Life satisfaction in parents after successful in vitro fertilization and natural pregnancy: a prospective longitudinal study from before pregnancy to 24 months after childbirth

Abstract

Changes in psychological variables in couples after successful in-vitro fertilization (IVF) have rarely been investigated. This paper describes follow-up assessments of a previously published study investigating changes in life satisfaction, stress and worry related to childbirth in couples undergoing IVF and those with natural pregnancy. Questionnaire data were obtained in 75 IVF couples and 70 couples with natural pregnancy before pregnancy, and at 6 and 12 months postpartum; follow-up data were recorded 18 and 24 months postpartum. IVF couples had less favourable baseline scores for all variables than those with natural pregnancy. Their life satisfaction increased, stress and worry decreased, during the first year postpartum. Couples with natural pregnancy reported transient worsening in all variables during this period. During follow-up, all variables remained largely stable in both groups; while life satisfaction and stress no longer differed between groups, worry was lower in IVF couples at month 24. Gender differences were small in both groups. In IVF couples, negative impacts of infertility may fully abate after childbirth; in naturally conceiving couples, initial negative changes in wellbeing are reversed and stabilized during the child´s first 2 years. Both groups may benefit from psychological support at different times during pregnancy and parenthood.

Keywords:

• In vitro fertilization
• pregnancy
• childbirth
• life satisfaction
• stress
• worry

Introduction

The birth of a child involves major changes in couples´ lives, necessitating substantial effort for readjustment (Nomaguchi & Milkie, 2003). Numerous studies have investigated wellbeing and life satisfaction in couples following childbirth, and mostly reported transient negative rather than positive changes (Nomaguchi & Milkie, 2020). This may be explained by increasing stress burden due to childcare responsibilities, sleep deprivation, partnership conflicts, or reduced time for leisure and social activities, for example (Knoester & Eggebeen, 2006; Pollmann-Schult, 2014). Research suggests that these demands outweigh rewarding aspects like positive feeling towards the new family member or personal growth (Nomaguchi, 2012; Umberson et al., 2010).

The above research was almost entirely concerned with couples having children through natural pregnancy; little is known about couples who become parents via medically assisted reproduction. This is important given that 10–15% of couples in the industrialized world experience infertility, and a large proportion of them opt for medical measures, particularly in-vitro fertilization (IVF), to fulfill their desire to have a child (Butler & Khanna, 2003; Evers, 2002; Ragni et al., 2005). In approximately 35% of these couples, the treatment is successful (Society for Assisted Reproductive Technology (SART), 2016). There is strong evidence of reduced life satisfaction and increased distress and symptoms of mental disorders in couples affected by infertility (Deka & Sarma, 2010; Fisher & Hammarberg, 2017). In contrast, research on changes in wellbeing and life satisfaction after successful IVF is scarce, particularly longitudinal studies (Klock & Greenfeld, 2000; Sydsjö et al., 2002).

This paper describes the results of follow-up assessments of a longitudinal study comparing changes in life satisfaction, stress and worry between couples with successful IVF and those having a child through natural pregnancy (Kiesswetter et al., 2022). While the first part of the study investigated the immediate effects of pregnancy and childbirth (period before pregnancy until 12 months after birth), this paper covers changes occurring during the entire study period (before pregnancy until 24 months after birth), with a particular focus on the second year after birth. At baseline, couples preparing for IVF reported lower life satisfaction, and higher stress burden and worry, than those awaiting natural pregnancy. The more negative situation of couples with IVF may be ascribed to stressors related to infertility and the psychological and physical burdens imposed by the treatment (Dooley et al., 2014; Gourounti et al., 2011). However, these couples exhibited a steep increase in life satisfaction, and reductions in stress and worry, at 6 and 12 months after birth. In couples experiencing natural pregnancy, life satisfaction strongly decreased 6 months after childbirth, but increased again by month 12, while stress and worry increased at month 6 and decreased at month 12. Regarding the different time courses of the variables in both groups, it may be that couples with IVF perceive the long-awaited pregnancy and childbirth as more rewarding overall than couples with natural pregnancy; in turn, this may ameliorate the effects of stressors related to the newborn child and negative aspects of the new role as parents.

Hypotheses pertaining to changes in couples with IVF and natural pregnancy during the second year after childbirth can be formulated based on setpoint theory, which posits that wellbeing fluctuates around an individually determined setpoint (Anusic et al., 2014; Diener et al., 2006; Ormel et al., 2017). While critical life events can cause substantial distress, readjustment processes foster a return of wellbeing to the baseline level during the years following the event. In couples with natural pregnancy, this is reflected by increasing life satisfaction and decreasing stress and worry at the end of the first year after birth. According to setpoint theory, stabilization of these variables at baseline level may be expected during the second year. However, in couples undergoing IVF, the situation is different. A longer-lasting burden due to infertility and acute stress during IVF treatment may cause life satisfaction to fall below the setpoint level. Therefore, it is unlikely that life satisfaction will return to this unfavourable state at baseline during the second year. Assuming a similar setpoint in infertile couples as in those with natural pregnancy, it may be hypothesized that life satisfaction, stress and worry reach similar plateaus in both groups during the second year after childbirth.

