Effectiveness of double ABCX-based psychotherapy for psychological distress among women undergoing in vitro fertilization-embryo transfer: a three-arm randomized controlled trial

Abstract

Objective

This study aimed to evaluate the effectiveness of double ABCX-based psychotherapy for psychological distress during in vitro fertilization-embryo transfer (IVF-ET) among a female group (FG), couple group (CoG) and control group (CG).

Methods

A total of 201 women undergoing their first IVF-ET cycle were randomized into three groups. The 6-session intervention was delivered at each visit to the IVF clinic. The primary outcomes were depression and anxiety, and the secondary outcomes included sleep quality, serum cortisol (nmol/L) levels and the clinical pregnancy rate, which were assessed before and after the intervention.

Results

The group-by-time effects were significant for depression, anxiety, sleep quality and serum cortisol levels, with larger effect sizes in the FG than in the CoG. There was no significant difference in the pregnancy rate among the three groups.

Conclusion

Psychotherapy effectively mitigated psychological distress, suggesting greater effectiveness for couples undergoing IVF couples than for women only. It is structured and easy to use during the IVF treatment cycle.

Trial registration number

https://register.clinicaltrials.gov (NCT03931187, retrospectively registered on April 23, 2019)

IMPACT STATEMENT

• Double ABCX-based psychotherapy, which was developed in this study, effectively mitigated depression, anxiety, sleep disturbance and increased cortisol levels, suggesting greater effectiveness for couples undergoing IVF-ET than women only.

• Double ABCX-based psychotherapy is structured and easy to use during the IVF-ET treatment cycle, thereby presenting a promising and feasible approach to improve care for couples or women with assisted reproductive technique-related stress.

Keywords:

• In vitro fertilization-embryo transfer
• double ABCX-based therapy
• psychosocial intervention
• depression
• anxiety

Introduction

Infertility, defined as the failure to have a clinical pregnancy after 12 months or more of regular unprotected sexual intercourse, is estimated to affect 1 in 10 couples globally [1]. Currently, in vitro fertilization-embryo transfer (IVF-ET), which now progressed into intracytoplasmic sperm injection (ICSI), is widely performed, with more than 1.6 million cases annually. Success rates have increased over time to approximately 25–30% per treatment cycle [2–4] and to 36.2% in the Netherlands and 36.4% in the United States over three cycles [5]. The IVF-ET treatment cycle is physically and emotionally exhausting [6]. Anxiety, depression, and sleep disturbance are common and are caused by high costs, uncertain success and a lack of knowledge [7,8]. Especially in China, where Confucianism is dominant, most couples hold the traditional idea that having a child is necessary, and thus, infertility leads to greater psychological distress [9]. Therefore, psychological support for women or couples undergoing IVF-ET is necessary.

Existing psychotherapy offered to women or couples undergoing IVF-ET can be categorized into cognitive behavioral therapy, relaxation techniques and self-guided psychotherapy. Cognitive behavioral therapy has a long tradition and usually involves emotional expression, psychoeducation, relaxation training, and cognitive reconstruction [10–13]. Relaxation techniques include progressive muscle relaxation [14], yoga [15], and mindfulness therapy [16]. Self-guided psychotherapy includes the maintenance of a gratitude diary [16] and expressive writing [17]. Few meta-analyses have confirmed the efficiency of psychotherapy, particularly cognitive–behavioral therapy, in reducing psychological distress [13,18].

Based on a review of the existing intervention research, there are still several research gaps. On the one hand, most interventions have not been integrated into the IVF-ET treatment process [18,19]. Psychosocial interventions should be incorporated into routine practice to ensure sustainability and acceptability [18,19]. On the other hand, only a few recent studies have focused on couples rather than women only [20–23], and previous research has not differentiated the effectiveness between women-only and couple-based interventions [24]. Recent research has found that couple dyadic interactions have a significant influence on the effectiveness of psychotherapy [25–27]. Studies of psychotherapy incorporated into routine practice and comparing the effectiveness between women-based and couple-based approaches are worthwhile.

