Table 1 Double-blind randomised controlled trials of gabapentin in bipolar disorder (BD).

Mokhber 2008 [39]

Parallel

DSM-IV dysphoric mania

Fixed dose gabapentin 900 mg/day (n = 18), carbamazepine 600 mg/day (n = 13) and lamotrigine 100 mg/day (n = 20)

No

8 weeks

MMPI-2

All treatment groups showed a significant within-group improvement in the mania and depression subscales of the MMPI-2 at 8 weeks. Gabapentin showed a significantly greater improvement on the depression symptoms subscale (50%) compared to lamotrigine (33.5%) and carbamazepine (13.6%). There were no significant between-group differences in improvement on the mania symptom subscale. The authors included only study completers in the efficacy analyses.

Frye 2000 [40]

Cross-over

DSM-IV refractory bipolar disorder (BD-I = 11; BD-II = 14) and unipolar depression (UP = 6)

Flexibly dosed gabapentin (mean dose phase 1 = 3987 mg), lamotrigine (274 mg), and placebo

Yes

6 weeks (per phase)

CGI-BP

Response rates (CGI-BP score of much or very much improved) observed during phase 1 was 50% for lamotrigine, 33% for gabapentin and 18% for placebo. Similar clinical response rates were seen across all three phases of the study. Only participants completing all three phases of the study were included in the analysis.

Pande 2000 [41]

Parallel

DSM-IV BD-I with manic/hypomanic (n = 82), or mixed symptoms (n = 35)

Flexible dose adjunctive gabapentin (n = 58) or placebo (n = 59) alongside adjunctive treatment with valproate (n = 57), lithium (n = 39), or both (n = 21).

Yes

10 weeks

YMRS and HAM-D

The placebo group showed a significantly greater improvement in the total YMRS change score compared to gabapentin. There was no significant between-group difference in the HAM-D change score. Secondary outcomes showed no significant differences in CGIC response rate or HAM-A change scores between gabapentin and placebo groups, demonstrating no evidence in favor of gabapentin as an adjunctive treatment in BD

Vieta 2006 [42]

Parallel

DSM-IV BD-I or BD-II in clinical remission (HAM-D ≤ 8 and YMRS ≤ 4)

Flexible dosed adjunctive gabapentin (n = 13) or placebo (n = 12)

Yes

1 year

CGI-BP

There was a significant difference in favor of gabapentin in baseline-to-endpoint (1 year) change in CGI-BP score (gabapentin -2.1, placebo -0.6) and in use of sleeping medication (PSQI-item 6, gabapentin -1.1 vs. placebo -0.6). There were no significant differences in YMRS, HAM-D, HAM-A, PSQI change scores, or time to recurrence of new mood episodes between gabapentin and placebo groups.