Although relevant, gender effects for the variables assessed in the study have not been analyzed thus far. Based on the literature, only minor differences between women and men may be expected in life satisfaction (Batz & Tay, 2018; Graham & Chattopadhyay, 2013; Pavot & Diener, 1993). While there is no convincing evidence of gender differences in stress burden and worry in naturally conceiving couples (Feinberg et al., 2019; Johansson et al., 2020), it was reported that distress due to infertility is greater in women than men, which may relate to differences in social pressure, stigmatization, and coping strategies (Greil et al., 2010). Moreover, women experience significant physical stress during invasive IVF treatment (Miller et al., 2019). Therefore, for IVF couples, we hypothesized that stress burden and worry are greater in women than men before pregnancy and childbirth, and that gender differences decrease across the observation period.

In summary, during the observation interval ranging from before pregnancy to 12 months after childbirth, couples with successful IVF reported large increases in life satisfaction, and decreases in stress burden and worry (Kiesswetter et al., 2022). In couples with natural pregnancy, an initial decrease in life satisfaction and decreases in stress and worry were seen, followed by a return toward baseline values. This paper focuses on follow-up measurements performed during the second year after childbirth. Based on setpoint theory, stabilization of life satisfaction, stress and worry was expected at a plateau, which essentially does not differ between couples with IVF and natural pregnancy. Higher levels of stress and worry were expected in women than men, especially in couples with IVF during the initial phase of the observation period.

Materials and methods

Study design

In this prospective longitudinal study, changes in psychological variables in couples with successful IVF and natural pregnancy were investigated. Data from the initial part of the observation period [three measurement occasions: before pregnancy (T1), 6 months after childbirth (T2), and 12 months after childbirth (T3)] have been published elsewhere (Kiesswetter et al., 2022). In this paper, results for the entire observation interval, including two additional measurement occasions during the second year after birth [18 months after birth (T4) and 24 months after birth (T5)] are delineated. Questionnaire data on the life satisfaction, stress burden, and levels of worry of couples who had a child after IVF treatment (IVF group) or without medical assistance (Natural Pregnancy group) were obtained.

Participants

In total, 145 couples participated in the study (75 in the IVF group and 70 in the Natural Pregnancy group). The participants of the IVF group were recruited via three IVF centres in Trentino - South Tyrol (Italy) that are part of the public health system. Six IVF attempts in Italy are covered by public health insurance. The participants of the Natural Pregnancy group were recruited via the Department of Gynaecology of the Hospital of Merano, and through personal connections. The participants were bilingual or German-speaking Italians, since German is the most widely spoken language in this region of northern Italy. This allowed us to recruit participants with a relatively homogenous cultural background.

Within the IVF group, one couple divorced after T2, another lost the child after T4, and another decided to not continue the study after T4. One couple in the Natural Pregnancy group lost the child after T3 (overall dropout rate of 2.68%).

Procedure

Medical specialists in gynaecology or andrology working in reference centres recruited the participants to the IVF group, informed them about the study procedure and collected sociodemographic data before the beginning of an IVF treatment. At this point, 237 couples were recruited; they all underwent IVF at the reference centres, which was successful in 78 couples. The life birth rate of 32.9% was slightly higher than reported in the literature for this age group (Scaravelli et al., 2020). In this group, a mean of 480.83 days (SD = 190.05 days) passed between completion of the questionnaires at T1 and the birth of the child. In total, 136 couples who stated that they were planning to have a child in the near future were recruited as possible participants for the Natural Pregnancy group. On average, 71 of these couples had a child 555.65 days (SD = 152.13 days) after T1. The time between T1 and birth was shorter in the IVF than Natural Pregnancy group (t(143) = −2.61, p = 0.01, d = 0.43).

The questionnaires were presented in paper-pencil format. At T1, they were presented to the couples by the doctors of the reference centres or the lead researcher (M.K.). At T2–T5, the couples were contacted by phone, received the questionnaires in a pre-franked envelope and returned them to the lead researcher. All questionnaires were checked after completion. In cases with missing responses to any of the items, the questionnaires were handed back to the participants, and they were asked to complete them. For this purpose, participants were initially informed by phone and then received the questionnaires once again by mail.

The study protocol was approved by the Ethics Committee of the Public Health Authorities of South Tyrol (Italy) and all subjects provided written informed consent prior to participation.

