A long shot? Affective temperaments predict adherence to pharmacotherapy during infertility treatment in a prospective longitudinal study.

Infertility, defined by the World Health Organisation (WHO) as the inability to achieve pregnancy after one year of regular, unprotected intercourse, is a significant global health concern, with an estimated lifetime prevalence of 17.5% (WHO 2023 ). The impact of infertility extends to mental, physical, sexual, and social aspects of couples’ lives (Bala et al. 2021 ; Nayeri et al. 2022 ; Kelley and Kingsberg 2024 ). Importantly, infertility is often associated with considerable psychological distress and may contribute to the development or worsening of psychiatric disorders, such as depression and anxiety (De Berardis et al. 2014 ). Therefore, the success of infertility treatment is multifaceted and influenced by various factors beyond the medical interventions themselves (Leeners et al. 2023 ; Wu et al. 2023 ). While medical treatments are pivotal for facilitating conception, adopting a healthy lifestyle is equally crucial for better outcomes. Evidence suggests that maintaining a balanced diet (Karayiannis et al. 2018 ; Sanderman et al. 2022 ; Noli et al. 2023 ), engaging in regular physical activity (Minas et al. 2022 ; Xie et al. 2022 ), and avoiding harmful habits, such as smoking and excessive alcohol consumption significantly enhance fertility (Hazlina et al. 2022 ). Given the complexity and duration of infertility treatments, adherence to both medical and lifestyle recommendations is essential for a positive outcome. Adherence involves the consistent implementation of therapeutic advice, prescribed medications, and necessary lifestyle changes to ensure successful outcomes (WHO 2023 ). In assisted reproductive technology (ART), this translates to the uptake of the ART cycles recommended by the doctor until pregnancy is achieved or until there is a recommendation to end treatment, as well as compliance with medication (Gameiro et al. 2013 ). Adherence is a highly complex behaviour influenced by a range of factors, including patient-related aspects, the healthcare team and system, prescribed therapy, characteristics of the disease, and socioeconomic circumstances (WHO 2003 ). Research indicates that adherence rates among infertile patients are not high, ranging between 30 and 50% (Guo et al. 2020 ; Ni et al. 2021 ), compared to adherence rates around 50% observed in other chronic somatic conditions (WHO 2003 ), although reported values may depend on measurement methods. This non-adherence undermines therapeutic outcomes, prolonging treatment durations and reducing pregnancy rates (Szabo et al. 2024 ). Understanding the factors that influence adherence is therefore critical for improving treatment outcomes in this population. Recent studies suggest that affective temperaments play a significant role in adherence and, consequently, in the outcomes of infertility treatments (Szabo et al. 2023 , 2024 ). Affective temperaments—categorised as depressive, cyclothymic, hyperthymic, irritable, and anxious—have a genetic background and represent the biological ‘cores’ of personality. These temperaments influence an individual’s energy levels, mood, emotional reactivity, and cognitive patterns (Kawamura et al. 2010 ). They not only affect how patients adhere to treatment recommendations but may also directly impact therapeutic outcomes (Szabo et al. 2022 ). For instance, depressive, anxious, and cyclothymic temperaments have been associated with reduced adherence and poorer outcomes in infertility treatments in our previous retrospective study (Szabo et al. 2023 , 2024 ). Recognising the role of affective temperaments is particularly important, given the emotional and psychological stress associated with prolonged treatment processes. This study aims to assess the impact of affective temperaments on medication adherence in women undergoing infertility treatments in a prospective, follow-up design.

