Table 1.

Overview of results from studies investigating HC-effects on psychological and neurophysiological mechanisms underlying mood regulation—emotion recognition and reactivity, reward processing, and stress response. Studies are listed in order of appearance in the text

	Research design	Sample size and HC method	Research modality	Task	Results		Main findings in HC-users
Emotion recognition
Pahnke et al. [21]	Cross-sectional	42 combined OC 53 NC (35 follicular, 18 luteal)	Behavioral	RMET	OC-users performed significantly worse in complex face recognition independent of emotional valence or type of OC (androgenic vs antiandrogenic).	↓	Impaired emotion recognition
Hamstra et al. [22]	Cross-sectional	49 combined OC 44 NC (21 early follicular, 23 luteal)	Behavior-genotype interaction	Facial expression recognition task, RMET	OC-users with MC haplotype 1 and 3 performed generally worse in face recognition task than luteal NC women (trend-level). OC-users with MC haplotype 2 recognized less positive characteristics in the RMET than luteal NC women.	↓	Impaired emotion recognition
Hamstra et al. [23]	Cross-sectional	44 combined OC 40 NC (11 follicular, 29 luteal)	Behavior-genotype interaction	Facial expression recognition task, emotional categorization, and memory task	OC-users showed worse recognition of anger independent of MC haplotype.	↓	Impaired emotion recognition of negative emotions
					OC-users with MC haplotype 1 and 3 recognized fearful and sad images significantly better and recalled more negative characteristics, but also had longer reaction times for detecting these emotions.	↑	Attention bias to negative emotions in MC haplotype 1 and 3 carriers
Hamstra et al. [24]	Cross-sectional	26 OC1 14 follicular NC	Behavioral	Facial expression recognition task	OC-users showed worse recognition of facial expressions depicting anger, as well as a trend for sadness and disgust, compared with NC women.	↓	Impaired emotion recognition of negative emotions
Hamstra et al. [25]	Longitudinal	57 combined OC (within subjects active vs. inactive pill phase) 39 NC (within subjects early follicular vs. luteal)	Behavior-genotype interaction	Facial expression recognition task, RMET	OC-users showed worse recognition of sadness and happiness (trend-level) and had shorter reaction times for detecting anger and happiness. OC-users recognized more positive characteristics in RMET	↔	Mixed findings: both impaired and enhanced emotion recognition
Radke et al. [26•]	Cross-sectional	25 combined OC (inactive phase) 30 combined OC (active phase) 18 NC	Behavioral	Affective responsiveness task, emotion recognition task, perspective-taking task	No differences in emotion recognition and perspective taking between OC-users and NC women.	↔	No differences in emotion recognition
					Increased accuracy for OC-users in active phase in affective responsiveness compared with OC-users in an inactive phase.	↑	Enhanced emotional reactivity in active OC phase
Emotional reactivity
Spalek et al. [27•]	Cross-sectional	1215 HC1 (majority OC) 954 NC	Behavioral	Picture rating task, picture memory task	HC-users rated emotional pictures (negative and positive) more emotionally intense and neutral images less arousing than NC women.	↑	Enhanced emotional reactivity
					HC-users remembered significantly more emotional pictures (positive and negative) than NC women which were mediated by valence/arousal ratings.	↑	Enhanced emotional memory
Gingnell et al. [28]	RCT	17 combined OC-starters 17 placebo-starters (both groups with previous mood-related side effects of OCs)	Task fMRI	Emotional face-matching task	No differences in face-matching accuracy.	↔	Similar ratings
					OC-starters had significantly more mood swings and depressed mood after 1 month of intake compared with pre-start and to the placebo group.	↑	Mood swings and depressed mood
					OC-starters had reduced BOLD response in the left insula, left MFG, and bilateral IFG compared with placebo and reduced BOLD response in bilateral IFG compared with pre-start.	↓	Blunted emotional reactivity
					Decreased habituation of amygdala in OC-starters compared with placebo between time points.	↑	Higher vigilance for negative emotional stimuli
Miedl et al. [29•]	Cross-sectional	23 combined OC 30 NC	Task fMRI	Traumatic vs neutral video clips	No differences in valence and arousal ratings.	↔	Similar ratings
					Enhanced BOLD responses in OC-users in the insula and dorsal ACC during traumatic vs. neutral clip viewing.	↑	Enhanced emotional reactivity for traumatic content
Merz et al. [30]	Cross-sectional	29 combined OC 30 luteal NC 39 men	Task fMRI, physiology	Fear acquisition and extinction	Slower habituation of SCR rates, correlated with increased BOLD signal in response to fear-evoking stimuli in OC-users compared with NC women in the right amygdala, right ACC, bilateral thalamus, and vmPFC.	↔ ↑	Similar emotional reactivity in the acquisition phase Increased emotional reactivity in the extinction phase
