Simvastatin Causes Changes in Affective Processes in Elderly Volunteers

Knashawn Morales, ScD; Marsha Wittink, MD; Catherine Datto, MD; Suzanne DiFilippo, RN; Mark Cary, PhD; Thomas TenHave, PhD; Ira R. Katz, MD, PhD

J Am Geriatr Soc. 2006;54(1):70-76. 

Abstract and Introduction

Objectives: To test for simvastatin-induced changes in affect and affective processes in elderly volunteers.
Design: Randomized, clinical trial.
Setting: The Geriatric Behavioral Psychopharmacology Laboratory at the University of Pennsylvania.
Participants: Eighty older volunteers, average age 70, with high normal/mildly elevated serum cholesterol.
Intervention: Simvastatin up to 20 mg/d or placebo for 15 weeks.
Measurements: Daily diary records of positive and negative affects and of events and biweekly measures of depressive symptoms. Affect ratings were obtained using the Lawton positive and negative affect scales; independent raters coded the valences of events.
Results: Thirty-one of 39 subjects assigned to placebo and 33 of 41 receiving simvastatin completed the study. During biweekly assessments, four subjects on simvastatin and one on placebo experienced depressive symptoms, as manifest by Center for Epidemiological Studies Depression scale scores greater than 16 (exact P = .36). Diary data demonstrated significant effects on affective processes. For positive affect, there was a significant medication-by-time interaction that reflected decreases in positive affect in subjects receiving simvastatin, greatest in those patients whose final total cholesterol levels were below 148 mg/dL. For negative affect, there were significant medication-by-event, and medication-by-event-by-time interactions, reflecting a time-limited increase in the apparent effect of negative events.
Conclusion: Simvastatin has statistically significant effects on affect and affective processes in elderly volunteers. The decrease in positive affect may be significant clinically and relevant to the quality of life of many patients.

The National Cholesterol Education Program's Clinical Practice Guidelines recently advised physicians to initiate treatment at lower thresholds for elevated cholesterol and to conduct more-intensive treatments for patients with elevated levels.[1] In this context, it is important to refine the evidence on the side effects, as well as the benefits, of these agents. In spite of long-standing questions about the issue, there is still controversy about whether statins and other cholesterol-lowering treatments can affect mood and behavior.

A series of studies have suggested that depression may be associated with low levels of plasma cholesterol,[2-9] possibly through an association between low cholesterol and changes in serotonin metabolism,[10] but the reported findings are difficult to interpret as a result of problems in controlling for the effects of medical illnesses and for the nutritional effects of depression.[11] Because of these potential confounders, findings from cholesterol-lowering treatments may be more informative. A pharmacoepidemiological study testing the hypothesis that prescription of 3-hydroxy-3 methyl glutaryl coenzyme A reductase inhibitors (statins) would be associated with subsequent prescriptions for antidepressants was negative,[12] but a cohort study suggested that individuals eating an antihyperlipemic diet and those taking simvastatin exhibited a greater rate of absence from work due to depression.[13]

More-recent studies have attempted to evaluate whether reducing cholesterol levels with statins may have an effect on mood or psychological well-being. Most of these studies have demonstrated no associations.[14-17] A recent investigation assessed the effect of long-term statin use on psychometric measures in an adult population with stable coronary artery disease and found that use of these agents was associated with lower risk of depression and related symptoms,[18] but selective dropout of patients with depression or those most vulnerable to it may have affected the findings. There have been suggestions that, because it is lipophilic, simvastatin may have a greater potential for causing affective toxicity than other statins, and a randomized clinical trial reported that simvastatin administration was associated with increased self-reports of depression.[19]

The potential affective toxicity of statin treatment remains a matter of major importance from neurobiological and public health perspectives.[20-22] To address this matter using methods complementary to those of previous research, a randomized clinical trial of simvastatin versus placebo in which the primary outcome measures were derived from multiple repeated self-reports of positive and negative affects and of daily events obtained by asking subjects to complete daily diaries in which they would report on their daily affects and events was conducted in elderly subjects. The methods for data acquisition and assessment were developed[23-31] as a strategy for improving time resolution in evaluating within-subject changes in affect and for allowing analyses of the effect of daily events on affect. In the context of the current research, the use of multiple repeated measures also serves to improve statistical power for detecting drug effects. The hypotheses tested in this study were that subjects receiving simvastatin would report changes in affect or affective processes manifest by increases in negative affect, decreases in positive affect, increases in the effect of negative events (on negative and positive affect), or decreases in the effect of positive events.

