Ao y Cognicion Revision Sistematica
Ao y Cognicion Revision Sistematica
Ao y Cognicion Revision Sistematica
Frontiers in Neuroendocrinology
journal homepage: www.elsevier.com/locate/yfrne
A R T I C L E I N F O A B S T R A C T
Keywords: Oral contraceptives (OCs) are widely used. While the physical impacts of OCs have been well researched, there is
Oral contraceptive increasing interest on potential impacts of OCs on brain, behaviour and cognition. We systematically reviewed
Hormonal contraceptive the literature to determine the influence of OCs on cognition, including neurocognition, social cognition and
Cognition
emotional processing. Inclusionary criteria were: (a) premenopausal females taking OCs; (b) a control group of
Neuropsychological
Social cognition
naturally cycling women or OCs users in their inactive (i.e. ‘sugar pill’) phase; and (c) at least one measure of
Emotion recognition performance on a neurocognitive or social cognitive task. The systematic review found that OC use was asso
Memory ciated with some differences in performance on all cognitive domains examined (with the exception of basic
auditory attention and psychomotor performance). Several factors were identified that are likely to modulate the
way OCs influence cognition, including task related factors, OC type and control group characteristics. Directions
for future research are highlighted.
https://doi.org/10.1016/j.yfrne.2022.101052
Received 10 May 2022; Received in revised form 19 December 2022; Accepted 21 December 2022
Available online 26 December 2022
0091-3022/© 2022 Published by Elsevier Inc.
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
such as stress (Herrera et al., 2020). The aim of this systematic review is 2.2. Risk of bias assessment
to provide an update and synthesis on the literature relating to OC use
and cognitive performance, including both neurocognitive domains as Risk of bias assessment was conducted for all studies that met the
well as social-emotional cognition, as well as to identify factors that criteria for review and results are included in Tables 1-10. Studies were
modulate the associations between OCs and cognitive performance. given a score out of nine, with higher scores indicating a lower risk of
bias, on the basis of three categories – selection, comparability, and
2. Methods outcome. Scores of 0–4 were deemed “unsatisfactory”, scores 5–6 clas
sified as “satisfactory; scores 7–8 classified as “good” and a score of 9
The systematic review was performed in accordance with the was classified as “very good”. A modified version of the Newcastle -
PRISMA guidelines (Page et al., 2021) and registered on PROSPERO Ottawa Quality Assessment Scale (Wells et al., 2000) (adapted for cross-
(2021 CRD42021284393). Observational, experimental and rando sectional studies) was used (see supplementary material) and two au
mised trials evaluating cognitive impacts of OCs published in English thors (I.N. and C.G.) performed risk of bias ratings. Where there was a
after 1980 were considered. Articles were identified by searching the discrepancy between ratings, a third author (A.L.) provided input.
electronic databases Medline and PSYCInfo from 1980 to March 2022. Studies rated as having an unsatisfactory risk of bias rating were further
The search consisted of titles and abstracts and used search terms noted in the results section to aid interpretation of findings.
relating to oral contraceptives, including synonyms, and cognition,
including synonyms (see supplementary material for the full search 3. Results
strategy). A manual review of reference lists, including those of previous
systematic reviews on the topic, was also conducted. 3.1. Search results
2.1. Selection criteria for systematic review The search yielded a total of N = 3911 records, of which N = 450
duplicates were removed. After title and abstract screening, N = 3366
Studies were included if they involved (a) a group of premenopausal articles were discarded because they did not meet inclusion criteria. Two
females taking oral contraceptive medications at the time of evaluation studies were discarded as the full text was unable to be retrieved. The
[studies which included females taking ‘hormonal contraceptives’ that remaining N = 130 studies were assessed in full-text. Of these, N = 80
included oral contraceptives were also included] and (b) either a sepa articles were excluded as they did not meet the selection criteria, leaving
rate control group of naturally cycling women; a ‘control’ group of oral 50 articles that met inclusion criteria for the systematic review (Fig. 1).
contraceptive users who performed repeated cognitive evaluation in the Studies were grouped and are presented according to the cognitive
active versus inactive (i.e. ‘sugar pill’) phases; or a group of women who domain assessed: Attention, Working Memory, Psychomotor abilities,
performed cognitive testing before using an OC (i.e. with a natural Mental arithmetic; Visual-Spatial cognition, Visual Learning and Mem
menstrual cycle) and then repeated cognitive testing after commencing ory, Verbal abilities; Verbal Learning and Memory, Executive Func
an OC; and (c) at least one clinical measure of cognitive performance tioning and Decision-making, and Social Cognition (separated into
assessing one or more of the following domains: attention, memory, categories of emotion recognition, emotional memory and empathy).
executive function, verbal or visual-spatial abilities, social cognition or Where data was available, lifestyle and experimental factors that were
emotional memory / emotion recognition. reported to influence the association between OCs and cognition were
Articles were excluded for one or more of the following reasons (a) further summarised, either within each cognitive domain or more
retrospective studies based on reported historical oral contraceptive use, generally if the factors appeared relevant to more than one cognitive
(b) studies specifically involving women with a medical illness that may domain.
impact cognitive functioning, (c) studies involving administration of
additional medications or study conditions with potential for cognitive 3.2. Attention
impact, unless meaningful analysis of baseline evaluations was reported
or (d) studies conducted prior to 1980 (due to changes in OCs and the Seven studies examined different aspects of attention, including
potential that findings from older studies may not be applicable to the auditory-verbal attention span (digit span), reaction time, visual atten
OCs used today) as well as studies with a sample size of less than 10 in tion as well as more complex auditory divided attention. As seen in
the OC group where a lack of an OC effect may reflect the fact that the Table 1, six of the seven studies that examined different aspects of
study was underpowered (see (Hampson, 2022). attention did not report an effect of OCs on attention. This included four
The OC or intervention included in this review could include com studies comparing OC users to naturally cycling women (Gogos, 2013;
bined estrogen-progesterone contraceptive medication (e.g., combined Mordecai et al., 2008; Kuhlmann and Wolf, 2005; Mihalik et al., 2009)
oral contraceptive pills) or progesterone-only contraception. If the (of which one study had an unsatisfactory risk of bias rating), as well as
length of time on OCs was reported in the original study, this data was three studies that compared OC users during their active to their inactive
reported. Where available, the dose and characteristics of the estrogen pill phases (Mordecai et al., 2008; Gurvich et al., 2020; Mihalik et al.,
and progestin were included in the summary table. Where the progestin 2009) (including one study with an unsatisfactory risk of bias rating)
names were provided, they were broadly classified as ‘androgenic’ that and one study that employed a cross-over design (Silber et al., 1987).
is, progestins that are structurally related to testosterone, or “anti- Pletzer et al. (Pletzer et al., 2014) examined whether OC use effects
androgenic” (Regidor, 2018; Sitruk-Ware and Nath, 2013). OC pill types attentional selection of global versus local components of stimuli in a
were further discussed based on androgenic, estrogenic, progestogenic, traditional Navon paradigm. They reported that interference (which is
glucocorticoid or anti-glucocorticoid activity where this information the delayed rejection of stimuli displaying targets at the non-attended
was available. level) was significantly higher in 25 OC users as compared to 16 natu
Control groups were defined by the menstrual cycle phase they were rally cycling women (the difference was greatest between OC users and
tested in, where available. Where the days of the menstrual cycle were naturally cycling women in the luteal phase). The OC users in this study
provided, the details were included in the systematic review. were on pills containing adrongenically active first, second or third
Studies were first screened based on their title and abstract and then generation progestins. Hence, the existing evidence suggests OC use is
further examined in full text for suitability. Author I.N. performed the unlikely to influence auditory verbal attention, but it is possible that OC
initial screening and C.G. checked and confirmed studies for inclusion. use influences higher level aspects of visual attention.
2
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
3.3. Working memory influence working memory when the task also involves sustained
attention.
Seven studies examined visual or verbal working memory (Table 2)
using the N-back task, continuous performance tasks, digit span back
wards, subtraction tasks or matching tasks. Overall, 6 of the 7 studies did 3.4. Psychomotor abilities
not demonstrate significant group differences between OC users and
naturally cycling women (Kuhlmann and Wolf, 2005; Islam et al., 2008; Seven studies examined the relationship between OC use and psy
Komnenich et al., 1978; Rosenberg and Park, 2002; Vranic and Hro chomotor abilities using tasks including the grooved pegboard task,
matko, 2008) or between OC users in their active and inactive pill phase tapping and reaction time tasks as well as coding/digit symbol tasks and
(Gurvich et al., 2020; Herrera et al., 2020; Komnenich et al., 1978; letter cancellation tasks (Table 3). Beck et al. (Beck et al., 2008)
Rosenberg and Park, 2002; Vranic and Hromatko, 2008) on tasks of demonstrated an OC advantage (in a sample of 45 using a wide range of
working memory (three of these studies has unsatisfactory risk of bias OC types) in relation to psychomotor performance, using a grooved
ratings). Gravelsins et al. (Gravelsins et al., 2021) found that a sample of pegboard task when comparing OC users to naturally cycling women
57 OC users (taking OCs containing ethynyl estradiol with a 2nd; 3 rd or (but no pill phase effect was reported when comparing OC users between
4th generation progestin) had a significantly higher proactive behav active and inactive pill phases). In a sample of 14 women taking
ioural index for reaction time on a continuous performance task (AX- androgenic OCs, Becker et al. (Becker et al., 1982) reported an OC
CPT) than 62 naturally cycling women, which was suggested to indicate disadvantage (i.e. lower tapping rate) on a tapping task using a cross-
a better maintenance of goal-relevant information in the OC group. over design; however, this study was rated as unsatisfactory in the risk
Hence, OC use does not appear to influence working memory, but may of bias ratings. The other five studies did report any OC effect on psy
chomotor skills (one of these had a risk of bias rating of unsatisfactory).
3
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 1
Studies included in the systematic review of attention and OC use.
Author, year OC sample (n) OC type Control sample study TASK results Risk of bias
natuRally cycling design score,
(n) category
Gogos N = 16 EE (0.02 mg, N = 2; 0.03 mg, N = N = 27Tested in Cross- Digit span, Coding No effect of OC 7, Good
(2013) Participants were 12; 0.035 mg, N = 2), combined follicular phase sectional
on OC for at least 2 with LNG (0.1 mg, N = 2; 0.125 (days 2–8, N = 14), Between-
months prior to mg, N = 4; 0.15 mg, N = 6); CPA (2 or mid-luteal phase group
testing mg, N = 1); DNG (2 mg, N = 1); (days 19–25, N =
DRSP (3 mg, N = 1); or NET (0.5 13) a
mg, N = 1)
N = 14 males
Gurvich N = 35 Tested in N = 18 androgenic:EE N/A control group Between- One-back No effect of OC 8, good
et al. both active (0.02, 0.03 or 0.05 mg) combined was inactive pill group and
(2020) (days 7–10) and with LNG (0.1, 0.125 or 0.150 mg) phase within-
inactive (days 3–5) N = 17 anti-androgenic:EE group
pill phases (0.02, 0.03 or 0.035 mg) combined
with CPA (2 mg, N = 5), DNG (2
mg, N = 1, varied dose n-1), DRSP
(3 mg, N = 9)Estradiol
(1.5 mg) combined with NOMAC
(2.5 mg), N = 1*
Kuhlmann N = 20 Tested in EE (0.02–0.035 mg) and a N = 27Tested in Cross- Digit span No effect of OC 5, satisfactory
and Wolf active pill phase progesterone derivative luteal phase sectional
(2005) (days 7–14) (days − 4 to − 8 Between-
before menses, N = group
14) or follicular
phase (days 2 to 4,
N = 13) a
Mihalik et al. N = 24 Tested in Not stated N = 12Tested once Between- Computer No effect of OC 4,
(2009) both inactive in follicular phase group and administered choice unsatisfactory
(days 3-5a) and (days 3–5) and once within- reaction time, go/no-
active pill phases in luteal phase (days group go task, symbol
− 4 to − 6 before match task
menses) a
Mordecai N = 20 N = 12 triphasic N = 16Tested once Between- Brief Test of No effect of OC 6, satisfactory
et al. Participants were EE (0.025–0.040 mg) and a in follicular phase group and Attention – complex
(2008) on OC for at least 6 progestin (days 2–4) and once within- auditory divided
months prior to N = 8 monophasic in midluteal phase group attention.
testing EE (0.03–0.035 mg) and a (days 20–22) b
Tested in both progestin
inactive (days 2–4)
and active (days
20–22) pill phases
Pletzer et al. N = 25 N = 6 NETA; N = 8 LNG; N = 11 NG N = 16Tested in Between Global and local Luteal women 5, satisfactory
(2014) Tested in active DSG. either follicular group processing using showed lower
phase. phase (before Navon stimuli global
ovulation) or luteal advantage than
phase (after OC users
ovulation)
b
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; NET = norethisterone; NETA = norethindrone acetate; NG = norgestimate; DSG = desogestrel a determined using count approach, b cycle phase
confirmed with radioimmunoassay of plasma estradiol and/or progesterone levels,; * = results reported in relation to pill type.
