Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Apr 2;52(2):e70132. doi: 10.1111/jmft.70132

Does Depression and Relationship Distress Predict Longer Couple Therapy Treatment in Naturalistic Settings?

Preston C Morgan 1,✉, Chi‐Fang Tseng 2, Courtney Doman 3, Yunyao Zhu 4
PMCID: PMC13409356  PMID: 41928491

ABSTRACT

The average couple therapy treatment length is 11.50 sessions in the United States, yet these estimates are supported by decades‐old research. We update this prior work by accounting for therapist effects and investigating clinical cut‐offs for depression and relationship distress. We used a sample of 417 mixed‐sex couples and 122 therapists from the Marriage and Family Therapy Practice Research Network. Using multilevel modeling in Mplus, couples had an average of 11.73 sessions with therapists explaining 7.40% of the variance. Having one partner meeting the relationship distress cut‐off as well as both partners meeting depression cut‐offs were associated with longer treatment. Additionally, therapists in the Western region of the United States and treatment after the COVID pandemic were associated treatment length. These findings can aid clinicians in providing treatment plans and clearer expectations for treatment length. Couple therapy modalities can consider briefer adaptations to meet shorter treatment lengths.

Keywords: couples therapy, depression, length of treatment, relationship distress, therapist effects

1. Introduction

How long are couples in couple therapy? Research consistently shows that couples typically attend between 9 and 12 sessions. Doherty and Simmons (1996) collected case information in 1994 from 526 therapists (across disciplines) from 15 states in the United States and found a median of 11.50 couple therapy sessions (Doherty and Simmons 1996). This is consistent with their earlier study of case information in 1992 from 76 Marriage and Family Therapists (MFT) in Minnesota that found couples attended an average of 11 sessions over 4 months (Simmons and Doherty 1995). Both of these studies defined the treatment length from completed cases, which means that these estimates reflect the average length of completed couple therapy. A more recent study of 90 MFT training therapists at a university‐based clinic in the Western region of the United States found an average of 9.71 sessions from data collected between 2005 and 2018 (Willis et al. 2021). Given the focus on dropouts, the length of treatment was evaluated across the entire caseload. As a result, the estimated treatment duration reflected average length treatment for all cases, regardless of whether they successfully completed treatment or dropped out.

The mental health treatment landscape has shifted since the 1990s. For example, therapy is both more publicly accepted and still under‐utilized by many groups of people in the United States (Bommersbach et al. 2022; Hall et al. 2019). Couple therapy alone is not generally supported by insurance policies, so couple therapy is often provided to self‐paying clients (Hubbard and Anderson 2022). While cost transparency has become increasingly relevant (e.g., the No Surprises Act that went into effect in 2022), understanding treatment length is also important for clinical and theoretical reasons. Updated estimates on couple therapy treatment length could aid clinicians in treatment planning and aid clients in managing their financial expectations. Perhaps, most importantly, the longer couples are in couple therapy, the more likely they are to experience better outcomes in therapy. Particularly, meaningful change in couple therapy may occur after a certain number of sessions. However, this is largely unclear for couple therapy models with few exceptions. For example, a minimum of 8 sessions is recommended for couples receiving Emotionally Focused Couple Therapy (Johnson 2020), up to 26 sessions for Integrative Behavioral Couple Therapy (Gurman 2008), and between 15 and 20 sessions for Gottman Method Couple Therapy (Gurman 2008).

Moreover, not all couples remain in couple therapy and some drop out of treatment prematurely. Although there is literature on dropout in couple therapy (e.g., Willis et al. 2021), which can be related, our focus is on predictors of how long couples are in couple therapy treatment in naturalistic settings. This highlights an important distinction between clinical trial samples and naturalistic setting samples. To increase internal validity, clinical trials develop selective samples and tight protocols to evaluate treatment effects; however, not all participants receive the full treatment. A closer examination of recent clinical studies of couple therapy for depression and relationship distress reveal discrepancies between the actual number of sessions attended and the initially proposed protocol for treatment duration. To illustrate, Baucom et al. (2018) noted an average of 10.85 sessions, with a range from 2 to 26 sessions, against a 10 to 20 weeks protocol. Alder et al. (2018) noted a 12‐week protocol but an actual average duration of 8 weeks. Denton et al. (2012) documented a 10.6‐week average compared to a 15‐week proposed protocol. Naturalistic studies, on the other hand, emphasize ordinary clinical therapy settings where therapists and participants are not preselected and aim for increased external and ecological validity, which is essential for understanding and mimicking real life interventions (Philips and Falkenström 2021; Verster et al. 2019). Hence, examining the length of treatment in naturalistic settings would provide real‐world estimates as well as estimates that relate with everyday couple therapy practices. There are three important gaps to address when investigating the length of couple therapy treatment: the severity of depressive symptoms and relationship distress, and therapist effects.