Methods

Recruitment between March 1999 and September 2001 led to entry of 80 healthy volunteers aged 65 and older who responded to newspaper advertisements, talks at senior centers, and other forms of outreach. The sample size was calculated to allow detection of effect sizes for change scores on repeated measures obtained at clinic visits (see below) of approximately one-third. Before study entry, consenting volunteers received a screening evaluation consisting of a physical examination, a semistructured psychiatric interview, and a screen of cognitive capacity. Subjects were required to be English speaking and to have at least 8 years of education. Inclusion also required plasma levels of total cholesterol at baseline between 160 and 240 mg/dL. General medical exclusion criteria were hospitalizations or significant medication changes within the previous month and significant laboratory abnormalities. Specific medical exclusions were central nervous system disease, active liver or kidney disease demonstrated through a review of systems or screening laboratory evaluations, and insulin-dependent diabetes mellitus. Psychiatric exclusions were cognitive impairment (Mini-Mental State Examination score <24); significant depressive symptoms (Center for Epidemiological Studies Depression scale (CES-D[32]) score >12); alcohol or substance abuse within the previous 5 years; and a history of a psychotic disorder, major depressive disorder, or bipolar disorder. Pharmacological exclusions were prior use of simvastatin and current use of medications capable of causing cognitive impairment or depression and medications with significant interactions with simvastatin. Potential subjects were also excluded if their performance was outside of the normal range on the tests that were administered to evaluate cognitive outcomes in other components of this research. Written informed consent was obtained before any study procedures. The institutional review board of the University of Pennsylvania approved the protocols. The research was supported by a grant from the National Institute of Mental Health (NIMH) and through the enabling resources of an NIMH Advanced Center for Intervention and Services Research and a Department of Veterans Affairs Mental Illness Research Education and Clinical Center. The sponsors had no role in the design, conduct, or analysis of or decision to publish these findings.

The study was a randomized, double-blind, placebo-controlled clinical trial. The process for obtaining informed consent occurred during a subject's first visit to the Geriatric Behavioral Psychopharmacology Laboratory of the University of Pennsylvania. The session included instructions on how to complete the daily affect diary and practice on cognitive test procedures used in other components of the research. The following week, baseline evaluations were conducted, and patients were randomized to increasing doses of simvastatin or placebo under double-blind conditions. They were seen every other week (at Weeks 1, 3, 5, 7, 9, 11, 13, and 15) to obtain cholesterol levels, to perform safety monitoring, and to collect diaries and at Weeks 7, 11, and 15 for psychological and cognitive testing. The volunteers were seen at various times of the day, with individual volunteers reporting to the laboratory at the same time each test session.

Subjects received simvastatin or placebo for 15 weeks. Thedosage was 5 mg/d for 2 weeks, then 10 mg/d for 2 weeks, and then 20 mg/d for the remainder of the study. To maintain the double blind, all medication was prepared in opaque identical-appearing red-and-blue gelatin capsules, which were sealed in blister cards with each individual dose identified on the packaging by the day and the time it was to be taken.

Each study visit included evaluations of mood and depressive symptoms using the CES-D[32] and probes for Diagnostic and Statistical Manual of Mental Disorders (DSM), Fourth Edition, symptoms of major depression based on the Structured Clinical Interview for DSM, nonpatient version.[33] Safety evaluations conducted at each study visit included structured probes for known side effects of simvastatin as well as open-ended inquiries. Other mood-related assessments included the confusion subscale from the Profile of Mood States,[34] a commonly used self-report instrument that covers a range of moods or affects, and the self-rated form of the Apathy Evaluation Scale,[35] an instrument designed to evaluate decreases in motivation. In addition, a battery of cognitive tests was administered; findings on cognitive outcomes will be reported separately.

At every visit, volunteers were given booklets that included forms for completing daily records of affects and space for recording daily events. Subjects were asked to complete the Lawton Positive Affect and Negative Affect rating scales at the end of each day.[29] The Lawton scales consist of five positive affect terms and five negative affect terms on which volunteers rate their affective state at the end of each study day on 5-point scales, with higher ratings corresponding to more-intense affects. The positive affect scale (range 5-25) comprises the affective states of energetic, warmth towards others, interested, happy, and content; the negative affect scale (range 5-25) comprises the affective states of annoyed, irritated, depressed, worried, and sad/blue. These assessment instruments were developed specifically for multiple repeated assessments of affect in older patients. The positive and negative affect scales have been validated in subjects living in the community and in residential care settings[23-31] and have been used in previous studies of drug effects.[36,37] Subjects were asked to record daily events that were outside of their usual routine. Independent raters evaluated the events using a "reasonable person" standard to establish their valences (whether they were positive, negative, or neutral). Events were rated as positive (e.g., "I went to dinner and a movie with friends"), negative (e.g., "I got in a fender-bender on the way home"), or neutral (e.g., "I picked up the dry cleaning"). Based on the reasonable-person standard, a visit from an adult child was considered positive, even though, in principle, it is possible that the child abused or exploited the subject. Two judges independently assigned valences and attained an agreement greater than 97%.

