Abstract
Importance
Over 50% of older adults with late-life major depressive disorder fail to respond to initial treatment with first-line pharmacologic therapy.
Objective
To assess typical patterns of response to an open-label trial of venlafaxine XR for late-life depression, and to evaluate which clinical factors are associated with the identified longitudinal response patterns.
Design
Group-based trajectory modeling was applied to data from a 12-week open-label pharmacologic trial conducted as part of the Incomplete Response in Late Life: Getting to Remission Study. Clinical prognostic factors, including domain-specific cognitive performance and individual depression symptoms, were examined in relation to response trajectories.
Setting
Specialty care.
Participants
Adults aged ≥60 years with current major depressive disorder (n=453).
Intervention
Open-label venlafaxine XR (titrated up to 300 mg/day) over 12 weeks.
Main Outcome Measure
Sub-groups exhibiting similar response patterns were derived from repeated measures of overall depression severity obtained using the Montgomery-Asberg Depression Rating Scale.
Results
Three sub-groups with differing baseline depression severity clearly responded to treatment: one group with the lowest baseline severity had a rapid response (15.23%), and distinct responses were also apparent among groups starting at moderate (23.84%) and higher (5.52%) baseline symptom levels. Three sub-groups had non-responding trajectories: two with high baseline symptom levels (totaling 35.98%), and one group that had moderate baseline symptomology (19.43%). Several factors were independently associated with having a non-responsive trajectory, including: greater baseline depression severity, longer episode duration, less subjective sleep loss, more guilt, and more work/activity impairment. Higher delayed memory (list-recognition) performance was independently associated with having a rapid response.
Conclusion and Relevance
Based on the observed trajectory patterns, late-life depression patients with high baseline depression severity are unlikely to respond after 12-weeks of venlafaxine XR care. However, high baseline depression severity alone may be neither a necessary nor sufficient predictor of treatment non-response. Additional strategies should be considered when treating patients with severe depression and/or the prognostic factors identified (including those identified in crude analyses) which may signal etiologic or phenotypic differences relevant to treating late-life depression. Future research is needed to address whether the identified factors moderate efficacy or are generally prognostic.
Over 50% of adults treated for late-life depression (LLD) fail to respond to first-line pharmacologic therapy1,2. High pre-treatment depression severity may be a powerful marker of first-line resistance, however past research (including our own1) linking baseline severity to treatment response may be limited by a simple and important methodological consideration. Specifically, use of a threshold-based definition of remission (i.e. having a Montgomery Asberg Depression Rating Scale3 (MADRS) score of ≤10 by week 12 of treatment) results in a situation where, depending on initial symptom levels, patients recovering at the exact same rate may or may not reach the remission cut-off within the study timeframe.
Group-based trajectory modeling provides a complementary approach to examine treatment response variability without relying on a pre-specified remission threshold or assuming the entire sample follows a single trajectory. This data-driven approach identifies sub-groups which share common patterns of change over time. Previous applications to clinical research focused on depression4–11 consistently demonstrate that group-based trajectory methods capture heterogeneity in the course of illness that would be neglected using traditional approaches.
To our knowledge, no prior study has utilized group-based trajectory methods to examine treatment response variability within a large sample of older adults undergoing solo pharmacologic treatment for LLD. Although both sophisticated and important, previous applications of group-based longitudinal methods to study treatment response have been either: large but not restricted to older adults (predominately including middle-aged adults)4–6, small plus not restricted to older adults7, or small/not restricted to solo pharmacologic treatment8,9. Compared with younger adults, factors including pre-treatment co-morbidity12 and polypharmacy13 may contribute to an altered course of pharmacologic treatment and response in LLD14–16. Variability in neuropsychological function among older adults also predicts treatment outcomes17–19.
We therefore applied group-based trajectory modeling to examine LLD response in the largest open-label trial of solo pharmacotherapy (venlafaxine XR) conducted among older adults to date. Given that this was an open-label trial delivered in a specialty care setting that included depression management support and monitoring, we did not aim to evaluate pharmacological efficacy or the causality of effect (i.e. via venlafaxine, monitoring, or placebo effects). Instead, we took an exploratory approach to describe the typical patterns of symptom change observed over a trial of pharmacological care for LLD. In addition, we assessed associations between previously identified prognostic factors (i.e. medical co-morbidity, depression symptom severity, and neuropsychological function) with data-derived response patterns. Given evidence suggesting variability in individual sleep8 or core depression symptoms20 might predict treatment outcomes, we also explored associations between individual symptom expression and response patterns.
METHODS
This report pertains to the initial phase of the Incomplete Response in Late-Life Depression: Getting to Remission (IRLGREY) Study which was a three-site open-label trial conducted to prospectively assess treatment response.
Participants
Details of recruitment have been described previously1. Participants were aged ≥60 (n=466) and had current nonpsychotic major depressive disorder as diagnosed by the Structured Clinical Interview for DSM-IV21 plus a MADRS score ≥15. Exclusions were: lifetime diagnosis of bipolar disorder, schizophrenia, schizoaffective, other psychotic disorders or current psychotic symptoms; clinical history of dementia or cognitive impairment as indicated by a score of <20 on the Mini Mental State Exam22; alcohol or substance abuse in the past 3 months; high suicide risk; unstable medical illness; or contraindication to venlafaxine XR or aripiprazole (which was administered in a subsequent trial). Participants were included in the analysis if they had complete outcome data from baseline and at least one follow-up visit (n=453). Participants provided informed consent and ethical approval was obtained from site institutional review boards.