The majority of studies indicate OC use does not affect psychomotor speed or accuracy of performance. Becker et al. (Becker et al., 1982)
performance. reported the number of correctly solved mental arithmetic problems was
lower with OC use in their cross-over design; although this study was
rated as unsatisfactory in the risk of bias ratings. Bradshaw et al.
3.5. Mental arithmetic
(Bradshaw et al., 2020) demonstrated that naturally cycling women
spent more time on mental arithmetic problems and solved more
Four studies examined the relationship between OC use and mental
problems correctly than their sample of 89 women taking hormonal
arithmetic task performance (Table 4). Only one of the four studies
contraceptives (the majority of this sample, n = 63, were taking OCs).
found no group difference between their sample of 16 OC users (OC type
Findings were interpreted to suggest that changes in perseverance
not reported) and 12 naturally cycling women on a counting span task
mediated the relationship between hormonal contraceptive use and
(Islam et al., 2008). The remaining studies found differences in either
4
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 2
Studies included in the systematic review of working memory and OC use.
Author, year OC sample OC type Control sample STUDY Task Results Risk of bias
natUrally cycling (n) DESIGN
Gurvich et al. N = 35 Tested in both active N = 18 androgenic:EE N/A control group was Between- Two-Back task No phase differences 8, good
(2020) (days 7–10) and inactive (0.02, 0.03 or 0.05 mg) combined inactive pill phase group and
(days 3–5) pill phases with LNG (0.1, 0.125 or 0.150 mg) within-
N = 17 anti-androgenic:EE group
(0.02, 0.03 or 0.035 mg) combined
with CPA (2 mg, N = 5), DNG (2 mg,
N = 1, varied dose n-1), DRSP (3 mg,
N = 9)Estradiol
(1.5 mg) combined with NOMAC
(2.5 mg), N = 1*
Gravelsins N = 57 N = 45 androgenicEE N = 62Tested once in Between- N-back, AX- No effect of OC or 7, good
et al. Participants were on OC for at (0.02 mg) combined with LNG (0.1 early follicular phase group and CPT, digit span time of pill ingestion
(2021) least 3 months prior to testing mg, N = 32); EE (0.03, 0.01 mg) (days 1–5) and once in within- and digit on N-back, digit span
Tested on two occasions – combined with LNG (0.15 mg, N = late follicular phase group ordering and digit ordering
1–2 h after pill ingestion and 1); EE (0.03 or 0.02 mg) combined (days 9–14) Y- Ax-CPT.
a
again, roughly 2 weeks later with DSG (0.15 mg, N = 7)EE OC users higher
approximately 24 h after pill (0.35 or 0.25 mg) combined with NG proactive control on
ingestion (0.18, 0.215, 0.25 mg, N = 5)EE AX-CPT.
(0.35 or 0.25 mg) combined with NG Interactions between
(0.25 mg, N = 3)EE OC use and COMT
(0.01 mg) combined with NETA (1 genotype
mg, N = 2)
N = 12 anti-androgenicEE
(0.035 mg) combined with CPA (2
mg, N = 4)EE
(0.03 mg) combined with DRSP (3
mg, N = 3)
Herrera et al. N = 20 Mixed HCs (N = 3 vaginal ring N/A control group was Within- N-back, with No phase differences 7, good
(2020) Participants were on OC for at included) inactive pill phase group and without
least 4 months prior to N = 16 androgenicEE stressor
testingTested in both inactive (0.02–0.035 mg) combined with:
(days 24–28) and active (days LNG (0.1 mg or 0.15 mg, N = 4);
8–21) pill phases NETA (0.1 mg or 0.15 mg, N = 5);
NET (0.1 mg, N = 1); ethynodiol
diacetate (0.1 mg, N = 1); DSG (0.15
mg, N = 2); or NG (0.25 mg, N = 3)
N = 1 anti-androgenicEE
(0.03 mg or 0.035 mg) combined
with DRSP (0.15 mg, N = 1)
N = 3 vaginal ringEE
(0.015 mg) combined with
etonogestrel (0.12 mg)
Islam et al. N = 16 Not stated N = 12Tested in early Between- Digit span No group differences 7, good
(2008) Participants were on OC for at luteal phase group and forward and
least 3 months prior to testing (1–4 days following within- back
Tested on two occasions ovulation) and menses group
during active pill phase (days 4–7) b,c
Kuhlmann N = 20 Tested during active EE (0.02 and 0.035 mg) and a N = 27Tested in luteal Between- Digit Span No group difference 5, satisfactory
and Wolf pill phase progesterone derivative phase group backwards
(2005) (days 7–14) (days − 4 to − 8 before
menses, N = 14) or
follicular phase (days
2 to 4, N = 13) a
Rosenberg N = 10 Not reported N=8 Between- Verbal working No group differences 3,
and Park Participants were on OC for at Tested on days 0, 7, 14 group and memory unsatisfactory
(2002) least 3 months prior to and 21c within-
testingTested in inactive (day group
0) and active (days 7, 14, 21)
pill phases
Vranic and N = 27 Tricyclic OCs N = 66Tested in early Between- Memory game No group or phase 3,
Hromatko N = 13 tested in active pill follicular group and – matching differences unsatisfactory
(2008) phase (weeks 1–3) (days 1–7, N = 29) or within- photos
N = 14 tested in inactive pill in luteal phase (N = group
phase 37) a
Note: HC = hormonal contraceptives; EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate;
CPA = cyproterone acetate; DRSP = drospirenone; NET = norethisterone; NETA = norethindrone acetate a determined using count approach; AX-CPT = continuous
performance task; a determined using count approach, b cycle phase confirmed with radioimmunoassay of plasma estradiol and progesterone levels, c cycle phase
determined by basal body temperature rise signifying ovulation luteal. * = results reported in relation to pill type (see Supplementary material for additional details
about Progestin types).
5
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 3
Studies included in the systematic review of psychomotor abilities and OC use.
Author, OC sample OC TYPE Control sample STUDY DESIGN Task RESULTS RISK OF BIAS
year natUrally cycling
(n)
Beck et al. N = 45 Tested in N = 18 monophasic N = 21 N = 46 Tested in Between-group Grooved OC were faster 7, good
(2008) inactive (days 3–7) or triphasic, with constant EE early follicular and within-group pegboard than naturally
active (days 11–14 or and three levels of phase (days 3–7), cycling. No phase
17–25) pill phases progestin; N = 2 triphasic peri-ovulatory effect reported.
Mixed HCs (N = 1 with constant progestin phase (days
patch) and three levels of EE; N = 11–14) and luteal
12 OCs but could not recall; phase (days
N = 1 patch contraceptive 17–25) a
Becker et al. N = 14 Tested every N = 14 androgenic EE 0.05 N = 14 Tested Cross-over and Reaction time, Lower tapping 4,
(1982) other day for one cycle mg with LNG 0.5 mg every other day for repeat measure tapping, line rate while taking unsatisfactory
one cycle tracing, OCs
Brown et al. N = 41 Tested on days Not stated N = 48 Tested on Between-group Psychomotor Mean 4,
(1984) 2, 8 or 14 of cycle days 2, 8 or 14 of reminiscence reminiscence unsatisfactory
cycle values remained
low in OC group
Genzel et al. N = 15 Participants N = 15 anti-androgenic EE N/A control group Within-group, Tapping task No phase 6, satisfactory
(2014) were on OC for at least (0.03 mg) combined with was inactive pill also exploring differences
2 months prior to DNG (2 mg) or CMA (2 mg) phase effect of sleep on Interaction with
testing Tested during off-line off-line
inactive and active consolidation consolidation
(second week of pill
cycle) pill phases
Gurvich N = 35 Tested in both N = 18 androgenic: EE N/A control group Between-group Chase test No group 8, good
et al. active (days 7–10) and (0.02, 0.03 or 0.05 mg) was inactive pill and within-group differences
(2020) inactive (days 3–5) pill combined with LNG (0.1, phase
phases 0.125 or 0.150 mg) N = 17
anti-androgenic: EE (0.02,
0.03 or 0.035 mg)
combined with CPA (2 mg,
N = 5), DNG (2 mg, N = 1,
varied dose n-1), DRSP (3
mg, N = 9) Estradiol (1.5
mg) combined with
NOMAC (2.5 mg), N = 1*
Kuhlmann N = 20 Tested in active EE (0.02–0.035 mg) and a N = 27 Tested in Between-group Letter No group 5, satisfactory
and Wolf pill phase (days 7–14) progesterone derivative luteal phase (days cancellation difference
(2005) − 4 to − 8 before task
menses, N = 14) or
follicular phase
(days 2 to 4, N =
13) a
Mihalik N = 24 Tested in both Not stated N = 12 Tested Between-group Visual motor No group or 4,
et al. inactive (days 3-5a) once in follicular and within-group processing phase differences unsatisfactory
(2009) and active pill phases phase (days 3–5) speed
and once in luteal (ImPACT)
phase (days − 4 to
− 6 before menses)
a
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; a ImPACT = The Immediate Postconcussion Assessment and Cognitive Test; a cycle phase confirmed with estrogen and progesterone metabolites from
urine samples or progesterone levels from blood samples, * = results reported in relation to pill type.
performance on maths problems. Pletzer et al. (Pletzer et al., 2014) 3.6. Visual-spatial abilities
found that OC users’ performance on a number bisection task, involving
mentally calculating the average of two numbers, was similar to natu Sixteen studies examined different aspects of visual-perceptual and
rally cycling women in the follicular phase, but differed from women in visual-spatial abilities (Table 5; with one study having an unsatisfactory
the luteal phase (this study was also rated as unsatisfactory in the risk of risk of bias rating). Twelve of these studies assessed mental rotation
bias ratings). Pletzer et al. (Pletzer et al., 2014) also looked at brain abilities and the majority of these studies did not report group differ
activation patterns (BOLD responses) while performing numerical tasks ences when all pill types were considered together; however, differences
and found OC users had different brain activation patterns to naturally emerged when pill type and/or menstrual cycle phase of the naturally
cycling women, that is OC users recruited different neural resources cycling women was considered (Bernal and Paolieri, 2022; Wharton
during mental calculations to naturally cycling women; however, OC et al., 2008; Griksiene and Ruksenas, 2011; Peragine et al., 2020).
users demonstrated a similar pattern of brain activation to men. More Griksiene et al. (Griksiene et al., 2018) was the only study that reported
specifically, in OC users (and in men), the default mode network deac differences at a group level between naturally cycling women and OC
tivation was more stable and less affected by modulations of the tasks users for mental rotation; however, this study only included women
than in naturally cycling women. In sum, it is possible OC use is asso using anti-androgenic OCs and results suggested that anti-androgenic
ciated with poorer mental arithmetic accuracy. OC users were less accurate in their mental rotation performance. Five
studies examined different aspects of visual attention and visual
perceptual skills. One study reported no group differences (Gogos, 2013)
6
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 4
Studies included in the systematic review of mental arithmetic and OC use.
