Empirical research in well-being intended as optimal psychological functioning (Ryan and Deci, 2001) has rapidly increased in recent years (e.g., Diener et al., 1999; Keyes et al., 2002; Seligman, 2011). Purpose, mastery, strong relationships, and self-acceptance (which are dimensions coherent with the ‘eudaimonic’ view of well-being, occurring when people’s life activities are most congruent or meshing with deeply held values and are holistically or fully engaged, e.g., Ryff, 1989; Keyes et al., 2002), as well as life satisfaction, presence of positive affect, and absence of negative affect (which are dimensions of a ‘hedonic’ view of well-being consisting of subjective happiness and concerning the experience of pleasure versus displeasure broadly construed to include all judgments about the good/bad elements of life; e.g., Diener, 2000) are most frequently used variables to assess well-being. However, evidence suggests that a positive state is more than the mere absence of negative mood or symptoms (Clark and Watson, 1991).

Even though research on well-being has traditionally focused on hedonic versus eudaimonic approach (Waterman, 1993), recent studies claimed that well-being could be best described as a multidimensional construct that includes the elements of both hedonic and eudaimonic perspectives (Keyes and Waterman, 2003; Gallagher et al., 2009; Robitschek and Keyes, 2009). In line with Karademas (2007), well-being has multiple components and positive well-being could be conceived as the cognitive and affective reactions to the perception of adequate personal characteristics and achievements, efficient interaction with the world and social integration, and positive progress in time.

Within this field of research, one of the most dominant issues concerns the degree of cross-situational consistency and stability across time that well-being possesses. Some researchers conceive well-being as a trait-like feature, which remains relatively stable across time and situations (Eid and Diener, 2004; Davern et al., 2007; Lucas and Diener, 2008), being closely linked to genetic predispositions and personality traits, in particular to extraversion, emotional stability and, to a lesser extent, consciousness (Diener and Seligman, 2002; Steel et al., 2008). On the contrary, other authors believe that well-being is a state-like characteristic, the levels of which may vary under specific life events and circumstances such as unemployment, higher income, graduation, and marriage (Reis et al., 2000; Blanchflower, 2001; Bostic and Ptacek, 2001; Lucas and Schimmack, 2009; Angner, 2010; Salinas-Jmenez et al., 2011). Nevertheless, studies focusing on the associations between malleable circumstances and well-being have mostly investigated changes in subjective well-being and have not used a multi-componential approach.

Recently, Hudson et al. (2017) showed that not only people’s global well-being (Schimmack and Oishi, 2005), as traditionally measured by self-reports, but also people’s experiential well-being, as measured by assessing lived experiences and momentary reports of well-being and including both cognitive and affective dimensions, is relatively stable over extended periods of time.

Gender differences literature have specifically addressed women’s well-being supporting the idea that women differ from men in the intensity of their affective experiences (Frank et al., 1991). Due to the emphasis on emotional experience within the specific social role that women have historically enacted (e.g., caretaker), they possibly have developed more appropriate skills to both recognize and express their own affect. A greater emotional sensitivity and expressiveness (Grossman and Wood, 1993; Hess et al., 2000) might justify the female (as compared to male) tendency to report more detailed representations of their own emotions (Wood et al., 1989; Barrett et al., 2000) both assessed through self-reports and objective measures (Domes et al., 2010), and higher level of happiness and well-being (Herzog et al., 1982).

Women’s well-being has been investigated in a life course perspective by focusing on the relationship between some major life events and subsequent perceived well-being. A number of studies, which were aimed at exploring the level of women’s well-being in correspondence with distinct life events (e.g., retirement: Kim and Moen, 2002; menopause: Dennerstein et al., 1994), reported a moderate stability even in the face of changing life circumstances and over long periods of time (see Diener et al., 1999). These results are consistent with the set-point theory suggesting that individuals tend to return to their baseline level of well-being even after significant life events (Headey and Wearing, 1989).

The stability of women’s well-being across short window of time has been scarcely investigated. Specifically, an interesting short period of time for monitoring women well-being is constituted by the menstrual cycle that may have important consequences on women’s psychological processes and behaviors (Neave, 2008; Durante et al., 2011, 2013; Pearson and Schipper, 2013; Iannello et al., 2015a, 2016). Among the psychological aspects that have been linked to hormonal changes during the menstrual cycle, particular emphasis has been placed on women emotional experience and self-representation. Concerning the first issue, even if some authors found changes in anxiety, depression, and responses to stress (De Ronchi et al., 2000; Clayton, 2008), these results related primarily to the pre-menstrual phase that occurs few days prior to menstruation. The current state of evidence shows little support for a specific premenstrual negative mood change occurring with any regularity in the general population (Romans et al., 2012).