Intervention
Participants were treated with Venlafaxine XR starting at 37.5 mg/day and titrated up to 300 mg/day following a study-standardized protocol for at least 12 weeks. Details on titration and use of other medications have been published previously1. Depression-specific psychotherapy was not provided, although pharmacological treatment was delivered with depression management support including clinical care focused on depression symptoms, treatment, suicidal ideation, countermeasures for adverse medication effects, and adherence.
Measures
Outcome
Depression symptom burden was measured over time as the total MADRS score assessed at study baseline, week 1, week 2, and every two weeks thereafter until the completion of the open-label study.
Baseline prognostic factors
Baseline assessments included the Hamilton Rating Scale for Depression (HRSD-17)23. To explore potential prognostic roles of both overall symptom burden and individual symptoms, the MADRS and HRSD-17 were examined as scored totals and individual items. Adequacy of prior antidepressant treatment was measured with the Antidepressant Treatment History Form (ATHF24; scores ≥3 indicate a prior adequate trial). Other clinical factors were: age at first episode onset, current episode duration (natural-log transformed), single vs. recurrent episodes, and reception of outside psychotherapy. The Brief Symptom Inventory (BSI)25 and Scale for Suicidal Ideation (SSI)26 were used to assess general anxiety and suicidal ideation, respectively. Medical comorbidity was assessed using the Cumulative Illness Rating Scale for Geriatrics (CIRS-G)27 (expressed as total and count). The physical function sub-section of the 36-item Medical Outcomes Study Short-From Health Survey (SF-36 Physical) was expressed as a total score28. Self-reported alcohol use was assessed as the number of drinks per week.
Baseline neuropsychological measures included the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS)29. Initially, we examined domain-specific RBANS summary scores. Among RBANS domains associated with trajectory group, we further examined the individual subtests that composed the relevant summary score. Subtest scores were analyzed using age-normed Z-scores. Two tasks from the Delis-Kaplan Executive Function System (D-KEFS)30 were also examined in this manner.
Statistical Methods
A semiparametric, group-based modeling strategy was used to classify the cohort into sub-groups based on identifying heterogeneous longitudinal polynomial trajectories31,32. We implemented this technique using PROC TRAJ32 in SAS version 9.3 (SAS Institute Inc., Cary, NC). For our implementation, we assumed that the error structure followed a censored normal distribution. We determined the number of groups and degree of polynomial in each of the trajectory groups using the Bayesian information criterion (BIC), which measures improvement in model fit gained by adding additional groups or shape parameters incorporating a penalty for added complexity. The BIC log Bayes factor approximation, defined as 2×[ΔBIC] (subtracting a less complex from a more complex model), has been shown to be a good approximation to the log Bayes factor criterion33 and was used to base selection of the number of trajectory groups which fit the data. A log Bayes factor approximation >10 is considered to be strong evidence in favor of the more complex model32. Solutions that included small (<5% of the sample) trajectory groups were rejected. The degree of the fitted polynomials was determined by examining BIC for all possible permutations of linear and quadratic trajectories. Cubic polynomials did not improve model fit as indicated by BIC. To assess model fit, we examined the average posterior probability of group membership (>70% recommended) and odds of correct classification (OCC; >5 considered adequate).
Weighted multinominal logistic regression was used to assess crude associations of prognostic factors with trajectory group membership. These regressions were weighted by the probability of group membership to account for measurement error introduced by the probabilistic nature of group assignment. The reference group was chosen to be the largest group with a distinct response. To assess whether individual depression symptoms were associated with trajectory group membership beyond a function of overall depression severity, we adjusted models including individual symptoms for total baseline depression severity (measured using the item’s scale’s total score). Models including neuropsychological variables were adjusted for age and education. Associations of group membership with prognostic factors that achieved crude statistical significance (defined as p<0.05) were selected as predictors in a maximum model. Automated backwards elimination (elimination threshold p>0.20) was then used to select uniquely predictive prognostic factors which were entered as baseline risk factors within the PROC TRAJ framework.
Results
Model selection and trajectory group description
Regardless of whether all groups were assigned linear, quadratic, or cubic trajectories, BIC consistently indicated the same optimal number of groups. Model fit indicated by BIC improved with additional trajectory groups (Table 1); however, a seven-group solution was rejected because it included a small (<5%) group. BIC log Bayes factor approximations indicated strong evidence that six groups fit the data better than five groups. The data was best fit by a combination of linear and quadratic trajectory groups. Average posterior probabilities were high for all six groups (range: 0.84–0.90) and the OCC were all well above 5 (minimum OCC=20.18). This trajectory model depicted heterogeneity in symptom course across all levels of total baseline severity (Figure 1). Groups were assigned names based on their relative starting position and/or response trajectory.
Table 1.