Author, OC sample OC TYPE Control sample natUrally STUDY Task RESULTS RISK OF BIAS
year cycling (n) DESIGN
Becker N = 14 Tested every Androgenic EE N = 14 Tested every other Cross-over Mental OC < NC (fewer correct in 4,
et al. other day for one cycle (0.05 mg) with LNG day for one cycle and repeat arithmetic OC group) unsatisfactory
(1982) (0.5 mg) measure
Bradshaw N = 89 Participants N = 63 OC; N = 7 N = 86 Phase not reported Between- Mental NC > HC. (NC spent more 6, satisfactory
et al. were on OC for at least hormonal IUDs/skin group arithmetic time and solved more
(2020) 2 months prior to implants; N = 2 problem correctly than
testing Mixed HC (N = NuvaRing; N = 13 HCs)
63 OCs) data missing
Islam et al. N = 16 Participants Not reported N = 12 Tested in early Between- Counting span No group differences 8, good
(2008) were on OC for at least luteal phase (1–4 days group and
3 months prior to following ovulation) and within-
testing Tested on two menses (days 4–7) b,c group
occasions during active
pill phase
Pletzer N = 14 Participants Not stated N = 16 Tested once in Between Number No difference in reaction 3,
et al. tested in active phase. follicular phase (days 0 to group bisection and time on number unsatisfactory
(2014) 5 days before ovulation) number comparison task. OC-users
and once in midluteal comparison pattern of performance
phase (3 days post- tasks resembles follicular phase
ovulation to 5 days before but not luteal phase.
menses) c
Note: NC = naturally cycling women; HC = Hormonal contraceptives; EE = ethinyl estradiol; LNG = levonorgestrel; DRSP = drospirenone; a cycle phase confirmed
with estrogen and progesterone metabolites from urine samples or progesterone levels from blood samples,b cycle phase determined by day count and urine ovulation
tests; * = results reported in relation to pill type.
on figure copy and line orientation tasks between 16 OC users (taking difference only reached statistical significance for the maze recall task.
mixed OCs) and 27 healthy controls. In two studies with larger samples Griksiene et al. (Griksiene et al., 2018) specifically looked at women
sizes, OC users were less accurate on visual perceptual tasks, whereby using anti-androgenic OCs and found that this group of OC users were
OC users (n = 57) were reported to be less accurate on a visual- less accurate than naturally cycling women, although responded faster,
perceptual task than a sample of 86 naturally cycling women, task on a mental rotation task. Of note, not all studies report an advantage for
(McFadden, 2000) and 89 hormonal contraceptive users (63 taking OCs) androgenic OC users. Griksiene and Ruksenas (Griksiene and Ruksenas,
spotted less differences than 86 naturally cycling women on a ‘spot the 2011) found that third generation OC users (with androgenic profiles
difference’ task (Bradshaw et al., 2020). Spatial attention asymmetry that have lower androgenic profiles than second generation pills but are
during active and inactive OC pill phases was examined by Cicinelli et al. not considered anti-androgenic) demonstrated longer reaction time in
(Cicinelli et al., 2011) using a line bisection task in 36 OC users. A mental rotation compared to naturally cycling participants; newer
leftward bias on the line bisection task is typically observed in right- generation OC users (anti-androgenic profiles) did not differ from
handed healthy, naturally cycling women and this reflects efficient naturally cycling women in reaction time or accuracy.
interhemispheric communications. They found that this leftward bias Most recently, Beltz et al. (Beltz et al., 2021) examined estrogenic,
was emphasized in right-handed women during their active OC pill androgenic and progestational activity in OCs and their relationship
phase, compared to a rightward bias during the inactive pill phase. The with mental rotation and self-perceived masculinity. They found an in
authors suggested that OC use has a positive effect on interhemispheric verse association between OC estrogenic and progestational activity in
connections (Cicinelli et al., 2011). From the existing literature, it ap their sample. They also reported that androgenic activity of OCs was not
pears that several factors contribute to the associations between OC use a significant moderator of the masculinity-spatial skills relation. In OC
and visual-spatial abilities (outlined in section 3.6.1.). users with low estrogenic activity and high progestational activity, there
was a positive association between masculinity and spatial skills. In their
3.6.1. Contributing factors to associations between OC use and visual- earlier study, in relation to estrogen dose, Beltz et al. (Beltz et al., 2015)
spatial abilities demonstrated that an increased dose of ethinyl estradiol was associated
with decreased mental rotation score in a group of monophasic and
3.6.1.1. Pill type. Wharton et al. (Wharton et al., 2008) ranked hor triphasic OC users.
monal contraceptives by overall androgenic activity and a correlation
revealed a direct positive association between pill androgenic activity 3.6.1.2. Task factors – Angular disparity. Increased angular disparity is
and performance on the mental rotation task. Similarly, when indicative of increased difficulty in mental rotation. Griksiene et al.
comparing OC users grouped according to pill generation, whereby third (Griksiene et al., 2018) demonstrated that men and naturally cycling
generation progestins have less androgenic properties than second women show increased reaction times, and decreased accuracy with
generation progestins (Sitruk-Ware, 2006), those taking second gener increased angular disparity, indicating longer processing time for
ation progestins performed better than controls on a mental rotation greater mental rotation of objects. In contrast, for OC users there was no
task, but those taking third generation progestins (norgestimate pro association between angular disparity and accuracy or reaction time
gestin) with low androgenic activity did not perform differently to (Griksiene et al., 2018). This was in line with their earlier study, which
controls (Beltz et al., 2015). In line with this finding, Gurvich et al. demonstrated that OC users generated more correct responses than
(Gurvich et al., 2020) examined visual-spatial task performance in naturally cycling women when the task was relatively easy; however, as
relation to pill androgenicity and observed similar pattern of improved task difficulty increased, OC users demonstrated less correct responses
visual-spatial performance in OC users with androgenic progestins than naturally cycling women (Griksiene and Ruksenas, 2011).
compared to anti-androgenic OC users across all three visual-spatial
tasks assessed—mazes, shapes, and maze recall. However, the
7
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 5
Studies included in the systematic review of visual-spatial abilities and OC use.
Author, year OC sample OC TYPE Control sample STUDY Task RESULTS Risk of bias
naturally cycling (n) DESIGN
Beltz et al. N = 148 N = 52 monophasicEE N = 87 Between- Mental mOC (2nd generation 8, good
(2015) Tested anytime during (0.01–0.03 mg) combined Phase not reported group rotation progestins) performed
active or inactive with NETA (1–1.5 mg) better than naturally
phase N = 41 triphasicEE cycling; EE pill dose
(0.025–0.035 mg) combined significant predictor of
with NG 0.18, 0.215, 0.25 mg* performance.
Remaining participants had an
OC with EE and either a 2nd,
3rd or 4th generation
progestin, but were not
analysed separately
Beltz et al. N = 229 EE combined with LNG (N = N = 141 Between- Mental No OC versus control 8, good
(2021) Participants were on 15); NETA (N = 71); NG (N = Phase not reported group rotation test difference in mean levels.
OC for at least 3 71); DRSP (N = 45); other Pill type moderated
months prior to testing progestins (N = 18) relation between self-
Coded and analysed according perceived masculinity
to estrogenic, progestational, and mental rotation skills
and androgenic activity
Bradshaw N = 89 N = 63 OC; N = 86 Between- Spot the Naturally cycling spent 6, satisfactory
et al. Participants were on N = 7 hormonal IUDs/skin Phase not reported group difference more time and spotted a
(2020) OC for at least 2 implants; greater number of
months prior to N = 2 NuvaRing; differences than HC
testingMixed HC N = 13 data missing Nuva
(N = 63 OCs) ring; n = 13 data missing
Cicinelli N = 36 N = 36EE N/A control group Visual line Y Changes in spatial 7, good
et al. Participants were on (0.02 mg) combined with was inactive pill bisection attention asymmetry in
(2011) OC for at least 6 DRSP (3 mg) phase OC use
months prior to testing
Gogos N = 16 EE (0.02 mg, N = 2; 0.03 mg, N = 27Tested in Cross- Figure copy No group differences 7, good
(2013) Participants were on N = 12; 0.035 mg, N = 2), follicular phase sectional and line
OC for at least 2 combined with LNG (0.1 mg, (days 2–8, N = 14), Between- orientation
months prior to testing N = 2; 0.125 mg, N = 4; 0.15 or mid-luteal phase group combined
mg, N = 6); CPA (2 mg, N = 1); (days 19–25, N =
DNG (2 mg, N = 1); DRSP (3 13) a
mg, N = 1); or NET (0.5 mg, N
= 1) N = 14 males
Gordon and N = 34 Combined estrogen/ N = 34Tested in Between- Mental No group or phase 5, satisfactory
Lee (1993) Tested three times, progesterone menses (days 2 or group and rotation and differences
days 2 or 3; days 3), follicular phase within- visual
10–14 and days 20–24 (days 10–14) and group perceptual
luteal phase (days tasks
20–24) b
Gurvich N = 35Tested in both N = 18 androgenic:EE N/A control group Between- Shapes; Mazes No group differences 8, good
et al. active (0.02, 0.03 or 0.05 mg) was inactive pill group and
(2020) (days 7–10) and combined with LNG (0.1, phase within-
inactive (days 3–5) pill 0.125 or 0.150 mg) group
phases N = 17 anti-androgenic:EE
(0.02, 0.03 or 0.035 mg)
combined with CPA (2 mg, N
= 5), DNG (2 mg, N = 1,
varied dose n-1), DRSP (3 mg,
N = 9)Estradiol
(1.5 mg) combined with
NOMAC (2.5 mg), N = 1*
Griksiene N = 22 N = 11 androgenic:EE N = 20Tested in Between- Mental No group or phase effects 8, good
and Participants were on (0.02–0.035 mg) combined follicular phase group and rotation test overall. NC responded
Ruksenas OC for at least 3 with gestodene, N = 6); DSG (days 2–5), luteal within- faster than androgenic
(2011) months prior to testing (N = 4); or NG (N = 1) phase (about 6 days group group
Tested 3 times to align N = 11 anti-androgenic:EE post-ovulation) and
with naturally cycling (0.02–0.035 mg) combined ovulatory phase
with DRSP (N = 10); or DNG
(N = 1)*
Griksiene N = 35 N = 35 anti-androgenic:EE N = 34Tested in Between- Mental NC were significantly 8, good
et al. Participants were on combined with DRSP follicular phase group rotation test more accurate than OC
(2018) OC for at least 3 (N = 23); CMA (N = 5); DNG (N = 13) or luteal group. Reaction time was
months prior to testing (N = 3); CPA (N = 2); or phase (n = 21) b, faster in OC than NC
estradiol combined with phases combined in group.
NOMAC (N = 2) analysis
Islam et al. N = 16 Not stated N = 12Tested in Between- Mental No group or phase effects 8, good
(2008) Participants were on early luteal phase group and rotation
OC for at least 3 (1–4 days following within-
months prior to testing ovulation) and group
Tested on two
(continued on next page)
8
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 5 (continued )
Author, year OC sample OC TYPE Control sample STUDY Task RESULTS Risk of bias
naturally cycling (n) DESIGN
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; NG = norgestimate; a determined using count approach with ovulation predictor kits or with a period tracking app, b cycle phase confirmed with
immunoassay of plasma estradiol and progesterone levels or saliva levels of free 17β-estradiol, free progesterone and free testosterone; * results reported in relation to
progestin class.
9
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 6
Studies included in the systematic review of visual learning and OC use.
Author, OC sample OC TYPE Control sample STUDY Task RESULTS Risk of bias
year naturally cycling DESIGN
(n)
Bianchini N = 26Tested during Triphasic N = 25Tested Between- Walking Corsi Significant Group 7, good
et al. active (days 20 – 21) and during early group test – effect: OC users
(2018) inactive follicular phase topographic learned the path in
(days 2–4) (days 4–5) and memory fewer trials than
Taking pill for at least 6 mid-luteal phase controls. No phase
months (days 20–21) effect
Genzel N = 15 N = 15 anti-androgenicEE N/A control Within- Procedural No phase differences 7, good
et al. Participants were on OC (0.03 mg) combined with DNG group was group motor learning
(2014) for at least 2 months prior (2 mg) or CMA (2 mg) inactive pill phase
to testingTested during
inactive and active (second
week of pill cycle)
pill phases
Gurvich N = 35Tested in both N = 18 androgenic: EE N/A control Between- Maze recall Effect of pill class on 8, good
et al. active (0.02, 0.03 or 0.05 mg) group was group and maze recall
(2020) (days 7–10) and inactive combined with LNG (0.1, 0.125 inactive pill phase within-
(days 3–5) pill phases or 0.150 mg) group
N = 17 anti-androgenic:EE
(0.02, 0.03 or 0.035 mg)
combined with CPA (2 mg, N =
5), DNG (2 mg, N = 1, varied
dose n-1), DRSP (3 mg, N = 9)
Estradiol
(1.5 mg) combined with
NOMAC (2.5 mg), N = 1*
Islam et al. N = 16 Not stated N = 12Tested in Between- Delayed No group or phase 8, good
(2008) Participants were on OC early luteal phase group and matching to differences
for at least 3 months prior (1–4 days within- sample
to testing following group
Tested on two occasions ovulation) and
during active pill phase menses (days
4–7) b,c
Mordecai N = 20 N = 12 triphasic N = 16Tested in Between- BVMT-R No group or phase 6, satisfactory
et al. Participants were on OC EE (0.025–0.04 mg) and a early follicular group and effects
(2008) for at least 6 months prior progestin phase within-
to testing N = 8 monophasic (days 2–4) and group
Tested in both inactive EE (0.03–0.035 mg) and a midluteal phase
(days 2–4) and active progestin (days 20–22) b
(days 20–22) pill phases
Mihalik N = 24Tested in both Not stated N = 12Tested Between- Visual memory No group differences 4,
et al. inactive once in follicular group and (ImPACT) unsatisfactory
(2009) (days 3-5a) and active pill phase within-
phases (days 3–5) and group
once in luteal
phase (days − 4 to
− 6 before
menses) a
Silber et al. N = 20 N = 11 monophasic EE N = 20 Cross-over Pattern memory No effect of OC 6, satisfactory
(1987) (0.03 mg) combined with LNG Tested in luteal test; Benton test
(0.15 mg, N = 7); EE (0.05 mg) phase a
combined with LNG (0.25 mg,
N = 3); or EE (0.05 mg)
combined with lynestrenol (2.5
mg, N = 1);
N = 9 triphasic EE
(0.03–0.04 mg) combined with
LNG (0.05–0.12 mg)
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; NG = norgestimate; BVMT-R Brief Visuospatial Memory Test Revised, ImPACT Immediate Postconcussion Assessment and Cognitive Test a determined
using count approach with ovulation predictor kits or with a period tracking app, b cycle phase confirmed with immunoassay of plasma estradiol and progesterone
levels or saliva levels of free 17β-estradiol, free progesterone and free testosterone; * results reported in relation to progestin class.
