Body Mass Index as a Moderator of Treatment Response to Ketamine for Major Depressive Disorder

Abstract

Purpose/Background:

Major depressive disorder (MDD) and obesity commonly co-occur. We sought to assess the impact of Body Mass Index (BMI) on the acute antidepressant effects of ketamine in patients with treatment resistant depression (TRD).

Methods/Procedures:

Post-hoc analyses were conducted from a multi-site, randomized, double-blind, placebo-controlled trial designed to assess the rapid-onset effects of intravenous ketamine. Patients (N=99) were randomized to a single dose administration of ketamine 0.1 mg/kg (n=18), ketamine 0.2 mg/kg (n=20), ketamine 0.5 mg/kg (n=22), ketamine 1.0 mg/kg (n=20), or active placebo, midazolam 0.045 mg/kg (n=19). Patients were stratified for BMI. For patients randomized to ketamine (N=80), BMI was assessed as a continuous variable and also categorically (obese, overweight, not obese/overweight (reference)). The primary outcome measure was the change on the 6-item Hamilton Depression Rating Scale (HAM-D6) 24 hours after treatment. Outcomes at day 3 were also assessed.

Findings/Results:

HAMD-6 change scores at 24 hours were inversely associated with BMI (−0.28±0.12, p=0.02). With BMI operationalized categorically, both obese (−4.15±1.41, p=0.004) and overweight (−1.99±1.14, p=0.08) categories were inversely related to HAMD-6 change score at 24 hours, statistically significant for the obese category, as compared with the reference group. Similar but weaker findings were observed at 72 hours post-infusion.

Implications/Conclusions:

Higher BMI and obesity were associated with a more robust acute antidepressant response to ketamine. This may have clinical relevance for a great number of patients who have both MDD and obesity.

Clinical Trial Registration:

NCT01920555.

Introduction

Major depressive disorder (MDD) and obesity co-occur and have been shown to have a deleterious impact on quality of life beyond simple additive effects. ¹ The association between obesity and depression appears to be bidirectional, with obesity predicting the onset of depression, and depression predicting subsequent obesity. ^2–5 Metabolic syndrome, lipid dysregulation, cardiovascular disease, and risk of death at an earlier age are elevated in individuals who have psychiatric disorders, including MDD. ^6,7 In addition to other variables, many psychiatric medications have weight gain and metabolic dysregulation as commonly experienced side effects. ⁸ The presence of obesity may indicate a risk of a more chronic course of MDD and may be accompanied by prefrontal neuroanatomical differences that predispose such risk. ⁹ Notably, treatment response to antidepressants may be impacted by baseline body mass index (BMI). Specifically, higher BMI may be associated with a lower or slower response to some antidepressant medications but not others ¹⁰, although this finding is inconsistent and has not been the primary focus of studies in which this is reported. ¹¹

Ketamine treatment for MDD has been increasingly studied and has become more commonly available and utilized. Predictors of treatment response would help determine when ketamine should be considered. Data are sparse as regards ketamine response and the impact of BMI and obesity. Previously in two other reports, higher BMI was demonstrated to be associated with a greater decrease in Hamilton Depression Rating Scale (HAM-D) scores.¹²,¹³ In that study, higher BMI was associated with greater improvement of depressive symptoms at 230 minutes and one day after a single infusion, although not at one week after the infusion.

In order to provide more data on a possible relationship between BMI and ketamine treatment response, the current analyses represent a secondary analysis of a multiple dose, placebo-controlled trial of ketamine for the acute treatment of MDD for which patients had not experienced an adequate response from antidepressant trials. The primary study has been reported in detail and demonstrated an acute superior antidepressant response to a single intravenous ketamine administration compared to active placebo. ¹⁴ Planned exploratory analyses did not demonstrate a sex difference in response rates, nor differences between pre- and postmenopausal women.¹⁵ The objective of these analyses was to assess the impact of BMI and obesity on the acute antidepressant effects of ketamine. We aimed to assess BMI as a continuous variable, as well as assessing the weight categories of obesity, overweight, and normal weight and the impact on acute antidepressant treatment response to ketamine.