BIC by number of groups for models with all linear exponents, and associated BIC log Bayes factor approximation
| Number of Groups | BIC | 2*ΔBIC (Evidence against Ho) |
|---|---|---|
| 1 | −11971.12 | |
| 2 | −11243.70 | 1454.84 |
| 3 | −11058.70 | 370 |
| 4 | −10978.19 | 161.02 |
| 5 | −10931.47 | 93.44 |
| 6 | −10888.79 | 85.36 |
| 71 | −10883.20 | 11.18 |
solution included a small (<5%) trajectory group and was rejected;
BOLD indicates selected solution
Abbreviations: BIC, Bayesian information criteria; Ho, null hypothesis
Figure 1.
Mean depressive symptom severity over 12-weeks of open-label treatment by data-derived trajectory group (Abbreviation: MADRS – Montgomery Asberg Depression Rating Scale)
Three groups exhibited clear changes in symptom severity overall and were all labeled “responding” trajectories (totaling 44.6% of the sample). Two started at relatively lower/moderate symptom levels and followed quadratic trajectories (15.23%, “rapid response”; 23.84%, “moderate, response”; see Table 2 for baseline and 12-week symptom levels). The third responding group had higher average baseline symptom levels which followed a linear trajectory (5.52%, “high, response”). The number of weeks until average symptom levels within these groups reached below typical remission criteria (i.e. MADRS≤10) appeared to differ (Figure 1).
Table 2.
Descriptive Patient Characteristics by Trajectory Group
| Responders | Non-responders | ||||||
|---|---|---|---|---|---|---|---|
| Rapid Response |
Mod., Response |
High, Response |
Mod., Mixed/Non- response |
High, Non- response 1 |
High, Non- response 2 |
p1 | |
| Percentage (n) in group | 15.23 (69) | 23.84 (108) | 5.52 (25) | 19.43 (88) | 24.28 (110) | 11.70 (53) | |
| Demographic factors | |||||||
| White, % (n) | 88.41 (61) | 88.89 (96) | 84 (21) | 89.77 (79) | 82.73 (91) | 96.23 (51) | 0.99 |
| Female % (n) | 69.57 (48) | 67.59 (73) | 76 (19) | 65.91 (58) | 60.91 (67) | 58.49 (31) | 0.50 |
| Age | 70.13 (8.37) | 70.81 (8.04) | 67.73 (5.40) | 69.45 (7.09) | 67.44 (5.92) | 67.41 (6.43) | 0.008 |
| BMI | 29.76 (6.54) | 29.48 (7.02) | 29.17 (6.34) | 29.78 (5.92) | 30.94 (7.38) | 29.86 (7.61) | 0.67 |
| Education (years) | 14.49 (3.16) | 14.71 (2.88) | 13.88 (2.85) | 14.82 (2.91) | 13.91 (2.33) | 14.26 (2.97) | 0.18 |
| Potential Prognostic Factors | |||||||
| Age at first episode | 47.48 (20.67) | 45.02 (22.15) | 35.28 (20.57) | 42.97 (21.63) | 38.88 (21.02) | 37.13 (21.39) | 0.02 |
| Episode duration (weeks)2, median (IQR) | 58 (24–160) | 52 (19.5–156) | 104 (26–210) | 80 (30–364) | 104 (34–277) | 104 (26.5–173) | 0.03 |
| Recurrent depression, % (n) | 72.46 (50) | 70.37 (76) | 84 (21) | 68.18 (60) | 68.18 (75) | 79.25 (42) | 0.48 |
| Adequate prior treatment (ATHF=> 3), % (n) | 60.87 (42) | 52.78 (57) | 48 (12) | 59.77 (52) | 66.67 (72) | 81.13 (43) | 0.02 |
| Number of drinks/week2 | 1.19 (2.96) | 0.70 (1.58) | 0.20 (0.65) | 0.85 (1.80) | 0.93 (2.67) | 1.23 (3.23) | 0.69 |
| SF-36 Physical Health | 45.40 (10.84) | 43.18 (10.9) | 43.53 (10.57) | 43.57 (12.09) | 40.98 (12.14) | 38.67 (11.69) | 0.05 |
| CIRS Count2 | 5.62 (2.23) | 6.44 (2.54) | 6.04 (2.01) | 6.13 (2.21) | 6.26 (2.44) | 5.94 (2.37) | 0.47 |
| CIRS Total2 | 8.94 (4.68) | 10.13 (4.58) | 9.56 (3.85) | 9.93 (3.96) | 10.09 (4.62) | 9.62 (4.65) | 0.71 |
| SIS Total | 1.83 (4.24) | 1.76 (3.74) | 1.28 (2.64) | 2.38 (4.63) | 2.38 (4.32) | 3.94 (5.68) | 0.10 |
| BSI Anxiety | 1.18 (0.85) | 1.36 (0.89) | 1.97 (0.92) | 1.26 (0.82) | 1.59 (0.94) | 2.07 (0.91) | <0.0001 |
| Total Depression Scores | |||||||
| Total baseline MADRS | 21.09 (4.12) | 25.87 (4.84) | 32.16 (4.06) | 24.09 (4.22) | 28.86 (4.64) | 32.96 (4.11) | <0.0001 |
| Total 12 week MADRS | 3.92 (5.59) | 5.63 (4.26) | 6.35 (4.05) | 13.49 (6.63) | 23.05 (4.67) | 30.84 (4.86) | <0.0001 |
For descriptive purposes, mean (SD) shown unless otherwise noted;
p values are based on Wald test of individual factors in separate, crude multinomial regression with trajectory group as the outcome;
p value computed after log transformation (mean and SD are not transformed);
BOLD indicates prognostic factor selected into the maximum model;
Abbreviations: Mod., Moderate; BMI, Body Mass Index; IQR, Interquartile range; ATHF, Antidepressant Treatment History Form; SF-36, Short Form-36; CIRS, Cumulative Illness Rating Scale for Geriatric; SIS, Scale for Suicidal Ideation; BSI, Brief Symptom Inventory; MADRS, Montgomery Asberg Depression Rating Scale;
Two groups did not exhibit clear changes in symptom severity and were labeled “non-response” (totaling 36% of the sample). These groups had high severity at baseline and followed linear trajectories without substantive change (24.3%, “high, non-response 1” and 11.7%, “high, non-response 2”). The last group had initially moderate symptom levels that followed a linear trajectory; while some symptom change was apparent (Figure 1), this group retained the majority of their average baseline symptoms at the end of the trial and was labeled “moderate, mixed/non-response” (19.43% of the sample).
