Table 1 Characteristics of included studies
From: Lithium use in childhood and adolescence, peripartum, and old age: an umbrella review
Study | Number of primary studies | Design of primary studies | Number of individuals treated with Li+ | Patients’ description | Specifically focused on lithium | Specific outcomes on Li+ | Meta-analytic findings on Li+ a | Findings on efficacy | Findings on safety | Conclusions |
|---|---|---|---|---|---|---|---|---|---|---|
Liu et al. 2011 | 7 | Open label, double-blind studies | 151 | Children (< 18 years) with BD | No | Efficacy of Li+ for paediatric BD | No | Open-label: response rates: 23%-55%; Double-blind: 1/3 found Li+ to induce long-term stabilisation | – | Limited data available |
Amerio et al. 2018 | 30 | RCTs, observational studies | 1320 | Children and adolescents (< 18 years) with BD treated with Li+ monotherapy or combined with other drugs | Yes | Safety and efficacy of Li+ for paediatric BD | No | Li+ monotherapy: efficacy for acute mania in up to 50% of patients; some evidence of long-term maintenance efficacy Combination therapy: some evidence for Li+ + APs and Li+ + DVP | Li+ generally safe in the short term. Most common AEs: gastrointestinal, polyuria or headache | Li+ reasonably safe and effective in children and adolescents with BD in the short-term |
Duffy et al. 2018 | 4 | RCTs | 176 | Children and adolescents (< 18 years) with BD experiencing a manic or mixed episode | Yes | Efficacy and tolerability of Li+ for paediatric mania | Yes | Li+ > plc and comparable to DVP, but < Risp for treating manic/mixed episodes and comorbid ADHD | Li+ not associated with serious AEs, and generally well tolerated with common AEs similar to those reported in adults | Limited data to suggest that Li+ is effective and tolerable for some forms of paediatric mania |
Pisano et al. 2019 | 19 | Open label, RCTs | 871 | Children and adolescents with BD and MDD | Yes | Efficacy and safety of Li+ for paediatric BD and MDD | No | Li + effective for manic symptoms, both in the acute phase and as maintenance strategy. Efficacy on depressive symptoms less clear | Generally, Li + resulted relatively safe | Li + is effective and well-tolerated in paediatric BD. Further evidences are needed for other clinical indications |
Yee et al. 2019 | 5 | RCTs, open-label | 143 | Youths (< 18 years) with BD treated for at least 6 months | No | Efficacy and safety of Li+ for the maintenance of juvenile BD lasting ≥ 24 weeks | Yes | Clinical response rate: 51.1% [CI = 0 to 164] | Mean AE risk: 23.9% [CI = 18.1 to 29.7] | Li+ may reduce long-term morbidity in juvenile BD. Limited data |
Peripartum and lactation | ||||||||||
Doucet et al. 2011 | 10 | Open, retrospective, prospective, studies, case reports | 73 | Women with PP mood episodes with psychotic features exposed to Li+ | No | Effectiveness of Li+ for the prevention or treatment of PP psychosis | No | 5 studies supported the prophylactic effect of Li+, 3 studies supported the use of lithium in treating PP | – | Preliminary evidence suggests Li+ to be effective for PP psychosis prevention and treatment |
Galbally et al. 2010 | 12 | Case–control, prospective studies, case series | 773 | Women exposed to Li+ during pregnancy | No | Pregnancy AEs and child developmental outcomes (safety) | No | – | Ebstein’s anomaly, prematurity and ↑ birth weight reported. Negative results in cohort-controlled studies | Limited data available |
Uguz et al. 2016 | 5 | Case series, case reports | 26 | Women with BD treated with Li+ during lactation | No | Infant AEs (safety) | No | – | Few clinically significant AEs reported | The incidence of AEs in infants exposed to Li+ is very low |
Pacchiarotti et al. 2016 | 6 | Case series, case reports | 35 | Women with BD treated with Li+ during lactation | No | Infant AEs (safety) | No | – | No clinically significant AEs reported | Available data supports the use of Li+ as during breastfeeding |
Haskey et al. 2017 | 2 | Retrospective cohort studies | 18 | Women exposed to Li+ during pregnancy | No | Child developmental outcomes (safety) | No | – | No AEs on cognitive development | Data on Li+ exposure are reassuring but are both limited and low quality |