Concerning self-representation, the association between menstrual cycle phases and self-esteem still appears inconsistent. Some authors found that self-esteem was lower in the premenstrual phase compared to other phases in women with premenstrual syndrome (Bloch et al., 1997; Taylor, 1999). In non-clinical populations Edmonds et al. (1995) did not find an association between self-esteem and menstrual cycle in college students, while recently Hill and Durante (2009) found that self-esteem was higher at the least fertile phase of the cycle (late luteal phase) and low around the 5 days before and 2 days after ovulation (the most fertile point of the cycle) in undergraduate students.

To date, researchers have principally focused on the implications of hormonal fluctuations across the menstrual cycle on changes in affective states and self-representation, whereas to our knowledge only one study addressed the implications for individual women’s well-being directly (Iannello et al., 2015b).

Starting from literature that sustains the stability of the whole system of well-being over long periods of time, this study aimed to verifying that the level of well-being remains stable across short window of time represented by menstrual cycle phases. We investigated well-being in high and low fertility phases through a multidimensional assessment that included: (1) psychological well-being concerning the fulfillment and self-realization of the individual in several components of positive psychological functioning; (2) self-esteem defined as the personal judgment of overall self-worth, recognized as one indicator of well-being (e.g., Baumeister et al., 2003; Mann et al., 2004); and (3) emotional self-efficacy beliefs that contribute to individuals’ psychological well-being by sustaining self-regulatory mechanisms (Bandura et al., 2003; Datu, 2012).

Twenty-five women among undergraduate students of the Department of Psychology at the Catholic University of the Sacred Heart of Milan were recruited. The eligibility for participation included (1) being between 20 and 30 years old, (2) having a regular menstrual cycle, (3) not being pregnant or nursing, and (4) not taking any hormonal contraceptives (e.g., pills, shots, patches, etc.). The mean age of participants was 22.4 years (SD = 1.9). Furthermore, we ensured that women had a level of emotional stability and extroversion consistent with normative data, as measured by the HEXACO personality inventory (Lee and Ashton, 2004), and excluded women with personality characteristics that could influence their perceived well-being (Costa and McCrae, 1980; Aghababaei and Arji, 2014). Specifically, women whose scores were above the 75th percentile on either emotional stability or extraversion were excluded from the study (N = 3).

Participants decided voluntarily to participate in the study and their anonymity was guaranteed. Each participant was asked to read and sign the consent form that confirmed their eligibility for participation and agreement to participate. Ethical compliance of the procedures was approved by the Ethical Committee at Department of Psychology of the Catholic University of Milan.

Each participant was asked to complete three sessions (see Figure 1). In the first session, participants were asked to complete a questionnaire about their menstrual cycle to obtain information needed to estimate both the expected high-fertility and low-fertility days. Based on the information women provided about their cycle, each participant was scheduled to come into the lab for two sessions, during an expected high-fertility day and during an expected low-fertility day. To confirm the high-fertility phase, participants were instructed to use a commercially available urine-based luteinizing hormone (LH) predictor test kit at home 2 days before the expected high-fertility session and to communicate the result of the test to the researchers via email. The testing session was confirmed following a positive test, which indicated ovulation would occur within 24–36 h. After a negative test, the high-fertility lab session was re-scheduled for the subsequent ovulatory cycle. Low-fertility sessions were scheduled 6 days or more after the LH surge (if high-fertility testing session took place first) or at least 3 days before menstrual onset (if low-fertility testing session took place first). During each session, women were administered the Psychological Well-Being Scales (PWB) (Ryff and Keyes, 1995), the Regulatory Emotional Self-Efficacy Scales (Caprara and Gerbino, 2001), and the Rosenberg Self-Esteem scale (Rosenberg, 1965). In both sessions, before completing the scales, participants were specifically asked to answer the questionnaires referring to the past 3 days (“Think about your last three days. Please, answer each item of the questionnaires referring to that short period of time”). The order of presentation of the three scales varied randomly in each session.

We hypothesized that the level of well-being remained stable across menstrual cycle phases (low vs. high fertility). To test this hypothesis, we used the paired sample t-test Bayes factor (BF) to assess a ratio between the likelihood of the data given null-hypothesis and the one given the alternative one (Liang et al., 2008; Masson, 2011; Nuzzo, 2014; Rouder, 2014).

Cauchy prior width, that we used in our computation, represents the scale of the Cauchy prior density on effect size under the alternative hypothesis; the default is 0.707. On the other hand, “error %” in our results indicates the error of the Gaussian quadrature integration routine used for the computation of the BF.

Paired sample t-test BF is universally considered a more robust statistical test compared to p-value. In a recent note the American Psychological Association (2010) considered the claims of American Statistical Association and stated the importance of using also BF, among others (Wasserstein and Lazar, 2016). In this study, the choice of BF is due to the need to compare repeated measurements under the hypothesis of similarity between them. The use of BF allowed us to determine whether a model in which similarity is considered is strongly better as compared to a model where differences are considered. BF, as computed, provides the likelihood ratio of such comparison. All analyses were conducted using JASP (JASP, 2014).