3.8.1. Contributing factors to associations between OC use and verbal generation progestin. The authors proposed that the greater progesto
abilities pill type genic activity associated with second generation progestins may be the
Progestin generation appeared to influence verbal abilities in the two driver of the improved fluency performance. Griksiene and Ruksenas
studies that specifically examined pill type. Beltz et al. (Beltz et al., 2015) (Griksiene and Ruksenas, 2011) also examined pill type and found that
did not find differences between naturally cycling women and OC users third generation (with androgenic activity) OC users generated signifi
on an expressional fluency task, but within the OC group, they found cantly less words (using a total combined scores for letter and semantic
that the users of monophasic OCs containing a second generation pro fluency) compared to naturally cycling and newer generation (anti-
gestins performed better than users of triphasic OCs containing a third androgenic) OC users.
10
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 7
Studies included in the systematic review of verbal abilities and OC use.
Author, year OC sample OC TYPE Control sample STUDY Task RESULTS Risk of bias
naturally cycling DESIGN
(n)
Beltz et al. N = 148 N = 52 monophasic EE N = 87 Between- Making mOC (2nd generation 8, good
(2015) Tested anytime (0.01–0.03 mg) combined with Phase not group sentences progestins) performed better
during active or NETA (1–1.5 mg) reported than tOC (3rd generation
inactive phase N = 41 triphasic EE progestins) but neither
(0.025–0.035 mg) combined with differed from naturally
NG 0.18, 0.215, 0.25 mg* cycling
Remaining participants had an
OC with EE and either a 2nd, 3rd
or 4th generation progestin, but
were not analysed separately
Bradshaw N = 89 N = 63 OC; N = 86 Between- Solvable and Naturally cycling women 6,
et al. Participants were N = 7 hormonal IUDs/skin Phase not group unsolvable spent more time on solvable satisfactory
(2020) on OC for at least 2 implants; reported anagrams and unsolvable anagrams,
months prior to N = 2 NuvaRing; but no differences in accuracy
testingMixed HC N = 13 data missing Nuva ring; n
(N = 63 OCs) = 13 data missing
Gogos N = 16 EE (0.02 mg, N = 2; 0.03 mg, N = N = 27Tested in Cross- RBANS No group differences 7, good
(2013) Participants were 12; 0.035 mg, N = 2), combined follicular phase sectional language
on OC for at least 2 with LNG (0.1 mg, N = 2; 0.125 (days 2–8, N = Between- tasks
months prior to mg, N = 4; 0.15 mg, N = 6); CPA 14), or mid-luteal group
testing (2 mg, N = 1); DNG (2 mg, N = 1); phase (days
DRSP (3 mg, N = 1); or NET (0.5 19–25, N = 13) a
mg, N = 1)
N = 14 males
Griksiene N = 22 N = 11 androgenic:EE N = 20Tested in Between- Verbal Third generation 8, good
and Participants were (0.02–0.035 mg) combined with follicular phase group and fluency (androgenic) OC users
Ruksenas on OC for at least 3 gestodene, N = 6); DSG (N = 4); (days 2–5), luteal within- generated significantly less
(2011) months prior to or NG (N = 1) phase (about 6 group words compared to naturally
testing N = 11 anti-androgenic:EE days post- cycling and new generation
Tested 3 times to (0.02–0.035 mg) combined with ovulation) and (anti-adrogenic) OC users
align with naturally DRSP (N = 10); or DNG (N = 1)* ovulatory phase
cycling
Mordecai N = 20 N = 12 triphasic N = 16Tested in Between- Verbal No group or phase differences 6,
et al. Participants were EE (0.025–0.040 mg) and a early follicular group and fluency satisfactory
(2008) on OC for at least 6 progestin phase within-
months prior to N = 8 monophasic (days 2–4) and group
testing EE (0.03–0.035 mg) and a midluteal phase
Tested in both progestin (days 20–22) b
inactive (days 2–4)
and active (days
20–22) pill phases
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; NG = norgestimate; RBANS – Repeatable Battery for the Assessment of Neuropsychological Status; EE = ethinyl estradiol; a determined using count
approach with ovulation predictor kits or with a period tracking app, b cycle phase confirmed with immunoassay of plasma estradiol and progesterone levels or saliva
levels of free 17β-estradiol, free progesterone and free testosterone.
3.9. Verbal learning and recall advantage on learning and recall of stories as compared to a naturally
cycling group in their peri-ovulatory phase (but the OC group did not
Ten studies examined the influence of OCs on verbal learning and differ to naturally cycling women in their follicular or luteal phases).
recall using word lists, verbal paired associate and story memory Hence, it is possible OCs provide a verbal memory advantage in relation
(Table 8). Four studies (Mordecai et al., 2008; Gurvich et al., 2020; to story memory.
Mihalik et al., 2009; Genzel et al., 2014) examined whether pill phase
effects verbal learning and recall; and only one of these studies reported
a phase effect (Mordecai et al., 2008), whereby OC users performed 3.10. Executive functions and decision making
better on a verbal list learning task during their active, as compared to
inactive phase. The remaining studies did not show a difference on Two studies examined executive function skills and two studies
verbal learning between active and inactive pill phases. Eight studies examined decision making (Table 9).
examined between-group effects of the OC and six of these studies did Of the two studies that examined executive function skills, one study
not find group differences on word list or paired associative word lists explored inhibitory control on a Go-NoGo task using a small (n = 15 per
(Mordecai et al., 2008; Wharton et al., 2008; Kuhlmann and Wolf, 2005; arm) randomized, placebo-controlled trial (Gingnell et al., 2016).
Mihalik et al., 2009; Silber et al., 1987; Plamberger et al., 2021) Findings showed that participants randomized to OCs significantly
(including one study with an unsatisfactory rating scale). Gogos (Gogos, improved their performance between the baseline and end of the single
2013) reported that their OC group scored higher on immediate and treatment cycle (Gingnell et al., 2016). This study also investigated brain
delayed recall using a score that combined story memory and word lists activity during response inhibition using event-related functional mag
memory, as compared to a group of males and a naturally cycling group netic resonance imaging (fMRI) and found that during the treatment
in follicular phase (but no difference to naturally cycling group in their cycle OC users displayed decreased activity in the right middle frontal
luteal phase). Peragine et al. (Peragine et al., 2020) reported an OC gyrus in comparison with placebo users, overall suggesting OC use may
represent reduced effort or increased efficacy in maintaining inhibitory
11
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 8
Studies included in the systematic review of verbal learning and recall and OC use.
Author, year OC sample OC TYPE Control sample STUDY Task RESULTS Risk of bias
naturally cycling DESIGN
(n)
Genzel et al. N = 15 N = 15 anti-androgenic EE N/A control group Within- Verbal paired No phase differences 7, good
(2014) Participants were on (0.03 mg) combined with was inactive pill group associate task
OC for at least 2 months DNG (2 mg) or CMA (2 mg) phase
prior to testingTested
during inactive and
active (second week of
pill cycle)
pill phases
Gogos (2013) N = 16 EE (0.02 mg, N = 2; 0.03 mg, N = 27Tested in Cross- RBANS OC group scored 7, good
Participants were on N = 12; 0.035 mg, N = 2), follicular phase sectional immediate and higher on immediate
OC for at least 2 months combined with LNG (0.1 mg, (days 2–8, N = 14), Between- delayed recall for and delayed recall as
prior to testing N = 2; 0.125 mg, N = 4; 0.15 or mid-luteal phase group word lists and compared to a group of
mg, N = 6); CPA (2 mg, N = (days 19–25, N = stories males and the
1); DNG (2 mg, N = 1); DRSP 13) a naturally cycling
(3 mg, N = 1); or NET (0.5 group in follicular
mg, N = 1) N = 14 males phase.
Gurvich et al. N = 35Tested in both N = 18 androgenic: EE N/A control group Between- Word list recall No phase differences 8, good
(2020) active (0.02, 0.03 or 0.05 mg) was inactive pill group and
(days 7–10) and combined with LNG (0.1, phase within-
inactive (days 3–5) pill 0.125 or 0.150 mg) group
phases N = 17 anti-androgenic: EE
(0.02, 0.03 or 0.035 mg)
combined with CPA (2 mg,
N = 5), DNG (2 mg, N = 1,
varied dose n-1), DRSP (3
mg, N = 9) Estradiol
(1.5 mg) combined with
NOMAC (2.5 mg), N = 1*
Kuhlmann N = 20Tested in active EE (0.02–0.035 mg) and a N = 27 Tested in Between- Word list No group difference 5, satisfactory
and Wolf pill phase progesterone derivative luteal phase group immediate and
(2005) (days 7–14) (days − 4 to − 8 cued recall
before menses, N =
14) or follicular
phase (days 2 to 4,
N = 13) a
Mordecai N = 20 N = 12 triphasic N = 16 Tested in Between- Word List - CVLT Significant phase 6, satisfactory
et al. Participants were on EE (0.025–0.04 mg) and a early follicular group and effect, better verbal
(2008) OC for at least 6 months progestin phase within- learning in OC active
prior to testing N = 8 monophasic (days 2–4) and group compared to inactive
Tested in both inactive EE (0.03–0.035 mg) and a midluteal phase phase.
(days 2–4) and active progestin (days 20–22) b
(days 20–22) pill
phases
Mihalik et al. N = 24Tested in both Not stated N = 12 Tested once Between- Verbal memory – No group of phase 4,
(2009) inactive in follicular phase group and word memory, differences unsatisfactory
(days 3-5a) and active (days 3–5) and once within- symbol match and
pill phases in luteal phase group three letters
(days − 4 to − 6 (ImPACT)
before menses) a
Peragine N = 55 N = 55 monophasic, N = 140 Tested in Between- Logical Memory OC group who took OC 7, good
et al. N = 25 who had taken estrogen and androgenic early follicular group (stories) WMS after cognitive testing
(2020) OC prior to testing progestin ohase had better verbal
N = 30 who had taken (days 3–5, N = 27); memory than
OC after testing ovulatory phase ovulatory group.
(days 10–14, N =
56); and mid-luteal
phase (7–10 days
prior to menses, N
= 57) a
Plamberger N = 19 EE with DNG or LNG N = 43 Tested in Between- Word list and No effect of OC 8, good
et al. follicular phase group word pair task
(2021) (N = 16) or luteal
phase (N = 27)b
Silber et al. N = 20 N = 11 monophasic EE N = 20 Cross-over Associative No effect of OC 6, satisfactory
(1987) (0.03 mg) combined with Tested in luteal Learning test
LNG (0.15 mg, N = 7); EE phase a
(0.05 mg) combined with
LNG (0.25 mg, N = 3); or EE
(0.05 mg) combined with
lynestrenol (2.5 mg, N = 1);
N = 9 triphasicEE
(continued on next page)
12
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 8 (continued )
Author, year OC sample OC TYPE Control sample STUDY Task RESULTS Risk of bias
naturally cycling DESIGN
(n)
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; NG = norgestimate;RBANS – Repeatable Battery for the Assessment of Neuropsychological Status; CVLT = California Verbal Learning Test, WMS LM =
Logical Memory subtest of the Wechsler Memory Scale, DRM = Deese–Roediger–McDermott task, VLMT = Verbal Learning and Memory Task, ImPACT = Immediate
Postconcussion Assessment and Cognitive Test.