Baseline sample characteristics of treatment groups were compared using t tests for continuous measures and chi-square tests for categorical measures. The analysis of outcomes, for biweekly measures and daily diary data, consisted of testing and estimating the regression effect of positive and negative events on daily positive and negative affect across time for the entire sample and within and between each treatment group.

Positive affect scores, defined by the Lawton Daily Affect Scale, were modeled as a continuous outcome, but negative affect scores were highly skewed, with most ratings for most subjects at the instrument's floor. Therefore, analyses required defining cutpoints and transforming negative affects scores into categorical binary outcome measures corresponding to the presence or absence of a day with significant negative affect. Because any single threshold could be viewed as arbitrary, negative affect scores were dichotomized using three different cutpoints (70th, 90th, and 95th percentiles, corresponding to scale scores >7, 9, and 11, respectively). The transformed data formed three binary outcome measures representing the absence or presence of a "bad" day characterized by low, moderate, or high thresholds of negative affect.

Linear and logistic mixed regression models were used depending on the type of outcome, continuous or binary, respectively. Each model included main effects for baseline positive and negative affect scores, time (days), time-varying events (positive events were scored as 1 if ≥1 positive events were reported for the day, and 0 if there were none; negative events were scored in an analogous manner), and treatment group (placebo or simvastatin). In addition, the models included two-way and three-way interactions between day, time-varying events (positive and negative separately), and treatment group. The models also included random intercepts and slopes for time. These models were fit using Proc Mixed for the linear mixed regression model and the Glimmix macro for the logistic mixed regression model. Both procedures were used within SAS version 8 (SAS Institute, Inc., Cary, NC). For the linear mixed regression model with positive affect as the outcome, slope main effects are reported in terms of the change in positive affect with respect to a unit change in a covariate. Main effects from the logistic mixed regression models for dichotomous negative affect are the change in log odds of high versus low negative affect with a one-unit change in a covariate. The hypothesis that subjects receiving simvastatin would report changes in affect or affective processes corresponded to tests of significance for the group, group-by-time, group-by-positive event, group-by-negative event, group-by-positive event-by-time, and group-by-negative event-by-time effects in the linear model with positive affect and the logistic models with categorical negative affect as the dependent variables. To assess the change in positive affect with final total cholesterol level, four linear mixed regression models were fit, each containing a binary indicator of cholesterol quartile. For example, one model for positive affect included main effects for baseline positive and negative affect scores, time (days), and the indicator for the first cholesterol quartile. In addition, the models included two-way interactions between time (days) and final total cholesterol group. Because there is a high degree of colinearity between group assignment and final cholesterol status, these models did not control for treatment group in the analysis of the effect of cholesterol status on outcome. Results were reported in terms of the change in positive affect per day.

Some investigators have used sophisticated monitoring technologies to record the times when subjects completed entries in daily diaries and have reported high degrees of variability and nonadherence with instructions. Consequently, they have suggested that there may be difficulties with subject adherence in diary studies and that a "parking lot effect," in which subjects may fill in diary entries immediately before turning them in to investigators, may compromise the timing of entries.[38-40] This potential problem was addressed through a number of safeguards, including collection of diaries every 2 weeks and tests of the association between diary entries and periodic telephone inquiries.[36] Further evidence of the validity of diary data in the elderly subjects who participate in this research within this program comes from the analysis of lag effects, as reported previously[37] and confirmed in (unreported) analyses of data from this study that indicate that the associations between each day's positive and negative affect and same-day events are robust but that 1- and 2-day lag effects are attenuated.

Results

Of the 80 subjects in this study, 41 were randomized to receive placebo and 39 to simvastatin. In both groups, subjects were cognitively intact and euthymic, average age 70, and with 14 to 15 years of schooling ( ). The only significant difference between groups at baseline was a clinically modestly higher total cholesterol level in those randomized to receive placebo. Over the course of the 3 months of double-blind randomization, there was a highly significant decrease in total cholesterol levels in the group assigned to simvastatin and no change in that assigned to placebo. There were no significant differences between treatment groups in adherence to study medication, early termination due to any causes, early termination due to depression, onset of depressive symptoms indicated by CES-D scores greater than 16, or side effects. Nor were there significant differences between groups in changes over time in analyses that included biweekly CES-D scores or self-report measures of apathy or confusion.