Infertility and IVF-ET treatment are stressful events for the whole family, where family function or marital quality affect daily dyadic interactions [25,27]. The double ABCX model was initially developed by [28], postulating that a stressor (A) interacts with family resources (B) and with the family’s appraisal of the event (C) to produce a crisis (X); the process was further developed by [29,30]. The double ABCX model, which means that the recursive ABCX procedure can involve cycles, has been used by therapists to guide the development of interventions for family adaptation to chronic health conditions among couples [31,32]. Changing family appraisal and promoting family coping may be a new but effective way to decrease psychological distress. Therefore, this study aimed to develop, evaluate and compare the effectiveness of double ABCX-based psychotherapy for psychological distress integrated into the IVF-ET treatment process among a female group (FG), couple group (CoG) and control group (CG). We aimed to test the primary hypothesis that double ABCX-based psychotherapy as an adjunct to usual care would be superior to usual care alone in improving psychological distress postintervention. The secondary hypothesis was that double ABCX-based psychotherapy would lead to greater improvement in the CoG than in the FG postintervention.

Methods

Study design

This was a single-center, three-armed, randomized, controlled trial reported based on the Consolidated Standards of Reporting Trials (CONSORT) guidelines. All procedures involving human subjects/patients were approved by the Ethical Institutional Review Board of the researchers’ university (2017055). Written informed consent was obtained from all patients.

Participants

Eligible women were recruited at the Infertility Care Center of a tertiary hospital in Changsha City, Hunan Province, China, between December 2018 and May 2019. The inclusion criteria were (1) infertility due to the female partner, (2) first-time IVF-ET with a long-acting long ovulation induction protocol in the middle luteal phase, (3) an age between 20 and 45 years old, (4) an education level above junior high school, (5) no biological or adopted children, and (6) transfer of biopsied embryos. The exclusion criteria were (1) undergoing ICSI, 2) the use of frozen gametes/embryos, (3) infertility due to the male partner, (4) diabetes, hyperthyroidism, hypothyroidism and other endocrine and metabolic diseases or treatment that could affect cortisol levels within six months, and (5) other professional psychological therapy related to depression or anxiety.

The sample size was calculated by Gpower 3.0. The MANOVA approach (repeated means, within-between interaction) was chosen, and the parameters were set as follows: two-tailed test, an effect size= 0.31 based on the mean treatment effects of combined psychological outcomes from a meta-analysis study [7], an α = 0.05, a power (1-β) = 0.9, a number of groups = 3, and a number of measurements =2. The minimum total sample size was calculated as 135. For an anticipated dropout rate of 30%, the total number of participants that needed to recruited at baseline was least 193.

Procedure, randomization and blinding

A total of 287 women were approached, and 201 women were randomly allocated to the couple group (CoG, n = 67), female group (FG, n = 67), and control group (CG, n = 67). Participants finished the baseline measurement on the day of downregulation (T₀) in a private room at the center. Downregulation was the beginning of long-acting long ovulation induction protocol in the middle luteal phase, and Gonadotropin-releasing hormone (GnRH) agonists were used before IVF-ET treatment to trigger egg development and ovulation. The FG and CoG received the first session of the intervention immediately after the evaluation on T₀. The postintervention evaluation at T₁ was conducted on the day of embryo transfer. The process of the trial is shown in Figure 1.

Figure 1. CONSORT flow chart of recruitment, exclusion, and dropout of participants.

Block randomization with a block length of 6 was applied in a 1:1:1 ratio. A research team member who did not interact with the participants independently generated the computerized random numbers of the blocks. The allocations to each block was provided in sealed envelopes to the principal investigator and then communicated to the participants. The trained nurse at the care center who was responsible for disseminating and collecting the questionnaire was blinded to the group allocation. The technician who tested serum cortisol levels was from the other testing department and was also blinded to the group allocation.

Intervention

The Medical Research Council (MRC) framework for developing and evaluating complex interventions was adopted to guide the development [33,34]. Three steps were taken successively: (1) identifying evidence though a literature review and our previously published qualitative and quantitative research findings; (2) choosing the appropriate family adaptation theories of the double ABCX-based model; and (3) modeling the process and outcome of the intervention. Furthermore, several experts in nursing, reproduction, and psychology were invited to give comments. The intervention program was revised and finalized after a pilot feasibility test of four cases.