EE = ethinyl estradiol; a determined using count approach with ovulation predictor kits or with a period tracking app, b cycle phase confirmed with immunoassay of
plasma estradiol and progesterone levels or saliva levels of free 17β-estradiol, free progesterone and free testosterone; * results reported in relation to progestin class.
cognitive control (Gingnell et al., 2016). The second study examining was, however, a pattern of improved performance in OC users with
executive function used a set-shifting task and did not observe any OC androgenic compared to anti-androgenic progestins on this maze task.
effects on task performance when comparing OC users during their Two studies examined decision making using the Decision Making Task,
active and inactive (i.e. sugar pill) phases (Gurvich et al., 2020). There an adaptation of the Iowa Gambling Test. One study did not report any
Table 9
Studies included in the systematic review of executive functions and decision making and OC use.
Author, OC sample (n) OC Details Control sample STUDY DESIGN Task RESULTS RISK OF
year natUrally cycling (n) BIAS
Note: EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC = nomegestrol acetate; CPA = cyproterone acetate; DRSP =
drospirenone; NG = norgestimate; DMT = Decision Making Task; an adaptation of the Iowa Gambling Test; * results reported in relation to progestin class.
13
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
group differences (Hamstra et al., 2017) and the second study found OC (both positive and negative) in a picture memory task for hormonal
use was associated with a reduction of risky decisions in trials with a low contraceptive users (of which 83 % were taking OCs) as compared to
probability to win high gains (Hamstra et al., 2015). women who were not taking hormonal contraceptive.
3.11. Social cognition – Emotion recognition and emotional memory 3.11.3. Emotional interference
Scheuringer et al. (Scheuringer et al., 2020) conducted a placebo
Within the social cognition domain, 20 studies were included that controlled RCT to determine whether OC use influenced emotional
examined OC use in relation to emotion recognition, emotional memory interference by negative and positive stimuli and found no interaction
as well as empathy (Table 10). These will be reported separately. between performance on the emotional verbal Stroop and OC use,
indicating that OCs did not evoke any differences in emotional inter
3.11.1. Emotion recognition ference to the three word categories.
Thirteen studies examined emotion recognition using either the
Reading the Mind in the Eyes Task, the Facial Expression Recognition 3.11.4. Empathy
Task, the Vienna Emotion Recognition Task or an experimental emotion Three studies examined whether OC use impacted empathy using the
recognition task. Five studies did not find OC group effects (Gamsa Multifaceted Empathy Test. None of the studies found an effect of OC use
khurdashvili et al., 2021; DeSoto and Gear, 2007; Radke and Derntl, on cognitive empathy; however, OC use was associated with lower af
2016; Shirazi et al., 2020; Gamsakhurdashvili et al., 2021) and Gurvich fective empathy in Strojny et al. (Strojny et al., 2021). Two separate
et al. (Gurvich et al., 2020) did not find a difference in a social-emotion studies [6; 62] reported that OC users demonstrated a negativity bias for
recognition task between the active and inactive pill phase, but they did affective empathy-related ratings, that is, they gave higher affective
report lower accuracy in a social emotional task for OC users with OCs empathy ratings towards people displaying negative emotion compared
containing an anti-androgenic progestin as compared to OC users with to people displaying positive emotions.
OCs containing androgenic progestins. The remaining findings were
mixed. In their large sample of hormonal contraceptive users (n = 520, 3.11.5. Gender of protagonist
of whom 83 % were using OCs), Spalek et al. (Spalek et al., 2019) re Two studies explored the gender of the person displaying emotions
ported that women using hormonal contraceptives rated the valence of during emotion recognition or empathy tasks (Gamsakhurdashvili et al.,
negative pictures as more negative and positive images as more positive 2021; Hamstra et al., 2014). Gamsakhurdashvili et al. (Gamsa
than women not using hormonal contraceptives. Hamstra and col khurdashvili et al., 2021) found that women taking OCs and naturally
leagues have conducted a series of studies where they have explored cycling women in their late luteal phase performed better at emotion
interactions between emotion recognition, OC use and mineralocorti recognition if the protagonist in the photo is female compared to male,
coid receptor genes (this is further discussed in section 3.12.v). In terms whereas naturally cycling women in the middle of their cycle (i.e.
of broader OC effects, Hamstra et al. (Hamstra et al., 2014) reported that around ovulation) perform equally well in emotion recognition for male
OC users recognized fewer expressions of anger, sadness and disgust, and female faces. Hamstra et al. (Hamstra et al., 2014) reported an
with faster reaction times on correctly recognized sadness and distrust interaction between OC use and the gender of the protagonist in the
trials; however, there were no differences between OC users and natu photo, whereby the recognition of disgust was better in female faces by
rally cycling women on recognizing happy emotions. The same group women who were not using OCs.
partially replicated this finding in two different samples, demonstrating
OC users to have poorer recognition of anger compared to naturally 3.11.6. Interactions with affective state.
cycling women (Hamstra et al., 2015) and in another study OC users to Only one study examined the interacting role of affect on emotion
be quicker to recognize anger and happiness as well as recognize fewer recognition (Kimmig et al., 2021). For OC users (and not naturally
expressions of happiness and sadness compared to naturally cycling cycling women) there was an inverse association between negative af
women (Hamstra et al., 2017). Pahnke et al. (Pahnke et al., 2018) re fective state and neutral face recognition of OC users, whereby the worse
ported that women using OCs were less accurate than naturally cycling their affective state, the more likely women misclassified a neutral
women when recognizing emotions, regardless of valence. They also expression as sadness or anger.
reported that emotion recognition was poorer in OC users, compared to
naturally cycling women, with particularly complex expressions 3.12. Additional factors noted to influence the OC and cognition
(Pahnke et al., 2018). relationship
14
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 10
Studies included in the systematic review of social cognition (including emotional memory and emotion recognition) and OC use.
Author, year OC sample (n) OC type Control sample study Task results Risk of bias
natUrally cycling (n) design score,
category
15
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 10 (continued )
Author, year OC sample (n) OC type Control sample study Task results Risk of bias
natUrally cycling (n) design score,
category
Kimmig et al. (2021) N = 30 Not stated N = 56 Tested in early Between- Short VERT-K No overall OC effect on 7, good
Participants were follicular phase group accuracy or response
on OC for at least (N = 30) or peri- time. Group by Emotion
6 months prior to ovulatory phase (N = interaction. In the OC
testingTested 26)b group only, negative
during active pill affective state was
phase associated with lower
(days 3–21 of recognition accuracy and
active pill intake) longer response times for
neutral faces
Merz (2017) N = 30 Monophasic N = 60 Tested in early Between- Recall of positive, No OC main effect. 6, satisfactory
Participants were EE (0.02–0.035 mg) follicular phase group negative and OC Group by Emotion ×
on OC for at least combined with (days 4–10, n = 30) or neutral words; stress (cortisol)
3 months prior to androgenic progestins luteal phase (3–9 days interaction with interaction.
testing LNG or DSG; or anti- before onset of next stress
Tested during androgenic progestins menstruation, n =
active phase CPA, DNG, DRSP or 30)b
CMA
Progestin types
included but numbers
not stated
Nielsen et al. (2011) N = 34 N = 5 triphasic N = 32 Between- Emotional story No overall effect of OC 6, satisfactory
Participants were N = 29 monophasicEE Phase not reported group recall Recall of gist versus
on OC for at least (0.015–0.035 mg), details differed between
1 month prior to progestin details not NC and OC groups
testing stated
Nielsen et al. (2013) N = 36 N = 11 triphasic N = 42 Tested in Between- Free recall of No overall OC effect of 6, satisfactory
Participants were N = 25 monophasic follicular phase group emotional images emotional image recall.
on OC for at least EE (0.015–0.03 mg), (N = 15) or luteal following a stress Recall of emotional
1 month prior to no progestin details phase (N = 27)a test images differed between
testing stated NC and OCs depending
Tested in either on stress response
active or inactive
phase, specific
phase at time of
testing was not
recorded
Nielsen et al. (2014) N = 49 N = 8 triphasic N = 60Tested in luteal Between- Free recall of OC users had blunted 5, satisfactory
Tested in either N = 41 monophasic phase group emotional stories stress hormone responses
active or inactive EE combined with a (days 15–30) and altered retention of
b
phase, specific progestin information from an
phase at time of emotional event
testing was not compared to NC
recorded
Pahnke et al. (2018) N = 42 N = 15 multiphasic N = 53 Tested in Cross- RMET OC users less accurate in 7, good
Includes OC users N = 27 monophasic follicular phase sectional emotion recognition than
in either active or Both androgenic (N = (days 0–14, N = 35) between- NC.No
inactive pill phase 21, including LNG, or luteal phase (days group differences between
– number in each DSG, NG) and anti- 15–28, N = 18) androgenic and anti-
b
phase not stated androgenic (N = 21, androgenic progestins
including CMA, CPA,
DNG, DRSP) progestin
types included*
Person and Oinonen N = 57 N = 14 multiphasic N = 39 Between- Emotional Spatial No overall Group 5, satisfactory
(2020) Participants were N = 43 monophasic Participants had not group Memory Test differences
on OC for at least No further detail used OC in 2 months
2 months prior to stated regarding prior to testingTested OC users had a higher
testing oestrogen or progestin in either menstrual, ratio of positive to
components periovulatory or negative items compared
luteal phase with nonusers
(numbers not stated)a
(continued on next page)
16
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Table 10 (continued )
Author, year OC sample (n) OC type Control sample study Task results Risk of bias
natUrally cycling (n) design score,
category
Radke and Derntl N = 55 Monophasic N = 18 Between- Short VERT-K, No Group differences on 6, satisfactory
(2016) Participants were Estrogen details not Phase not reported group perspective emotion recognition or
on OC for at least stated taking, affective perspective-taking. OC
3 months prior to In the active pill phase responsiveness users in active pill phase
testingTested in group: N = 13 had increased accuracy
active androgenic progestins for affective
(N = 30) or (LNG), N = 17 anti- responsiveness compared
inactive (N = 25) androgenic progestins to inactive pill phase.
pill phase (DNG, DRSP, CMA)
In the inactive pill
phase group: N = 9
androgenic progestins
(LNG, DSG, NET), N =
16 anti-androgenic
progestins (DNG,
DRSP, CMA, CPA)
Scheuringer et al. N = 36 Estradiol (1.5 mg) N = 32 Between- Emotional Stroop No effect of OC 8, good
(2020) Randomized to combined with Randomized to group, RCT
OC group NOMAC (2.5 mg) placebo groupTested
Took OC for 3 in follicular phase
consecutive 24/4 (n = 20) or in luteal
treatment cycles, phase (n = 12)b
then tested at end
of these 3 cycles
Shirazi et al. (2020) N = 203 OC brands and EE dose N = 192 Between- RMET No group differences 7, good
Tested twice, 1 recorded for only N = Tested in both group
week apart 125 participants periovulatory phase
Included both and luteal phasea
monophasic and
triphasic formulations
Progestin details not
stated
Spalek et al. (2019) N = 520 HC Not recorded N = 1649 Between Picture rating HC rated valence of 7, good
83 % using an OC, Cycle phase not Group (valence rating) negative and positive as
reported and Picture more extreme than NC.
Memory HC recalled more
emotional (positive and
negative)
images than NC.