1.1. Severity of Depressive Symptoms and Relationship Distress

Our rationale for examining depressive symptoms and relationship distress is rooted in the Marital Discord Model of Depression (MDMD; Beach and Cassidy 1991) that was developed to describe how couples experience depressive symptoms and suggest a clinical model to intervene with the couple. Specifically, MDMD posits that the primary driver of depressive symptoms in couples is relationship discord. For example, when couples experience relationship distress, they're at a higher risk of experiencing depressive symptoms. However, if couples handle that distress by enhancing their relationship—improving their overall levels of relationship satisfaction, it buffers the risk of depressive symptoms. Theoretically, if couples experience greater symptoms, they may also be in couple therapy longer to treat these higher‐level symptoms. Relatedly, if couples experience greater relationship satisfaction, they may be in couple therapy less because their enhanced relationship buffers depressive symptoms, which may result in fewer sessions. Since the development of the MDMD, several systematic reviews have highlighted the empirical support for couple therapy as an effective treatment for depression and relationship distress (Barbato and D'Avanzo 2020; Wittenborn et al. 2022.). Hence, using MDMD as our theoretical support, we focus on depressive symptoms and relationship distress as key predictors of how long couples remain in couple therapy.

In clinical research, clinical cut‐offs are often used to evaluate depression and relationship distress. Clinical cut‐offs are important to investigate because with higher symptom severity, the longer therapy may be needed to be for symptoms to reduce to normal functioning. In the Doherty and Simmons study (1996), severity of client problems was associated with a longer length of treatment, but it did not distinguish between individual, couple, or family therapy. However, a depression diagnosis (according to the DSM‐III‐R) was not associated with an increase in sessions (Simmons and Doherty 1995). Although the literature on depression and relationship distress predicting length of couple therapy is nearly non‐existent, research is mixed on whether the severity of symptoms is related to the length of individual therapy. A systematic review of individual therapies suggests that mild‐to‐moderate depression requires 4–6 sessions (e.g., low‐intensity treatments) while more severe symptoms would require 4–26 sessions (e.g., high‐intensity treatments) (Robinson et al. 2020). However, studies note that diagnostic and symptom severity may be inadequate for accurately predicting treatment duration (Mueller and Pekarik 2000) and was associated with varied treatment lengths depending on treatment type: cognitive behavioral therapy, interpersonal psychotherapy, imipramine with clinical management, and placebo with clinical management (Elkin et al. 1995). Together, these studies highlight the importance of further research on whether symptom severity or clinical cut‐offs are related to treatment length. Yet, assessing depression and relationship distress as predictors of treatment length for couple therapy are notable gaps in this literature. Understanding these associations could help clinicians better tailor interventions to couples' needs and allow couples to understand potential duration of therapy to set realistic expectations for therapy.

1.2. Therapist Effects

Therapists may influence the length of treatment beyond client clinical cut‐offs and individual characteristics (i.e., therapist effects). A notable study examined therapist effects for individual psychotherapy with a large sample of 5828 clients and 158 therapists and found therapists explained less than 1% of the variance in client outcomes (Goldberg et al. 2016). Specifically, they found that the length of treatment differed by whether the therapist was a high‐performing therapist or a low‐performing therapist. Moving from individual therapy to couple therapy, a recent study examined reasons for couples prematurely dropping out of couple therapy and found that those reasons included discomfort with therapy (e.g., therapy was too stressful) and at home (e.g., made things at home too uncomfortable), and external factors like finances and availability (Codecà et al. 2025). Only 14% of the sample of 160 participants who dropped out of couple therapy for dissatisfaction with their therapist (Codecà et al. 2025). These participants attended 2–10 sessions before dropping out of couple therapy. We were unable to identify any study that examined therapist effects of couple therapy treatment length. However, two studies examined therapist effects on dropouts in couple therapy (Owen et al. 2014; Willis et al. 2021), and we focus on Willis et al. (2021), due to its larger sample of couples and therapists. With 90 therapists, at a university‐based clinic in the Western region of the United States, who treated 1192 couples from 2005 to 2018, this study found some therapists had great retention of clients and others did not. Specifically, therapist effects accounted for 9.4% of the variance in couples dropping out of couple therapy. Interestingly, couple therapy treatment was an average of 9.71 sessions, slightly lower than the average of 11.5 sessions (Doherty and Simmons 1996). However, the sample of therapists was smaller and located to one clinic instead of across the United States. Since dropout is related to the treatment length, but also a separate construct, this study highlights the importance of examining therapist effects.

1.3. Demographic and Treatment‐Related Characteristics

In addition to depressive symptoms, relationship distress, and therapist effects, there are other constructs that need to be accounted for when examining treatment length in couple therapy. These include common demographic characteristics such as income, age, cohabitation status, and race, as these factors have been associated with treatment engagement (e.g., Andersen and Newman 2005). Other factors such as current and previous treatment experiences, treatment progress, and whether therapy occurred during or after COVID‐19 may influence couples' expectations and motivation to remain in therapy (American Psychological Association 2022; McLaren et al. 2023). Additionally, “pressure to attend therapy” has been associated with a higher risk of early dropout (Codecà et al. 2025). Finally, contextual factors such as residence in the Western region of the United States, as regional differences in treatment availability and access may influence clients' capacity to remain in therapy.

1.4. Present Study

Together, this study aims to update the studies by Doherty and Simmons in the 1990s. To accomplish this, we used a clinical sample of couples in couple therapy largely at Marriage and Family Therapy University outpatient clinics across the United States. Particularly, we operationalized the length of treatment as the length of treatment for all caseloads, regardless of whether they completed treatment or not (e.g., Willis et al. 2021) because this reflects the real‐world settings for therapists. Next, we extended the current literature on couple therapy treatment length by investigating depression and relationship distress cut‐offs as predictors of treatment length while accounting for therapist effects as well as demographic and treatment‐related characteristics. Specifically, we aim to examine two research questions:

RQ1: What was the average length of treatment for couples in couple therapy, accounting for therapist effects?