Associations of prognostic factors with trajectory group
Clinical characteristics are presented by trajectory group for descriptive purposes only (Table 2). We also present crude associations of neuropsychological test performance and group membership (Supplementary Table 1); in these models, better performance in the delayed memory domain was related to higher odds of being in the “rapid response” group (all comparisons are with the “moderate, response” group); better performance in the language domain was related to lower odds of membership in the three non-responding groups.
In the final multivariable-adjusted model (Table 3), longer episode duration was related to 54% higher odds of being in the “high, non-response 1” group, as well as 40% higher odds of being in the “moderate, mixed/non-response” group. Greater education was related to 17% decreased odds of being in the “high, non-response 1” group.
Table 3.
Adjusted odds (95% confidence intervals) from the final multivariable model predicting trajectory group
| Responders | Non-responders | |||||
|---|---|---|---|---|---|---|
| Rapid Response | Mod., Response |
High, Response | Mod., Mixed/Non- response |
High, Non- response 1 |
High, Non- response 2 |
|
| Age | 1.01 (0.95–1.09) | Ref | 0.96 (0.88–1.04) | 1.02 (0.96–1.08) | 0.96 (0.91–1.02) | 0.98 (0.90–1.07) |
| Education | 0.83 (0.68–1.01) | Ref | 0.89 (0.73–1.08) | 0.91 (0.78–1.07) | 0.87 (0.76–0.99) | 0.93 (0.77–1.12) |
| Baseline MADRS | 0.56 (0.46–0.69) | Ref | 1.43 (1.21–1.69) | 0.76 (0.65–0.88) | 1.13 (1.00–1.27) | 1.54 (1.30–1.83) |
| Reduced Sleep | 0.83 (0.59–1.17) | Ref | 0.99 (0.62–1.59) | 0.72 (0.54–0.96) | 0.69 (0.53–0.91) | 0.62 (0.42–0.91) |
| Work/activity impairment | 1.29 (0.63–2.65) | Ref | 1.68 (0.57–4.95) | 1.51 (0.75–3.02) | 0.97 (0.55–1.70) | 3.65 (1.22–10.99) |
| Feelings of Guilt | 0.62 (0.32–1.19) | Ref | 1.47 (0.81–2.69) | 1.13 (0.67–1.90) | 1.49 (0.99–2.25) | 2.16 (1.23–3.79) |
| List Recognition | 2.22 (1.18–4.20) | Ref | 1.05 (0.62–1.77) | 1.02 (0.68–1.53) | 0.99 (0.70–1.42) | 0.67 (0.43–1.05) |
| Adequate prior Rx | 2.12 (0.73–6.16) | Ref | 0.34 (0.11–1.07) | 2.06 (0.84–5.04) | 1.84 (0.86–3.90) | 2.81 (0.89–8.90) |
| Episode duration (log) | 1.11 (0.78–1.60) | Ref | 1.25 (0.84–1.86) | 1.40 (1.04–1.87) | 1.54 (1.20–1.96) | 1.32 (0.93–1.88) |
Estimates were also adjusted for study site;
Bold indicates p<0.05;
Abbreviations: Mod. – Moderate, MADRS – Montgomery Asberg Depression Rating Scale, Rx – pharmacological treatment
Higher baseline depression severity was associated with greater odds of being in the two “high” treatment non-response groups as well as the “high, response” group (Table 3). Besides severity, no other baseline factors were associated with membership in the “high, response” group. Higher baseline depression severity was associated with reduced odds of being in both the “rapid response” and “moderate, mixed/non-response” groups.
Higher levels of the MADRS “reduced sleep” item was associated with lower odds of being in all three non-response groups. Greater work/activity impairment was associated with over three times the odds of being in the “high, non-response 2” group, and higher levels of guilt were associated with over twice the odds of being in this group. The association between guilt and greater odds of membership in the “high, non-response 1” group was non-significant (p=0.06).