Strojny et al. (2021) N = 38 Monophasic N = 45 Tested in Between- Multifaceted NC higher emotional 8, good
EE (less than0.05 mg) either follicular phase group Empathy Test empathy than OC. No
Progestin details not (n = 19), luteal phase effect on cognitive
stated (n = 18) or empathy
periovulatory phase
(n = 6)
Phase was unclear for
n = 2 participants due
to irregular cycle.a
Notes: OC – oral contraceptive; NC - naturally cycling; EE = ethinyl estradiol; LNG = levonorgestrel; CPA = cyproterone acetate; DNG = dienogest; NOMAC =
nomegestrol acetate; CPA = cyproterone acetate; DRSP = drospirenone; NG = norgestimate;; RCT = randomized-controlled trial; RMET = Reading the Mind in the
Eyes Task, WMT = Word categorization and memory task, EMT = Emotional categorization and memory task; FERT = Facial expression recognition task; VERT-K =
Vienne Emotion Recognition Task; IPANAT = Implicit Positive and Negative Affect test a determined using count approach with ovulation predictor kits or with a
period tracking app, b cycle phase confirmed with immunoassay of plasma estradiol and progesterone levels or saliva levels of free 17β-estradiol, free progesterone and
free testosterone;* results reported in relation to progestin class.
significantly higher level of memory performance and motor perfor 2020). Results demonstrated similar n-back working memory perfor
mance from baseline to retest (four hours post learning session), led the mance after stress in both active and inactive phases; however, larger
authors to suggest that OC use may enhance off-line memory consoli changes in cortisol from before to after stress exposure predicted better
dation. Plamberger et al. (Plamberger et al., 2021) examined whether working memory performance in the active pill phase, but not the
memory consolidation overnight differed between OC and naturally inactive pill phase.
cycling women and reported women using OCs showed a highly sig Three studies examined stress effects on word retrieval tasks. Kuhl
nificant change in memory performance overnight (whereas naturally mann and Wolf (Kuhlmann and Wolf, 2005) investigated whether
cycling women in the follicular phase did not). cortisol administration differentially affected memory retrieval in
women using or not using OCs. OC users did not show a cortisol-induced
3.12.3. Cortisol interactions and role of stress retrieval impairment, as compared to naturally cycling women who did
Seven studies investigated the interacting effects of cortisol admin show a significant impairment in memory retrieval following cortisol
istration or stress on cognitive performance in the context of OC use. One administration. Espin et al. (Espin et al., 2013) also did not show any
study investigated whether stress exposure differentially affected difference in performance on a verbal memory task in OC users who
working memory during active versus inactive OC phases (Herrera et al., were exposed to a pre-learning psychosocial stressor (Trier Social Stress
17
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Test) as compared to OC users not exposed to the psychosocial stressor domains examined (with the exception of basic auditory attention and
(of note naturally cycling women also failed to show an effect of psy psychomotor performance). Extending previous research, this system
chosocial stress on memory performance; however, men did improve atic review identified several factors that appear to influence the asso
their memory performance under stress). Merz (Merz, 2017) reported ciation between OC use and cognition. These include within domain,
that pre-encoding stress enhanced recall of neutral (but not negative or task related factors; characteristics of the ‘control’ group including the
positive words) in women taking OCs, as compared to naturally cycling menstrual cycle phase of a naturally cycling women versus comparing
women and men. active and inactive pill phases; genetic factors; the role of stress, and
Two studies investigated emotional memory and stress. Nielsen et al. variation in the types of OCs, specifically related to the androgenic,
(Nielsen et al., 2013) investigated emotional memory in the context of estrogenic and progestogenic actions of the OC.
stress hormone release (cortisol and norepinephrine) in response to a
Cold Pressor Stress. Results showed that compared to naturally cycling 4.1. Task related factors
women, OC users had blunted stress hormone responses to emotional
images and a Cold Pressor Stress. Stress hormone (norepinephrine) One area where OCs may differ in how they influence cognitive
release at encoding and post-training stress exposure (and associated performance within a domain is whether the task involves a semantic
cortisol release) was also differentially associated with emotional component. Within the verbal memory domain, there was a pattern of
memory, as well as recall for the gist versus detail of emotional stories, findings that suggested OC use to have a benefit on story memory, in the
depending on OC status (Nielsen et al., 2013; Nielsen et al., 2014). absence of a clear benefit on list memory. In the emotional memory
Unlike naturally cycling women, who experience enhanced recall of domain, there was some evidence that recall type (recall of details versus
details and gist from an emotional story if they experience a post- recall of gist) may explain some of the heterogeneity at the domain level
learning stressor, OC users did not demonstrate an effect of post- and more specifically, recall of gist (i.e. semantic themes) was associated
learning cortisol release on memory for gist or peripheral details from with a positive effect of OC use. Hence, a possible pattern is that OCs
emotional stories (Nielsen et al., 2014). may have more of an influence on semantic tasks. In relation to the
neural correlates of verbal word list versus more semantic based mem
3.12.4. Timing of OC ingestion ory, studies have suggested the former tasks rely more on the integrity of
Two studies reported on the timing of OC ingestion relative to executive functions and frontal lobes, whereas semantic tasks rely more
cognitive performance. Peragine et al. (Peragine et al., 2020) investi on the integrity of the temporal lobes (Riello et al., 2021; Vonk et al.,
gated time of OC ingestion on a mental rotation task and a story memory 2019). This differential association with types of memory has also been
task. There was no effect of timing of OC ingestion for story memory, but discussed in the context of sex differences and associations with estra
for mental rotation, OC users who had ingested an OC within two hours diol (Schultheiss et al., 2021); hence future research should at a mini
prior to performing the mental rotation task showed poorer performance mum report sub-scores for letter or list-based tasks, as compared to
relative to males. In contrast, OC users who were tested during their semantic tasks.
active pill phase, but had not ingested their OC just prior to cognitive Task complexity might also be a relevant factor to consider in the
testing did not differ from males. Gravelsins et al. (Gravelsins et al., association between OC use and cognition. In the context of mental
2021) examined timing of pill ingestion on working memory perfor rotation, increased angular disparity is indicative of increased difficulty.
mance and failed to see a difference when comparing working memory Griksiene and colleagues (Griksiene and Ruksenas, 2011; Griksiene
1–2 h and 24 h after pill ingestion. et al., 2018) have demonstrated that men and naturally cycling women
show increased reaction times, and decreased accuracy with increased
3.12.5. Genetic factors angular disparity, indicating longer processing time for greater mental
Three studies examined how genetic factors interact with OC use and rotation of objects. In contrast, for OC users there was no association
cognitive performance. Hamstra et al. (Hamstra et al., 2015) demon between angular disparity and accuracy or reaction time, which led the
strated an interaction between OC use and mineralocorticoid receptors authors to suggest OC users were utilizing different strategies to com
(MR). Specifically, carriers of MR haplotype 1 or 3 are more sensitive to plete the task to men and naturally cycling women. That is, rather than
the impact of OCs on the recognition of sad and fearful faces and on relying solely on the imagined rotation of an object, OC users may
emotional memory, whereas carriers of MR haplotype 2 were not implement other kinds of analytical strategies to complete more chal
(Hamstra et al., 2015). In line with these findings, OC users with MR lenging mental rotation tasks. Griksiene and colleagues further sug
haplotype 1/3 performed worse in the facial expression recognition task gested that the OC use may result in changes in brain activation in
(using a total score of combined emotions) than MR haplotype 1/3 regions relating to either decision making (increased activation in pre
carriers in the mid-luteal phase of the menstrual cycle (Hamstra et al., frontal regions) or mental rotation (suppressed activation of parietal
2016). regions) and this may underpin differences in the approach to tasks.
An interaction between OC use and the single nucleotide poly Our systematic review also identified one study that suggested that
morphism (SNP, Val158Met, rs4680) of the gene coding for catechol-o- OC use may alter one’s approach to a cognitive task, specifically sug
methyltransferase (COMT; the protein that degrades dopamine at syn gesting OC users performed tasks with less perseverance, as measured by
apses in the prefrontal cortex) was demonstrated in relation to working time on task and associated diminished task performance, on simple
memory performance (Gravelsins et al., 2021). Specifically, OC users visual-spatial tasks as well as more complex mental arithmetic and
demonstrated a more pronounced effect of COMT, whereby met/met verbal word problem solving (anagrams) tasks (Bradshaw et al., 2020).
carriers (associated with higher prefrontal dopamine concentrations) The authors suggested decreased perseverance on simple and complex
had significantly higher accuracy or working memory performance than tasks associated with OC use may reflect decreased ability to exercise
val/val (lower prefrontal dopamine concentrations); this COMT effect cognitive control during problem solving (Bradshaw et al., 2020).
was not observed in naturally cycling women. Hence, findings across studies involving assessment of task difficulty or
approaches to task suggest that OC use may influence cognitive control
4. Discussion and decision making processes. Imaging findings suggest OC use may
alter both the structure and function of the brain in regions relevant to
Findings from this systematic review indicates that OC use does not decision making, cognitive control and potentially motivation, such as
clearly improve or impair cognitive performance at a domain level, the amygdala, prefrontal cortex and cingulate gyrus (Bronnick et al.,
consistent with previous research (Warren et al., 2014); however, OC 2020). Task difficulty has also been described as the most important
use was associated with differences in performance on all cognitive modulating factor in the association between menstrual cycle related sex
18
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
hormone fluctuations and cognitive performance (Bernal and Paolieri, working memory performance. Duration of OC use in OC users, as well
2022). There is a need for further investigative work on how OC use as whether there had been previous OC use in the naturally cycling
influences ones’ approach to a task, by measuring task difficulty, as well groups was either not reported or, where reported, the information was
as exploring factors such as how OC use influences cognitive control, not detailed enough to be considered in a meaningful way in this review.
decision making, motivation and perseverance. Structural brain changes in grey matter volume have been associated
with duration of OC use and grey matter volumes in the hippocampus
4.2. Control group related factors have been related to duration of previous OC use (Pletzer et al., 2015).
Several studies used within-group designs to examine active versus
Findings from our systematic review identified variability in the 1) inactive pill phases on cognitive performance. While most studies did
the menstrual cycle phase of the control sample and 2) the reliability of not report phase effects, there was one report of an OC advantage in the
measurement of menstrual cycle phase. This is relevant because the active versus inactive phase for verbal list learning (Mordecai et al.,
control group provides the reference group for assessing OC related ef 2008) and one report of an OC advantage in the active versus inactive
fects on cognition and while there was not a consistent pattern of find phase for affective responsiveness, that is, a measure of emotion
ings, there was evidence to suggest menstrual cycle phase is an recognition. Gingnell et al. (Gingnell et al., 2016) also reported an
important consideration. For example, OC users’ performance on verbal improvement in inhibitory control performance in women during an
learning and memory as well as mental rotation demonstrated the active pill phase as compared to a baseline assessment that was con
greatest difference to a peri-ovulatory group (Peragine et al., 2020), in ducted during a pre-treatment cycle (i.e. no OC). Hence, it remains
comparison to the follicular or luteal phase groups. Research on neu unclear whether there is a time dependent effect of OC use on cognitive
rocognitive changes across the menstrual cycle is not consistent performance and this should be investigated in future research.
(Sundström-Poromaa and Litwack, 2018; Le et al., 2020), with the most Different OC formulations have differing estrogenic, androgenic and
robust findings in mental rotation performance, where performance is progestogenic profiles, depending on the dose and type of estrogen, the
potentially better in the early follicular phase (with lower estradiol dose and type of progestin, which differentially interact with the pro
levels) as compared to the mid-luteal phase (Bernal and Paolieri, 2022; gesterone receptor as well as other steroid hormone receptors, including
Peragine et al., 2020). Research relating to emotional processing across the androgen receptor, the estrogen receptor, the glucocorticoid recep
the menstrual cycle is more consistent, with poorer emotion recognition tor or the mineralocorticoid receptor (Sitruk-Ware, 2006), as well as the
accuracy and enhanced emotional memory during the luteal phase, regime (monophasic, biphasic, triphasic or quadriphasic) (De Leo et al.,
potentially driven by changes in progesterone levels (Le et al., 2020; I. 2016). Our systematic review findings suggested pill type is likely to
Sundstrã¶M Poromaa, and M. Gingnell, 2014). Menstrual cycle research influence the association between OC use and cognitive performance.
often compares performance during the early follicular phase (low Positive associations or group differences indicated that OC pills with
estradiol / low progesterone) and the mid-luteal phase (high estradiol / higher androgenic activity were associated with better performance on
high progesterone), with less research examining cognitive performance visual-spatial tasks, particularly the mental rotation task (Wharton et al.,
during the peri-ovulatory phase (high estradiol / low progesterone): it 2008; Gurvich et al., 2020; Griksiene et al., 2018; Beltz et al., 2015)
has been recently suggested to understand the influence of sex hormones (with the exception of Griksiene and Ruksenas (Griksiene and Ruksenas,
on cognition across the menstrual cycle it is important to include at least 2011) as well as better performance on facial affect discrimination
these three phases (Bernal and Paolieri, 2022). Similarly, study designs (Gurvich et al., 2020). While few studies have examined other cognitive
comparing OC users to naturally cycling women should at the very least domains, it has been reported that verbal learning and memory, atten
accurately report menstrual cycle phase of the control group and ideally tion, executive function and psychomotor ability may not be related to
compare OC use to different menstrual cycle phases. Furthermore, of the OC androgenicity (Gurvich et al., 2020). Ethinyl estradiol dose has also
studies that did report cycle phase, many studies relied on self-report or been inversely associated with mental rotation (Beltz et al., 2015).
count methods to confirm cycle phase (i.e. counting days from the onset Androgenic and anti-androgenic progestins have been demonstrated to
of menses). Only some studies used menstrual cycle tracking apps differentially modulate brain structure (Pletzer et al., 2015). It is likely
together with examination of hormone levels from blood or saliva that different OC formulations differentially influence cognitive per
samples to confirm cycle phase. As recommended by Hampson formance in a task dependent manner and further investigation is
(Hampson, 2020), future studies should confirm cycle phase of the needed to better understand this area.
