I. Executive Summary

To determine the long-term mental health improvement of participants in Affair Recovery's programs, data from 20,797 unique participants across the four Affair Recovery programs were analyzed across time, from program enrollment through follow-up time points extending to 36 months after program completion. Three critical mental health outcomes tracked in this analysis included symptoms of anxiety, depression, and PTSD.

Infidelity precipitates a profound mental health crisis for all individuals affected; the severe trauma experienced by the betrayed spouse is often recognized as Betrayal Trauma or Complex PTSD. The symptoms of this distress and trauma overlap to a large extent with the critical mental health outcomes measured in this study. Utilizing latent growth curve models (LGCMs) to analyze the trajectories of symptom change, this study determined that Affair Recovery's participants experienced highly beneficial and significant improvements in their symptoms.

Results indicated Affair Recovery's participants achieved long-term, statistically significant improvement (positive change) in symptoms across all measured outcomes during the program and follow-up periods. The magnitude of symptom improvement also was clinically significant; effect sizes (the magnitude of change) ranged from moderate (ES=0.51) to very large (ES=1.16). Crucially, a comprehensive set of covariates, including the number of betrayals, was generally found not to moderate the overall rate of improvement, demonstrating that significant symptom improvement occurred across a large and diverse group of participants.

II. Study Design and Results by Program

Study Design

The analysis tracked three critical mental health outcomes using validated psychological assessments: the Burns Anxiety Inventory for anxiety, the Beck Depression Inventory for depression, and the PCL-5 PTSD checklist for post-traumatic stress symptoms. This measure is particularly relevant as it assesses core symptoms associated with Betrayal Trauma and Complex PTSD. Confidence in the validity of the findings is based on the study's large, longitudinal dataset encompassing 20,797 unique participants across all programs that were followed for up to 60 months from program completion.

The analysis utilized latent growth curve models (LGCMs) to model the trajectory of scores from program enrollment through available follow-up time points. Time points included in the analysis were pre-test, post-test, 3-month, 6-month, 9-month, 12-month, 24-month, 36-month, and 60-month to the extent that sufficient data were available. Missing data at these time points were handled using full-information maximum likelihood estimation (FIML), which requires at least 5% of data to be present across time points. LGCMs allowed for symptom change to be modeled as linear for the pre- and post-program periods and non-linear at later follow-up points.

Effect sizes quantify the strength of a program's effectiveness by calculating the magnitude of the difference between scores at two time points. Generally, an effect size of 0.20 is considered small, 0.50 moderate, and 0.80 large. Effect sizes were computed using only participants who provided responses at both pre-test and the latest time point included in the LGCMs.

The study included participants in the following Affair Recovery programs:

• EMS Online "Emergency Marital Seminar Online": 13-week online, group-based course for couples
• EMS Weekend: 3-day weekend intensive with 6-week online, group-based aftercare
• Harboring Hope: 13-week online, group-based course for betrayed spouses
• Hope For Healing: 17-week online, group-based course for unfaithful spouses

Symptom Improvement Over Time

The table below summarizes the effect sizes (magnitude of symptom improvement) and longevity of the effects for participants in all four programs.

The finding that participants demonstrated continued, statistically significant symptom reduction through at least 24 months, with effects lasting up to 36 months for core outcomes, demonstrated that participants in the Affair Recover programs experienced long-lasting benefits. The large effect size observed for Harboring Hope on PTSD symptoms (ES=1.16) highlighted the magnitude of symptom relief participants experienced and the maintenance of that improvement over at least two years.

III. Testing the Effects of the Programs for Participants with Differing Characteristics

To test whether the observed improvement was consistent across participants with differing characteristics, a comprehensive set of 15 covariates—ranging from demographic variables to specific affair characteristics—was tested to determine whether they moderated the rate of improvement over time in the Latent Growth Curve Models. Findings indicated that participants' symptom improvement did not depend to a statistically significant degree on the set of situational and demographic covariates tested in the models. As a result, participation in the programs appeared to be equally effective at reducing symptoms of anxiety, depression, and PTSD regardless of differences in participant characteristics or the specific details of the affair.

Covariates tested that were generally found not to moderate the rate of improvement across the programs included:

• Number of betrayals: The number of prior affairs (mean number of betrayals ranged from 2.33 to 3.09 across programs) did not generally affect the rate of symptom decline across all four programs.
• Gender: Gender did not significantly moderate the rate of improvement across any outcome.
• Military status: Whether one or both spouses were in the military at the time of the affair did not moderate the rate of improvement.
• Involvement in a faith community: Involvement in a faith community before starting the program generally did not moderate the rate of improvement.
• Personal importance of religious faith: Personal importance of faith before starting the program generally did not moderate the rate of improvement.
• Involvement in other recovery organizations: Involvement in other recovery organizations generally did not moderate the rate of improvement.

IV. Moderators of Improvement Trajectories

While the participants' outcomes largely were not affected by specific demographic or affair characteristics, a few variables consistently moderated the rate at which participants improved, that is, the speed of improvement. The consistent moderators across all programs were:

• Initial Symptom Severity: Initial scores on measures of anxiety, depression, and PTSD consistently moderated rate of improvement. Participants with lower initial scores generally demonstrated a more rapid rate of decline in their symptoms compared to those with higher initial severity.
• Time Since Affair Disclosure: The effect of this moderator varied by program type. In the programs for couples (EMS Online, EMS Weekend), recent disclosure was often associated with a faster rate of improvement in anxiety symptoms, while in individual programs (Hope for Healing, Harboring Hope), those who learned about the affair longer ago generally had a faster rate of improvement in depression and PTSD symptoms.

V. Conclusion

Overall, the sustained, statistically significant and clinically significant improvements observed in Affair Recovery's program participants demonstrated that their mental health substantially improved during and after their participation in the program. Crucially, this positive trajectory holds true across diverse demographics and differing betrayal histories.