Hwang et al. [31]	Cross-sectional	16 combined OC 32 NC (16 high estradiol, 16 low estradiol) 37 men	Task fMRI	Fear conditioning, extinction and recall procedures	Reduced BOLD response during fear conditioning in the insular cortex, MCC, amygdala, and hypothalamus in OC-users compared with high estradiol NC women.	↓	Blunted emotional reactivity during fear conditioning
					No differences between OC-users and NC women for unconditioned fear, fear extinction, and recall.	↔	Similar emotional reactivity in fear extinction and recall
Armbruster et al. [32]	Cross-sectional	35 combined OC 35 NC (within-subjects early follicular vs. late luteal)	Physiology (SCR, startle magnitude)	Acoustic startle response task during image presentation	No difference in valence and arousal ratings between OC-users and NC women.	↔	Similar ratings
					OC had blunted startle magnitudes and SCR, especially for negative images.	↓	Blunted emotional reactivity
Reward processing
Petersen et al. [33]	Cross-sectional	44 combined OC 46 NC	Structural MRI	–	OC-use associated with lower cortical thickness in lateral OFC and posterior cingulate cortex.	↓	Lower cortical thickness in reward-related region
Bonenberger et al. [34]	Cross-sectional	12 OC1 12 NC	Task fMRI	Monetary incentive task	Enhanced BOLD response during monetary reward expectation in anterior insula and inferior PFC in OC-users compared with NC women in the follicular phase.	↑	Enhanced reward response
Arnoni-Bauer et al. [35]	Cross-sectional	12 combined OC 20 NC	Task fMRI	Visual food cues	OC-users show similar BOLD response as NC women in the luteal phase but greater BOLD response as NC women in the follicular phase in reward response (amygdala, putamen) and decision-making regions (PFC).	↔↑	Similar to enhanced reward response
Scheele et al. [36]	Cross-sectional	21 HC (16 combined OC, 5 IUS) 19 NC	Task fMRI	Attractiveness rating under single oxytocin nasal dose	Oxytocin increased attractiveness ratings of the partner’s face in NC women but not in HC-users. Reduced BOLD response in striatal reward regions (NAcc, VTA) in OC-users compared with NC women.	↓	Blunted reward response
Jakob et al. [37•]	Longitudinal	38 combined OC (within-subjects active vs. inactive pill phase) 30 NC (within-subjects early vs. late follicular)	Behavior-hormone-genotype interaction	Probabilistic reinforcement learning task	Decrease in ability to avoid punishment with rising estradiol levels in 9RP carriers NC women, no such behavioral variations in OC-users according to DAT1-genotype differences or intake phase.	↔	No genotype interaction for reward responses
Stress response
Merz et al. [38]	Cross-sectional	30 combined OC 60 NC	Behavioral, physiological	SECPT	Blunted cortisol response in OC-users compared with NC women after stress exposure.	↓	Blunted stress response
Barel et al. [39]	Cross-sectional	20 combined OC 17 NC	Behavioral, physiological	TSST	Blunted cortisol response in OC-users compared with NC women after stress exposure.	↓	Blunted stress response
Nielsen et al. [40]	Cross-sectional	49 combined OC 60 NC	Behavioral, physiological	Emotional recall after CPS	Blunted cortisol after stress exposure, paralleled by weaker performance in emotional recall response in OC-users compared with NC women.	↓	Blunted stress response
Mordecai et al. [41]	Longitudinal	39 combined OC (within-subjects active vs. inactive pill phase) 40 NC (within-subjects follicular vs. luteal)	Behavioral, physiological	Emotional recall after TSST	Blunted cortisol response and worse emotional recall for negative words in OC-users (similar for both active and inactive pill phase) compared with NC women after stress exposure.	↓	Blunted stress response
					Higher baseline cortisol levels in OC-users compared with NC women.	↑	Higher baseline cortisol
Hertel et al. [42••]	Cross-sectional	74 OC (70 combined OC, 4 progestin-only) 159 NC	Structural MRI, physiological	–	Higher baseline cortisol levels and reduced hippocampal gray matter in OC-users compared with NC women.	↑ ↓	Higher baseline cortisol Reduced hippocampal volume

¹No information stated on which specific HC/OC-type used in the study

ACC anterior cingulate cortex, BOLD blood oxygen level-dependent, CPS cold-pressor stress, DAT-1 dopamine transporter, fMRI functional magnetic resonance imaging, HC hormonal contraceptive, IFG inferior frontal gyrus, MC mineralocorticoid, MCC middle cingulate cortex, MFG middle frontal gyrus, NAcc nucleus accumbens, NC naturally cycling, OC oral contraceptive, OFC orbitofrontal cortex, PFC prefrontal cortex, RCT randomized, placebo-controlled trial, RMET Reading the Eyes in the Mind Test, SECPT socially evaluated cold-pressor test, SCR skin conductance response, TSST Trier social stress test, vmPFC ventromedial prefrontal cortex, VTA ventral tegmental area