Poels et al. 2018 | 9 | Cohort studies, case reports | 107 | Women exposed to Li+ during pregnancy | Yes | Neurodevelopmental consequences for the child (safety) | No | – | Clinical cohort studies: Li+ associated with normal development. Case reports: Most (4/5) reported at least one AE | Interpretation is difficult due to study heterogeneity |
Imaz et al. 2019 | 13 | Case reports, case series | 39 | Women treated with Li+ during lactation | Yes | AEs or developmental outcomes in infants (safety) | No | – | 80% of breasted infants showed no AE 20.% showed at least one AE | Heterogeneity and low-moderate quality of studies |
Newmark et al. 2019 | 12 | Case reports | 37 | Women treated with Li+ during lactation | Yes | Infant AEs (safety) | No | – | AEs were reported in 9.4% of breastfed infants | Limited data |
Fornaro et al. 2020 | 29, 13 for meta-analysis | Case–control, cohort, and interventional | 2,622 | Women with BD treated with Li+ during pregnancy and the PP compared to unexposed control subjects (both women in the general population and patients with affective disorders not exposed to Li+) | Yes | Relapse prevention (efficacy); risk of any malformation (safety) | Yes | Li+ > no Li+ in relapse prevention (OR = 0.16, 95%CI = 0.03–0.89) | Li+ associated with ↑ risk of any congenital anomaly (OR = 1.81, 95%CI = 1.35 to 2.4); cardiac anomalies (OR = 1.86, 95% CI = 1.16 to 2.96); spontaneous abortion (OR = 3.77, 95% CI = 1.15 to 12.39). Compared only with unexposed mood disorder patients, significant results only for spontaneous abortion and cardiac anomalies (in 1st trimester). No association with preterm birth and low birth weight | Li+ exposure-associated risk at any time during pregnancy is low; ↑ risk for 1st-trimester or higher-dosage exposure |
Uguz 2020 | 9 | Any design, including case reports and case series | 142 | Women with BD exposed to Li+ during pregnancy and the PP | No | Relapse prevention (efficacy) | No | BD recurrence rates: Pregnancy: 22.7% PP: 20.3% | – | Li+ effective in preventing new mood episodes in BD during the perinatal period |
Advanced age | ||||||||||
Ross 2008 | 3 | Prospective, RCTs | 38 using Li+ followed by discontinuation | Patients > 65 years with MDD treated with Li+ augmentation to AD, followed by Li+ discontinuation | Yes | Relapse prevention (efficacy) | No | Recurrence rates after discontinuation: 50% relapse over ∼6 months follow-up | – | Risk of relapse in elderly patients whose Li+ augmentation treatment for MDD is discontinued. Limited data |
Cooper et al. 2011 | 5 | Open-label, RCTs | 64 | Patients > 55 years with treatment resistant MDD treated with Li+ augmentation | No | Response to treatment (efficacy) | Yes | The overall response rate for Li+ augmentation was 42% (95% CI = 21–65) | – | Replicated evidence for Li+ augmentation as effective in treatment resistant MDD |
Rej et al. 2012 | 10 | RCTs, case–control studies, retrospective, cross-sectional, descriptive | 835 | Patients > 65 years using Li+ | Yes | Renal AEs (safety) | No | – | ARF incidence: 1.5% per person-year, CRF prevalence: 1.2% to 34%; The prevalence of NDI:1.8% to 85% | No evidence to suggest that Li+ should be avoided in elderly patients |
De Fazio et (2017) | 15 | Retrospective, prospective, RCTs | 701 | Patients > 50 years with BD | Yes | Efficacy and safety of Li+ in the treatment and prevention of the mania | No | In the 2 RCTs: Li+ > plc and ≥ other mood stabilizers; Recent retrospective/ prospective studies: average 72% positive response to Li+ | Less AEs at lower doses | Evidence suggests that Li+ is effective and well tolerated; limited evidence |
Sun et al. 2018 | 37 | Case reports | 38 | Patients > 65 years using Li+ with BD or MDD | Yes | Li+ toxicity in elderly (safety) | No | – | Most common AEs: neurotoxicity, renal and cardiovascular toxicity. Precipitating factors: polypharmacy, comorbidity, high Li+ concentration | Lower doses of Li+ should be used in the elderly |