The double ABCX-based relaxation program was designed as a single- or couple-based program, consisting of psychoeducation, experience sharing and discussion, relaxation exercises, and skills (see Figure 2). The program included six sessions, and each session lasted approximately 1 h, of which 30 min involved audio-guided relaxation for the women. The facilitator conducted each session face-to-face and individually for each involved participant. We designed the sessions to be administered at each IVF-ET time point when the participants visited the clinic. The intervention was delivered to women only (FG) or to the couple as a whole (CoG). When conducting interventions for women in the FG, their partners waited outside the room or at their living place. For the CoG, the 67 women were asked to invite their partners to participate in the discussion, experience sharing and skills practice, which were designed for mutual involvement compared with the activities in the FG.

Figure 2. Intervention process of three groups across the study.

Figure 3. Mean change in the three groups before and after intervention.

In the intervention, A indicated IVF-ET treatment, B indicated coping and resources within or outside the family, C indicated negative cognition of IVF treatment, and X indicated positive perceptions. The detailed description of each session is listed below and in Figure 2.

• (A) Session 1 involved an overview of infertility, IVF-ET, and the intervention program, which facilitated the participants’ expression of their feelings of confusion regarding the event/stressor.

• (B: within the family) Session 2 involved sharing experiences during IVF-ET, discussing coping resources and strategies, and improving ineffective coping within the family. This targeted dyadic coping and family resources.

• (B: outside the family) Session 3 included sharing experiences from past days and encouraged the utilization of external resources. This targeted family coping and external resources.

• (C) Session 4 included discussion of negative cognition of oneself and one’s partner and reflected inappropriate appraisals of infertility and IVF-ET. This targeted inappropriate perceptions of the event.

• (Double ABC) Session 5 allowed the participants to express their feelings regarding ovum retrieval and encouraged them to seek marital support, which targeted enhanced dyadic communication.

• (X) Session 6 was the conclusion of the program and encouraged positive perceptions. The intervention was implemented in a meeting room at the center by the first author, who is a qualified psychological counselor and has ten years of experience in midwifery.

The active CG received routine care, including relaxation therapy, and received a booklet with an overview of infertility and IVF-ET at the beginning of the intervention. They were also provided a 30-min relaxation program during each session, similar to the IG.

Measures

Sociodemographic data from several participants were collected at baseline. We selected depression and anxiety as indicators of psychological distress and as the primary outcomes and sleep disturbance, the serum cortisol level, and the pregnancy rate as indicators of physical adaptation and as the secondary outcomes. Both the primary and secondary outcomes were measured for women in each group.

Primary outcomes

Depression, a primary outcome, was measured by the Self-rating Depression Scale (SDS). This 20-item scale was developed by Zung in 1965, and a validated Chinese version has been widely used to assess depression [35]. The standardized score is determined by multiplying the total score of 20 items by 1.25, with a cutoff of 50. The Cronbach’s alpha coefficient (ἀ) in this study was 0.75.

Anxiety was assessed by the Self-rating Anxiety Scale (SAS). The 20-item scale was developed by Zung, and a validated Chinese version has been widely used to assess anxiety [36]. The standardized score is determined by multiplying the total score of 20 items by 1.25, with a cutoff of 50. The Cronbach’s alpha coefficient (ἀ) in this study was 0.74.

Secondary outcomes

Sleep disturbance was measured by the Self-Rating Scale of Sleep (SRSS), a widely used and validated tool [37]. It contains ten five-point items with a total score ranging from 10 to 50, with a higher score reflecting poorer sleep quality. The Cronbach’s alpha coefficient (ἀ) was 0.73.

The level of serum cortisol was selected as the endocrine indicator to reflect stress in women undergoing IVF [10]. Three-milliliter fasting venous blood samples were collected from 8:00 am to 9:00 am. The blood samples were sent to the laboratory of a tertiary hospital immediately after collection. Serum cortisol was tested using a cortisol determination kit (Siemens Healthcare Diagnostics Inc., USA) for direct chemiluminescence immunoassay using an ADVIA Centaur XP immunoassay system.