RQ2: Does meeting depression cut‐off and relationship distress cut‐off associate with the length of couple therapy treatment in naturalistic settings while accounting for therapist effects?

2. Methods

We used data from the Marriage and Family Therapy Practice Research Network (MFT‐PRN; Johnson et al. 2017), which is a secondary data collection of clinical data from settings across the United States from 2019 to 2023—98% were from on‐site outpatient clinics at Marriage and Family Therapy programs and 2% were from private practices. Adult clients were recruited at each setting and consented to the MFT‐PRN. Consented clients completed continuous assessments prior to each session via online assessments. The institutional review board approved this study of secondary data (IRB‐22‐456). This clinical data is from the PRN‐MFT (Johnson et al. 2017) and was not publicly available. This study was not preregistered. We report on how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study.

We began with 1415 couples, which were limited to those who had a therapist id (n = 224 therapists; n = 1137 couples) and consented to participation in the MFT‐PRN. Next, we limited the data to both partners who completed demographics, depression, and relationship distress measures at session 1. One couple that identified as single was removed. This resulted in 465 couples and 125 therapists. Most of the couples received therapy in the Western region of the United States (81%), with some in the Midwest (7%) and Southern regions (12%). Although all couples reported themselves to be committed to each other, 20% reported that they were not living together. Some couples were parents (39%) with an average number of 2.52 (SD = 1.83) kids. Partners were in their late 20 s, the majority identified as White (83%), had income between $40,000 to $49,999, nearly half had a bachelor's degree or higher (47%), most identified as heterosexual sexual orientation (92%) as well as male or female (99%). Information on specific therapeutic interventions and reasons for dropout or termination were not available. See Table 1 for details on sample characteristics.

Table 1.

Sample characteristics.

M or % SD Range
Age 28.62 9.44 18–72
Female 51.75% 0, 1
Race
White 81.66% 0, 1
Hispanic, Latino, or Spanish 6.30% 0, 1
Black or African American 0.47% 0, 1
Asian 1.90% 0, 1
American Indian or Alaskan Native 0.59% 0, 1
Native Hawaiian or Other Pacific Islander 0.35% 0, 1
Middle Eastern or North African 0.11% 0, 1
Some other race 0.23% 0, 1
Multiracial 8.33% 0, 1
US Census Region
West 80.95% 0, 1
Midwest 6.90% 0, 1
South 12.15% 0, 1
Education
Junior High School or Less 0.47% 0, 1
GED/High School 27.02% 0, 1
Vocational/Technical School 5.11% 0, 1
Associate Degree (2 years) 19.88% 0, 1
Bachelor Degree (4 years) 36.19% 0, 1
Graduate Professional Degree 11.30% 0, 1
Household Income
Less than $10,000 12.14% 0, 1
Between $10,000 and $19,999 18.21% 0, 1
Between $20,000 and $29,999 13.92% 0, 1
Between $30,000 and $39,999 10.47% 0, 1
Between $40,000 and $49,999 9.28% 0, 1
Between $50,000 and $59,999 7.26% 0, 1
Between $60,000 and $69,999 5.59% 0, 1
Between $70,000 and $79,999 3.45% 0, 1
Between $80,000 and $89,999 4.28% 0, 1
Between $90,000 and $99,999 3.33% 0, 1
Greater than $100,000 12.02% 0, 1
Open in a new tab

Note: These are characteristics at the couple level meaning that characteristics represent the average or percentage for each couple.

3. Measures

3.1. Length of Treatment

Length of treatment was a count variable and was created by identifying the maximum number of couple therapy sessions attended per couple.

3.2. Depression

Depression was measured by the Patient Health Questionnaire‐9 (PHQ‐9; Kroenke et al. 1999). Nine items assessed how often over the past 2 weeks they had been bothered by depression criteria that ranged from “Little interest or pleasure in doing things” to “Thoughts that you would be better off dead, or of hurting yourself”. Cronbach's alpha was 0.85 for partner 1 and 0.88 for partner 2 at session 1. All items were rated on a scale of 0 (Not at all) to 3 (Nearly every day) with summed scores ranging from 0 to 28 with a clinical cut‐off score of 10 for depression (Manea et al. 2012). We coded two binary couple‐level variables of depression: one partner who met the cut‐off for depression and another for both partners meeting the cut‐off for depression.

3.3. Relationship Distress

Relationship distress was measured by the Couple Satisfaction Index (CSI‐16; Funk and Rogge 2007). Sixteen items assessed the romantic relationship ranging from “In general, how often do you think that things between you and your partner are going well?” to “How well does your partner meet your needs?”. All items were summed together for scores that ranged from 0 to 81, with a cut‐off of 51.5 for relationship distress (Funk and Rogge 2007). Cronbach's alpha was 0.98 for both partners at session 1. We coded two binary couple‐level variables of relationship distress: one partner who met the cut‐off for relationship distress and another for both partners meeting the cut‐off for relationship distress.