Questionnaires

Life satisfaction was quantified using the Life Satisfaction Questionnaire (LSQ) [Fragebogen zur Lebenszufriedenheit] (Fahrenberg et al., 2000). The LSQ includes 10 scales representing the major life domains: Health, Work and employment, Financial position, Leisure time, Marriage and partnership, Relationship with own children, Own person, Sexuality, Friends and relatives, and Own flat. However, since only a small proportion of the sample had children, the Relationship with own children scale was not used. The 70 items comprising the questionnaire relate to the present time and are answered on a 7-point Likert scale ranging from 1 (I don’t agree) to 7 (I fully agree). In addition to the scores for each individual scale, a total life satisfaction score can be calculated. In the computation or this sum score, the Work and employment, Marriage and partnership, and Relationship with own children scales are not considered. This procedure was chosen by the authors of the questionnaire because a relatively high number of individuals are unable to respond to all items of these scales due to their particular life situation. Higher scores on each scale indicate higher satisfaction. In this study, the Cronbach’s alpha values of the scales ranged between 0.81 and 0.96 for T1, 0.76 and 0.95 for T2, 0.73 and 0.95 for T3, 0.72 and 0.92 for T4, and 0.73 and 0.93 for T5.

The German version of the Perceived Stress Questionnaire (PSQ) (Fliege et al., 2001) was used to record stress levels. The PSQ was designed to quantify perceived stress using four scales, each with five items (4-point Likert scale: 1= never, 2 = sometimes, 3 = often, 4 = mostly). The Worry scale assesses worry and fear about the future, as well as feelings of hopelessness and depression. The Tension scale deals with arousal, fatigue, and the inability to relax, the Joy scale with positive feelings, namely happiness, vitality, and security, and the Demands scale with perceptions of environmental stressors, such as time pressure and overwork. In this study, only the sum score of the questionnaire was included in the analysis. Higher scores denote higher stress levels. The Cronbach’s alpha was 0.90 for T1, 0.89 for T2, 0.88 for T3, 0.85 for T4, and 0.89 for T5.

The Penn State Worry Questionnaire (PSWQ) (Meyer et al., 1990) is a well-established measure of self-reported worry, which is often associated with distress and negative emotional states. In this version of the PSWQ, 15 items pertaining to the last week are answered using a 6-point Likert scale (0 = never, 5 = almost always). Higher scores indicate greater worry. The Cronbach’s alpha was 0.89 for T1, 0.87 for T2, 0.89 for T3, 0.87 for T4, and 0.87 for T5.

Data analysis

As the study groups significantly differed in age (see Results section), age was controlled in the analysis of the questionnaire data. In a first step, linear regression models were computed for all variables, with age as a predictor and LSQ subscale and sum scores, PSQ subscale and sum scores, and PSWQ score as dependent variables. Standardized residuals – which are independent of age – resulting from the regression equations were computed (Cohen et al., 2003). All analyses described below were conducted with the residuals as dependent variables.

Changes in the questionnaire scores of both study groups across the five measurement occasions were analyzed using repeated-measures ANOVA. Due to behaviour-influencing factors that have to be assumed to exist for both partners of a couple, and thus the interdependence of self-reports, gender groups were considered as paired rather than independent samples (Kenny et al., 2006). Therefore, the ANOVAs included gender (women vs. men) and measurement occasion (T1–T5) as repeated-measures factors, and study group (IVF group vs. Natural Pregnancy group) as the between-subjects factor. The residuals of the LSQ sum score, PSQ sum score, and PSWQ score served as dependent variables (see Supplemental Material for ANOVAs pertaining to the LSQ and PSQ subscales). The Greenhouse–Geisser correction was applied. The original degrees of freedom and corrected p values are provided.

The effect sizes of the ANOVAs are indicated by partial Eta squared (η²_p), where values ≥0.01, ≥0.06 and >0.14 denote small, medium and large effect sizes, respectively (Cohen, 1988). Cohen´s d is reported as the effect size index for post-hoc t-tests. d values between 0.2 and 0.5 indicate small effects, while values between 0.5 and 0.8 indicate medium effects and those exceeding 0.8 indicate large effects (Cohen, 1988). Alpha was set at 0.05 in all analyses. Given the possibility of Type I error inflation related to multiple statistical testing, the use of a lower significance threshold was considered. However, this would have substantially reduced the power of the tests, i.e. increased the chance of Type II errors and reduced the probability of detecting actual effects (Rothman, 2014). Rather than focusing solely on statistical significance, we also used effect size as an indicator of the relevance of effects.

For 12 couples in the Natural Pregnancy group, the T4 (month 18) and T5 (month 24) measurements took place during the COVID-19 pandemic (March 2020–March 2021). To account for possible bias due to pandemic-related factors, the same analyses were performed without these 12 couples; the results of these analyses are presented in the Supplemental Material (Table S3 and Figures S14–S16).