Clinical pregnancy was defined as a gestational sac detected by B-ultrasound examination four weeks after embryo transfer.

Treatment fidelity and feasibility

To ensure the homogeneity of the intervention, the facilitator for every participant in the FG and CoG was the first author. All groups received identical information about infertility, IVF-ET, and relaxation to eliminate nonspecific treatment influences on the outcomes. Feasibility was assessed by the dropout rate, as this represented how acceptable the intervention was to the participants.

Statistical analysis

Data were input and analyzed using SPSS 20.0. The chi-square test and ANOVA were used to compare sociodemographic differences in variable scores among the three groups. We conducted generalized estimation equation (GEE) analysis to examine the outcome variables among the three groups, adjusting for sociodemographic variables if there was a significant difference at baseline. Missing data were handled using the last observation carried forward method according to the ‘intention to treat’ (ITT) principle, and a per-protocol (PP) analysis on complete data was performed as a sensitivity analysis. Cohen’s d represented the effect size, with values of 0.2 to 0.5 indicating a small effect size, 0.5 to 0.8 indicating a medium effect size, and above 0.8 indicating a large effect size [38].

In addition, intervention effects regarding the primary outcomes were reported in terms of two rates: (1) clinically significant depressive/anxiety symptoms (based on a cutoff of ≥50) and (2) a minimal clinically important difference (MCID) for sum scores between measurement points [39]. The MCID was determined by the distribution-based method of standard deviation (SD). The current study employed an SD of 0.5 at baseline because an SD of 0.5 was found to approximate the upper limit of human mental discrimination ability and is the most described formula to calculate MCID values [40]. Finally, the average MCID values were found to be 4 points for depression and 3 points for anxiety. To compare the proportion of participants with these rates as well as the pregnancy rate across groups, chi-square tests were performed.

Results

Study flow and treatment feasibility

A total of 201 participants were randomly assigned to the FG, CoG and CG. Figure 1 shows the study flow chart. At baseline, except for age, there was no significant difference in sociodemographic and outcome variables among the three groups, as shown in Table 1.

Table 1. Socio-demographic variables and baseline outcomes among three groups.

	Total	FG (n = 67)	CoG (n = 67)	CG (n = 67)	χ^2 /F	P
Socio-demographic variables
Education level, n (%)
Junior high school	89 (44.2)	29 (43.3)	34 (50.7)	26 (38.8)	2.30	0.68
Senior high school	70 (34.8)	23 (34.3)	20 (30.0)	27 (40.3)
College above	42 (20.9)	15 (22.4)	13 (19.4)	14 (20.1)
Occupation, n (%)
Self-employed	93 (46.2)	30 (44.8)	35 (52.2)	28 (41.8)	4.62	0.32
Part time employed	59 (29.4)	16 (23.9)	19 (28.4)	24 (35.8)
Full employed	49 (24.4)	21 (31.3)	13 (19.4)	15 (22.4)
Monthly income (¥) , n (%)
<2000	101 (50.2)	35 (52.2)	35 (52.2)	31 (36.3)	1.30	0.85
2000-3000	83 (41.3)	26 (38.8)	28 (41.8)	29 (43.3)
>3000	17 (8.5)	6 (9.0)	4 (6.0)	7 (10.4)
Per capita housing area (m^2) , n (%)
<30	48 (23.9)	17 (25.4)	15 (22.4)	16 (23.9)	4.23	0.37
30-40	64 (31.8)	17 (25.4)	20 (29.9)	27 (30.3)
>40	89 (44.3)	33 (39.2)	32 (47.7)	24 (35.8)
Number of embryo transferred, n (%)
3	24 (11.9)	7 (10.4)	12 (17.9)	5 (7.4)	3.69	0.15
2	177 (88.1)	60 (89.6)	55 (82.1)	62 (92.6)
Causes of infertility, n (%)
Fallopian tubes related	176 (87.6)	58 (86.6)	58 (86.6)	60 (89.6)	0.83	0.93
Ovarian related	16 (8.0)	5 (7.4)	6 (9.0)	5 (7.4)
Other reasons	9 (4.4)	4 (6.0)	3 (4.4)	2 (3.0)
Age (year)	30.02 ± 4.02	31.01 ± 3.73	30.20 ± 4.46	28.85 ± 3.57	5.15	0.01
Marriage length (year)	5.92 ± 4.12	6.31 ± 4.21	6.14 ± 4.72	5.29 ± 3.29	1.16	0.31
Length of infertility (year)	4.39 ± 2.49	4.23 ± 2.29	4.16 ± 2.68	4.77 ± 2.50	1.19	0.30
Outcomes at baseline
Depression	52.35 ± 6.95	51.84 ± 6.34	51.77 ± 6.37	53.43 ± 8.00	1.22	0.30
Prevalence of depression, n (%)	120 (59.7)	41(61.1)	39 (58.2)	40 (59.7)	0.12	0.93
Anxiety	50.67 ± 5.75	50.80 ± 6.41	50.18 ± 5.54	51.05 ± 5.30	0.39	0.67
Prevalence of anxiety, n (%)	102 (50.7)	33 (49.2)	33 (49.2)	36 (53.7)	0.35	0.83
Sleep disturbance	21.55 ± 3.94	21.61 ± 3.73	22.04 ± 4.83	21.43 ± 3.50	0.77	0.46
Cortisol (nmol/L)	627.31 ± 25.58	628.35 ± 31.98	627.58 ± 17.74	630.31 ± 35.20	0.05	0.94