  Study Participants' Baseline Characteristics, Disposition, and Side Effects

Characteristic, Disposition, or Side Effect Placebo
(n=41)
Simvastatin
(n=39)
Age, mean 70.9 70.5
Education, years, mean 14.2 15.0
Mini-Mental State Examination score, mean 28.6 28.7
Male, % 56.1 41.0
African American, % 14.6 15.4
Married, % 46.3 38.4
Baseline CES-D score, mean 3.2 3.5
Baseline positive affect, mean±SD 3.6±0.5 3.6±0.6
Baseline negative affect, mean±SD 1.2±0.2 1.2±0.3
Baseline total cholesterol mg/dL, mean±SD* 224.6±42.2 205.5±24.7
Final total cholesterol mg/dL, mean±SD 220.5±35.9 151.8±27.5
Adherence <80%, % 4.9 15.4
Any CES-D score >16, % 2.4 10.3
Early termination, % 17.1 20.5
For CES-D score >16 2.4 5.1
For other subjective symptoms 7.3 10.3
For intercurrent illnesses 4.9 5.1
For withdrawn consent 2.4 0.0
With reported side effects, % 39.0 41.0
Gastrointestinal symptoms 24.4 20.5
Aches, pains 14.6 10.3
Tiredness 4.9 15.4

*t = 2.39; P = .02.

t = 9.54; P < .001.

SD = standard deviation; CES-D = Center for Epidemiological Studies Depression subscale.

The findings on diary data presented here are derived from intention-to-treat analyses that considered all available data on all subjects regardless of dropout status. Findings on the subgroups of subjects who completed the full study (31/39 subjects assigned to placebo and 33/41 receiving simvastatin) were similar. As described above, diary data on positive affect were analyzed with random-effects mixed models that considered group assignment, time, positive events, negative events, and the two- and three-way interactions ( ). The findings that are relevant to the effects of medication on positive affect are those that include the group variable. The only significant effect is a group-by-time interaction indicating a greater decrease over time in positive affect in individuals who received simvastatin than in those who received placebo. The effect was greatest in the subgroups with lower final levels of total cholesterol, and it was significant only in the quartile with the lowest levels (below 148, ). Thus, the diary findings demonstrate that administration of simvastatin is associated with time-dependent, cholesterol-related decreases in self-reports of positive affect.

  Linear Mixed-Effects Models: Estimates of Treatment and Events Effects on Positive Affect

Parameter Estimate (Standard Error) P-value
Group 0.1068 (0.0557) .06
Time 0.000112 (0.000711) .88
Positive events 0.3394 (0.0528) <.001
Negative events −0.1936 (0.0465) <.001
Group by time −0.00240 (0.00183) .04
Positive events by time −0.00154 (0.00075) .04
Negative events by time −0.00030 (0.00076) .69
Positive events by group −0.06683 (0.06854) .33
Negative events by group −0.00191 (0.06180) .98
Positive events by group by time 0.000819 (0.000989) .41
Negative events by group by time 0.000285 (0.000285) .79

Effect is the average change in positive affect score with one unit change in the covariate holding all others constant. Results are adjusted for baseline positive and negative affect. The full model included baseline positive affect, baseline negative affect, group (coded 0 for placebo and 1 for simvastatin), time (in days), positive events (coded 0 for days without positive events and 1 for days with positive events), negative events (coded 1 for days without negative events and 1 for days with negative events), group by time, positive events by time, negative events by time, group by positive events, group by negative events, group by positive events by time, and group by negative events by time.

  Final Cholesterol Levels and Changes in Positive Affect over Time

Quartile Cholesterol (mg/dL) Simvastatin/ Placebo Group Slope (Change/Day) 95% Confidence Interval t P-value
1 115-147 19/0 −0.00284 −0.00549 to −0.00019 −2.13 .04
2 148-174 14/6 −0.00135 −0.00290-0.00021 −1.72 .09
3 175-226 5/15 −0.00042 −0.00265-0.00182 −0.37 .71
4 227-305 1/19 −0.00073 −0.00279-0.00133 −0.71 .48

The model for these calculations included baseline positive and negative affect and cholesterol but not positive or negative events.

It is possible to use the sample standard deviation in positive affect at baseline as a metric to allow a preliminary estimate of the possible clinical significance of this decrease in positive affect; specifically, it is possible to take 0.5 sample standard deviations as a threshold that defines when within-individual changes can be viewed as comparable with between-individual sources of variation and when they are likely to reflect clinically significant differences.[41] The decrease in positive affect in the entire simvastatin group was 0.13 baseline sample standard deviation units at 30 days, 0.25 units at 60 days, and 0.38 units at 90 days, although it is more informative to use individual-subject random-effect coefficients to estimate the number of individuals with decreases in positive affect of at least 0.5 sample standard deviations. At 30 days, such changes occurred in 3% of subjects assigned to simvastatin and none assigned to placebo (chi-square (X 2)=3.198; P = .07); at 60 days, the corresponding figures were 25.6% and 7.5% (X 2=4.727; P = .04); and at 90 days, they were 30.8% and 10% (X 2=5.274; P = .03).