In total, our prospective cohort study included 179 women (25–45 years of age) who underwent infertility treatment in the Assisted Reproduction Centre of the Department of Obstetrics and Gynaecology of Semmelweis University in Budapest, Hungary, between March 2022 and September 2024. Forty percent of the participants struggled with infertility for more than 2 years, 65% of the participants were primary infertile, 11% already had at least one child from a previous pregnancy, while 24% reported not reaching live birth but had pregnancies ending in miscarriage(s). The mean age of our cohort was 35.27 ± 4.39 years, and the mean BMI was 24.67 ± 5.65 kg/m 2 . Fifty-eight percent of the women were diagnosed with some kind of chronic somatic disease other than infertility, typically with various problems of carbohydrate metabolism, thyroid function problems, endometriosis, or a combination of these. Thirty-five percent of them had emotional or psychological difficulties in their medical history. Sixty percent of the participants had to take medication regularly during the infertility treatment, their medication adherence score ranging between 0.5 and 8, with an average of 6.61 ± 1.49, 33% falling into the high adherence category (MMAS-8 = 8). In terms of affective temperaments, the average scores in the current infertile population were as follows: cyclothymic 25.55 ± 4.00, depressive 28.56 ± 3.00, anxious 33.44 ± 5.34, irritable 25.28 ± 3.94, and hyperthymic 31.80 ± 3.77. Regarding further demographic parameters, 76% of the participants had a tertiary education, 23% secondary, while only 1% primary education. Forty-nine percent lived in the capital, 16% in other cities, and 36% in smaller towns or villages. Seventy-eight percent owned a house, while 22% rented. Fifty-six percent reported that they had more than enough income on a monthly basis, 38% earned just enough, and 5.6% earned less than enough. Based on linear regression analyses, higher cyclothymic, depressive, anxious, and irritable affective temperament scores predicted significantly poorer medication adherence (cyclothymic: β  = −0.122, p  < 0.001, depressive: β  = −0.178, p  < 0.001, irritable: β  = −0.114, p  = 0.002, anxious: β  = −0.071, p  = 0.08). Table 1 summarises the results of the univariate regression analyses of affective temperaments on medication adherence. Results of uni- and multivariate regression models showing the effect of affective temperaments on adherence to fertility treatment. β : regression coefficient; SE : standard error; BMI: body mass index. The upper table summarises the results of univariate regression analyses using change in TEMPS-A scores as exposure variables and change in MMAS score as outcome variable. The lower table shows the results of multivariate regression analyses using change in TEMPS-A scores, age, BMI, socio-economic and clinical history parameters as predictors and MMAS-8 scores as outcome variables. Affective temperaments were included in the multiple regression analyses separately. The bold values in the table represent significant findings. Based on multivariate regression analyses, the effect of cyclothymic, depressive, anxious, and irritable affective temperament scores on medication adherence remained significant when sociodemographic and medical factors were included in the model (cyclothymic: β  = −0.170, p  < 0.001, depressive: β  = −0.183, p  = 0.001, irritable: β  = −0.140, p  = 0.003, anxious: β  = −0.086, p  = 0.011). None of the sociodemographic or medical factors had any significant effect on adherence. Table 1 summarises the results of the multivariate regression analyses of the effect of affective temperaments, sociodemographic, and medical factors on medication adherence. Interaction analysis revealed that education did not have a significant moderating effect on the relationship between any affective temperament and medication adherence. Similarly, age did not significantly moderate the association between cyclothymic, depressive, irritable, or hyperthymic temperaments and medication adherence. However, increased age intensified the negative impact of anxious affective temperament on medication adherence ( β  = −0.015, p  = 0.024). The main result of the interaction analysis is presented in Figure 1 . Interaction between anxious affective temperament and age in predicting adherence. Note: Age was measured as a continuous moderator in the interaction analysis; it was categorised into ‘younger’ and ‘older’ groups in the figure for illustrative purposes only.

