Research report
Divalproex, lithium and suicide among Medicaid patients with bipolar disorder

https://doi.org/10.1016/j.jad.2007.07.014Get rights and content

Abstract

Background

Suicide completion and attempted suicide are major concerns for people with bipolar disorder. Studies in the private sector have suggested that lithium treatment may be superior to divalproex therapy with regard to minimizing suicidal behavior among individuals with bipolar disorder. However, few data are available regarding Medicaid patients diagnosed with bipolar disorder.

Methods

Subjects were 12,662 Oregon Medicaid patients diagnosed with bipolar disorder and treated with medication between 1998 and 2003. Outcomes measures were completed suicide and emergency department visits for suicide attempts (including non-fatal poisoning). Cox proportional hazards models were used to adjust for demographics, co-morbidity, and concurrent psychotropic medication use.

Results

Divalproex was the most common mood stabilizer (used by 33% of subjects) followed by gabapentin (32%), lithium (25%), and carbamazepine (3%). There were 11 suicide deaths and 79 attempts. Adjusted hazard ratios (versus lithium users) for suicide attempts were 2.7 for divalproex users (p < 0.001), 1.6 for gabapentin users (not significant) and 2.8 for carbamazepine users (not significant). For suicide deaths, the adjusted hazard ratios were 1.5 for divalproex users (not significant), 2.6 for gabapentin users (p < 0.001), and not available for carbamazepine users.

Limitations

It should be noted that subjects were not assigned at random to medication use, data on prior suicide attempts were not available, medication use was measured by automated pharmacy records, and duration of mood stabilizer utilization may have been brief.

Conclusions

Lithium may have a protective effect with regard to suicide attempts among Medicaid patients with bipolar disorder. It remains unclear whether or not lithium protects these patients against completed suicide.

Introduction

Completed suicide and attempted suicide are major concerns for people with bipolar disorder (Baldessarini et al., 2006, Baldessarini and Tondo, 2003, Goodwin, 1999, Müller-Oerlinghausen et al., 2002. In the absence of treatment, approximately one person per hundred individuals with bipolar disorder completes suicide annually and about four per hundred attempt suicide (Baldessarini et al., 2006). These risks are approximately one hundred-fold higher (completed suicide) and ten-fold greater (suicide attempts) than those for the general population (Baldessarini et al., 2006).

Although the data are not unequivocal (Coryell et al., 2001; Gelenberg, 2001) there is evidence that lithium may markedly reduce the risk of completed suicide and suicide attempts among individuals with affective conditions such as bipolar disorder (Baldessarini et al., 2006, Baldessarini et al., 2003, Baldessarini and Tondo, 2003, Müller-Oerlinghausen, 2001, Schou, 1999, Tondo et al., 2001). Summarizing numerous studies, Baldessarini et al. (2006) suggested that lithium treatment reduces the risk of completed suicide by roughly nine-fold (to around one death per thousand persons per year) and lowers the risk of suicide attempt about four-fold (to about one attempt per hundred persons per year) among people with affective conditions (chiefly bipolar disorder).

Over the last several years in the United States anticonvulsant medications have overtaken lithium for treatment of people with bipolar disorder (Blanco et al., 2002, Fenn et al., 1996, Goodwin, 1999, Goodwin et al., 2003) with divalproex being especially popular. In a large study of some twenty thousand health maintenance organization members with bipolar disorder, Goodwin et al. (2003) found that risks of completed suicide and of suicide attempts were lower during treatment with lithium versus treatment with divalproex. For example, the suicide death rates were 0.7 per thousand persons per year during lithium use versus 1.7 per thousand persons per year during divalproex use. Similarly, a randomized trial of 378 individuals with affective disorder followed for over two years suggested that lithium (versus carbamazepine) may have a protective effect with regard to suicide completion and suicide attempts (Thies-Flechtner et al., 1996). On the other hand, in a small study of 140 private practice patients with bipolar disorder followed for an average of about two years, Yerevanian et al. (2003) found no differences in suicide attempts between users of lithium versus patients taking antiepileptic medications (chiefly divalproex). Moreover, there was only one suicide death in the Yerevanian et al. (2003) study (in a patient who was off lithium).

It should be noted that both the Goodwin et al. (2003) and the Yerevanian et al. (2003) studies involved persons with bipolar disorder treated in the private sector. Little is known about relationships between medication use and suicidal behavior among individuals with bipolar disorder served chiefly in the public sector. In particular, there are few (if any) data about Medicaid patients with this condition. Yet the joint state-federal Medicaid program is a key health insurer for poor and/or disabled individuals in the United States who have long-lasting mental conditions such as bipolar disorder. For example, the National Co-morbidity Survey Replication showed that in the United States from 2001 through 2003 some 26% of individuals with lifetime diagnosis of Type I bipolar disorder were Medicaid patients at the time of the interview versus only 9% of the general population.

In addition, the Goodwin et al. (2003) study included individuals who switched medication (for example from lithium to divalproex). Moreover, subjects who switched medication could contribute time under observation to, for example, both lithium and divalproex (Goodwin et al., 2003). It has been pointed out that switching medications complicates interpretation of the data (Yerevanian et al., 2004). Also, it has been suggested that individuals who discontinue lithium (especially if lithium is discontinued rapidly) may be at an especially high risk for suicide (Bowden, 2000). An alternative approach is to restrict the study to the initial (or index) episode of medication treatment during the observation period (Goodwin et al., 2004).