Diary data on negative affect were highly skewed, with most normal subjects reporting no negative affects on most days. Accordingly, as described above, data on negative affect were analyzed using logistic mixed-effects analyses in which categorical "bad days" were defined in terms of cutpoints corresponding to the 70th (low threshold), 90th (moderate threshold), and 95th (high threshold) percentiles of scores; these cutpoints correspond approximately to 1 bad day every 3 days, 1 every 2 weeks, and 1 every month, respectively ( ). The findings that are relevant to the effects of medication on negative affect are those that include the group variable. There were group-by-negative event interactions for the moderate and high thresholds for the definition of bad days. Transforming the logistic regression coefficients (and standard errors) of ( ) into odds ratios (ORs), the OR for a bad day on a day with a negative event (relative to no negative event) was 4.66 (95% confidence interval (CI)=2.82-7.73) for placebo and 9.72 (95% CI=6.65-14.19) for simvastatin for the moderate threshold; the ORs were 5.46 (95% CI=3.08-9.66) and 16.44 (95% CI=10.80-25.03), respectively, for the high threshold. There were also group-by-negative event-by-time interactions for all of the thresholds. Considering these two effects together suggests that simvastatin caused a transient increase in the apparent effect of negative events on the probability of experiencing a bad day that is apparent early after drug administration, but then it becomes attenuated over time. The combined time-dependent coefficients for the interaction are (0.206-0.017 x days), (0.734-0.027 x days), and (1.103-0.023 x days), for the low, moderate, and high threshold definitions of bad days. Thus, the simvastatin-related increases in the effect of negative events are no longer present after approximately 15, 27, and 48 days for the low, moderate, and high thresholds, respectively.

  Logistic Mixed-Effects Models: Estimates of Treatment and Events Effects on Negative Affect

Parameter Negative Affect Percentile
70th 90th 95th
Estimate (Standard Error) P-value
Group −0.2165 (0.2820) .44 −0.4318 (0.3358) .20 −0.5646 (0.4481) .21
Time −0.00867 (0.00430) .04 −0.0129 (0.00512) .01 −0.00900 (0.00574) .12
Positive events −0.4328 (0.2201) .049 −0.6989 (0.2788) .01 −0.8803 (0.3423) .01
Negative events 10.7971 (0.2448) <.001 10.5398 (0.2575) <.001 10.6966 (0.2913) <.001
Group by time 0.00085 (0.00609) .89 0.00363 (0.00718) .61 0.00315 (0.00812) .70
Positive events by time 0.00391 (0.00479) .41 0.00075 (0.00581) .90 0.00206 (0.00733) .78
Negative events by time 0.01474 (0.00562) .009 0.01572 (0.00547) .004 0.00824 (0.00564) .14
Group by positive events −0.1255 (0.2940) .67 0.3812 (0.3520) .28 0.3480 (0.4278) .42
Group by negative events 0.2059 (0.3094) .51 0.7342 (0.3198) .02 10.1034 (0.3602) .002
Group by time by positive events −0.00168 (0.00631) .79 −0.00148 (0.00745) .84 −0.01410 (0.00969) .15
Group by time by negative events −0.01695 (0.00689) .01 −0.02650 (0.00685) <.001 −0.02298 (0.00733) .002

Coefficients represent the natural logarithm of the odds ratios for the change in the probability of a day with significant negative affect with one unit change in the covariate holding all others constant. Parameters were calculated in separate analyses for three different definitions of what constitutes a day with significant affect: cutpoints >7 corresponding to the 70th percentile of negative affect ratings and the experience of a "bad" day approximately once every 3 days; >9 corresponding to the 90th percentile and the experience of a "bad" day approximately once every 2 weeks; and >11 corresponding to the 95th percentile and the experience of a "bad" day approximately once every month. Results were adjusted for baseline positive and negative affect. The full model includes baseline positive affect, baseline negative affect, group (coded 0 for placebo and 1 for simvastatin), time (in days), positive events (coded 0 for days without positive events and 1 for days with positive events), negative events (coded 0 for days without negative events and 1 for days with negative events), group by time, positive events by time, negative events by time, group by positive events, group by negative events, group by positive events by time, and group by negative events by time.