Each standard deviation higher in list recognition performance (a subtest from the RBANS delayed memory domain) was associated with over twice the odds of being a “rapid responder”. The association between higher list-recognition performance and lower odds of being in the “high, non-response 2” group was non-significant (p=0.08).
Using the fitted parameters (as odds ratios in Table 3, and as β values in Table 4), and values for a patient’s pre-treatment characteristics, the probability of any patient belonging to each of the identified groups can be calculated following the equation described in Table 4.
Table 4.
Final model parameters and method to calculate the probability of membership in each of the six groups using pre-treatment factors
| Constant | Age | Education1 | Episode duration2 |
Adequate prior Rx |
Baseline MADRS |
Feelings of Guilt |
Work/activity impairment |
Reduced Sleep |
List Recognition3 |
|
|---|---|---|---|---|---|---|---|---|---|---|
| Rapid Response | 13.64234 | 0.01175 | −0.16811 | 0.1313 | 0.73722 | −0.55257 | −0.45662 | 0.18697 | −0.16743 | 0.80786 |
| High, response | −5.71627 | −0.0457 | −0.12917 | 0.0129 | −0.56455 | 0.33875 | 0.20344 | −0.01773 | −0.07557 | −0.01645 |
| Moderate, mixed/non-response | 4.72826 | 0.01463 | −0.0803 | 0.32928 | 0.74684 | −0.25408 | 0.07296 | 0.35628 | −0.34077 | 0.00213 |
| High, non-response 1 | 1.52122 | −0.0489 | −0.14872 | 0.34814 | 0.72161 | 0.12487 | 0.31311 | −0.15492 | −0.38422 | 0.00966 |
| High, non-response 2 | −11.74549 | −0.0433 | −0.10488 | 0.05454 | 1.43566 | 0.43544 | 0.43275 | 0.80958 | −0.5105 | −0.43492 |
Note that adjustment for study site in Table 3 was added for face validity and did not substantively alter findings; here study site is not adjusted for to facilitate use of this model for patients not located at one of the three IRLGREY study sites; All variables are defined as stated in the methods section, including:
Education in years;
Episode duration (in weeks) natural log transformed; and
Sub-test of the RBANS as age-normed z-score;
To calculate the estimated probability of membership in each group: (A) for each row above, raise e to the power of (constant + (Age*(value above))…+(List recognition*(value above))), (B) the estimated probability of membership in each group is equal to the number calculated for each row in step A, divided by (1+the sum of all numbers (one for each row) calculated in step A), (C) the probability of being in the “moderate, response” group (reference) is equal to 1 minus the sum of probabilities of being in the five other groups (from step B). Taken from equation #3 in Jones, Nagin, & Roder (2001) Sociological Methods & Research, 29(3), 374–393.
Abbreviations: Mod. – Moderate, MADRS – Montgomery Asberg Depression Rating Scale, Rx – pharmacological treatment;
DISCUSSION
Our study provides a novel description of the typical longitudinal patterns of depressive symptom change observed among older adults undergoing open-label pharmacological treatment for LLD. High baseline depression severity was generally related to having a non-responsive trajectory over time. Nevertheless, a small group of patients (“high, response”) did have a favorable trajectory, despite starting the trial among the most severely depressed. On the other hand, the “moderate, mixed/non-response” group began treatment with relatively moderate symptomology that did not respond to treatment robustly. Taken together, these findings suggest that while patients with a high baseline symptom burden are unlikely to respond to an initial trial of venlafaxine XR, treatment outcomes are likely jointly determined by several factors in addition to depression severity, including symptom expression, episode duration, and cognitive function.
A particularly innovative aspect of our study is the exploratory examination of individual symptoms in relation to response trajectories. Higher levels of guilt and greater work/activity impairment were both strongly and independently associated with only the “high, non-response 2” group. In contrast, greater levels of subjective sleep loss were associated with reduced odds of being in all of the non-responding groups. These findings suggest that the severity of these symptoms may be prognostic markers of response to pharmacological treatment for LLD; however, because this was an open-label trial (that was not placebo-controlled), our findings do not support concluding that these symptoms moderate venlafaxine efficacy. Also note that in the final model, three other sleep symptoms (early, middle, and late insomnia) from the HRDS-17 were not associated with trajectory group membership. Future placebo-controlled studies including objective sleep-wake assessments are needed to understand the neurobiological basis and roles of specific sleep characteristics in depression treatment response.
Consistent with prior analyses1, in our final model, prolonged episode duration was associated with non-response. Prolonged episode duration may thus contribute to treatment resistance, or alternatively, prolonged episode duration might be a consequence of other underlying resistance factors. The current analysis cannot resolve whether episode duration is a cause or consequence of treatment resistance. Nevertheless, episode duration should be considered when initiating a treatment strategy, and the early detection and treatment of depression may be advantageous.
We found that better performance on a delayed memory word list recognition test was associated with having a rapid response. This finding suggests that, despite older age and depressive illness, relative preservation of retention ability, which is associated with hippocampal function, may facilitate a rapid response. This hypothesis is consistent with prior evidence linking hippocampal volume to LLD treatment outcomes34,35. In crude models, we found worse performance on a semantic fluency test was associated with being in all three of the non-responsive groups. Mild Cognitive Impairment (MCI), although not formally adjudicated in the present study, is common among older adults with LLD36 and may explain this finding. Future studies tracking the role of MCI in the development of depression can determine whether semantic fluency deficits mark a distinct phenotype that (by nature of being a downstream resultant of cognitive impairment) will hinder response to traditional depression treatment.
