A total of 644 patients with manic episodes who were treated with lithium carbonate at the study hospital between January, 2022 and June, 2022 were retrospectively included using convenience sampling. The study was conducted in accordance with the Declaration of Helsinki and the study protocol was approved by the Ethics Committee of Wuhan Wudong Hospital (approval number: WDYY-LL-2021-12, Date: December 30, 2021).

The inclusion criteria were as follows: (1) patients who met the diagnostic criteria for manic episodes outlined in the Chinese Classification and Diagnostic Criteria for Mental Disorders, 3rd Edition [12], (2) patients with a total Bech–Rafaelsen Manic Rating Scale score $\ge$6 points with no obvious impulsive behavior, and (3) patients aged $\ge$18 years.

The exclusion criteria were as follows: (1) patients who had used psychiatric drugs 2 weeks prior to admission, (2) patients who used long-acting injections, a combination of multiple mood stabilizers or antipsychotics, or antidepressants, (3) patients with acute or chronic nephritis, severe cardiovascular disease, acute infection, organic diseases of the central nervous system, and/or clear renal insufficiency.

Sample size was determined using G*Power 3.1 software (Heinrich-Heine-University Düsseldorf, Düsseldorf, Germany), assuming a medium effect size (f = 0.25), using an $\alpha$-value of 0.05 and a power of 0.80 with four groups for analysis of variance, which resulted in a required sample size of 180. The total sample size of 644 patients in this study far exceeds this requirement.

The Bech-Rafaelsen Mania Rating Scale (BRMS) [3] is a clinical tool used to assess the severity of manic symptoms. It is widely applied in the diagnosis and evaluation of treatment efficacy for bipolar disorder (manic-depressive illness) and other psychiatric disorders. Below is detailed information about this scale:

Scale Structure: The BRMS consists of 11 core items, which assess manic symptoms in various aspects of patients, including mood, speech, flight of ideas, activity level, sleep disturbance, irritability, and increased sexual interest. In addition, the scale collaboration group has added 2 supplementary items (hallucinations and delusions), bringing the total number of items to 13.

Scoring Method: Scoring System: A five-point rating scale is used. Each item is scored based on the severity of the symptom, ranging from 0 (no symptoms) to 4 (extremely severe symptoms). Scoring Basis: The rating is completed by evaluators through interviews with patients, observations, and inquiries to family members or ward staff. Time Frame: The initial rating is usually based on the patient’s performance over the past week, and subsequent ratings are generally conducted within 2 to 6 weeks.

Total Score Calculation: The total score is obtained by summing up the scores of all items, with a range of 0 to 44 points.

Interpretation of Total Score: 0–5 points: No significant manic symptoms. 6–10 points: Mild manic symptoms. 11–14 points: Hypomania. 15–24 points: Moderate mania. 25–44 points: Severe mania. The higher the total score, the more severe the manic symptoms. Changes in the total score before and after treatment can reflect therapeutic efficacy, with a larger difference indicating better therapeutic outcomes.

Assessors: Typically, the assessment is conducted by clinically trained doctors or psychiatric nurses. Professional Requirements: Assessors need to have certain clinical experience in psychiatry and training in scale usage to ensure the accuracy and consistency of scoring. Reliability: Statistical data from the domestic scale collaboration group show that the BRMS has high inter-rater reliability (r = 0.97–0.99), indicating good reliability. Validity: The total score of the BRMS has a strong correlation with the clinical judgment of the severity of mania (r = 0.92), suggesting high validity.

Patients were given oral lithium carbonate (H10900013, Enhua, Xuzhou, Jiangsu, China) at an initial dose of 0.9–1.5 g per day for 1 week that was reduced to a maintenance dose of 0.6–0.9 g per day thereafter. The specific dosage within this range for each patient was determined by their treating psychiatrist based on individual clinical factors. The drug was administered immediately after meals and at dinner (at 17:00). Patients were not allowed any other mood stabilizers, antipsychotics, angiotensin-converting enzyme inhibitors, non-steroidal anti-inflammatory drugs, diuretics, phenytoin, or certain antibiotics during the study period.

All enrolled patients received a common diet provided by the hospital, with a total energy intake of about 10,460 kJ/day. The target energy intake of 10,460 kJ/day was selected based on the findings of a systematic review by McKenzie et al. (2022) [13]. While we aimed for all patients to meet this intake, individual variations in appetite and food consumption occurred.

Serum lithium concentration was measured 7 days after initiation of the maintenance dose. Two mL venous blood was drawn 12 hours after the last dose, to represent the 12-hour standard serum lithium concentration [14]. Serum lithium concentrations were measured using the electrode method (Caretium XI-921ET, Shenzhen, Guangdong, China).