Note: Income is valued in Chinese currency unit, with ¥2,000 RMB≈$314 USD; Advanced age >35 years may result in later oocyte quantity, quality, and embryo/fetal developmental capacity, resulting in 3 embryos transferred versus 2 under 35 years of age.

No adverse events were reported during the study. The mean session attendance did not differ between the FG (4.71 ± 0.80) and CoG (4.65 ± 0.78). The dropout rate was approximately 23%, and the data showed that there were no differences among the groups regarding the drop-out rate.

Primary outcomes: depression and anxiety

After adjusting for several covariates, group effects, time effects and group-by-time effects were obtained from the GEE model, as summarized in Table 2, the mean change in the three groups before and after intervention was in Figure 3, and the effect sizes between and within groups are summarized in Table 3. The prevalence and minimal clinically important differences (MCID) for depression and anxiety are shown in Table 4.

Table 2. Group effect, time effect and group-by-time effect of continuous outcomes.

	Time effect		Group effect		Group × Time effect
Variable	Wald χ^2	P	Wald χ^2	P	Wald χ^2	P
Depression	15.63	0.00	1.48	0.22	9.64	0.00
Anxiety	29.25	0.00	11.61	0.00	31.87	0.00
Sleep	47.07	0.00	38.52	0.00	71.12	0.00
Cortisol (nmol/L)	2.90	0.08	4.21	0.04	7.32	0.01

Note: *p < 0.05, **p < 0.01.

Table 3. Changes and effect size between group comparisons for continuous outcomes.

		M(SD)			Mean difference (95% CI)		Cohen d (95% CI)
		FG (n = 67)	CoG (n = 67)	CG (n = 67)	FG-CG	CoG-CG	FG-CG	CoG-CG
Depression	T0	51.84 (6.34)	51.77 (6.37)	53.43 (8.00)	−1.58(−3.95, 0.77)	−1.65 (−4.08, 0.76)
	T1	49.85 (6.96)	47.78 (6.10)**	55.95 (8.68)	−5.95 (-8.47, −3.43)**	−8.11 (-10.62, −5.58)**	0.77 (0.40, 1.10) **	1.08 (0.71, 1.44) **
Anxiety	T0	50.80 (6.41)	50.18(5.54)	51.05(5.30)	−0.24 (-2.21, 1.73)	−0.85 (−2.68, 0.96)
	T1	48.56 (6.41)	47.79 (4.99)**	55.59 (7.27)**	−7.01 (−9.16, −4.86)**	−7.77 (−9.92, −5.63)**	1.02 (0.65, 1.37) **	1.25 (0.86, 1.60) **
Sleep	T0	21.61(3.73)	22.04(4.83)	21.43 (3.50)	0.17 (-1.50, 1.85)	0.60 (−1.06, 2.28)
	T1	20.67(3.06)	17.34(2.20)**	29.15(4.19)**	−8.47 (-9.82 − 7.13)**	−11.80 (−13.14, −10.46)**	2.30 (1.66, 2.89) **	3.52 (2.93,4.09) **
Cortisol (nmol/L)	T0	628.35(31.98)	627.58(17.74)	630.31(35.20)	−1.96 (-14.05, 10.12)	−2.72 (−14.81, 9.36)
	T1	675.14(57.08)**	654.69(46.20)**	711.21(49.97)**	−36.07 (-57.22, −14.92)**	−56.51 (−77.66, −35.34)**	0.67 (0.26, 0.96) **	1.17 (0.81, 1.50) **