Materials and Methods

Overview

The relationship between BMI and response to ketamine was conducted through exploratory post-hoc analyses from a multi-site, randomized, placebo-controlled trial designed to assess the rapid-onset antidepressant effects of adjunctive ketamine infusion therapy for treatment resistant depression (TRD), and to assess the efficacy of differential dosing. ¹⁴ In brief, the study was a randomized, placebo-controlled trial of intravenous ketamine or active placebo administered (midazolam infusion) in addition to ongoing, stable, and adequate antidepressant therapy. The study was approved by the site IRBs and NIMH Data Safety and Monitoring Board. BMI was calculated as body weight (kilograms) divided by height (meters) squared. Prior to randomization, patients were stratified for BMI into two groups (BMI ≤30 or BMI >30), and then blocked randomized into treatment groups. Patients were randomized to one of five treatment arms in a 1:1:1:1:1 distribution: a single dose of ketamine 0.1 mg/kg (n=18), a single dose of ketamine 0.2 mg/kg (n=20), a single dose of ketamine 0.5 mg/kg (n=22), a single dose of ketamine 1.0 mg/kg (n=20), or a single dose of active placebo, midazolam 0.045 mg/kg (n=19).

For exploratory analyses pertaining to BMI, participants were categorized based on BMI categories of underweight, normal weight, overweight and obese as per World Health Organization Guidelines. As there was only one participant classified as “underweight,” we grouped them in with the “normal weight” individuals to form a “not overweight/obese” category.

Subjects

Male and female outpatients between the ages of 18 and 70 years were eligible for the study if they were experiencing a major depressive episode (MDE) of at least eight weeks’ duration, and had a diagnosis of MDD, as defined by the Diagnosis and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision. Eligible subjects were confirmed to have TRD during the current MDE, defined as failure to achieve response (≥50% improvement) after at least two adequate treatment courses of antidepressant therapy. Remote raters confirmed that patients met inclusion criteria. For a complete summary of the inclusion/exclusion criteria used in this protocol, and a CONSORT diagram, please refer to the primary report. ¹⁴

Methods and Study Medication

Participants were screened between 7 to 28 days, during which eligibility was determined, prohibited medications were discontinued, and eligible subjects progressed to the baseline visit. Participants were stratified by body mass index (BMI) (≤30 and >30), and were block randomized into five treatment arms. N= 99 subjects were assigned to each of five arms in a 1:1:1:1:1 fashion; four treatment groups that received IV ketamine at different single-dose administrations (0.1 mg/kg, 0.2 mg/kg, 0.5 mg/kg, and 1.0 mg/kg respectively), and an active control group that received a single dose of IV midazolam (0.045 mg/kg). At the baseline visit (Day 0), subjects received either ketamine or midazolam by continuous infusion for 40 minutes.

Assessments

Remote trained raters of the primary outcome measure were blinded to treatment assignment. The primary outcome measure was the change on the 6-item Hamilton Depression Rating Scale (HAM-D6) at 24 hours after treatment. ^16–19 This version was selected as the primary outcome measure due to sensitivity for the detection of acute changes. ¹⁷

Statistical analyses

Analyses were conducted with BMI operationalized both continuously and categorically. For the categorical variable, cutoffs were obtained from the World Health Organization and Centers for Disease Control and Prevention (CDC): BMI >30 was defined as “obese,” BMI < 30 and ≥25 was defined as “overweight,” and BMI < 25 was “not obese or overweight.”

To assess the relationship between BMI and response to ketamine, we utilized a general linear regression model using PROC GLM in which HAMD-6 change score (Day 1-baseline) was the dependent variable, and HAMD-6 score at baseline, ketamine treatment group, sex, and BMI were independent variables. We additionally assessed for interactions between treatment group and BMI. Since the interaction terms were not significant, we did not include them in the final models. We included a random effect for site (six sites). Ketamine treatment group was a four-level group variable to account for different ketamine dosages (0.1, 0.2, 0.5, 1.0 mg/kg) assigned to participants. Four regressions were run in order to assess HAMD-6 change scores for Days 1 and 3 (Day 3-baseline) as well as two different ways of measuring BMI—continuously and categorically, as described above. We ran similar models within the placebo group. All analyses were performed using SAS 9.4 statistical software.

Results

For Day 1 outcomes, N=80 participants were included in analyses, as N=99 were randomized in the clinical trial, 80 of whom were randomized to one of four doses of ketamine, and 19 were randomized to placebo (midazolam). Results presented for this report were restricted to participants randomized to ketamine. When analyzed separately, there was not a significant relationship between BMI and HAMD-6 response among those randomized to placebo (results not shown). Baseline data including demographics and BMI are presented in Table 1. When we compared demographics across weight categories, the only variable that was significantly different was sex. Therefore, sex (male/female) was included as a covariate in the analyses.

Table 1.