Contrary to prior evidence37,38, the current analysis did not detect associations between physical co-morbidity and treatment outcomes. However, the current analysis did not include a full medical history or biological assessments. We did detect crude associations between perceived physical functioning, anxiety symptoms, age of onset, and pharmacological treatment history with trajectory group membership; however, these associations were attenuated in multivariable models, suggesting shared variance between these factors, those retained in the final model (i.e. work/activity impairment, depression severity, episode duration), and response.
Several limitations should be noted, including the need to replicate the observed response patterns in other samples. IRLGREY consisted of older adult out-patients and these findings cannot necessarily generalize to middle-aged/younger patients or older in-patients. Sufficient monitoring, dose titration, and depression management support were provided to minimize the possibility that inadequate pharmacological care would confound our analysis; in addition, the patterns and frequency of response cannot necessarily generalize to other settings, e.g. where monitoring/support are inadequate and attrition is more common. Analyses were prognostic and could not determine which factors, if any, moderate venlafaxine XR efficacy. We did not examine several other potentially important sources of variability in treatment response including biological and social-contextual factors. Still, the comprehensive set of possible prognostic factors examined was large which potentially increased the risk of type 1 error; the prognostic associations detected require confirmation in other samples. The severity of reduced sleep was associated with trajectory grouping, however reliance on a single self-reported item and not a validated sleep questionnaire eliminates the possibility of comparison with prior studies, included prior depression treatment research that included objective sleep assessments (e.g.39). Nevertheless, our results are consistent with one recent study that found slower treatment responses among a patient sub-group with relatively fewer subjective sleep complaints and more guilt/activity impairment40.
Strengths of our study include application of group-based trajectory modeling to the largest open-label pharmacologic trial for LLD conducted to date. The large cohort provided increased statistical power compared with prior studies which may have enabled detection of distinct response patterns across the full range of pre-treatment LLD severity. Another strength of our study is the multivariable analysis of a comprehensive set of prognostic factors, including both previously investigated (e.g. depression severity) and novel response markers (e.g. individual depressive symptoms).
In conclusion, we found that even in a specialty care setting, patients with severe pre-treatment depression regularly fail to respond to first-line pharmacotherapy. Nevertheless, since we observed response in the presence of high baseline severity, and treatment non-response in its absence, high depression severity alone may not be a completely accurate predictor of treatment response. Several other clinical prognostic factors likely contribute to the risk of treatment non-response, and variability in these prognostic factors potentially reflect important differences in MDD stage or phenotype. Intensive short term-monitoring and/or additional treatment strategies may be important for patients with severe depression, prolonged episode duration, and/or the specific LLD characteristics/co-morbidities discussed above. Future investigations should examine the biological basis of these prognostic markers and test whether these factors moderate the efficacy of first-line pharmacological treatments. Isolating the specific psychological and neurobiological circuits whose dysfunction moderates treatment efficacy will greatly improve our understanding of the development, detection, and treatment of LLD.
Supplementary Material
At a glance.
Among a large group of adults with late-life major depressive disorder (LLD), we found substantial heterogeneity in symptom change over 12 weeks of open-label Venlafaxine XR treatment.
High pre-treatment depression severity was associated with having a non-responsive trajectory, however we also observed that a small group of patients with high pre-treatment depression severity did appear to meaningful respond over 12 weeks.
A group of patients that did not have high pre-treatment depression severity did not have a clear and distinct response to treatment.
Several clinical prognostic factors (in addition to depression severity) were associated with having a particular (responsive or non-responsive) trajectory of symptom change, including: depressive episode duration, individual depression symptom expression, and performance on specific neuropsychological tests.
Acknowledgments
Supported by the three R01s at Pittsburgh (R01 MH083660), Washington University in St. Louis (R01 MH083648), and the Center for Addiction and Mental Health, Toronto (R01 MH 083643), Center Core grant P30 MH90333, and the UPMC Endowment in Geriatric Psychiatry. Dr. Butters served as a consultant for GlaxoSmithKline for whom she participated in diagnostic consensus conferences related to a clinical trial. Dr. Lenze received support from the Taylor Family Institute for Innovative Psychiatric Research. SFS has been supported by Research Training grant T32 AG000181 and T32 MH019986, and thanks his doctoral committee for their contribution to this work, as well as Bobby L. Jones, Ph.D., for his expert consultation regarding the statistical methods used.
Footnotes
SFS had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Trial Registration: clinicaltrials.gov identifier NCT00892047.