Results

The numbers of couples included in each stage of the study are displayed in the flow diagram of Figure 1. The demographic characteristics of the participants are summarized in Table 1. The IVF group was significantly older than the Natural Pregnancy group; this was the case for both the women and the men. Moreover, there was a significant difference in parity, where more couples from the IVF group were having their first child. No significant group differences were observed for marital status, gender of the child, income, duration of education or age of the children on the five measurement occasions (Table 1).

Figure 1. Flow diagram of the prospective longitudinal study with the numbers of couples, and the numbers of and reasons for dropouts, at each stage of the study.

Table 1. Demographic characteristics of the participants; statistics of the group comparisons.

	IVF group		Natural Pregnancy group
	n		n
Marital status					χ^2	p	V
Married	20 Couples		22 Couples		0.80	0.37	0.05
Living together	55 Couples		48 Couples
Gender of the child					χ^2	p	V
Male	30		33		0.75	0.39	0.07
Female	45		37
Parity					χ^2	p	V
Primipara	65		50		5.12	0.02	0.19
Multipara	10		20
Income (€)					U	p	z
<15,000	25		16		9447	0.11	−1.61
15,000–30,000	63		88
30,000–45,000	48		28
45,000–70,000	11		7
>70,000	3		1
	IVF group		Natural Pregnancy group		t[143]	p	d
	M	SD	M	SD
Men: age (years)	37.07	5.40	33.71	4.88	3.92	<0.001	−0.65
Women: age (years)	33.73	3.66	30.43	4.15	4.98	<0.001	−0.85
Men: duration of education (years)	14.88	4.55	14.33	3.67	0.80	0.43	−0.13
Women: duration of education (years)	15.19	4.07	15.89	3.09	−1.16	0.25	0.19
Age of child (T2) (months)	6.08	0.49	6.03	0.48	0.64	0.52	−0.10
Age of child (T3) (months)	12.03	0.49	12.03	0.17	−0.3	0.98	0.00
Age of child (T4) (months)	18.00	0.16	17.99	0.21	0.65	0.52	−0.05
Age of child (T5) (months)	24.00	0.23	23.99	0.22	0.27	0.79	−0.04

Note: M: mean; SD: standard deviation; d: Cohen´s d; χ²: Chi-square; V: Cramér’s V; U: Mann–Whitney U value; z: z value.

All statistical results presented in the following are based on analyses using residuals derived from linear regression. However, in Figures 2–5, the original data are displayed. Table 2 shows the results of repeated-measures ANOVAs of the LSQ sum score, PSQ sum score and PSWQ score. The results of post-hoc group comparisons at the five measurement occasions are provided in Table 3.

Figure 2. Changes in Life Satisfaction Questionnaire (LSQ) sum score over time for the IVF group and the Natural Pregnancy group (T1: before pregnancy, T2: 6 months after birth, T3: 12 months after birth; T4: 18 months after birth; T5: 24 months after birth; bars denote standard errors of the mean); higher scores indicate higher life satisfaction.

Figure 3. Changes in Perceived Stress Questionnaire (PSQ) sum score over time for the IVF group and the Natural Pregnancy group (T1: before pregnancy, T2: 6 months after birth, T3: 12 months after birth; T4: 18 months after birth; T5: 24 months after birth; bars denote standard errors of the mean); higher scores indicate greater stress levels.

Figure 4. Changes in Penn State Worry Questionnaire (PSWQ) sum score over time for the IVF group and the Natural Pregnancy group (T1: before pregnancy, T2: 6 months after birth, T3: 12 months after birth; T4: 18 months after birth; T5: 24 months after birth; bars denote standard errors of the mean); higher scores indicate greater worry.

Figure 5. Changes in Penn State Worry Questionnaire (PSWQ) sum score over time for the men and women (T1: before pregnancy, T2: 6 months after birth, T3: 12 months after birth; T4: 18 months after birth; T5: 24 months after birth; bars denote standard errors of the mean); higher scores indicate greater worry.

Table 2. Results of repeated-measures ANOVAs of the Life Satisfaction Questionnaire sum score (LSQ), Perceived Stress Questionnaire sum score (PSQ) and Penn State Worry Questionnaire score (PSWQ); main effects of time, group and gender and interaction effects between these factors are indicated.