  Logistic Mixed-Effects Models: Estimates of Treatment and Events Effects on Negative Affect

Parameter Negative Affect Percentile
70th 90th 95th
Estimate (Standard Error) P-value
Group −0.2165 (0.2820) .44 −0.4318 (0.3358) .20 −0.5646 (0.4481) .21
Time −0.00867 (0.00430) .04 −0.0129 (0.00512) .01 −0.00900 (0.00574) .12
Positive events −0.4328 (0.2201) .049 −0.6989 (0.2788) .01 −0.8803 (0.3423) .01
Negative events 10.7971 (0.2448) <.001 10.5398 (0.2575) <.001 10.6966 (0.2913) <.001
Group by time 0.00085 (0.00609) .89 0.00363 (0.00718) .61 0.00315 (0.00812) .70
Positive events by time 0.00391 (0.00479) .41 0.00075 (0.00581) .90 0.00206 (0.00733) .78
Negative events by time 0.01474 (0.00562) .009 0.01572 (0.00547) .004 0.00824 (0.00564) .14
Group by positive events −0.1255 (0.2940) .67 0.3812 (0.3520) .28 0.3480 (0.4278) .42
Group by negative events 0.2059 (0.3094) .51 0.7342 (0.3198) .02 10.1034 (0.3602) .002
Group by time by positive events −0.00168 (0.00631) .79 −0.00148 (0.00745) .84 −0.01410 (0.00969) .15
Group by time by negative events −0.01695 (0.00689) .01 −0.02650 (0.00685) <.001 −0.02298 (0.00733) .002

Coefficients represent the natural logarithm of the odds ratios for the change in the probability of a day with significant negative affect with one unit change in the covariate holding all others constant. Parameters were calculated in separate analyses for three different definitions of what constitutes a day with significant affect: cutpoints >7 corresponding to the 70th percentile of negative affect ratings and the experience of a "bad" day approximately once every 3 days; >9 corresponding to the 90th percentile and the experience of a "bad" day approximately once every 2 weeks; and >11 corresponding to the 95th percentile and the experience of a "bad" day approximately once every month. Results were adjusted for baseline positive and negative affect. The full model includes baseline positive affect, baseline negative affect, group (coded 0 for placebo and 1 for simvastatin), time (in days), positive events (coded 0 for days without positive events and 1 for days with positive events), negative events (coded 0 for days without negative events and 1 for days with negative events), group by time, positive events by time, negative events by time, group by positive events, group by negative events, group by positive events by time, and group by negative events by time.

Discussion

The results reported here confirm the hypothesis that simvastatin administration can lead to adverse effects on affect and affective processes. The effects include a time-dependent decrease in positive affect and a transient effect on the effect of negative events on negative affect. The latter effects correspond to an initial increase in the subjects' sensitivity to the negative events or hassles of everyday life that is then attenuated over the course of a number of weeks.

Although it is statistically significant, the decrease in positive affect over time due to simvastatin may not be of a magnitude that was clinically significant in the group as a whole, but the number of individuals who experienced decreases of a degree that was likely to constitute a clinically significant difference was greater for those receiving simvastatin than placebo. The overall effect of simvastatin on negative affect appears to involve a time-limited increase in the probability that negative events will lead to significant negative affect, especially for the moderate and high thresholds. Although the statistical models suggest that the negative effect of simvastatin on these parameters is no longer significant after long-term administration of the drug, further research on the effect of chronic treatment will be needed to probe for longer-term effects.

The findings reported here warrant discussion from a number of perspectives. First, there are questions about the relationship between these observations and long-standing questions about depression as an adverse effect of statins and other cholesterol-lowering treatments. It is possible that the affective toxicity observed with diary data reflects mechanisms that lead to the onset of clinical depression in the vulnerable individuals who may be most likely to experience statin-related depression.[42] From this perspective, it is possible that the suggestions in the literature about the associations between cholesterol-lowering treatments and clinical depression may represent the elusive "tip of an iceberg," reflecting other less-pervasive effects on affective processes that are operative far more frequently. Second, there may be questions about the interpretation of the relationships between affects and events. Although it may be intuitively reasonable to interpret these associations as a reflection of the effect of events on affects, this interpretation could be challenged. An alternative explanation is that the recollection and report of events and affects could be sensitive to subjects' affect at the time that entries were recorded. Third, the fact that significant findings emerged from this study in spite of ongoing controversy about the affective toxicity of statins suggests the power of the methods, including the use of daily diaries and mixed-effects models for statistical analysis.

There must also be questions about the mechanisms that underlie the findings reported here. It is not clear whether they reflect effects attributable to simvastatin as a specific medication, to the class of statins, or to all effective cholesterol-lowering interventions. Moreover, the literature suggests that statins may have effects related to the inhibition of cholesterol synthesis in the brain as well as the periphery[43-45] and that they may have direct physiological effects on blood vessels and other tissues independent of those effects that are mediated by lowering plasma cholesterol.[46-49] Therefore, it is not clear whether the affective outcomes related to simvastatin follow directly from decreases in plasma cholesterol or whether the affective outcomes and the changes in serum cholesterol reflect separable processes. Finally, there must be questions about generalizability to other statins and other treatments. Of the statins in clinical use, simvastatin (as well as lovastatin) is relatively lipophilic and able to cross the blood-brain barrier, whereas others are more hydrophilic.[50] Hence, depending upon the mechanisms that are operative, simvastatin may be more likely than other agents to cause central nervous system effects. These questions are not just relevant to the basic neurobiology of affect, they also represent problems for clinical management of patients for whom significant affective symptoms arise during treatment with simvastatin or related agents. From a pragmatic perspective, current knowledge is not adequate to recommend any specific intervention for patients who experience decreased positive affect, increased reactivity to negative events, or depression while taking simvastatin. Alternative clinical approaches include changing to a different statin that is less likely to cross the blood-brain barrier, changing to a different class of cholesterol-lowering medication, or addition of an antidepressant or other treatment directed toward the mood-related symptoms.