Women attending their initial consultation at the Assisted Reproduction Centre of the Department of Obstetrics and Gynaecology at Semmelweis University between March 2022 and September 2024 were invited to participate in this study via email ( n  = 2037) Out of them, n  = 419 agreed to participate, n  = 22 disagreed to participate, n  = 1596 did not respond, the overall response rate of the baseline assessment being 22%. Eligibility criteria included: being between 18 and 45 years old (in line with age requirements for state-subsidized assisted reproductive services in Hungary), actively trying to conceive, fluent in Hungarian, voluntarily consenting to the study, and completing both the baseline and follow-up assessments six months after starting treatment. Prior pregnancies—whether natural or assisted—were not used to determine eligibility. After removing repeated and incomplete entries, a final cohort of 179 participants was analysed. The dropout rate between the baseline and follow-up assessment was 57%. Demographic, medical, and psychological data were obtained through self-administered questionnaires and from medical records. Voluntary participation was confirmed by written informed consent from the patients. The study protocol received ethical approval from the Scientific and Research Ethics Committee of the Hungarian Ministry of Health Medical Research Council (IV/1568-1/2022/EKU) and adhered to the ethical principles outlined in the Declaration of Helsinki. Further details on recruitment and inclusion and exclusion criteria in our studies can be found in our previous paper reporting retrospective results from a similar study design (Szabo et al. 2024 ). Details on participants’ age, education level, place of residence, living conditions, financial situation, body mass index (BMI), history of psychiatric disorders, duration of infertility, chronic illnesses, and reproductive history (including pregnancies, miscarriages, and births) were collected either directly from patients via self-report or extracted from medical documentation. Affective temperaments were evaluated following the initial consultation, before any therapeutic intervention, using the validated Hungarian version of the Temperament Evaluation of Memphis, Pisa, Paris, and San Diego (TEMPS-A) questionnaire, a 110-item self-report tool that evaluates five temperament dimensions: depressive, cyclothymic, irritable, anxious, and hyperthymic (Akiskal et al. 2005 ; Rózsa et al. 2006 , 2008 ). Participants responded to each item with ‘yes’ (1 point) or ‘no’ (0 points). Internal scale reliability in the original Hungarian validation ranged from good to excellent (Cronbach’s alpha: depressive = 0.63; cyclothymic = 0.81; irritable = 0.79; anxious = 0.84; hyperthymic = 0.78) (Rózsa et al. 2006 ). In our current sample, scale reliability remained strong: depressive = 0.65; cyclothymic = 0.83; irritable = 0.82; anxious = 0.85; hyperthymic = 0.77. Six months after beginning infertility treatment, participants completed the Morisky Medication Adherence Scale (MMAS-8) to assess their adherence with prescribed medication (Berlowitz et al. 2017 ; Bress et al. 2017 ). This 8-item tool includes seven binary (‘yes’/’no’) items and one Likert-scale item, with higher total scores indicating better adherence. The MMAS-8 demonstrated excellent internal reliability in its validation study (Cronbach’s alpha = 0.84). Internal reliability in the present study was found to be acceptable (Cronbach’s alpha = 0.66). Treatment strategies varied according to individual patient needs and underlying conditions. Most protocols involved ovarian stimulation based on standardised regimens (TEGGO et al. 2020 ), followed by in vitro fertilisation (IVF) or intracytoplasmic sperm injection (ICSI). Adjunct treatments, such as oral medications, were used depending on the infertility aetiology. Given the heterogeneity of approaches and meta-analytic evidence suggesting no particular method consistently predicts success (Van Loendersloot et al. 2010 ), the type of infertility treatment was treated as a constant in our analysis. Continuous data are reported as means with standard deviations and ranges, while categorical variables are presented as frequencies and percentages. Initial analyses involved simple linear regression to determine whether affective temperament traits predicted adherence outcomes. A subsequent multivariate linear regression included additional factors, such as demographic and clinical variables (age, education, residence, housing, financial status, BMI, psychiatric and somatic illness history, infertility duration, prior pregnancies, miscarriages, and births). Additionally, to explore whether the impact of affective temperaments on adherence is moderated by age and education, interaction analyses were conducted. Due to multicollinearity among temperament dimensions, each was entered separately into the models. Regression outcomes are presented as standardised coefficients ( β ), standard errors ( SE ), and p -values. Statistical significance was set at p  < 0.05, and analyses were conducted in R Statistical Software (version 4.4.1, Vienna, Austria). A statistical power analysis was conducted using G∗Power (Faul et al. 2009 ). Based on a medium anticipated effect size (0.15), a significance level of 0.05, and a power of 95%, the minimum required sample size for simple linear regression with one predictor was 89.