	LSQ			PSQ			PSWQ
	F (df)	p	η^2p	F (df)	p	η^2p	F (df)	p	η^2p
Time	131.05 (4,572)	<0.001	0.48	57.71 (4,572)	<0.001	0.29	15.25 (4,572)	<0.001	0.10
Group	2.57 (1,143)	0.11	0.02	0.01 (1,143)	0.91	0.00	0.25 (1,143)	0.62	0.00
Gender	0.043 (1,143)	0.84	0.00	2.22(1,143)	0.14	0.02	0.78 (1,143)	0.38	0.01
Time × Group	108.38 (4,572)	<0.001	0.43	9.78 (4,572)	<0.001	0.06	8.70 (4,572)	<0.001	0.06
Time × Gender	1.42 (4,572)	0.24	0.01	1.87 (4,572)	0.14	0.01	3.47 (4,572)	0.02	0.02
Group × Gender	1.04 (1,143)	0.31	0.01	0.11 (1,143)	0.75	0.00	1.32 (1,143)	0.25	0.01
Time × Group × Gender	2.34 (4,572)	0.08	0.02	2.53 (4,572)	0.06	0.02	0.28 (4,572)	0.83	0.00

Note. Bold p values indicate significant effects; LSQ: Life Satisfaction Questionnaire; PSQ: Perceived Stress Questionnaire; PSWQ: Penn State Worry Questionnaire; η²_p: partial eta squared.

Table 3. Results of post-hoc t-tests comparing the Life Satisfaction Questionnaire sum score (LSQ), Perceived Stress Questionnaire sum score (PSQ) and Penn State Worry Questionnaire score (PSWQ) between the two study groups (IVF group vs. Natural Pregnancy group) at all measurement occasions.

	LSQ			PSQ			PSWQ
	t (288)	p	d	t (288)	p	d	t (288)	p	d
T1	−14.52	<0.001	1.71	3.72	<0.001	0.48	3.58	<0.001	0.42
T2	4.98	<0.001	0.58	−2.27	0.01	0.27	0.23	0.82	0.20
T3	2.19	0.02	0.25	1.15	0.25	0.13	0.54	0.59	0.06
T4	0.56	0.29	0.07	−0.93	0.35	0.11	0.84	0.40	0.10
T5	0.19	0.43	0.02	−1.42	0.16	0.17	−2.23	0.03	0.26

Note. d: Cohen´s d; LSQ: Life Satisfaction Questionnaire; PSQ: Perceived Stress Questionnaire; PSWQ: Penn State Worry Questionnaire; bold p values indicate significant effects.

The main research question of the study involved changes in life satisfaction, stress burden and worry in couples undergoing IVF, and in those with natural pregnancy across the period before pregnancy and 24 months after childbirth. Moreover, gender effects on the assessed variables were analyzed. The ANOVA for the LSQ sum score revealed a main effect of time (large effect size), reflecting overall improvement of self-reported life satisfaction across the five measurement occasions for the entire sample (Figure 2). Moreover, a time by group interaction effect arose (large effect size), indicating that the time courses differed between the two groups. According to post-hoc testing, the LSQ sum score was lower in the IVF than the Natural Pregnancy group at T1 (large effect size). The score increased in the IVF group and decreased in the Natural Pregnancy group between T1 and T2, which led to a higher value in the former group at T2 (medium effect size). The LSQ sum score progressively increased between T2 and T5 in both study groups. While it was still higher in the IVF than Natural Pregnancy group at T3 (small effect size), the group difference disappeared at T4 and T5.

In the ANOVA of the PSQ sum score, a main effect of time was seen (large effect size), reflecting an overall decrease of perceived stress across the observation interval for the entire sample (Figure 3). A time by group interaction effect also arose (medium effect size). Post-hoc testing revealed a higher PSQ sum score in the IVF than Natural Pregnancy group at T1 (small effect size). In the IVF group, the score progressively decreased between measurement occasions T1 and T5. In contrast, in the Natural Pregnancy group it increased between T1 and T2 and exceeded the value of the IVF group at T2 (small effect size). The score decreased again thereafter and did not differ from the values of the IVF group at T3–T5. The results pertaining to the LSQ and PSQ subscales are provided in the Supplemental Material (Tables S1 and S2; Figures S1–S13).

A time effect (large effect size) was also seen for the ANOVA of the PSWQ score, suggesting that worry decreased across the five measurement occasions for the entire sample (see Figure 4). Moreover, a time by group interaction effect was seen (medium effect size). According to post-hoc testing, the PSWQ score was higher in the IVF than Natural Pregnancy group at T1 (small effect size). The PSWQ score progressively decreased across measurement occasions T1–T5 in the IVF group. The Natural Pregnancy group exhibited only minor changes in this score. The score slightly increased between T1 and T2, decreased between T2 and T3, and increased again between T3 and T5. The PSWQ score of the Natural Pregnancy group exceeded that of the IVF group at T5 (small effect size).

The ANOVA for the PSWQ score also revealed a time by gender interaction (small effect size), denoting a difference in the time course of worry between women and men (Figure 5). The PSWQ score progressively decreased between T1 and T3 in both genders. Between T3 and T4, it further decreased in women but increased in men, leading to a higher value in the latter group at T4 (small effect size; Table 4). Although the PSWQ score was still somewhat higher in men, the group difference was no longer significant at T5. The three-way interaction (time by group by gender) of the ANOVA did not reach significance, suggesting that the gender differences in time course were similar between the IVF and Natural Pregnancy groups.