control group measured by diaries or menstrual cycle tracking apps
together with examination of hormone levels from blood or saliva 4.4. Lifestyle and experimental factors that may modulate the OC-
samples. cognition association
4.3. OC related factors One lifestyle and/or experimental factor that may contribute to how
OCs influence cognition is the role of stress. In their recent meta-analysis,
Our systematic review identified several factors relating to the type Gervasio et al. (Gervasio et al., 2022) reported that OC users, relative to
and timing of OC use that may modulate the association between OCs naturally cycling women, have a blunted cortisol response in response to
and cognitive performance. Timing of pill ingestion, relative to cognitive psychosocial stress as well as a higher baseline level of cortisol. The
assessment, and length of time people have been taking OCs were not authors suggested it is possible that OC use may induce a physiological
reported in many studies. OC formulations containing ethinyl estradiol state that has similarities to chronic stress and this, in addition to the
reach peak plasma levels, similar to plasma concentrations of estradiol blunted stress response, may have subsequent influences on aspects of
during the mid-follicular ranges, within two hours after ingestion and cognition. Our systematic review findings suggest that stress exposure
taper to low levels thereafter (Stanczyk et al., 2013). Of the two studies has a differential influence on cognitive performance in OC users, rela
that have considered timing of OC ingestion relative to cognitive per tive to naturally cycling women in the areas of working memory (Her
formance, Peragine et al. (Peragine et al., 2020) found no effect of timing rera et al., 2020); word retrieval (Kuhlmann and Wolf, 2005; Merz,
of OC ingestion for story memory, but did find an effect of OC timing for 2017) and emotional memory (Nielsen et al., 2013; Nielsen et al., 2014)
mental rotation. OC users who had ingested their pill within two hours (although not all studies reported different responses in OC users who
of the mental rotation task showed poorer performance relative to were exposed to pre-learning psychosocial stressor (Espin et al., 2013).
males, but those who had a longer time between OC ingestion and the Kuhlmann and Wolf (Kuhlmann and Wolf, 2005) investigated whether
mental rotation task did not differ from males. Gravelsins et al. (Grav cortisol administration differentially affected memory retrieval in
elsins et al., 2021) did not see an influence of timing of pill ingestion on women using or not using oral contraceptives. Results indicated that OC
19
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
users appear to be less sensitive to acute cortisol elevations than natu and cognitive reserve/intelligence on cognitive performance (this is
rally cycling women. Hence, it appears that OC use may diminish stress particularly relevant for between-subjects research on cognition).
effects on some cognitive tasks, particularly in relation to memory. Future research also needs to consider that cognition and emotion pro
Our systematic review findings also included two studies that found cessing may fluctuate across the menstrual cycle in the absence of the
OC use affects sleep spindles and memory consolidation (Genzel et al., pill (i.e. including peri-ovulatory studies) and therefore take into
2014; Plamberger et al., 2021). There is extensive research demon consideration the best ‘control’ group(s) for between-subject designs.
strating the positive effects of sleep on memory consolidation (e.g. While the ideal approach to studying the effects of OCs on cognition is a
(Klinzing et al., 2019) and some preliminary research suggesting OC use within-subject design that would involve comparing women to them
may influence sleep (Baker et al., 2019). Our systematic review findings selves on and off OCs, this design is complicated and time-consuming. A
suggest in women using OCs, offline, sleep-dependent memory consol better understanding of how OCs influence cognitive performance and
idation may be improved relative to naturally cycling women. This ef the factors that modulate this association will not only advance our
fect appears to be observed when OC users are compared to women in general understanding of how (synthetic) sex hormones influence
the follicular phase, but not the luteal phase of their menstrual cycle, cognition, but also help provide information so females can make
and it is suggested that the beneficial effects of endogenous or synthetic informed decisions about contraceptive choices.
sex hormones on memory consolidation are mediated via the effects of Author Contributions;
progesterone (or progestin) on sleep (Plamberger et al., 2021). Future CG and I.N.: initial draft. A.L: analysis. All authors: interpretation of
research should further explore these potentially beneficial effects of OC data, revision, final approval, and agreement to be accountable for all
use on memory consolidation. aspects of the work.
4.5. Individual factors that may modulate the OC-cognition association Declaration of Competing Interest
Our systematic review identified three studies that suggest genetic The authors declare that they have no known competing financial
factors may interact with OC use to influence cognitive performance. A interests or personal relationships that could have appeared to influence
common functional mineralocorticoid receptors (MR) haplotype block, the work reported in this paper.
based on two single nucleotide polymorphisms: MR2G/C (rs2070951)
and MR-I180V (A/G, rs5522) (van Leeuwen et al., 2011) appears to Data availability
interact with OC use to influence emotional information processing.
Carriers of the MR haplotype 1 and 3 appear to be more sensitive to the No data was used for the research described in the article.
depressogenic impact of OCs and demonstrate better recall of negative
information and better recognition of sadness and fear in facial ex Appendix A. Supplementary data
pressions, compared to carriers of MR haplotype 2 (Hamstra et al., 2015;
Hamstra et al., 2014; Hamstra et al., 2016). An interaction between OC Supplementary data to this article can be found online at https://doi.
use and the single nucleotide polymorphism (Val158Met, rs4680) of the org/10.1016/j.yfrne.2022.101052.
gene coding for catechol-o-methyltransferase (COMT; the protein that
degrades dopamine at synapses in the prefrontal cortex) has also been References
demonstrated in relation to working memory performance, whereby OC
users demonstrate a more pronounced effect of COMT on working Baker, F.C., Sattari, N., de Zambotti, M., Goldstone, A., Alaynick, W.A., Mednick, S.C.,
2019. Impact of sex steroids and reproductive stage on sleep-dependent memory
memory performance (Gravelsins et al., 2021). Research exploring how consolidation in women. Neurobiol Learn Mem 160, 118–131.
genetic factors interact with OC use to influence cognition is still in its Beck, K.D., McLaughlin, J., Bergen, M.T., Cominski, T.P., Moldow, R.L., Servatius, R.J.,
infancy and future studies are needed in this field. 2008. Facilitated acquisition of the classically conditioned eyeblink response in
women taking oral contraceptives. Behav Pharmacol 19, 821–828.
Becker, D., Creutzfeldt, O.D., Schwibbe, M., Wuttke, W., 1982. Changes in physiological,
5. Conclusions and future directions EEG and psychological parameters in women during the spontaneous menstrual
cycle and following oral contraceptives. Psychoneuroendocrinology 7, 75–90.
Beltz, A.M., Moser, J.S., 2020. Ovarian hormones: a long overlooked but critical
There are several limitations that should be taken into account when contributor to cognitive brain structures and function. Ann N Y Acad Sci 1464,
interpreting the findings of this systematic review, including the limited 156–180.
number of studies that reported details on pill type; the studies that Beltz, A.M., Hampson, E., Berenbaum, S.A., 2015. Oral contraceptives and cognition: A
role for ethinyl estradiol. Hormones and Behavior 74, 209–217.
either failed to report menstrual cycle phase of naturally cycling women, Beltz, A.M., Loviska, A.M., Kelly, D.P., Nielson, M.G., 2021. The Link Between
or collapsed menstrual cycle phase for the purpose of analysis; the Masculinity and Spatial Skills Is Moderated by the Estrogenic and Progestational
number of studies that reported total scores for cognitive domains Activity of Oral Contraceptives. Front Behav Neurosci 15, 777911.
Bernal, A., Paolieri, D., 2022. The influence of estradiol and progesterone on
without reporting sub-domain results (e.g. collapsing letter fluency and
neurocognition during three phases of the menstrual cycle: Modulating factors.
semantic fluency), limiting the capacity to draw conclusions about task Behavioural Brain Research 417, 113593.
specific factors. Our overall findings suggest that OCs may improve Bianchini, F., Verde, P., Colangeli, S., Boccia, M., Strollo, F., Guariglia, C., Bizzarro, G.,
cognitive performance within the domains of verbal memory and vi Piccardi, L., 2018. Effects of oral contraceptives and natural menstrual cycling on
environmental learning. BMC Womens Health 18, 179.
suospatial abilities, as well as possibly decrease performance in relation Bimonte-Nelson, H.A., Bernaud, V.E., Koebele, S.V., 2021. Menopause, hormone therapy
to mental arithmetic accuracy and emotion recognition. However, there and cognition: maximizing translation from preclinical research. Climacteric 24,
are several factors that may modulate the influence of OCs on cognition. 373–381.
Bradshaw, H.K., Mengelkoch, S., Hill, S.E., 2020. Hormonal contraceptive use predicts
These included task related factors, such as reporting results at a task decreased perseverance and therefore performance on some simple and challenging
level (not domain level) and considering the role of task difficulty. cognitive tasks. Horm Behav 119, 104652.
Control group factors, such as considering and accurately reporting the Bronnick, M.K., Okland, I., Graugaard, C., Bronnick, K.K., 2020. The Effects of Hormonal
Contraceptives on the Brain: A Systematic Review of Neuroimaging Studies. Front
menstrual cycle phase of the naturally cycling group. OC related factors, Psychol 11, 556577.
including the duration of OC use, the timing of OC ingestion and the Brown, I.S., Forand, A.Q., Payne, R.B., 1984. Hormonal influences on psychomotor
likely differential impact of OCs on cognition depending on the OC reminiscence. Percept Mot Skills 58, 383–389.
Cicinelli, E., De Tommaso, M., Cianci, A., Colacurci, N., Rella, L., Loiudice, L.,
formulations and the subsequent differential estrogenic, androgenic and Cicinelli, M.V., Livrea, P., 2011. Oral contraceptive therapy modulates hemispheric
progestogenic profile. Future research should also consider individual asymmetry in spatial attention. Contraception 84, 634–636.
factors across both control and OC groups, such as the potential inter
acting or indirect effects of genetic polymorphisms, sleep, stress, mood
20
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
De Leo, V., Musacchio, M.C., Cappelli, V., Piomboni, P., Morgante, G., 2016. Hormonal Mihalik, J.P., Ondrak, K.S., Guskiewicz, K.M., McMurray, R.G., 2009. The effects of
contraceptives: pharmacology tailored to women’s health. Hum Reprod Update 22, menstrual cycle phase on clinical measures of concussion in healthy college-aged
634–646. females. J Sci Med Sport 12, 383–387.
DeSoto, C., Gear, D.C., 2007. Investigating the Role of Hormones in Theory of Mind. Mordecai, K.L., Rubin, L.H., Maki, P.M., 2008. Effects of menstrual cycle phase and oral
North American Journal of Psychology 9, 535–544. contraceptive use on verbal memory. Horm Behav 54, 286–293.
Espin, L., Almela, M., Hidalgo, V., Villada, C., Salvador, A., Gomez-Amor, J., 2013. Acute Nielsen, S.E., Ertman, N., Lakhani, Y.S., Cahill, L., 2011. Hormonal contraception usage
pre-learning stress and declarative memory: impact of sex, cortisol response and is associated with altered memory for an emotional story. Neurobiol Learn Mem 96,
menstrual cycle phase. Horm Behav 63, 759–765. 378–384.
Gamsakhurdashvili, D., Antov, M.I., Lubke, K.T., Pause, B.M., Stockhorst, U., 2021. The Nielsen, S.E., Segal, S.K., Worden, I.V., Yim, I.S., Cahill, L., 2013. Hormonal
role of olfaction and sex-hormone status in empathy-related measures. Physiol Behav contraception use alters stress responses and emotional memory. Biol Psychol 92,
230, 113289. 257–266.
Gamsakhurdashvili, D., Antov, M.I., Stockhorst, U., 2021. Sex-hormone status and Nielsen, S.E., Ahmed, I., Cahill, L., 2014. Postlearning stress differentially affects memory
emotional processing in healthy women. Psychoneuroendocrinology 130, 105258. for emotional gist and detail in naturally cycling women and women on hormonal
Gamsakhurdashvili, D., Antov, M.I., Stockhorst, U., 2021. Facial Emotion Recognition contraceptives. Behav Neurosci 128, 482–493.
and Emotional Memory From the Ovarian-Hormone Perspective: A Systematic Page, M.J., McKenzie, J.E., Bossuyt, P.M., Boutron, I., Hoffmann, T.C., Mulrow, C.D.,
Review. Front Psychol 12, 641250. Shamseer, L., Tetzlaff, J.M., Akl, E.A., Brennan, S.E., Chou, R., Glanville, J.,
Genzel, L., Baurle, A., Potyka, A., Wehrle, R., Adamczyk, M., Friess, E., Steiger, A., Grimshaw, J.M., Hrobjartsson, A., Lalu, M.M., Li, T., Loder, E.W., Mayo-Wilson, E.,
Dresler, M., 2014. Diminished nap effects on memory consolidation are seen under McDonald, S., McGuinness, L.A., Stewart, L.A., Thomas, J., Tricco, A.C., Welch, V.A.,
oral contraceptive use. Neuropsychobiology 70, 253–261. Whiting, P., Moher, D., 2021. The PRISMA 2020 statement: an updated guideline for
Gervasio, J., Zheng, S., Skrotzki, C., Pachete, A., 2022. The effect of oral contraceptive reporting systematic reviews. BMJ 372, n71.
use on cortisol reactivity to the Trier Social Stress Test: A meta-analysis. Pahnke, R., Mau-Moeller, A., Junge, M., Wendt, J., Weymar, M., Hamm, A.O.,
Psychoneuroendocrinology 136, 105626. Lischke, A., 2018. Oral Contraceptives Impair Complex Emotion Recognition in
Gingnell, M., Bannbers, E., Engman, J., Frick, A., Moby, L., Wikström, J., Sundström- Healthy Women. Front Neurosci 12, 1041.