3.4. Controls

To be consistent with the analyses, all of the controls were coded at the couple‐level. We controlled for demographic (income, age, cohabitation, both partners identified as White), treatment‐related (prior and current treatment experiences, average treatment progress, pressure to attend therapy, sessions started during or after COVID‐19 pandemic), and contextual factors (Western region of the United States) that may influence treatment length. Pressure to attend therapy was measured with a single item (“How much did someone pressure you to come for therapy?”) rated from 1 (Not at all) to 5 (Very pressured). Treatment progress was measured by three items where clients identified their three biggest problems respectively and rated the progress on each problem from 1 (Problem is much worse) to 7 (Problem is solved/much better). Scores were then averaged across the duration of treatment to create an average treatment progress variable for each partner. Binary controls were Western region of the United States, at least one partner is in current individual therapy, both partners have had previous mental health treatment, both partners identified as White, sessions that started during the COVID‐19 pandemic, and sessions that started after the COVID‐19 pandemic. The continuous variables were averages of both partners' responses: age (years), pressure to attend therapy, and average treatment progress, treatment progress (see Table 2 for details).

Table 2.

Descriptive information and bivariate correlations of variables.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
1. Treatment Length —
2. RDC: 1 Partnera 0.09 —
3. RDC: 2 Partnersa 0.05 −0.26* —
4. DC: 1 Partnerb 0.00 0.07 0.08 —
5. DC: 2 Partnersb 0.13* 0.10* 0.22* −0.26* —
6. White: 1 Partnerc 0.01 0.00 −0.01 0.04 −0.06 —
7. White: 2 Partnersc −0.01 −0.09 0.03 −0.04 0.04 −0.78* —
8. Income 0.05 0.11* 0.15* −0.04 −0.04 −0.03 0.02 —
9. Age 0.09 0.12* 0.25* 0.00 −0.01 −0.05 0.03 0.57* —
10. Pressured 0.03 0.06 0.01 −0.06 0.04 0.07 −0.09 −0.02 −0.02 —
11. Current Treatmente 0.02 0.02 0.06 0.05 0.06 −0.01 0.00 −0.10 −0.05 0.00 —
12. Previous Treatmentf 0.12* 0.08 0.02 0.05 0.09 0.07 −0.05 0.04 0.06 0.01 0.22* —
13. Average Treatment Progressg −0.04 −0.03 −0.33* −0.15* −0.21* 0.01 0.00 −0.01 −0.17* −0.05 0.02 −0.17* —
14. Western Regiong 0.20* −0.08 −0.09 −0.03 −0.06 −0.04 0.06 −0.03 −0.03 −0.05 −0.13* −0.06 0.20* —
15. Living Togetherh 0.17* 0.16* 0.13 0.08 0.06 −0.01 −0.02 0.30* 0.27* 0.08 −0.23* −0.07 −0.08 0.13* —
16. During COVIDi 0.13* −0.01 −0.16* −0.08 −0.11* −0.06 0.07 0.09 −0.02 −0.03 −0.01 0.01 0.07 0.16* 0.04 —
17. After COVIDh −0.18* −0.03 0.09 0.05 0.05 0.06 −0.07 0.04 0.03 0.01 0.03 0.03 −0.11* 0.01 −0.02 −0.53* —
M or % 11.83 0.19 0.23 0.34 0.11 0.19 0.72 4.93 28.36 1.34 0.29 0.52 4.97 0.81 0.80 0.68 0.11
SD 10.16 — — — — — — 3.15 9.28 0.69 — — 0.67 — — — —
Range 2–69 1, −1 1, −1 1,−1 1, −1 1,−1 1,−1 1–11 18–67 1–5 1, −1 1,−1 2‐7 1,−1 1,−1 1,−1 1,−1
Open in a new tab

Note: All variables were coded at the couple level. a = Reference group is neither partner met the depression cut‐off. b = Reference group is neither partner met the relationship distress cut‐off. c = Reference group is neither partner reported to be White or Caucasian. d = Pressure to attend therapy was measured by a single item that was rated from 1 (Not at all) to 5 (Very pressured). e = Reference group is neither partner is in current individual therapy. f = Reference group is One or neither partner had previous mental health treatment. g = The three items that assessed treatment progress were averaged across the duration of treatment to create an average treatment progress variable for each partner. h = Reference group is all other regions in the United States. i = Reference group is sessions started prior to the onset of the COVID‐19 pandemic based on the WHO.

Abbreviations: DC = depression cut‐off; RDC = relationship distress cut‐off.

*

p < 0.05.

3.5. Analytic Plan

We conducted the analytic plan in the following steps. First, descriptive and bivariate correlations were run in R (R Core Team 2024). To test the research questions, we used Multilevel Modeling in Mplus (Muthén and Muthén 1998‐2017). Because the outcome (length of treatment) is a couple‐level variable, we used a two‐level model where couples (level 1) were nested within therapists (level 2). Specifically, couples were accounted for by a couple id variable and therapists were accounted for by a therapist id variable. Consequently, all variables were created to be couple‐level as described above. We used multilevel modeling (TYPE = TWOLEVEL) to test research question 1. A base multilevel model estimated the mean and variances within couples and between therapists. Second, we built upon the base model to test research question 2. Due to length of treatment being a count variable we expected the distribution to be skewed. Although the skewness was 1.70 (between ‐3 and 3) and kurtosis was 3.40 (between ‐10 and 10), a visual inspection of the distribution—as recommended by Kline (2023)—revealed a notable right tail skew of treatment length. This was accounted for with a negative binomial distribution (i.e., COUNT = length of treatment (nb)). We then added the predictors at level 1 (all variables except for Western region) and level 2 (only Western region). For interpretation, we exponentiated the parameters of the model to calculate the incident rate ratios by MODEL CONSTRAINT commands. Restricted maximum likelihood was used to handle missing data. Model fit was evaluated by AIC, where the lower AIC values indicated a more parsimonious model.

