symptoms during the
post-partum period as related to use of oral contraceptives. Br.
Med. J. 2, 453–455.
Nilsson, A., Jacobson, L., Ingemanson, C.A., 1967. Side-effects of
an oral contraceptive with particular attention to mental
symptoms and sexual adaptation. Acta Obstet. Gynecol. Scand.
Olusi, S.O., Fido, A.A., 1996. Serum lipid concentrations in
patients with major depressive disorder. Soc. Biol. Psychiatry
Paige, K., 1971. Effects of oral contraceptives of affective
fluctuations associated with the menstrual cycle. Psychosom.
Med. 33, 515–537.
Patten, S., Love, E., 1993. Can drugs cause depression? A review
of the evidence. J. Psychiatry Neurosci. 18, 92–102.
Rouse, P., 1978. Premenstrual tension: a study using the Moos
Menstrual Questionnaire. J. Psychosom. Res. 22, 215–222.
Royal College of General Practitioners, 1974. Oral Contraceptives
and Health. Pitman Medical, London.
Roy-Byrne, P., Rubinow, D., Gold, P., Post, R., 1984. Possible
antidepressant effect of oral contraceptives: case report. J. Clin.
Psychiatry 45, 350–352.
Schiele, F., Vincent-Viry, M., Fournier, B., Starck, M., Siest, G.,
1998. Biological effects of eleven combined oral contraceptives
on serum triglycerides, gamma-glutamyltransferase, alkaline
phosphatase, bilirubin and other biochemical variables. Clin.
Chem. Lab. Med. 36, 871–878.
Sheehan, D., Sheehan, K., 1976. Psychiatric aspects of oral
contraceptive use. Psychiatr. Ann. 6, 500–508.
Sherwin, B., 1996. Hormones, mood, and cognitive functioning in
postmenopausal women. Obstet. Gynecol. 82 (Suppl.), 20s–
Silbergeld, S., Brast, N., Noble, E., 1971. The menstrual cycle: a
double-blind study of symptoms, mood and behaviour, and
biochemical variables using Enovid and placebo. Psychosom.
Med. 33, 411–428.
Sivin, I., Mishell, D.R., Darney, P., Wan, L., Christ, M., 1998.
Levonorgestrel capsule implants in the United States: A 5-year
study. Obstet. Gynecol. 92, 337–344.
Slap, G., 1981. Oral contraceptives and depression: impact,
prevalence, and cause. J. Adolesc. Health Care 2, 53–64.
Smith, S., Waterhouse, B.D., Chapin, J.K., Woodward, D.J., 1987.
Progesterone alters GABA and glutamate responsiveness: a
possible mechanism for its anxiolytic action. Brain Res. 400,
Special Advisory Committee on Reproductive Physiology, 1994.
Oral Contraceptives. Health Canada.
Stefanick, M.L., 1999. Estrogen, progestogens and cardiovascular
risk. J. Reprod. Med. 44, 221–226.
Suarez, E.C., 1999. Relations of trait depression and anxiety to
low lipid and lipoprotein concentrations in healthy young adult
women. Psychosom. Med. 61, 273–279.
Sutker, P.B., Libet, J.M., Allain, A., Randall, C., 1983. Alcohol
use, negative mood states, and menstrual cycle phases. Al-
cohol. Clin. Exp. Res. 7, 327–331.
Thase, M.E., Howland, M.E., 1995. Biological processes in
depression: An updated review and integration. In: Beckham,
E.E., Leber, W.R. (Eds.), Handbook of Depression, 2nd Edi-
tion. Guilford Press, New York, pp. 213–279.
Trussell, J., Kost, C., 1987. Contraceptive failure in the United
States: a critical review of the literature. Stud. Fam. Plann. 18,
Tuiten, A., Panhuysen, G., Koppeschaar, H., Fekkes, D., Pijl, H.,
Frolich, M., Krabbe, P., Everaerd, W., 1995. Stress, serotoner-
gic function, and mood in users of oral contraceptives. Psycho-
neuroendocrinology 20, 323–334.
Vessey, M.P., McPherson, K., Lawless, M., Yeates, D., 1985. Oral
contraception and serious psychiatric illness: absence of an
association. Br. J. Psychiatry 146, 45–49.
Walker, A., Bancroft, J., 1990. Relationship between premenstrual
symptoms and oral contraceptive use: a controlled study.
Psychosom. Med. 52, 86–96.
Warner, P., Bancroft, J., 1988. Mood, sexuality, oral contracep-
tives and the menstrual cycle. J. Psychosom. Res. 32, 417–427.
Watson, D., Tellegen, A., 1985. Toward a consensual structure of
mood. Psychol. Bull. 98, 219–235.
Wiegratz, I., Jung-Hoffman, C., Gross,W., Kuhl, H., 1998. Effects
K.A. Oinonen, D. Mazmanian / Journal of Affective Disorders 70 (2002) 229–240
of two oral contraceptives containing ethinyl estradiol and
gestodene or norgestimate on different lipid and lipoprotein
parameters. Contraception 58, 83–91.
Wilcoxon, L., Schrader, S., Sherif, C., 1976. Daily self-reports on
activities, life events, moods, and somatic changes during the
menstrual cycle. Psychosom. Med. 38, 399–417.
Wynn,V., Adams, P.W., Folkard, J., 1975. Tryptophan, depression
and steroidal contraception. J. Steroid Biochem. 6, 965–970.