Table 4. Results of post-hoc t-tests comparing the Penn State Worry Questionnaire (PSWQ) sum scores between genders at all measurement occasions.

	PSWQ
	t (288)	p	d
T1	−0.48	0.63	0.06
T2	−0.48	0.63	0.06
T3	0.40	0.69	0.05
T4	2.25	0.03	0.26
T5	1.40	0.16	0.17

Note. d: Cohen´s d; PSWQ: Penn State Worry Questionnaire; bold p values indicate significant effects.

The same analyses were performed without the 12 couples for whom the T4 (month 18) and T5 (month 24) measurements took place during the COVID-19 pandemic. The analyses excluding these couples yielded essentially the same results as those of the entire sample (for detailed information, see Supplemental Material Table S3 and Figures S14–S16).

Discussion

This study investigated changes in life satisfaction, stress burden and worry related to the birth of a child in couples with successful IVF treatment and those with natural pregnancy. In the first part of the study, which covered the period before pregnancy to 12 months after childbirth, a markedly lower LSQ sum score, and a higher PSQ sum score and PSWQ score, were seen in couples undergoing IVF treatment than those awaiting natural pregnancy (Kiesswetter et al., 2022). Time courses of all assessed variables substantially differed between the groups during the period immediately following childbirth. Couples with natural pregnancy showed a marked decrease in the LSQ sum score, and increases in the PSQ sum score and PSWQ score, at month 6; all scores tended to return to baseline at month 12 after childbirth. In contrast, a progressive increase in the LSQ sum score, and decreases in the PSQ sum score and PSWQ score, arose at months 6 and 12 in couples with successful IVF. The present paper focused on follow-up assessments conducted during the second year after childbirth. As hypothesized, all scores stabilized during this period. While the group differences in the LSQ and PSQ sum scores disappeared, the PSWQ score was higher at month 24 in couples with IVF than those with natural pregnancy. As a secondary research question, gender effects on the variables were explored. Unexpectedly, these effects were small overall in both groups. Different time courses between women and men were only seen for the PSWQ score, where men exhibited higher scores than women at month 18 after childbirth.

Previous longitudinal studies investigating changes in psychological variables in couples with IVF and natural pregnancy were limited by relatively short observation intervals. For example, Sydsjö et al. (2002) recorded psychological and medical data at two measurement occasions, i.e. during pregnancy (between the 15th and 20th gestational week) and 12 months after childbirth. Couples with natural pregnancy, but not those with successful IVF, exhibited a decrease in marital satisfaction between the two measurement occasions. Klock and Greenfeld (2000) reported that self-esteem increased, while anxiety decreased, from the 12th to 28th gestational week in women with IVF but not in those with natural pregnancy. Moreover, Gameiro et al. (2011) observed a decrease in marital congruence between the gestational week and 4 months after childbirth in couples with IVF and natural pregnancy.

The findings of the present study illustrate that it is essential to monitor psychological variables over longer periods after childbirth. In accordance with our main hypothesis, life satisfaction no longer differed between couples with IVF and natural pregnancy during the second year of the child´s life. This finding can be discussed in the context of setpoint theory, which postulates that wellbeing tends to return to baseline during the years following a critical life event (Anusic et al., 2014; Diener et al., 2006; Ormel et al., 2017). Studies testing this prediction reported mixed results, where the degree of re-adjustment depends on the nature of the life event (Anusic et al., 2014; Diener et al., 2006; Ormel et al., 2017). While wellbeing does not necessarily return to the pre-event level after disability or widowhood, this seems to apply mostly to events like marriage and childbirth (Anusic et al., 2014). The return of self-reported life satisfaction to baseline at month 12 seen in this study, and the stabilization thereof during the second year post-partum in naturally conceiving couples, may be explained by successful coping with the demands of the new family situation and requirement to care for the infant (Lawrence et al., 2008). Certainly, positive feelings towards the baby facilitate adjustment to the new role and help the parents deal with daily challenges (Cast, 2004; Luhmann et al., 2012). Moreover, the intensity of the demands may decrease during the second half of the infant´s first year of life. For example, changes in sleep-wake rhythm and food intake that occur after the 6th month in most infants are associated with a reduction of physical demands on the parents (Dias et al., 2018; Grueger, 2013). The pattern of changes in reported stress and worry may be explained in a similar way. While changes in the social role, partnership interaction, responsibilities and daily physical stressors may increase stress and worry immediately after childbirth, the couples´ resources seem to be sufficient to adjust successfully, such that life satisfaction is restored and stabilized.