4. Results

A summary of descriptive information and correlations of the variables are provided here, see Table 2 for further details. The average length of treatment was 11.83 sessions (SD = 10.17, Median = 8, Range = 2–69). Of the couples who met the cut‐off for relationship distress, 18% had one partner and 23% had both partners meeting the cut‐off. Similarly, of the couples who met the cut‐off for depression, 34% had one partner and 11% had both partners meeting the cut‐off. Bivariate correlations revealed that both partners meeting the depression cut‐off was associated with higher length of couple therapy treatment (r = 0.13).

4.1. Length of Couple Therapy Treatment

For research question 1, the base model of treatment length (AIC = 3141.95), accounting for the variance within couples and between therapists revealed that couple therapy was on average 11.75 sessions with a large variance within couples (σ 2  = 95.86) and a smaller variance across therapists (σ 2  = 7.73). Specifically, the intraclass correlation was 0.074, meaning that 7.43% of the variance in treatment length was explained by therapists. Treatment length ranged from 2 to 69 sessions with 5.48% of the sample having 2 sessions and 0.24% having 69 sessions.

4.2. Multilevel Models

For research question 2, the base model accounted for a negative binomial distribution and predictor variables were added to the model. A significant dispersion (b = 0.42, p < 0.000) suggested that a negative binomial was an appropriate test compared to a Poisson distribution. This model also had a lower AIC (AIC = 2868.872; ΔAIC = −273.07), suggesting a more parsimonious model compared to the base model. We summarize results here with a p < 0.05 (see Table 3). Both partners meeting the depression cut‐off was associated with a 0.16 higher expected log count of the number of sessions for couple therapy. This means that having both partners who met the depression cut‐off was associated with 17% more couple therapy sessions than couples with neither partner meeting the depression‐cut off. Only one partner meting the relationship distress cut‐off was associated with a 0.11 higher expected log count of the number of sessions for couple therapy. Both partners with relationship distress cut‐off and one partner with depression cut‐off were not associated with the length of couple therapy treatment. We then added the control variables, which was also a more parsimonious model (AIC = 2481.63; ΔAIC = −387.24). The previous associations remained the same. However, starting couple therapy after the COVID‐19 pandemic years was associated with lower expected log counts of couple therapy sessions. At the therapist level, being in the Western region of the United States was associated with a higher expected log count of couple therapy sessions. None of the remaining controls were associated with the length of couple therapy treatment.

Table 3.

Multilevel model of length of treatment as outcome.

Within Predictors only Full model
b SE p exp(b) b SE p exp(b)
1 Partner meets relationship distress cut‐off 0.11* 0.05 0.020 1.12 0.12* 0.05 0.011 1.12
2 Partners meet relationship distress cut‐off 0.05 0.05 0.294 1.06 0.07 0.05 0.156 1.08
1 Partner meets depression cut‐off 0.01 0.04 0.800 1.01 0.01 0.04 0.785 1.01
2 Partners meet depression cut‐off 0.16* 0.06 0.016 1.17 0.15* 0.07 0.031 1.16
Average income 0.00 0.02 0.914 1.00
Previous treatment 0.03 0.04 0.380 1.03
1 Partner identifies as White −0.05 0.07 0.454 0.95
2 Partners identify as White −0.03 0.06 0.637 0.97
Average age 0.01 0.01 0.393 1.01
Pressure to attend therapy 0.04 0.05 0.384 1.04
Average treatment progress −0.06 0.07 0.394 0.94
Living together 0.13 0.07 0.051 1.14
During COVID‐19 pandemic years 0.05 0.06 0.395 1.05
After COVID‐19 pandemic years −0.23* 0.07 0.000 0.79
Dispersion 0.42* 0.04 < 0.000 0.35* 0.03 < 0.000
Between
Western Region 0.28* 0.05 < 0.000 1.32
Intercept
Residual variance 0.10* 0.03 0.004 0.03 0.02 0.108
Open in a new tab
*

p < 0.05.

5. Discussion

Results revealed several findings that contribute to the literature and aid clinicians understanding in factors related with couples engaging in couple therapy longer than average. First, for our first research question, results revealed that the average length of couple therapy treatment was 11.73 sessions. This was consistent with the Doherty and Simmons work (1995, 1996), which supports the average length of couple therapy treatment in outpatient sessions to just shy of 12 sessions. Additionally, Doherty and Simmons' work had a sample of licensed clinicians, however, our findings are from a sample of MFT training therapists at outpatient clinics. Although training therapists are indeed less experienced than the licensed clinicians, they receive regular supervision, which has been shown to support treatment fidelity and clinical effectiveness even among novice therapists (Bernard and Goodyear 2019). Given that difference, it is interesting that the average length of couples therapy is nearly identical despite 30 years passing and quite different clinical settings. Furthermore, the estimates are similar, yet the definitions of treatment length were different with Doherty and Simmons works defining treatment length using completed cases only, whereas our study included both completed and incomplete cases. Importantly, data was not available to determine whether couples dropped out of treatment, so future research can distinguish the average length of completed treatment with those who drop out of treatment.