Concentration/dose ratio values were calculated as follows: steady-state lithium plasma concentration/oral dose.

Uric acid, blood glucose (GLU), and triglyceride (TG) levels were measured using an automatic biochemical analyzer (Mindray BS820, Shenzhen, Guangdong, China).

Blood pressure (BP) was measured daily during hospitalization. Patients were considered to have hypertension if they had a pre-existing diagnosis of hypertension or if they consistently showed an elevated BP (systolic BP $\ge$130 mmHg or BP diastolic $\ge$80 mmHg) for at least 3 consecutive days during their hospital stay. Patients were divided into groups based on the following criteria: a systolic BP $\ge$130 mmHg was considered high; a systolic BP 130 mmHg was considered non-high; a diastolic BP $\ge$80 mmHg was considered high; a diastolic BP 80 mmHg was considered non-high; patients with serum UA $\ge$420 µmol/L were considered hyperuricemic; patients with a serum UA 420 µmol/L were considered non-hyperuricemic; patients with a serum GLU $\ge$6.11 mmol/L were considered hyperglycemic; patients with a serum GLU 6.11 mmol/L were considered non-hyperglycemic; a serum TG $\ge$2.3 mmol/L was considered high; and a serum TG 2.3 mmol/L was considered non-high. Blood glucose, TG, and UA levels were measured upon admission and regularly throughout the hospitalization period as part of standard clinical care.

Analysis was completed using SPSS v. 17.0 statistical software (SPSS Inc., Chicago, IL, USA). A normality test was performed using the Kolmogorov–Smirnov method. Measurement data describing normality were expressed as mean $\pm$ standard deviation ($\overline{x}$ $\pm$ s). Independent sample t-tests were used for comparisons between groups. A Pearson’s correlation analysis was used to assess relationships between variables.

In addition to univariate analyses, the team performed multiple linear regressions to assess the relationship between C/D value and the study’s variables of interest (systolic BP, diastolic BP, GLU, TGs, and UA) while controlling for potential confounders, such as age, sex, and body mass index (BMI).

Statistical significance was established at p 0.05.

Of the 644 patients included in this study, 321 were men and 323 were women. The mean age was 30.70 $\pm$ 8.57 years, the mean BMI was 22.37 $\pm$ 3.21 kg/m², and the mean course of disease was 21.46 $\pm$ 7.32 years. The mean Bech-Rafaelsen Mania Rating Scale score at admission was 22.3 $\pm$ 6.5, indicating moderate to severe manic symptoms. Detailed demographic and clinical characteristics are shown in Table 1.

The mean blood lithium C/D value for all patients in this study was 0.832 $\pm$ 0.248 mmol$\cdot$L^-1$\cdot$g^-1$\cdot$d. Comparisons of serum lithium concentrations and C/D values between different groups are shown in Table 2. Notably, C/D values in the high systolic BP group (0.868 $\pm$ 0.265) were significantly lower than those in the non-high systolic BP group (0.954 $\pm$ 0.337; t = –2.845, p = 0.007). Concentration/dose ratio values in the hyperuricemic group (0.923 $\pm$ 0.279) were also significantly lower than those in the non-hyperuricemic group (0.939 $\pm$ 0.350; t = 0.615, p = 0.000).

Correlation analysis results between different physiological indicators, serum lithium concentrations, and C/D values are presented in Table 3. Systolic BPs were negatively correlated with C/D values (r = –0.232; p = 0.001), as were UA levels (r = –0.114; p = 0.013). No significant correlations were found for any other indicators.

To control for potential confounding factors, we performed a multiple linear regression analysis with C/D value as the dependent variable and systolic BP; diastolic BP; GLU, TG, and UA levels; age; sex; and BMI as independent variables. The results confirmed that systolic BP ($\beta$ = –0.215; p = 0.003) and UA ($\beta$ = –0.109; p = 0.018) were independently associated with C/D value, even after adjusting for other factors. $\beta$ represents the standardized regression coefficient.

Throughout the study, patients were monitored for signs of lithium toxicity. While some patients experienced mild side effects, such as nausea or tremors, no cases of severe lithium toxicity were observed.

This study has several limitations that should be considered. First, as an observational study, we cannot establish causal relationships between the observed physiological factors and lithium C/D value. Second, while we controlled for several potential confounding factors, there may be other unmeasured variables influencing our results. Third, our study was conducted at a single center, which may limit its generalizability to other populations.

Future research should focus on elucidating the mechanisms underlying the observed relationships between systolic BP, UA level, and lithium C/D value. Prospective studies examining how changes in these physiological factors over time affect lithium pharmacokinetics would be particularly valuable. Additionally, clinical trials investigating whether personalized dosing strategies based on these factors improve treatment outcomes would be an important next step in translating these findings into clinical practice.