Note: *p < 0.05, **p < 0.01; * on MD indicates the significant difference within the group from T₀ toT₁.

Table 4. Prevalence and minimal clinically important difference (MCID) of depression and anxiety, and pregnancy rate at post-intervention.

	FG (n = 67)	CoG (n = 67)	CG(n = 67)	χ^2	P
Prevalence of depression, n (%)	31 (46.2)a	22 (32.8)a	50 (74.6)b	24.41	<0.01
Prevalence of anxiety, n (%)	26 (38.8)a	22 (32.8)a	51 (76.1)b	29.49	<0.01
MCID on depression
Improvement, n (%)	20 (29.9)a	28 (41.8)a	8 (11.9)b	26.71	<0.01
No change, n (%)	35 (52.2)a	36 (53.7)a	36 (53.7)a
Deterioration, n (%)	12 (17.9)a	3 (4.5)b	23 (34.3)a
MCID on anxiety
Improvement, n (%)	24 (35.8)a	24 (35.8)a	5 (7.5)b	48.58	<0.01
No change, n (%)	35(52.2)a	40 (59.7)a	30 (44.8)a
Deterioration, n (%)	8 (11.9)a	3 (4.5)a	32 (47.8)b
Pregnancy rate, n (%)	39 (58.2)	42 (62.8)	30 (44.7)	4.78	0.09

Note: If the same letter exists (a or b), there is no statistical significance between the two units compared in parallel.

For depression, the time effect and group-by-time effect were significant. More specifically, after the intervention, the prevalence of depression in the FG [−5.95 (-8.47, −3.43)] and CoG [−8.11 (−10.62, −5.58)] was significantly lower than that in the CG group. There was no significant change in the FG and CG across time points, but the prevalence of depression significantly decreased in the CoG. The prevalence of depression in the FG and CoG at postintervention was significantly lower than that in the CG, and the deterioration proportion of depression in the CoG was significantly lower than that in the FG and CoG.

For anxiety, the group effect, time effect and group-by-time effect were all significant. After the intervention, the prevalence of anxiety in the FG [−7.01 (−9.16, −4.86)] and CoG [−7.77 (−9.92, −5.63)] was significantly lower than that in the CG. Anxiety did not significantly decrease in the FG across time points but significantly decreased in the CoG and increased in the CG. The prevalence of anxiety in the FG and CoG at postintervention was significantly lower than that in the CG. The improvement proportions in the FG and CoG were significantly higher than those in the CG, while the deterioration proportions were significantly lower than those in the CG.

Combining the effect sizes, prevalence, and MCID values shown in Tables 2–4, the intervention effectively prevented depression and anxiety from worsening in the FG and mitigated depression in the CoG compared with the CG.

Secondary outcomes: sleep disturbance, serum cortisol levels, and pregnancy rates

Regarding sleep disturbance, the group effect, time effect and group-by-time effect were all significant. After the intervention, the prevalence of sleep disturbance in the FG [−8.47 (−9.82-7.13)] and CoG [−11.80 (−13.14, −10.46)] was significantly lower than that in the CG. The prevalence of sleep disturbance did not significantly decrease in the FG across time points but significantly decreased in the CoG and increased in the CG. Thus, the intervention effectively prevented sleep disturbance from worsening in the FG and promoted sleep disturbance in the CoG.