Second, both partners meeting the depression cut‐off predicted longer couple therapy treatment. This is notable considering that only 11% of couples had both partners who met the depression cut‐off. From a Marital Discord Model of Depression perspective, these findings support that higher levels of depressive symptoms are associated with longer treatment, but our findings add nuance with the notion that it only applies to when both partners experience cut‐off levels of depressive symptoms. This longer treatment likely reflects both the time needed to reduce depression symptoms and the need to address relationship satisfaction, emotional support, and problem‐solving (Whisman and Beach 2012), impacted when depression is present for both partners. Additionally, longer treatment is recommended to decrease relapse of depression symptoms in individuals (Thase 1999) and may be the case for both the couple's depression symptoms and relational dynamics. However, there appears to be no difference in the length of treatment when couples have only one partner who met the depression cut‐off, which was more common in our sample, 34% of couples. These findings encourage future research to identify strategies for couple's therapy to address unique dual depression dynamics.

Third, unlike depression, only one partner needs to meet the relationship distress cut‐off to be in couple therapy longer. These types of couples comprised 18% of our sample. From a Marital Discord Model of Depression perspective, this finding supports the notion that higher levels of relationship distress are associated with longer treatment. What is interesting is that when only one partner, not both partners, meets the relationship distress are they in couple therapy longer. This suggests that there was no difference in how long couples were in couple therapy when both met or did not meet the relationship distress cut‐off. It is possible that couples are in therapy longer when one partner has relationship distress because both partners may differ in how they see their problems. Consequently, it may take longer to address these differences in couple therapy. These findings encourage future research to identify strategies for couple's therapy to address unique relational dynamics for when one partner is in relational distress.

Fourth, couples who started couple therapy after the COVID pandemic began were in less couple therapy compared to couples before the COVID pandemic. This finding is notable and can represent a massive shift in couple therapy being provided through teletherapy (e.g., Hardy et al. 2021). It remains unknown if couples are in couple therapy less because of online therapy and would have stayed longer in therapy in‐person. Although teletherapy is as effective as in‐person therapy, there can be a stronger therapeutic alliance in in‐person therapy (Bradford et al. 2024). Although we did not have available data to distinguish between teletherapy and in person, future research can explore why couples are in couple therapy less after COVID.

Fifth, therapist effects were an important aspect of the length of the treatment. Our analyses revealed that therapists accounted for 7.40% of the variance in treatment length, consistent with the 9.4% reported in prior research (Willis et al. 2021). The remaining 93.60% of the variance in couples therapy treatment length is explained by couple factors. Additionally, therapists in the Western region of the United States had longer treatment compared to couples in other regions. Mental health attitudes may vary from region to region with the Western region being more open to couple therapy treatment. We encourage future research to examine possible explanations for this regional effect on couple therapy treatment length.

5.1. Clinical Implications

Our study found that couples attended an average of 11.73 sessions in outpatient MFT clinics across the United States. This finding has practical implications for therapists, as it enables them to set realistic treatment plans tailored to couples' needs. Such tailoring can occur by assessing the levels of relationship distress and depression at intake. When one partner experiences relationship distress at intake, and when both partners are experiencing depression at intake, therapists can anticipate a longer duration of couple therapy. This insight is critical for therapists to set realistic treatment goals and provide couples with an accurate estimate of expected treatment length and costs. Beyond treatment planning, the average length of treatment for couple therapy suggests a possible revision to couple therapy modalities. Many current models, such as the Gottman Method Couple Therapy and Integrative Behavioral Couple Therapy, plan for couples therapy to be more than 20 sessions (Gurman 2008). Although Emotionally Focused Couple Therapy has an 8 sessions minimum, longer treatment is expected (Johnson 2020). These models may be designed for much longer treatments than many couples actually participate in. This raises several questions about current couple therapy designs. For example, if many couples attend less than an average of 11 sessions, should models incorporate interventions specifically tailored for brief couple therapy? Some couple therapy models take 3–4 sessions as assessment, could the assessment phase be condensed? We hope these questions start an important dialogue on couple therapy models and possible revisions.

5.2. Limitations

There are several limitations to these findings. We were unable to account for therapist characteristics (e.g., age), specific therapeutic models or interventions used in session, as well as whether these couples were seeking couple therapy for depression and relationship distress. Similarly, we used self‐reported measures with established clinical cut‐offs for depression and relationship distress, so these findings do not reflect the therapists' actual diagnoses. Racially, our sample was predominantly White, and 8% of our couples identified as gay or lesbian. Our sample included a low average annual income and were in their late 20 s. Future research can replicate these findings with more diverse and older samples. Financial cost is an important issue for couple therapy, many of the clinical settings in this sample are onsite outpatient clinics that offer reduced rates or sliding fee scales. These findings may be different for private practice and agencies that rely on insurance or higher self‐pay rates for couple therapy.

6. Conclusion

Clinicians and clients for couple therapy can expect couple therapy to be an average of 11.73 sessions. They can also expect treatment length to be longer when both partner's meet the cut‐off for depression, when both have prior mental health treatment, if they live together, and live in the Western region of the United States. These findings inform clinicians in the treatment planning process as well as informing clients of the potential cost of couple therapy. This length of treatment also suggests that couples may be in treatment far less than expected for common couple therapy modalities and adaptations for briefer treatments may be considered.