Therefore, the present study examined relationships between suicidal behavior and medication use for Medicaid patients with diagnoses of bipolar disorder. The objective was to compare rates of completed suicide and suicide attempts among those Medicaid patients using lithium, divalproex, and/or other anticonvulsant medications, with emphasis on the initial (or index) episode of medication use during the study period.

Section snippets

Methods

Subjects were identified from Oregon state Medicaid and mental health databases. Inclusion criteria were (a) enrollment in the Oregon Medicaid program, (b) diagnosis of bipolar disorder (either Type I or Type II) between 1998 and 2003, and (c) receipt of at least one dispensing of lithium, divalproex, carbamazepine, gabapentin, or other anticonvulsant medication (e.g., lamotrigine or oxcarbazepine). Subjects were excluded if there was more than one diagnosis of schizophrenia at any time during

Results

As shown in Table 1, subjects were 12,662 Oregon Medicaid patients receiving services for bipolar disorder between 1998 and 2003. As would be expected in a Medicaid population, the patients were two-thirds female. The average age was 39. There were 81 suicide attempts and 12 completed suicides in the study population during the years of the project.

Table 1 also shows that the most frequently used mood stabilizer was divalproex (33% of subjects) with gabapentin the next most used anticonvulsant

Discussion

These data suggest that lithium was associated with reduced risk of attempted suicide among Medicaid patients with bipolar disorder. It remains unclear whether or not lithium is associated with reduced risk of completed suicide. The elevated rate of completed suicide among gabapentin users may well be related to prescription of this medication for people with chronic pain (in addition to bipolar disorder) who could be at very high risk for suicide. Although the present study used diagnosis

Role of funding source

Funding for this study was provided by Abbott Laboratories, which had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Conflict of interest

This work was supported in part by Abbott Laboratories. B.H.M. currently receives research grant funding from Eli Lilly. The authors have no other conflicts of interest.

References (23)

  • Müller-OerlinghausenB. et al.

    Bipolar disorder

    Lancet

    (2002)
  • YerevanianB.I. et al.

    Lithium, anticonvulsants, and suicidal behavior in bipolar disorder

    J. Affect. Disord.

    (2003)
  • BaldessariniR.J. et al.

    Suicide risk and treatments for patients with bipolar disorder

    JAMA

    (2003)
  • BaldessariniR.J. et al.

    Lithium treatment and suicide risk in major affective disorders: update and new findings

    J. Clin. Psychiatry

    (2003)
  • BaldessariniR.J. et al.

    Suicide in bipolar disorder: risks and management

    CNS Spectr.

    (2006)
  • BenderK.J.

    NCDEU report I: bipolar, stress and panic disorders

    Psychiatric Times XV

    (1998)
  • BlancoC. et al.

    Trends in the treatment of bipolar disorder by outpatient psychiatrists

    Am. J. Psychiatry

    (2002)
  • BornC. et al.

    Newer prophylactic agents for bipolar disorder and their influence on suicidality

    Arch. Suicide Res.

    (2005)
  • BowdenC.L.

    The ability of lithium and other mood stabilizers to decrease suicide risk and prevent relapse

    Curr. Psychiatry Rep.

    (2000)
  • BowdenC. et al.

    Pharmacotherapy and risk of suicidal behaviors among patients with bipolar disorder [Letter]

    JAMA

    (2004)
  • BowdenC.L. et al.

    A randomized, placebo-controlled 12-month trial of divalproex and lithium in treatment of outpatients with bipolar I disorder

    Arch. Gen. Psychiatry

    (2000)
  • Cited by (65)

    • Lithium and suicide prevention in mood disorders and in the general population: A systematic review

      2020, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      An 8-years cohort study (Goodwin et al., 2003a), compared the risk of SA and SC during Li treatment with VPA and CBZ in a sample of 20,623 BD patients (7121 taking lithium), founding that the risk of SA or SC was 1.5–3 times higher during periods of treatment with VPA than during periods of treatment with Li (p = 0.03), with a greater risk of SA resulting in hospitalization during periods of treatment with CBZ (HR = 2.9, p < 0.001). A 5-years observational study (Collins and McFarland, 2008) detected that risk of SC was significantly greater (2.6 times) among patients treated with gabapentin (GBP) than in patients treated with Li (p < 0.001) and that risk of SA increased among VPA users (2.7) compared to Li users (HR 2.7, p < 0.001). Furthermore, it was found the lowest rate of SC (0.78 per thousand person-years) in the Li users group.

    • Divalproex and its effect on suicide risk in bipolar disorder: A systematic review and meta-analysis of multinational observational studies

      2019, Journal of Affective Disorders
      Citation Excerpt :

      Ultimately, 6 articles were included into the current meta-analysis (Table 1) (Ahearn et al., 2013; Collins and McFarland, 2008; Gibbons et al., 2009; Goodwin et al., 2003; Toffol et al., 2015; Tsai et al., 2016). Regarding the chosen studies, in brief, 4 studies compared the suicidal risks between treatment with divalproex and no medication treatment (Ahearn et al., 2013; Goodwin et al., 2003; Toffol et al., 2015; Tsai et al., 2016) (participants treated with divalproex = 3268, participants not treated with medication = 2870, total mean age = 41.7, total mean female proportion = 37.1%), and 4 studies compared the suicidal risks between divalproex and carbamazepine (Collins and McFarland, 2008; Gibbons et al., 2009; Goodwin et al., 2003; Tsai et al., 2016) (participants with divalproex = 10,676, participants with carbamazepine = 1381, total mean age = 38.5, total mean female proportion = 60.9%). The modified version of the Downs and Black quality checklist was used to evaluate the quality of the selected studies.

    View all citing articles on Scopus
    View full text