Regarding serum cortisol levels, the group effect and group-by-time effect were significant. After the intervention, the blood cortisol levels of the FG [−36.07 (−57.22, 14.92)] and CoG [−56.51 (−77.66, −35.34)] were significantly lower than those of the CG. Serum cortisol levels significantly increased in all three groups across time points, and the intervention effectively reduced cortisol levels.

Regarding the pregnancy rate, although the rate was higher in the FG and CoG, there was no significant difference among the three groups.

Per-protocol analysis

We performed GEE models based on the PP principle as a sensitivity analysis, and the results showed no difference between the ITT and PP analyses, with all effect sizes being larger in the PP analysis than in the ITT analysis.

Discussion

This study evaluated and compared the effectiveness of double ABCX-based psychotherapy for psychological distress among women undergoing IVF. Our intervention was more structured and relatively short compared to previous interventions, and it was embedded into six clinical visits for IVF-ET treatment. The dropout rate was approximately 23%, which was slightly lower than that of previous cognitive behavioral therapy (approximately 30%) [10–13] and self-guided psychotherapy (approximately 28%-30%) interventions [16,17] and comparable to that of psychotherapy incorporated into routine practice [16,41]. With a completion rate of 67% for the assigned mindfulness practice [42], the rates of home relaxation practice in the FG and CoG were acceptable, at 72% and 77%, respectively. The combination of health education and psychological counseling is conducive to IVF-ET treatment [19]. However, China’s infertility centers have not yet developed and integrated structured psychotherapy into routine care. Therefore, psychotherapy incorporated into routine practice has the potential to ensure sustainability and acceptability, which produced a satisfactory effect size in this study, as discussed below.

The main results indicated that double ABCX-based relaxation psychotherapy effectively prevented depression, anxiety and sleep disturbance from worsening in the FG and mitigated depression, anxiety and sleep disturbance in the CoG, which was consistent with previous cognitive behavioral therapy interventions [10–13]. We developed our intervention using rigorous methodology under the MRC framework, of which similar components were targeted at distorted thoughts closely related to depressive symptoms [11], stress control related to anxiety [43], and relaxation techniques related to physical sleep disturbance [14]. The components were organized using a double ABC model, which was structured and theory-based and may have contributed to the medium and large effect sizes in this study. A structured intervention may be easier for psychological consultants or clinical health professionals to manage and may ensure effectiveness in future implementation.

The intervention effectively reduced serum cortisol levels in the FG and CoG compared with the CG. Cortisol increased during IVF-ET treatment [44], which was in accordance with the active control group in this study. The mean cortisol level at baseline was relatively high in our study (625 nmol/L) compared with that in US participants at baseline (278 nmol/L), German women undergoing IVF-ET (364 nmol/l) and Chinese participants at baseline (250 nmol/L) [44–46]. This may partially explain the large effect size in our study, as the participants were in urgent need of intervention. Cortisol is usually examined as a parameter of ovarian function and as a potential prognostic indicator for the outcome of IVF-ET [46]. A recent review concluded that psychotherapy could modify the activity of stress-related brain regions and result in a decline in stress biomarkers [41]. However, previous evidence on the role of cortisol in relation to IVF-ET outcomes is mixed. A review found that while some studies indicated that higher cortisol may be associated with more favorable IVF-ET outcomes, another study indicated that lower cortisol levels may be conducive to IVF-ET success [47]. Further examining the study designs of these studies, we found that the test time points were different. The study with the largest sample of 264 females across multiple time-point tests showed that lower levels of cortisol at the time of oocyte retrieval and pregnancy testing were significantly associated with successful pregnancy [44]. This indicated that lower levels of cortisol at the time point close to the final process of IVF-ET were associated with more favorable fertility outcomes; thus, our intervention to reduce serum cortisol levels was clinically meaningful.