From the outset, the situation of couples who become parents through IVF is clearly different from those with natural pregnancy. The former group often goes through a long period of uncertainty and suffering, which causes a substantial decline in well-being (Kiesswetter et al., 2020; Ragni et al., 2005). In these couples, the crucial life event may not be the birth of the child but rather the diagnosis of infertility, which for many affected individuals is perceived as a disability (Goldschmidt et al., 2003). According to the literature, adjustment to disability is a relatively slow process, and in many cases wellbeing does not return to the pre-event state (Anusic et al., 2014; Lucas, 2007). As couples for whom the IVF intervention failed were not included in this study, no conclusions regarding adjustment to persistent infertility can be drawn. However, in our couples with successful IVF, childbirth was followed by a marked improvement of life satisfaction and less stress and worry; this persisted into the second year of life of the child. These couples also face the typical challenges associated with the birth of a new family member, but it seems that these demands are outweighed by the positive experience of birth of a long-desired child and relief of the burdens of infertility and IVF treatment.

It is important to note that indicators of life satisfaction and stress burden did not differ between our IVF and natural pregnancy couples during the second year after childbirth. In couples with IVF, these variables were not assessed before the diagnosis of infertility, so it cannot be definitively concluded that they returned to the initial state (i.e. the individual setpoint). Nevertheless, the observation that life satisfaction and stress reached similar plateaus to those seen in the Natural Pregnancy group supports the notion that the negative experiences of infertility fully abate during the second year after childbirth. Couples with IVF even reported less worry 24 months after childbirth than those with natural pregnancy. It may be that these couples often experience a long period of distress, which may promote the acquisition of strategies to deal with a wide variety of burdens, such that they may be less likely to be affected by the typical worries related to caring for a young child (Fido & Zahid, 2004; Jordan & Revenson, 1999; Ying et al., 2017).

Differences between women and men in the assessed variables were small overall. No gender effects or interactions between time and gender arose for the LSQ and PSQ sum scores. Studies on gender differences in life satisfaction have reported inconsistent results (Graham & Chattopadhyay, 2013; Helliwell & Putnam, 2004; Jovanović, 2019). Findings pertaining to gender differences in parental well-being are also contradictory; while some studies suggested a more positive impact of parenthood in women (Baranowska & Matysiak, 2012; Clark et al., 2008; Clark & Georgellis, 2013), others reported similar impacts between mothers and fathers (Angeles, 2010; Pollmann-Schult, 2014). Regarding stress, some studies reported a higher stress burden after childbirth in women than men (Cowan et al., 1985; Lu, 2006). However, according to more recent research, stress substantially increases in both men and women, where the impacts of particular stressors rather than stress levels per se may differ between genders (Allan et al., 2021; Feinberg et al., 2019). For example, fathers were more affected by interference of childcare with work, whereas mothers tended to suffer more from partner conflicts and emotional stressors (Feinberg et al., 2019; Johansson et al., 2020). By definition, gender roles have changed in the recent past (Morgenroth & Ryan, 2018). Traditionally, fathers were regarded as the primary breadwinners and had limited involvement in parenting; childcare was stereotypically the mother’s responsibility. In modern society, greater participation of fathers in childcare, and of mothers in professional life, may have mitigated differences in perceived stress between the genders. The lack of a gender difference in stress in our IVF group at the first measurement occasion (i.e. before IVF treatment) is particularly surprising. As aversive physical effects of IVF are limited to women, one may expect them to experience greater distress during the period of anticipation and preparation for the procedure than their partners. However, our finding supports the notion that men are also extremely emotionally involved in IVF treatment and the potential outcome of the intervention (Boivin & Schmidt, 2005; Connolly et al., 1993; Laffont & Edelmann, 1994). Interestingly, fathers reported greater worry than mothers 18 months after childbirth. The size of this effect was small, and interpretations must remain speculative. At this age, most children have learned to walk, such that they can move more around more quickly and independently, which may be associated with heightened perceptions of danger of the parents. It may be that fathers are particularly involved in activities involving the child outside the home, such that they may experience increased worry related to the typical activities that children of this age take part in.