Poromaa, I., 2016. The effect of combined hormonal contraceptives use on brain Patel, S.A., Frick, K.M., Newhouse, P.A., Astur, R.S., 2022. Estradiol effects on spatial
reactivity during response inhibition. Eur J Contracept Reprod Health Care 21, memory in women. Behav Brain Res 417, 113592.
150–157. Peragine, D., Simeon-Spezzaferro, C., Brown, A., Gervais, N.J., Hampson, E., Einstein, G.,
Gogos, A., 2013. Natural and synthetic sex hormones: effects on higher-order cognitive 2020. Sex difference or hormonal difference in mental rotation? The influence of
function and prepulse inhibition. Biol Psychol 93, 17–23. ovarian milieu. Psychoneuroendocrinology 115, 104488.
Gordon, H.W., Lee, P.A., 1993. No difference in cognitive performance between phases of Person, B., Oinonen, K.A., 2020. Emotional Memory in Oral Contraceptive Users:
the menstrual cycle. Psychoneuroendocrinology 18, 521–531. Negative Stimuli Are More Forgettable. Psychol Rep 123, 2282–2304.
Gravelsins, L., Duncan, K., Einstein, G., 2021. Do oral contraceptives affect young Plamberger, C.P., Van Wijk, H.E., Kerschbaum, H., Pletzer, B.A., Gruber, G.,
women’s memory? Dopamine-dependent working memory is influenced by COMT Oberascher, K., Dresler, M., Hahn, M.A., Hoedlmoser, K., 2021. Impact of menstrual
genotype, but not time of pill ingestion. PLoS One 16, e0252807. cycle phase and oral contraceptives on sleep and overnight memory consolidation.
Griksiene, R., Ruksenas, O., 2011. Effects of hormonal contraceptives on mental rotation J Sleep Res 30, e13239.
and verbal fluency. Psychoneuroendocrinology 36, 1239–1248. Pletzer, B., Kronbichler, M., Nuerk, H.C., Kerschbaum, H., 2014. Hormonal
Griksiene, R., Monciunskaite, R., Arnatkeviciute, A., Ruksenas, O., 2018. Does the use of contraceptives masculinize brain activation patterns in the absence of behavioral
hormonal contraceptives affect the mental rotation performance? Horm Behav 100, changes in two numerical tasks. Brain Res 1543, 128–142.
29–38. Pletzer, B., Petasis, O., Cahill, L., 2014. Switching between forest and trees: opposite
Gurvich, C., Warren, A.M., Worsley, R., Hudaib, A.R., Thomas, N., Kulkarni, J., 2020. relationship of progesterone and testosterone to global-local processing. Horm Behav
Effects of Oral Contraceptive Androgenicity on Visuospatial and Social-Emotional 66, 257–266.
Cognition: A Prospective Observational Trial. Brain Sci 10. Pletzer, B., Kronbichler, M., Kerschbaum, H., 2015. Differential effects of androgenic and
Hampson, E., 2020. A brief guide to the menstrual cycle and oral contraceptive use for anti-androgenic progestins on fusiform and frontal gray matter volume and face
researchers in behavioral endocrinology. Horm Behav 119, 104655. recognition performance. Brain Research 1596, 108–115.
Hampson, E., 2022. Oral contraceptives in the central nervous system: Basic Radke, S., Derntl, B., 2016. Affective responsiveness is influenced by intake of oral
pharmacology, methodological considerations, and current state of the field. Front contraceptives. Eur Neuropsychopharmacol 26, 1014–1019.
Neuroendocrinol 68, 101040. Regidor, P.A., 2018. The clinical relevance of progestogens in hormonal contraception:
Hamstra, D.A., De Rover, M., De Rijk, R.H., Van der Does, W., 2014. Oral contraceptives Present status and future developments. Oncotarget 9, 34628–34638.
may alter the detection of emotions in facial expressions. Eur Riello, M., Frangakis, C.E., Ficek, B., Webster, K.T., Desmond, J.E., Faria, A.V., Hillis, A.
Neuropsychopharmacol 24, 1855–1859. E., Tsapkini, K., 2021. Neural Correlates of Letter and Semantic Fluency in Primary
Hamstra, D.A., de Kloet, E.R., van Hemert, A.M., de Rijk, R.H., Van der Does, A.J., 2015. Progressive Aphasia. Brain sciences 12, 1.
Mineralocorticoid receptor haplotype, oral contraceptives and emotional Rosenberg, L., Park, S., 2002. Verbal and spatial functions across the menstrual cycle in
information processing. Neuroscience 286, 412–422. healthy young women. Psychoneuroendocrinology 27, 835–841.
Hamstra, D.A., de Kloet, E.R., Tollenaar, M., Verkuil, B., Manai, M., Putman, P., Van der Sahlberg, B.-L., Landgren, B.-M., Axelson, M., 1987. Metabolic profiles of endogenous
Does, W., 2016. Mineralocorticoid receptor haplotype moderates the effects of oral and ethynyl steroids in plasma and urine from women during administration of oral
contraceptives and menstrual cycle on emotional information processing. contraceptives. Journal of Steroid Biochemistry 26, 609–617.
J Psychopharmacol 30, 1054–1061. Scheuringer, A., Lundin, C., Derntl, B., Pletzer, B., Sundstrom Poromaa, I., 2020. Use of
Hamstra, D.A., de Kloet, E.R., Quataert, I., Jansen, M., Van der Does, W., 2017. an estradiol-based combined oral contraceptives has no influence on attentional bias
Mineralocorticoid receptor haplotype, estradiol, progesterone and emotional or depressive symptoms in healthy women. Psychoneuroendocrinology 113,
information processing. Psychoneuroendocrinology 76, 162–173. 104544.
Herrera, A.Y., Velasco, R., Faude, S., White, J.D., Opitz, P.C., Huang, R., Tu, K., Schultheiss, O.C., Köllner, M.G., Busch, H., Hofer, J., 2021. Evidence for a robust,
Mather, M., 2020. Brain activity during a post-stress working memory task differs estradiol-associated sex difference in narrative-writing fluency. Neuropsychology 35,
between the hormone-present and hormone-absent phase of hormonal 323–333.
contraception. Neurobiol Stress 13, 100248. Shirazi, T.N., Rosenfield, K.A., Cardenas, R.A., Breedlove, S.M., Puts, D.A., 2020. No
Islam, F., Sparkes, C., Roodenrys, S., Astheimer, L., 2008. Short-term changes in evidence that hormonal contraceptive use or circulating sex steroids predict complex
endogenous estrogen levels and consumption of soy isoflavones affect working and emotion recognition. Horm Behav 119, 104647.
verbal memory in young adult females. Nutr Neurosci 11, 251–262. Silber, M., Almkvist, O., Larsson, B., Stock, S., Uvnas-Moberg, K., 1987. The effect of oral
Kimmig, A.S., Bischofberger, J.A., Birrenbach, A.D., Drotleff, B., Lammerhofer, M., contraceptive pills on levels of oxytocin in plasma and on cognitive functions.
Sundstrom-Poromaa, I., Derntl, B., 2021. No Evidence for a Role of Oral Contraception 36, 641–650.
Contraceptive-Use in Emotion Recognition But Higher Negativity Bias in Early Sitruk-Ware, R., 2006. New progestagens for contraceptive use. Hum Reprod Update 12,
Follicular Women. Front Behav Neurosci 15, 773961. 169–178.
Klinzing, J.G., Niethard, N., Born, J., 2019. Mechanisms of systems memory Sitruk-Ware, R., Nath, A., 2013. Characteristics and metabolic effects of estrogen and
consolidation during sleep. Nat Neurosci 22, 1598–1610. progestins contained in oral contraceptive pills. Best Pract Res Clin Endocrinol
Komnenich, P., Lane, D.M., Dickey, R.P., Stone, S.C., 1978. Gonadal hormones and Metab 27, 13–24.
cognitive performance. Physiological Psychology 6, 115–120. Spalek, K., Loos, E., Schicktanz, N., Hartmann, F., de Quervain, D., Stier, C., Milnik, A.,
Kuhlmann, S., Wolf, O.T., 2005. Cortisol and memory retrieval in women: influence of 2019. Women using hormonal contraceptives show increased valence ratings and
menstrual cycle and oral contraceptives. Psychopharmacology (Berl) 183, 65–71. memory performance for emotional information. Neuropsychopharmacology 44,
Le, J., Thomas, N., Gurvich, C., 2020. Cognition, The Menstrual Cycle, and Premenstrual 1258–1264.
Disorders: A Review. Brain Sci 10. Stanczyk, F.Z., Archer, D.F., Bhavnani, B.R., 2013. Ethinyl estradiol and 17β-estradiol in
Lewis, C.A., Kimmig, A.S., Zsido, R.G., Jank, A., Derntl, B., Sacher, J., 2019. Effects of combined oral contraceptives: pharmacokinetics, pharmacodynamics and risk
Hormonal Contraceptives on Mood: A Focus on Emotion Recognition and Reactivity, assessment. Contraception 87, 706–727.
Reward Processing, and Stress Response. Curr Psychiatry Rep 21, 115. Strojny, J., Domes, G., Fischbacher, U., von Dawans, B., 2021. The modulation of social
McFadden, D., 2000. Masculinizing effects on otoacoustic emissions and auditory evoked behavior and empathy via oral contraceptives and female sex hormones.
potentials in women using oral contraceptives. Hear Res 142, 23–33. Psychoneuroendocrinology 131, 105250.
Merz, C.J., 2017. Contribution of stress and sex hormones to memory encoding. I. Sundstrã¶M Poromaa, and M. Gingnell, Menstrual cycle influence on cognitive function
Psychoneuroendocrinology 82, 51–58. and emotion processing—from a reproductive perspective. Frontiers in
Neuroscience 8 (2014).
21
C. Gurvich et al. Frontiers in Neuroendocrinology 69 (2023) 101052
Sundström-Poromaa, I., 2018. Chapter Thirteen - The Menstrual Cycle Influences Vranic, A., Hromatko, I., 2008. Content-specific activational effects of estrogen on
Emotion but Has Limited Effect on Cognitive Function. In: Litwack, G. (Ed.), working memory performance. J Gen Psychol 135, 323–336.
Vitamins and Hormones. Academic Press, pp. 349–376. Warren, A.M., Gurvich, C., Worsley, R., Kulkarni, J., 2014. A systematic review of the
United Nations, Trends in Contraceptive Use Worldwide 2015 (ST/ESA/SER.A/349), impact of oral contraceptives on cognition. Contraception 90, 111–116.
Population Division, Department of Economic and Social Affairs, 2015. G. Wells, B. Shea, D. O’Connell, j. Peterson, V. Welch, M. Losos, and P. Tugwell, The
van Leeuwen, N., Bellingrath, S., de Kloet, E.R., Zitman, F.G., DeRijk, R.H., Kudielka, B. Newcastle–Ottawa Scale (NOS) for Assessing the Quality of Non-Randomized Studies
M., Wüst, S., 2011. Human mineralocorticoid receptor (MR) gene haplotypes in Meta-Analysis. (2000).
modulate MR expression and transactivation: Implication for the stress response. Wharton, W., Hirshman, E., Merritt, P., Doyle, L., Paris, S., Gleason, C., 2008. Oral
Psychoneuroendocrinology 36, 699–709. contraceptives and androgenicity: influences on visuospatial task performance in
Vonk, J.M.J., Rizvi, B., Lao, P.J., Budge, M., Manly, J.J., Mayeux, R., Brickman, A.M., younger individuals. Exp Clin Psychopharmacol 16, 156–164.
2019. Letter and Category Fluency Performance Correlates with Distinct Patterns of
Cortical Thickness in Older Adults. Cereb Cortex 29, 2694–2700.
22