References

  1. Alder, M. C. , Dyer W. J., Sandberg J. G., Davis S. Y., and Holt‐Lunstad J.. 2018. “Emotionally‐Focused Therapy and Treatment as Usual Comparison Groups in Decreasing Depression: A Clinical Pilot Study.” American Journal of Family Therapy 46, no. 5: 541–555. 10.1080/01926187.2019.1572478. [DOI] [Google Scholar]
  2. American Psychological Association . (2022, November). Psychologists Struggle to Meet Demand Amid Mental Health Crisis: 2022 COVID‐19 Practitioner Impact Survey. https://www.apa.org/pubs/reports/practitioner/2022-covid-psychologist-workload.
  3. Andersen, R. , and Newman J. F.. 2005. “Societal and Individual Determinants of Medical Care Utilization in the United States.” Milbank Quarterly 83, no. 4. 10.1111/j.1468-0009.2005.00428.x. [DOI] [PubMed] [Google Scholar]
  4. Barbato, A. , and D'Avanzo B.. 2020. “The Findings of a Cochrane Meta‐Analysis of Couple Therapy in Adult Depression: Implications for Research and Clinical Practice.” Family Process 59, no. 2: 361–375. 10.1111/famp.12540. [DOI] [PubMed] [Google Scholar]
  5. Baucom, D. H. , Fischer M. S., Worrell M., et al. 2018. “Couple‐Based Intervention for Depression: An Effectiveness Study in the National Health Service in England.” Family Process 57, no. 2: 275–292. 10.1111/famp.12332. [DOI] [PubMed] [Google Scholar]
  6. Beach, S. R. , and Cassidy J. F.. 1991. “The Marital Discord Model of Depression.” Comprehensive Mental Health Care 1, no. 2: 119–136. [Google Scholar]
  7. Bernard, J. M. , and Goodyear R. K.. 2019. Fundamentals of Clinical Supervision (6th ed.). Pearson.
  8. Bommersbach, T. J. , Rosenheck R. A., and Rhee T. G.. 2022. “Racial and Ethnic Differences in Suicidal Behavior and Mental Health Service Use Among US Adults, 2009–2020.” Psychological Medicine 53, no. 12: 5592–5602. 10.1017/S003329172200280X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bradford, A. B. , Johnson L. N., Anderson S. R., et al. 2024. “Call Me Maybe? In‐Person vs. Teletherapy Outcomes Among Married Couples.” Psychotherapy Research 34, no. 5: 611–625. 10.1080/10503307.2023.2256465. [DOI] [PubMed] [Google Scholar]
  10. Codecà, L. , Anderson S. R., Christensen J. K., Jackson J. B., and D'Aniello‐Heyda C.. 2025. “Client Perspectives of Premature Discontinuation in Couple Therapy.” Journal of Marital and Family Therapy 51, no. 4: e70061. [DOI] [PubMed] [Google Scholar]
  11. Denton, W. H. , Wittenborn A. K., and Golden R. N.. 2012. “Augmenting Antidepressant Medication Treatment of Depressed Women With Emotionally Focused Therapy for Couples: A Randomized Pilot Study.” Journal of Marital and Family Therapy 38, no. s1: 23–38. 10.1111/j.1752-0606.2012.00291.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Doherty, W. J. , and Simmons D. S.. 1996. “Clinical Practice Patterns of Marriage aand Family Therapists: A National Survey of Therapists and Thier Clients.” Journal of Marital and Family Therapy 22, no. 1: 9–25. 10.1111/j.1752-0606.1996.tb00183.x. [DOI] [Google Scholar]
  13. Elkin, I. , Gibbons R. D., Shea M. T., et al. 1995. “Initial Severity and Differential Treatment Outcome in the National Institute of Mental Health Treatment of Depression Collaborative Research Program.” Journal of Consulting and Clinical Psychology 63, no. 5: 841–847. 10.1037//0022006x.63.5.841. [DOI] [PubMed] [Google Scholar]
  14. Funk, J. L. , and Rogge R. D.. 2007. “Testing the Ruler With Item Response Theory: Increasing Precision of Measurement for Relationship Satisfaction With the Couples Satisfaction Index.” Journal of Family Psychology 21, no. 4: 572–583. 10.1037/08933200.21.4.572. [DOI] [PubMed] [Google Scholar]
  15. Goldberg, S. B. , Hoyt W. T., Nissen‐Lie H. A., Nielsen S. L., and Wampold B. E.. 2016. “Unpacking the Therapist Effect: Impact of Treatment Length Differs for High‐And Low‐Performing Therapists.” Psychotherapy Research 28, no. 4: 532–544. [DOI] [PubMed] [Google Scholar]
  16. Gurman, A. S. (Ed.). 2008. Clinical Handbook of Couple Therapy (4th ed.). The Guilford Press.
  17. Hall, G. C. N. , Kim‐Mozeleski J. E., Zane N. W., et al. 2019. “Cultural Adaptations of Psychotherapy: Therapists' Applications of Conceptual Models With Asians and Asian Americans.” Asian American Journal of Psychology 10, no. 1: 68–78. 10.1037/aap0000122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hardy, N. R. , Maier C. A., and Gregson T. J.. 2021. “Couple Teletherapy in the Era of COVID‐19: Experiences and Recommendations.” Journal of Marital and Family Therapy 47, no. 2: 225–243. 10.1111/jmft.12501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hubbard, A. K. , and Anderson J. R.. 2022. “Understanding Barriers to Couples Therapy.” Journal of Marital and Family Therapy 48: 1147–1162. 10.1111/jmft.12589. [DOI] [PubMed] [Google Scholar]