Following our previous meta-analysis showing that affective temperaments significantly influence medication adherence in various somatic and psychiatric disorders (Szabo et al. 2022 ) and retrospective findings concerning the significant effect of affective temperaments in predicting infertility treatment success (Szabo et al. 2023 ) and the role of adherence to dietary recommendations as a mediating factor (Szabo et al. 2024 ), our present study aimed to investigate the influence of affective temperaments on medication adherence during ART treatment in a prospective, follow-up design. We confirmed our previous findings regarding the impact of affective temperaments and extended them specifically to ART treatments by showing that higher cyclothymic, depressive, anxious, and irritable affective temperament scores predicted significantly poorer medication adherence. This relationship remained significant even after taking into consideration sociodemographic and medical factors, such as age, education, income, BMI, and obstetric history, which, unlike temperamental factors, did not influence adherence on their own. Furthermore, considering sociodemographic factors, we found that education did not influence the effect of affective temperaments on adherence; however, there was a significant interaction with age in the case of anxious temperament, where the older the participants were, the greater the decrease in medication adherence associated with anxious affective temperament. Our results are the first to confirm that affective temperaments may affect adherence to prescribed pharmacotherapeutic interventions also among women undergoing infertility treatment, and suggest that, in this way, they possibly influence ART outcomes, and have a larger impact than sociodemographic and medical factors. The course of chronic diseases strongly depends on self-management, including self-regulatory processes, such as leading a healthy lifestyle and taking medicine as prescribed (Barlow et al. 2002 ). While extensive literature exists on the impact of adherence to lifestyle recommendations on intervention outcomes in women receiving ART, limited information is available on medication adherence in the infertility context (Ni et al. 2023 ). This is, however, a crucial issue, since failure to take medications correctly significantly decreases the likelihood that the medications will have the intended effect on hormonal and other processes, ultimately decreasing the chance of treatment success. The first and only systematic review looking at medication adherence during IVF to date included three studies and found that adherence to oral medication as first-line therapy is similar to general rates of adherence, around 50% in chronic diseases (Mahoney et al. 2019 ). Our MMAS-8 results fall in the average range detected in other studies on medication adherence in chronic ailments (Al-Hajje et al. 2015 ; Plakas et al. 2016 ; Moon et al. 2018 ; Zhang et al. 2021 ), thus, while confirming the rate of medication adherence detected in previous studies in chronic illnesses, also reflect the generalisability of our findings. Affective temperaments, as matrices of characteristic emotions, cognitions, attitudes, and reactions, also largely influence the long-term course of several somatic diseases, with non-hyperthymic temperaments typically associated with worse health outcomes (Favaretto et al. 2024 ). Their role in the prognosis of somatic illnesses may, among other factors, be related to their role in affecting mental states and behaviours influencing the motivation and ability to comply with medical recommendations and adhere to prescribed medications (Szabo et al. 2022 ), which in turn influences treatment outcome (WHO 2023 ). Considering this low rate of medication adherence, especially in the light of such a long-term and costly treatment with significant burden on the somatic and mental health of participants, understanding psychological and behavioural underpinnings of lower adherence as well as those which may pinpoint pathways of improving adherence in a personalised way are very important to improve efficacy, and thereby decrease health-related and social burdens as well as costs. A consideration of the key components of affective temperaments, including emotion regulation, cognitions, and attitudes, offers the possibility for pinpointing those components that, in given patients, are related to behaviours underlying lack of sufficient adherence, and which can be improved by targeted interventions, including psychoeducation, psychological counselling, training, or support. Beyond identifying the impact of affective temperaments on adherence to medications, we also looked at how key sociodemographic factors that have been shown previously to influence adherence interact with affective temperaments, including level of education (Teppo et al. 2022 ; Al-Tarawneh et al. 2023 ) and age (Krivoy et al. 2015 ; Kim et al. 2019 ; Ge et al. 2023 ). While