Sample Characteristics (N=80)

	Obese N=12 M±SD or N(%)	Overweight N=30 M±SD or N(%)	Not obese or overweight N=38 M±SD or N(%)	P	Full Sample N=80 M±SD or N(%)
Age	51.20±10.36	46.86±12.27	44.27±12.92	0.2318	46.28±12.41
Gender (% fem)	5 (41.67)	8 (26.67)	25 (65.79)	0.0053**	38 (47.50)
HAMD-6 score at Baseline	13.00±2.26	12.28±2.12	12.79±1.73	0.4476	12.63±1.96
Ketamine Group (mg/kg)				0.6983
1.0	4 (33.33)	8 (26.67)	8 (21.05)		20 (25.0)
0.5	5 (41.67)	6 (20.00)	11 (28.95)		22 (27.50)
0.2	2 (16.67)	8 (26.67)	10 (26.32)		20 (25.0)
0.1	1 (8.33)	8 (26.67)	9 (23.68)		18 (22.50)
Dose	55.27±32.91 Range: 7.45–98.90	37.7±31.79 Range: 7.03–100.70	26.92±23.03 Range: 4.92–81.40	0.0106*	35.21±29.44 Range: 4.92–100.70
BMI
Overall Group	32.07±1.49 Range: 30.23–34.48	27.37±1.40 Range: 25.09–29.92	21.68±1.92 Range: 18.43–24.96	<0.0001	25.37±4.20 Range: 18.43–34.48
Females	32.97±1.41 Range: 31.23–34.48	27.61±1.51 Range: 25.09–29.72	21.24±1.93 Range: 18.43–24.96	<0.0001	24.13±4.68 Range: 18.43–34.48
Males	31.43±1.26 Range: 30.23–33.80	27.28±1.38 Range: 25.21–29.92	22.52±1.66 Range: 18.96–24.73	<0.0001	26.50±3.39 Range: 18.96–33.80

Note: one person had a BMI of <18.5 (underweight), they were included in the “not obese or overweight” category;

^*p<0.05

^**p<0.01

Results are summarized in Table 2. In a regression with HAMD-6 change score from baseline to 24 hours post-treatment as the dependent variable and baseline HAMD-6, treatment group, and BMI measured continuously as the independent variables, HAMD-6 change score at 24 hours was inversely associated with BMI (−0.28±0.12, p=0.02) (R-Square = 0.18). When a similar regression was run, but with BMI operationalized categorically (obese, overweight, not obese/overweight (reference)), both obese (−4.15±1.41, p=0.004) and overweight (−1.99±1.14, p=0.08) categories were inversely related to HAMD-6 change score at 24 hours, with a statistically significant (p<0.05) association for the obese category relative to the non obese/overweight reference group (R-Square = 0.23).

Table 2.

Relationship between BMI and change in HAMD-6 score after administration of ketamine (N=80)

	Day 1 β± Standard Error	Day 3 β± Standard Error
BMI (continuous)	−0.28±0.12 p=0.02*	−0.21±0.12 p=0.09
BMI (categorical)
Not overweight/obese (ref.)	--	--
Overweight	−1.99±1.14 p=0.08	−1.17±1.14 p=0.31
Obese	−4.15±1.41 p=0.004**	−3.15±1.44 p=0.03*

Analyses adjusted for baseline HAMD-6 score, ketamine group, and sex;

^*p<0.05

^**p<0.01

On Day 3, N=74 participants were included in analyses, as 6 of the participants randomized to ketamine dropped out after Day 1. In a regression with HAMD-6 change score at 72 hours post-treatment as the dependent variable and baseline HAMD-6, treatment group, and BMI measured continuously as the independent variables, HAMD-6 change score at 72 hours was inversely related to BMI but the association was not statically significant (−0.21±0.12, p=0.09) (R-Square = 0.19). When a similar regression was run, but with BMI operationalized categorically, both obese (−3.15±1.44, p=0.03) and overweight (−1.17±1.14, p=0.31) categories were inversely related to HAMD-6 change score at 72 hours, with a statistically significant (p<0.05) association for the obese category, relative to the non obese/overweight reference group (R-Square = 0.21).

Discussion

In these analyses, a higher BMI and weight categorization as obese were associated with a greater acute improvement from a single dose administration of IV ketamine. This finding is consistent with two previous reports in which greater antidepressant response to ketamine was positively correlated with greater BMI.^12,13 The finding that continuous and categorical analyses of BMI were associated with antidepressant response, if consistently found in other studies, might help assist treatment decisions for overweight and obese patients with MDD. At the least, these results support that higher body mass index or fat mass may be a variable associated with response to ketamine and do not appear to lead to poorer outcomes with ketamine. These findings also have the potential to generate further inquiry into the mechanism of action of ketamine and the role of obesity in depression and its treatment. Previous studies have not shown consistently that patients respond differentially to other antidepressant therapies based on BMI. ^10,20,21

These results must be interpreted with caution until they are better understood. It remains unclear of why higher BMI would affect response to IV ketamine. It is unclear whether the observed effect is due to the pharmacokinetic differences that would be influenced by body fat mass, as ketamine is a lipophilic drug, or whether there is a more complicated relationship between obesity and response to treatment accounted for by other variables.