  20. Johnson, L. N. , Miller R. B., Bradford A. B., and Anderson S. R.. 2017. “The Marriage and Family Therapy Practice Research Network (MFT‐PRN): Creating a More Perfect Union Between Practice and Research.” Journal of Marital and Family Therapy 43, no. 4: 561–572. 10.1111/jmft.12238. [DOI] [PubMed] [Google Scholar]
  21. Johnson, S. 2020. The Practice of Emotionally Focused Couples Therapy: Creating Connection (3rd ed.). Routledge. 10.4324/9781351168366. [DOI]
  22. Kline, R. B. 2023. Principles and Practice of Structural Equation Modeling. Guilford Publications. [Google Scholar]
  23. Kroenke, K. , Spitzer R. L., and Williams J. B. W.. 1999. “Patient Health Questionnaire‐9 (PHQ‐ 9).” APA PsycTests. 10.1037/t06165-000. [DOI] [Google Scholar]
  24. Manea, L. , Gilbody S., and McMillan D.. 2012. “Optimal Cut‐Off Score for Diagnosing Depression With the Patient Health Questionnaire (PHQ‐9): A Meta‐Analysis.” Canadian Medical Association Journal 184, no. 3: E191–E196. 10.1503/cmaj.110829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. McLaren, T. , Peter L. J., Tomczyk S., Muehlan H., Schomerus G., and Schmidt S.. 2023. “The Seeking Mental Health Care Model: Prediction of Help‐Seeking for Depressive Symptoms by Stigma and Mental Illness Representations.” BMC Public Health 23, no. 1: 69. 10.1186/s12889-022-14937-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mueller, M. , and Pekarik G.. 2000. “Treatment Duration Prediction: Client Accuracy and Its Relationship to Dropout, Outcome, and Satisfaction.” Psychotherapy: Theory, Research, Practice, Training 37, no. 2: 117–123. 10.1037/h0087701. [DOI] [Google Scholar]
  27. Muthén, L. K. , and B. O. Muthén. 1998. ‐2017. Mplus User's Guide. Eighth Edition. Los Angeles, CA: Muthén & Muthén No Surprises Act, Pub. L. No. 116–260, div. BB, 134 Stat. 2758 (2022).
  28. Owen, J. , Duncan B., Reese R. J., Anker M., and Sparks J.. 2014. “Accounting for Therapist Variability in Couple Therapy Outcomes: What Really Matters?” Journal of Sex & Marital Therapy 40, no. 6: 488–502. 10.1080/0092623X.2013.772552. [DOI] [PubMed] [Google Scholar]
  29. Philips, B. , and Falkenström F.. 2021. “What Research Evidence Is Valid for Psychotherapy Research?” Frontiers in Psychiatry 11: 625380. 10.3389/fpsyt.2020.625380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. R Core Team . 2024. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org.
  31. Robinson, L. , Delgadillo J., and Kellett S.. 2020. “The Dose‐Response Effect in Routinely Delivered Psychological Therapies: A Systematic Review.” Psychotherapy Research 30, no. 1: 79–96. 10.1080/10503307.2019.1566676. [DOI] [PubMed] [Google Scholar]
  32. Simmons, D. S. , and Doherty W. J.. 1995. “Defining Who We Are and What We Do: Clinical Practive Patterns of Marriage and Family Therapists in Minnesota.” Journal of Marital and Family Therapy 21, no. 1: 3–16. 10.1111/j.1752-0606.1995.tb00134.x. [DOI] [Google Scholar]
  33. Thase, M. E. 1999. “Long‐Term Nature of Depression.” supplement, Journal of Clinical Psychiatry 60 Suppl 14, no. 14: 3–9. [PubMed] [Google Scholar]
  34. Verster, J. C. , van de Loo A. J. A. E., Adams S., et al. 2019. “Advantages and Limitations of Naturalistic Study Designs and Their Implementation in Alcohol Hangover Research.” Journal of Clinical Medicine 8, no. 12: 2160. 10.3390/jcm8122160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Whisman, M. A. , and Beach S. R. H.. 2012. “Couple Therapy for Depression.” Journal of Clinical Psychology 68, no. 5: 526–535. 10.1002/jclp.21857. [DOI] [PubMed] [Google Scholar]
  36. Willis, K. L. , Miller R. B., Anderson S. R., Bradford A. B., Johnson L. N., and Yorgason J. B.. 2021. “Therapist Effects on Dropout in Couple Therapy.” Journal of Marital and Family Therapy 47, no. 1: 104–119. 10.1111/jmft.12473. [DOI] [PubMed] [Google Scholar]
  37. Wittenborn, A. K. , Woods S. B., Priest J. B., et al. 2022. “Couple and Family Interventions for Depressive and Bipolar Disorders: Evidence Base Update (2010‐2019).” Journal of Marital and Family Therapy 48, no. 1: 129–153. 10.1111/jmft.12569. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Marital and Family Therapy are provided here courtesy of Wiley

RESOURCES