In summary, the findings reported here support the hypothesis that simvastatin can lead to adverse effects on affect and affective processes. Increasing evidence for the benefits of statins and other cholesterol-lowering treatments suggests the importance of further research targeted toward recognizing these symptoms and developing interventions to ameliorate them as components of strategies to improve acceptance of and adherence to more-widespread use of statins.


CLICK HERE for subscription information about this journal.

References

  1. Grundy SM, Cleeman JI, Merz CN et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110: 227-239.

  2. Engleberg H. Low serum cholesterol and suicide. Lancet 1992;339: 727-729.

  3. Muldoon MF, Manuck SB. Ischemic heart disease and cholesterol. Safety of cholesterol reduction remains in doubt. BMJ 1994;308: 1104-1105.

  4. Muldoon M. Injury, death and cholesterol. Ann Intern Med 1994;121: 719-720.

  5. Ryman A. Cholesterol, violent death, and mental disorder. BMJ 1994;309: 421-422.

  6. Brown SM, Salive ME, Harris TB et al. Low cholesterol concentrations and severe depressive symptoms in elderly people. BMJ 1994;308: 1328-1332.

  7. McCallum J, Simons L, Simons J et al. Low serum cholesterol is not associated with depression in the elderly: Data from an Australian community study. Aust N Z J Med 1994;24: 561-564.

  8. Boston P, Drusun S, Reveley MA. Cholesterol and mental disorder. Br J Psychiatry 1996;169: 682-689.

  9. Steegmans P, Hoes A, Bak AA et al. Higher prevalence of depressive symptoms in middle-aged men with low serum cholesterol levels. Psychosom Med 2000;62: 205-211.

  10. Steegmans P, Fekkes D, Hoes AW et al. Low serum cholesterol concentrations and serotonin metabolism in men. BMJ 1996;312: 221.

  11. Law M, Thompson S, Wald MJ. Assessing possible hazards of reducing serum cholesterol. BMJ 1994;308: 373-379.

  12. Lindberg G, Hallas J. Cholesterol-lowing drugs and antidepressants -- a study of prescription symmetry. Pharmacoepidemiol Drug Saf 1998;7: 399-402.

  13. Boumendil E, Tubert-Bitter P. Depression-induced absenteeism in relation to antihyperlipidemic treatment: A study using GAZEL cohort data. Epidemiology 1995;6: 322-325.

  14. Stewart R, Sharples K, North FM et al. Long-term assessment of psychological well-being in a randomized placebo-controlled trial of cholesterol reduction with pravastatin. Arch Intern Med 2000;160: 2129-2134.

  15. Dujovne CJ, Chremos AN, Pool JL et al. Expanded Clinical Evaluation of Lovastatin (EXCEL) study results, IV. Additional perspectives on the tolerability of lovastatin. Am J Med 1991;91(Suppl B): 25S-30S.

  16. Downs J, Oster G, Santanello NC. HMG CoA reductase inhibitors and quality of life. JAMA 1993;269: 3107-3108.

  17. Muldoon MF, Barger SD, Ryan CN et al. Effects of lovastatin on cognitive function and psychological well being. Am J Med 2000;108: 538-546.

  18. Hyyppa MT, Kronholm E, Virtanen A et al. Does simvastatin affect mood and steroid hormone levels in hypercholesterolemic men? A randomized double-blind trial Psychoneuroendocrinology. 2003;28: 181-194.

  19. Young-Xu Y, Chan KA, Liao JK et al. Long-term statin use and psychological well-being. J Am Coll Cardiol 2003;42: 690-697.

  20. Lett HS, Blumenthal JA, Babyak MA et al. Depression as a risk factor for coronary artery disease. Evidence, mechanisms, and treatment. Psychosom Med 2004;66: 305-315.

  21. Pratt LA, Ford DE, Crum RM et al. Depression, psychotropic medication, and risk of myocardial infarction. Prospective data from the Baltimore ECA follow-up. Circulation 1996;94: 3123-3129.

  22. Ariyo AA, Haan M, Tangen CM et al. Depressive symptoms and risks of coronary heart disease and mortality in elderly Americans. Circulation 2000;102: 1773-1779.