References
- 1.Joel I, Begley AE, Mulsant BH, et al. Dynamic Prediction of Treatment Response in Late-Life Depression. The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry. 2013 Feb 6; doi: 10.1016/j.jagp.2012.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Thomas L, Mulsant BH, Solano FX, et al. Response speed and rate of remission in primary and specialty care of elderly patients with depression. The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry. 2002 Sep-Oct;10(5):583–591. [PubMed] [Google Scholar]
- 3.Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. The British journal of psychiatry : the journal of mental science. 1979 Apr;134:382–389. doi: 10.1192/bjp.134.4.382. [DOI] [PubMed] [Google Scholar]
- 4.Stulz N, Thase ME, Klein DN, Manber R, Crits-Christoph P. Differential effects of treatments for chronic depression: a latent growth model reanalysis. Journal of consulting and clinical psychology. 2010 Jun;78(3):409–419. doi: 10.1037/a0019267. [DOI] [PubMed] [Google Scholar]
- 5.Uher R, Mors O, Rietschel M, et al. Early and delayed onset of response to antidepressants in individual trajectories of change during treatment of major depression: a secondary analysis of data from the Genome-Based Therapeutic Drugs for Depression (GENDEP) study. The Journal of clinical psychiatry. 2011 Nov;72(11):1478–1484. doi: 10.4088/JCP.10m06419. [DOI] [PubMed] [Google Scholar]
- 6.Uher R, Muthen B, Souery D, et al. Trajectories of change in depression severity during treatment with antidepressants. Psychol Med. 2010 Aug;40(8):1367–1377. doi: 10.1017/S0033291709991528. [DOI] [PubMed] [Google Scholar]
- 7.Hunter AM, Muthen BO, Cook IA, Leuchter AF. Antidepressant response trajectories and quantitative electroencephalography (QEEG) biomarkers in major depressive disorder. Journal of psychiatric research. 2010 Jan;44(2):90–98. doi: 10.1016/j.jpsychires.2009.06.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dew MA, Reynolds CF, 3rd, Houck PR, et al. Temporal profiles of the course of depression during treatment. Predictors of pathways toward recovery in the elderly. Archives of general psychiatry. 1997 Nov;54(11):1016–1024. doi: 10.1001/archpsyc.1997.01830230050007. [DOI] [PubMed] [Google Scholar]
- 9.Gildengers AG, Houck PR, Mulsant BH, et al. Trajectories of treatment response in late-life depression: psychosocial and clinical correlates. Journal of clinical psychopharmacology. 2005 Aug;25(4 Suppl 1):S8–S13. doi: 10.1097/01.jcp.0000161498.81137.12. [DOI] [PubMed] [Google Scholar]
- 10.Cui X, Lyness JM, Tang W, Tu X, Conwell Y. Outcomes and predictors of late-life depression trajectories in older primary care patients. The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry. 2008 May;16(5):406–415. doi: 10.1097/JGP.0b013e3181693264. [DOI] [PubMed] [Google Scholar]
- 11.Kuchibhatla MN, Fillenbaum GG. Trajectory classes of depression in a randomized depression trial of heart failure patients: a reanalysis of the SADHART-CHF trial. The American journal of geriatric pharmacotherapy. 2011 Dec;9(6):483–494. doi: 10.1016/j.amjopharm.2011.09.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Center for Financing A, and Cost Trends, AHRQ. Household Component of the Medical Expenditure Panel Survey. 2005 [Google Scholar]
- 13.Hajjar ER, Cafiero AC, Hanlon JT. Polypharmacy in elderly patients. The American journal of geriatric pharmacotherapy. 2007 Dec;5(4):345–351. doi: 10.1016/j.amjopharm.2007.12.002. [DOI] [PubMed] [Google Scholar]
- 14.Montgomery SA. Late-life depression: rationalizing pharmacological treatment options. Gerontology. 2002 Nov-Dec;48(6):392–400. doi: 10.1159/000065502. [DOI] [PubMed] [Google Scholar]
- 15.Shanmugham B, Karp J, Drayer R, Reynolds CF, 3rd, Alexopoulos G. Evidence-based pharmacologic interventions for geriatric depression. The Psychiatric clinics of North America. 2005 Dec;28(4):821–835. viii. doi: 10.1016/j.psc.2005.09.012. [DOI] [PubMed] [Google Scholar]
- 16.Whyte EM, Dew MA, Gildengers A, et al. Time course of response to antidepressants in late-life major depression: therapeutic implications. Drugs & aging. 2004;21(8):531–554. doi: 10.2165/00002512-200421080-00004. [DOI] [PubMed] [Google Scholar]
- 17.Morimoto SS, Gunning FM, Kanellopoulos D, et al. Semantic organizational strategy predicts verbal memory and remission rate of geriatric depression. International journal of geriatric psychiatry. 2012 May;27(5):506–512. doi: 10.1002/gps.2743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Pimontel MA, Culang-Reinlieb ME, Morimoto SS, Sneed JR. Executive dysfunction and treatment response in late-life depression. International journal of geriatric psychiatry. 2012 Sep;27(9):893–899. doi: 10.1002/gps.2808. [DOI] [PubMed] [Google Scholar]
- 19.Potter GG, Kittinger JD, Wagner HR, Steffens DC, Krishnan KR. Prefrontal neuropsychological predictors of treatment remission in late-life depression. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology. 2004 Dec;29(12):2266–2271. doi: 10.1038/sj.npp.1300551. [DOI] [PubMed] [Google Scholar]