no interaction with education was found, there was a significant interaction between anxious affective temperament and age, where increasing age worsened the negative effect of anxious temperament on adherence. Affective temperaments, as measured by the TEMPS-A, have been shown to be relatively stable over time (Kawamura et al. 2010 ). A large-scale review found that age only influenced TEMPS-A scores in the case of depressive temperament in women (Vázquez et al. 2012 ). Previous results on the relationship between age and medication adherence have been significant but mixed, suggesting that, while there is a relationship, this may be moderated and mediated by various factors. Our findings help to refine our view on the complex way affective temperaments may impact treatment adherence and draw attention to the contributing role of affective temperaments in the context of increasing age. Further studies are warranted to elucidate why anxious temperament increases nonadherence with growing age. As already suggested above, our results may have important implications for clinical practice. Assisted reproduction is a complex and intricate process, mainly influenced by biological factors but, to a not insignificant extent, also by psychological and behavioural components. While illness beliefs, health status perceptions, previous experiences, and psychological states are all at play in patient adherence behaviours (De Pasquale et al. 2016 ), affective temperaments, as demonstrated in this study, also represent important contributors, likely due to their impact on emotional reactivity, cognitions, and related behaviours which in turn significantly modify motivation and ability to engage in such lengthy processes requiring strict adaptation and endurance of mental and physical burdens and consequences. While other factors, such as depression and anxiety, due to their state-like nature, and also their proneness to be influenced by sexual hormones, are likely to fluctuate both as a function of treatment and as a consequence of previous failure and the prospect of a difficult treatment, stable traits, such as affective temperaments may be useful to predict behaviours and psychological reactions during treatment, which may, in turn, influence medication adherence and eventually treatment outcome. While affective temperaments are non-modifiable, they can be used to predict behaviours and emotional reactions with a known negative impact on adherence, and preventive measures and interventions to counteract such consequences can be designed accordingly in a personalised way. Thus, with careful pre-intervention evaluation and screening, healthcare professionals can offer patients personalised support programs, involving psychoeducation, coaching, psychological support, emotion regulation techniques, and motivation techniques in addition to psychotherapy where needed, and thus maximise ART success. Considering that these interventions are also costly, implementing psychological screening to identify patients where they are most needed and are likely to be most effective also significantly increases cost-effectiveness. Translating all this into real-world clinical practice: routine screening for affective temperaments in infertility care could help identify women at higher risk of non-adherence due to stress or emotional burden. Based on these profiles, tailored interventions—such as psychological counselling, stress-management, and improved patient–doctor communication—could be applied to support adherence. Integrating psychologists into the multidisciplinary fertility team would facilitate such personalised care, potentially reducing dropout and improving pregnancy outcomes. One strength of the present study lies in investigating affective temperaments in a population whose adherence to medication has not yet been extensively studied, especially in relation to temperaments. Further strong points of the study are its prospective design and acceptable sample size. Our study also has limitations that should be considered when interpreting our findings, mainly that all our measurements are based on self-reports, which gives rise to significant reporting bias. Also, a limitation of this study is the lack of a formal assessment of intellectual disability, which may have gone undetected in some participants and could have influenced the results. To conclude, we have demonstrated that affective temperaments influence infertility treatment adherence, representing an important facet of the psychological ingredients of active patient cooperation in a widespread condition of public health significance. By screening for affective temperament profiles, it would be possible to identify patient groups at high risk of drug non-adherence and then to aid adherence by applying patient-tailored treatment, including psychological interventions. This could ultimately increase the chances of successful pregnancies among women undergoing in vitro fertilisation.