  23. Winter L, Lawton MP, Casten RJ et al. The relationship between external events and affect states in older people. Int J Aging Hum Dev 2000;50: 85-96.

  24. Lawton MP, Winter L, Kleban MH et al. Affect and quality of life: Objective and subjective. J Aging Health 1999;11: 169-198.

  25. Lawton MP, De Voe MR, Parmelee P. Relationship of events and affect in the daily life of an elderly population. Psychol Aging 1995;10: 469-477.

  26. Lawton MP, Kleban MH, Dean J. Affect and age. Cross-sectional comparisons of structure and prevalence. Psychol Aging 1993;8: 165-175.

  27. Lawton MP, Kleban MH, Rajagopal D et al. Dimensions of affective experience in three age groups. Psychol Aging 1992;7: 171-184.

  28. Kleban MH, Lawton MP, Nesselroade JR et al. The structure of variation in affect among depressed and nondepressed elders. J Gerontol 1992;47: P190-P198.

  29. Lawton MP, Kleban MH, Dean J et al. The factorial

  30. Lawton MP, DeVoe MR, Parmelee P. Relationship of events and affect in the daily life of an elderly population. Psychol Aging 1995;10: 469-477.

  31. Lawton MP, Parmelee PA, Katz IR et al. Affective states in normal and depressed older people. J Gerontol B Psychol Sci Soc Sci 1996;51B: P309-P316.

  32. Radloff LS. The CES-D Scale: A self-report depression scale for research in the

  33. First MB, Spitzer RL, Gibbon M et al. Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), Clinician Version.Washington, DC: American Psychiatric Press, 1997.

  34. McNair DM, Lorr M, Droppleman LF. Profile of Mood States (POMS).San Diego, CA: Edits/Educational and Industrial Testing Service, 1981.

  35. Marin RS, Biedrzycki RC, Firinciogullari S. Reliability and validity of the Apathy Evaluation Scale. Psychiatry Res 1991;38: 143-162.

  36. Katz IR, Morales K, Datto C et al. Probing for affective side effects of drugs used in geriatric practice: Use of daily diaries to test for effects of metoclopramide and naproxen. Neuropsychopharmacology 2005 Apr 27; [Epub ahead of print].

  37. Furlan PM, Kallan MJ, TenHave T et al. SSRIs do not cause affective blunting in healthy elderly volunteers. Am J Geriatr Psychiatry 2004;12: 323-330.

  38. Schwartz JE, Stone AA. Strategies for analyzing ecological momentary assessment data. Health Psychol 1998;17: 6-16.

  39. Stone AA, ed. The Science of Self-Report.Mahwah, NJ: Lawrence Erlbaum Associates, 2000.

  40. Stone AA, Shiffman S, Schwartz JE et al. Patient compliance with paper and electronic diaries. Control Clin Trials 2003;24: 182-199.

  41. Cella D, Bullinger M, Scott C et al. Group vs individual approaches to understanding the clinical significance of differences or changes in quality of life. Mayo Clin Proc 2002;77: 384-392.

  42. Jacobs D, Blackburn H, Higgins M et al. Report of the conference on low blood cholesterol: Mortality associations. Circulation 1992;86: 1046-1060.

  43. Bjorkhem I, Meaney S. Brain cholesterol: Long secret life behind a barrier. Arterioscler Thromb Vasc Biol 2004;24: 806-815.

  44. Locatelli S, Lutjohann D, Schmidt HH-J et al. Reduction of plasma 24S-hydroxycholesterol (cerebrosterol) levels using high-dosage simvastatin in patients with hypercholesterolemia: Evidence that simvastatin affects cholesterol metabolism in the human brain. Arch Neurol 2002;59: 213-216.

  45. Lutjohann D, Stroick M, Bertsch T et al. High doses of simvastatin, pravastatin, and cholesterol reduce brain cholesterol synthesis in guinea pigs. Steroids 2004;69: 431-438.

  46. Bonetti PO, Lerman LO, Napoli C et al. Statin effects beyond lipid lowering -- are they clinically relevant? Eur Heart J 2003;24: 225-248.

  47. Stancu C, Sima A. Statins: Mechanism of action and effects. J Cell Mol Med 2001;5: 378-387.

  48. Laufs U. Beyond lipid-lowering: Effects of statins on endothelial nitric oxide. Eur J Clin Pharmacol 2003;58: 719-731.

  49. Wierzbicki AS, Poston R, Ferro A. The lipid and non-lipid effects of statins. Pharmacol Ther 2003;99: 95-112.

  50. Hamelin BA, Turgeon J. Hydrophilicity/lipophilicity. Relevance for the pharmacology and clinical effects of HMG-CoA reductase inhibitors. Trends Pharmacol Sci 1998;19: 26-37.