- 20.Sakurai H, Uchida H, Abe T, et al. Trajectories of individual symptoms in remitters versus non-remitters with depression. Journal of affective disorders. 2013 Jul 22; doi: 10.1016/j.jad.2013.06.035. [DOI] [PubMed] [Google Scholar]
- 21.First MB, Spitzer Robert L, Gibbon Miriam, Williams Janet BW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P) New York: Biometrics Research, New York State Psychiatric Institute; 2002. [Google Scholar]
- 22.Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. Journal of psychiatric research. 1975 Nov;12(3):189–198. doi: 10.1016/0022-3956(75)90026-6. [DOI] [PubMed] [Google Scholar]
- 23.Hamilton M. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry. 1960;23:56–62. doi: 10.1136/jnnp.23.1.56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Oquendo MA, Baca-Garcia E, Kartachov A, et al. A computer algorithm for calculating the adequacy of antidepressant treatment in unipolar and bipolar depression. The Journal of clinical psychiatry. 2003 Jul;64(7):825–833. doi: 10.4088/jcp.v64n0714. [DOI] [PubMed] [Google Scholar]
- 25.Derogatis LR, Melisaratos N. The Brief Symptom Inventory: an introductory report. Psychol Med. 1983 Aug;13(3):595–605. [PubMed] [Google Scholar]
- 26.Beck AT, Kovacs M, Weissman A. Assessment of suicidal intention: the Scale for Suicide Ideation. Journal of consulting and clinical psychology. 1979 Apr;47(2):343–352. doi: 10.1037//0022-006x.47.2.343. [DOI] [PubMed] [Google Scholar]
- 27.Miller MD, Paradis CF, Houck PR, et al. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry research. 1992 Mar;41(3):237–248. doi: 10.1016/0165-1781(92)90005-n. [DOI] [PubMed] [Google Scholar]
- 28.Ware J. SF-36 Health Survey. Manual and Interpretation Guide, 2. Boston, MA: 1997. [Google Scholar]
- 29.Randolph C, Tierney MC, Mohr E, Chase TN. The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): preliminary clinical validity. Journal of clinical and experimental neuropsychology. 1998 Jun;20(3):310–319. doi: 10.1076/jcen.20.3.310.823. [DOI] [PubMed] [Google Scholar]
- 30.Delis DCKE, Kramer JH. Dellis Kaplan Executive Function System Examiner’s Manual. The Psychological Corporation; 2001. [Google Scholar]
- 31.Roeder K, Lynch KG, Nagin DS. Modeling Uncertainty in Latent Class Membership: A Case Study in Criminology. Journal of the American Statistical Association. 1999;94(447):766–776. [Google Scholar]
- 32.Jones BL, Nagin DS, Roeder K. A SAS Procedure Based on Mixture Models for Estimating Developmental Trajectories. [February 1, 2001];Sociological Methods & Research. 2001 29(3):374–393. [Google Scholar]
- 33.Kass RE, Raftery AE. Bayes Factors. [1995/06/01];Journal of the American Statistical Association. 1995 90(430):773–795. [Google Scholar]
- 34.Sheline YI, Disabato BM, Hranilovich J, et al. Treatment course with antidepressant therapy in late-life depression. The American journal of psychiatry. 2012 Nov 1;169(11):1185–1193. doi: 10.1176/appi.ajp.2012.12010122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Hsieh MH, McQuoid DR, Levy RM, Payne ME, MacFall JR, Steffens DC. Hippocampal volume and antidepressant response in geriatric depression. International journal of geriatric psychiatry. 2002 Jun;17(6):519–525. doi: 10.1002/gps.611. [DOI] [PubMed] [Google Scholar]
- 36.Reynolds CF, Butters MA, Lopez O, et al. Maintenance Treatment of Depression in Old Age: A Randomized, Double-blind, Placebo-Controlled Evaluation of the Efficacy and Safety of Donepezil Combined with Antidepressant Pharmacotherapy. Archives of general psychiatry. 2011;68(1):51–60. doi: 10.1001/archgenpsychiatry.2010.184. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Reynolds CF, 3rd, Dew MA, Pollock BG, et al. Maintenance treatment of major depression in old age. The New England journal of medicine. 2006 Mar 16;354(11):1130–1138. doi: 10.1056/NEJMoa052619. [DOI] [PubMed] [Google Scholar]
- 38.Alexopoulos GS, Kiosses DN, Murphy C, Heo M. Executive Dysfunction, Heart Disease Burden, and Remission of Geriatric Depression. Neuropsychopharmacology. 2004;29(12):2278–2284. doi: 10.1038/sj.npp.1300557. 09/01/online. [DOI] [PubMed] [Google Scholar]
- 39.Troxel WM, Kupfer DJ, Reynolds CF, 3rd, et al. Insomnia and objectively measured sleep disturbances predict treatment outcome in depressed patients treated with psychotherapy or psychotherapy-pharmacotherapy combinations. The Journal of clinical psychiatry. 2012 Apr;73(4):478–485. doi: 10.4088/JCP.11m07184. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Buhler J, Seemuller F, Lage D. The predictive power of subgroups: an empirical approach to identify depressive symptom patterns that predict response to treatment. Journal of affective disorders. 2014 Jul;163:81–87. doi: 10.1016/j.jad.2014.03.053. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.