Although the clinical pregnancy rates at four weeks after embryo transfer were not significantly different, the ratio was slightly higher in the FG and CoG than in the CG. Several meta-analyses did not reach a consensus on the pregnancy rate [13,18]. The reasons behind the heterogeneity could be the design of the interventions (e.g. duration, content, format, delivery time points) and the characteristics of the included participants (e.g. good/bad prognosis, infertility etiologies) [41]. The influence that stress may have on chronic and acute IVF outcomes is still vague [47], and we were not able to follow-up on successful delivery. Hence, we cannot know if there is a long-term effect, and the evidence remains inconclusive.

Finally, our study revealed that the effect sizes of both primary and secondary outcomes, except the pregnancy rate, were larger in the FG than in the CoG. Couples undergoing IVF-ET tend to have different coping strategies; women are more likely to seek social support, connect with others and express their feelings, while men are more likely to use problem solving [48]. Even though instructing men to think about and express their innermost feelings and thoughts may be overly emotionally taxing [17], the participation of the male partner may give the female partner a greater sense of support, realistic mutual belief exchange and support seeking during the intervention. Marital dyadic interaction may enable individuals to reflect upon their stress, initially amplifying distress but eventually facilitating recovery [25], which may explain the findings that interventions for couples may be more beneficial than interventions for woman only.

Strengths, limitations and implications

In this study, we used the MRC framework to guide the development of our program. The implementation process followed a rigorous methodology to ensure that the results were credible. We initially compared the effectiveness between female-only and couple-based interventions, which is the first trial to do so to our knowledge.

There are still several limitations. First, as we included only participants who were undergoing their first IVF-ET cycle and who had education levels above junior high school, this may limit the generalizability of our findings to individuals with low education levels. Future studies should include couples experiencing two or more IVF-ET or ICSI cycles and test the effectiveness of the intervention. In addition, although different male/female infertility problems necessitate different treatment procedures ranging from pharmacotherapy and surgical-based approaches to assisted reproductive techniques [41], this intervention may also be applicable to those who have also experienced psychological distress caused by other types of infertility treatment [7,8]. Future studies can revise the time point and adapt the content of this intervention based on different treatments to test its applicability and effectiveness in patients with different types of infertility. Second, we did not measure the outcome variables of male partners because they were not routinely cared for and screened during IVF-ET. In the future, further analysis of the dynamics between female and male partners is recommended. Third, due to the relatively small sample size, we did not examine whether there were any differences in the different subgroups. Fourth, the intervention was delivered by a single, nonblinded qualified psychiatric nurse, and the nurse’s experience may have influenced the effect sizes of the intervention. Future research should involve different facilitators to test the effectiveness of the intervention. Finally, this study was conducted in a single center; therefore, the external validity may be limited, and a multicenter study is needed.

Conclusion

This three-armed, randomized clinical trial found that double ABCX-based psychotherapy effectively mitigated depression, anxiety, sleep disturbance and increased cortisol levels among women undergoing IVF-ET, suggesting greater effectiveness for couple-based psychotherapy than women-only psychotherapy. The intervention is structured and easy to use during IVF-ET treatment cycles, thereby presenting a promising and feasible approach to improve care for couples or women with assisted reproductive technique-related stress.

Authors’ contributions

All authors have participated in the study and have read and approved the submitted version of the manuscript. The authors declare they have no conflicts of interest. The authors abide by the copyright terms and conditions of Elsevier and the Australian College of Midwives. D. C. And A. W. were responsible for design, implementation, and writing. W. Z., Y. G., and S.Y. participated in data collection, analysis, and reporting. J.Z. was responsible for design, supervision, writing and date interpretation.

Ethical statement

All procedures involving human subjects/patients were approved by the Ethical Institutional Review Board of the researcher’s university on May 2017 (2017055). Written informed consent was obtained from all patients.

Patient or public contribution

Patients, service users, caregivers or members of the public were not involved in the design or conduct of the study, in the analysis or interpretation of the data, or in the preparation of the manuscript.

Acknowledgments

We are very grateful to the participants who participated in the trial and all clinicians in the center who supported the study. We thank all the clinicians and experts in offering their feedback during developing the intervention.

Disclosure statement

The authors have no conflicts of interest to declare.

Data availability statement

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.