Ketamine is metabolized to norketamine by cytochrome P450 (CYP) 3A and 2B6 enzymes.²² Rate of clearance of CYP3A substrates may be influenced by age, and at least some substrates may be influenced by sex steroids. ²³ While we previously did not find sex differences in response to treatment in this study, there may be subtle but clinically relevant effects of age and sex on body fat distribution and drug metabolism that are not fully captured by BMI as a crude variable.¹⁵ BMI remains a commonly utilized variable in clinical research, but may not adequately capture body composition to fully account for the effects of obesity.²⁴ Therefore, it remains unknown at this time whether the findings suggest pharmacokinetic differences that would lead to variable levels of ketamine based on BMI, or whether there are more pharmacodynamic processes that would account for the differences in treatment response.

There are several proposed mechanisms that may contribute to the comorbidity between depression and obesity and might explain a preferential response to ketamine’s antidepressant effects among patients with higher BMIs. Both obesity and MDD are conditions that are associated with high degrees of inflammation, and ketamine has been demonstrated to also have anti-inflammatory effects. ²⁵ Anti-inflammatory effects of some antidepressants have been demonstrated in response to antidepressants in animals and humans, although it is not clear that antidepressant effects are due to, or solely due to, acute anti- inflammatory actions. ^26–28 A more specific hypothesis is that adipokine dysregulation is present in individuals with both MDD and obesity and may also impact ketamine as well as other antidepressant responses. ^29–32 Co-occurrence may be predisposed by specific polymorphisms related to the FTO gene (fat mass and obesity associated) gene, in which expression of obesity is increased in individuals with depression in comparison to controls without depression. ³³

Antidepressant medications are not typically given in doses based on body weight, and data do not consistently demonstrate a rationale for doing so. However, ketamine in this study was specifically given in doses based on total body weight, and for each group dose assignment, it was given in mg/kg based on the patient’s total body weight. It is not known whether a different metric should be used for individuals at higher or lower weight ranges, or whether efforts should be taken to base dosing on lean body mass rather than total body mass.

The strengths of this report include the use of data from a randomized, active placebo-controlled, double-blind design, along with rigorous assessments of efficacy and safety. Ratings of the primary outcome were done by blinded, remote trained raters. BMI was collected during study visits by measurements of height and weight, rather than self report. Another strength of this report is the growing relevance of understanding ketamine treatment response, as ketamine and related compounds become more commonly accessible and utilized.

There are important limitations to this study. The limitations include the single dose administration and inability to generalize to serial dosing of ketamine as it is often used clinically in repeated administrations. We are additionally not able to generalize to other modes of ketamine administration such as the intranasal, oral, or transdermal route. Also, ketamine was given concomitantly with patients’ ongoing antidepressants, and it is not known if ketamine monotherapy would have produced similar findings. Another important limitation is that we do not know if higher doses based on higher body weights and body fat percentages resulted in higher peripheral ketamine levels or greater central nervous system concentrations. In fact, there are major limitations to this and other studies related to how body composition is assessed, and how it is considered in both research and clinical settings in the absence of systematic data pertaining to the use of most drugs.³⁴ The extant literature does not adequately report upon ketamine pharmacokinetics relative to obesity and body mass, and some antidepressants, for example, vortioxetine, have been recently demonstrated to have pharmacokinetic parameters such as washout half-lives vary by BMI category.²⁴ Finally, the study was not designed to assess the impact of body weight on treatment outcome, and the number of patients in each weight category was small.

In summary, we found that in this sample, higher BMI as a continuous variable and categorization in the obese/overweight categories were associated with a more robust acute antidepressant response to ketamine. This may have clinical relevance for a great number of patients who have both MDD and obesity, if this is demonstrated in longer-term studies of ketamine treatment. Thus far, studies do not consistently demonstrate that monoamine-based antidepressants are more efficacious in patients with higher BMI ^20,35, so the finding that ketamine may be particularly robust in individuals who have higher BMIs, or at the very least as helpful as for those in the normal range, deserves follow up study. Future research should elucidate how BMI factors into treatment response and dose selection.