A limitation of this study was that some constructs that should be taken into account when interpreting the findings (e.g. gender roles) were not assessed; as such, the explanations provided must remain hypothetical. Although age was controlled for in the statistical analysis, the older age of the IVF than Natural Pregnancy group must also be seen as a restriction. However, this also applies to previous studies comparing IVF and natural pregnancy couples (Gibson et al., 2000; Sydsjö et al., 2008). In addition, the interval between the first two measurement occasions (i.e. before pregnancy and 6 months after childbirth) was longer, and the number of couples that already had children was higher, in the Natural Pregnancy group than IVF group. The presence of one or more older children in the family may have multiple effects. On one hand, this may intensify everyday physical and psychological demands, increasing the couples' stress burden and worries and reducing life satisfaction. On the other hand, previous experience with the challenges of a newborn child may facilitate adjustment to the new situation and thus enable faster return of psychological variables to the initial state. Various sample characteristics potentially relevant to the research question were not recorded, including the mental and physical health of the couples, the duration of their relationship, the strength of their desire to have a child, religiosity, resilience and social support. Possible effects of dropout should also be considered. Among the four couples who had to be excluded from the study, one couple in the IVF group and one couple in the Natural Pregnancy group lost the child during the observation interval, and one couple from the IVF group divorced. By definition, these experiences constitute critical life events that can strongly reduce life satisfaction and increase stress and worry. Therefore, distortion of the results due to the exclusion of these couples cannot be ruled out. However, the dropout rate (2.68%) was lower than that reported in other follow-up studies on changes in psychological variables after childbirth (Durtschi et al., 2017; Hogström et al., 2012; Sydsjö et al., 2008), such that these effects may be relatively small. All couples from the IVF group were recruited at Italian public IVF centres. In Italy, six IVF attempts are financed by the public sector. Therefore, the generalizability of the results to countries where couples have to bear the full costs of IVF treatment may be restricted. It is feasible that in these countries, financial burden due to the treatment constitutes an additional stressor having a potential negative impact on life satisfaction. The relief experienced by couples after successful IVF may be even greater than in Italian couples, which in turn might result in even stronger positive effects on life satisfaction. Another limitation relates to recruitment of the Natural Pregnancy group from a single hospital and through personal connections, which restricts the representativeness of the sample and generalizability of the findings. Finally, for 12 couples in the Natural Pregnancy group, the last two measurement occasions were during the COVID-19 pandemic. However, the analysis excluding these couples yielded essentially the same results as that of the entire sample (for detailed information, see Supplemental Material).

In conclusion, this follow-up study confirmed that, despite differences in the initial situation (before pregnancy) and time courses during the first year of the child’s life, relevant psychological variables reached similar plateaus during the second year between couples with successful IVF and those with natural pregnancy. While differences in life satisfaction and stress between the groups disappeared during the second year, a tendency toward lower levels of worry was seen in couples with IVF at the end of the observation period. The positive changes and subsequent stabilization of all variables seen in these couples suggest that the negative psychological impacts of infertility may be fully compensated after childbirth.

Finally, some practical and clinical implications of our findings may be discussed. The observations suggest that psychological support may be helpful for parents with IVF and natural pregnancy at different times during pregnancy and parenthood. In couples with IVF, psychological interventions may be most important before and during the treatment, when stress burden and worry are high, and life satisfaction is markedly reduced. Such interventions may include psychoeducation, stress management, relaxation techniques or cognitive restructuring (Belevska, 2015; Czabała & Miedziun, 2016). Education of healthcare professionals working in IVF centres should foster awareness of the psychological burdens of infertile couples, and the stress related to the treatment. Moreover, it is essential that couples are involved in all medical decisions during the treatment process; an active participation can increase couples’ perception of control over the procedure, and thus reduce intervention-related stress and worries (Kiesswetter et al., 2020). Couples with natural pregnancy may benefit from psychological measures after birth when the psychological burden increases. Such interventions could aim to enhance the couples´ adjustment to their new situation and ability to deal with the demands of daily life, for example by providing coping strategies and methods taken from couple therapy, such as communication training or the strengthening of mutual support (Randall et al., 2010). Work of professionals in maternity care and child healthcare also requires detailed knowledge about the psychological changes during the initial period of parenthood. For example, in Italy prenatal classes mostly focus on medical aspects of the birth, such as the different birth positions, pain control and surgical interventions (forceps, ventouse or C-section) and care provided to the infant, like breastfeeding, diapering and umbilical care (Südtiroler Sanitätsbetrieb, 2023). In other countries, like the United Kingdom, in these courses there is a stronger focus on psychological changes and challenges faced by parents (National Institute for Health & Care Excellence, 2021). During the second year of a child´s life, wellbeing seems to stabilize in both groups of parents; thus, psychological counselling may only be necessary for couples affected by specific psychosocial challenges. By definition, the needs of parents and small children in burdensome situations should be recognized by policymakers, and sufficient healthcare resources should be provided to ensure adequate support. Support may range from low-threshold services, such as help with administrative issues or domestic assistance, to support by specialized professionals in cases of severe psychosocial difficulties or family crises. Various pilot projects have been launched by the Italian government to provide help to parents of infants; however, evaluations thereof suggests that current measures are far from sufficient to meet the families´ demands, and easily accessible psychosocial support is clearly warranted (Ladurner et al., 2020). The implementation of such measures may help to minimize the risk of psychological and health problems for families and limit the impact on public health.

Ethical approval

The study was approved by the Ethics Committee of the Public Health Authorities of South Tyrol, Italy. Informed consent was obtained from all individual participants included in the study.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Manuel Kiesswetter, Erik Danay and Stefan Duschek. The first draft of the manuscript was written by Manuel Kiesswetter and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Acknowledgment

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript. The authors have no relevant financial or non-financial interests to disclose.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The research data of the study is available to the public via the repository Open Science Framework (OSF: osf.io/7u4qp/).