Functional dyspepsia (FD) is a common, chronic syndrome characterized by repetitive and bothersome gastrointestinal symptoms without an organic disease,¹ and categorized into 2 subtypes, postprandial distress syndrome (PDS) and epigastric pain syndrome (EPS).^1-3 The prevalence of FD varies worldwide,^4,5 and the prevalence of FD was 10.3% in a recent nationwide study in South Korea following the Rome III criteria.⁶ The study showed a prevalence of 7.3% for the PDS subtype and 5.5% for the EPS subtype; 25% of patients were categorized as having an overlap syndrome.⁶

Several hypotheses have been proposed as the pathophysiological mechanisms of FD, including gastroduodenal dysmotility, acid exposure, visceral hypersensitivity, autonomic/central nervous system dysfunction and Helicobacter pylori infection, but they seems to be multifactorial and have not been fully understood.^7,8 In addition, functional gastrointestinal diseases are known to differ according to sex and gender, that the incidence, manifestation and suggested mechanism of the disease are different between males and females.⁹ FD and irritable bowel syndrome are more common in females worldwide,¹⁰ and females are more likely to have severe symptoms and coexistent anxiety or depression, with higher risk of delayed gastric emptying and constipation. These differences are partially explained by the role of sex hormones such as estrogen, increased visceral sensation and lower hyperthalamic-pituitary-adrenal activation against stress in females.⁹ In this perspective, recent Korean guideline for FD¹¹ and review article¹² suggested that pharmacological therapy including prokinetics for the PDS subtype and acid-suppressive drugs for the EPS subtype in FD patients. However, there are patients who do not respond to current treatment due to various pathophysiology of the disease.⁴ Therefore, natural product-derived drugs targeting the underlying gastric and duodenal inflammation have been investigated.¹³ In addition, as FD symptoms are sometimes refractory natural product-derived drugs could be more attractive because the efficacy and the safety of traditional natural product are nearly proven and it could be used for mild FD.

Dolichos lablab Linne (D. lablab L.), also known as white bean (hyacinth bean, ‘백편두’), is a widely distributed vine annual plant of the family Fabaceae, that has been cultivated as a foodstuff in Asian countries.¹⁴ It has been traditionally used for its antispasmodic, digestive, and stomach-stimulating properties.^15,16 In a previous study conducted on rats, the extract of D. lablab L. was found to be effective in improving stress-induced gastric mucosal injury by reducing inflammation, protecting mucins, and balancing regeneration, proliferation, and growth factors.¹⁷ Based on these results, we hypothesized that D. lablab could have potential therapeutic effects against FD products with multiple modes of action, of which effects could be different depending on sex or FD subtype. Therefore, the aim of this study is to assess the efficacy and safety of NOVAponin (250 mg of D. lablab L. extract powder per tablet; NOVAWells Co, Ltd, Cheongju, Korea) according to sex and subtypes in mild FD.

Study Design

One hundred and forty-eight subjects aged over 19 years with FD who do not need continuous medical treatment were recruited between June 2020 and December 2022. FD was defined in accordance with the Rome IV criteria as one or more symptoms of bothersome postprandial fullness, early satiation, epigastric pain, or burning sensation for at least 3 months. The symptom onset at least 6 months prior to diagnosis, and there should be no evidence of structural disease. Baseline screening tests including esophagogastroduodenoscopy were conducted 2 weeks before the first scheduled administration. Eligibility was based on the inclusion and exclusion criteria, and those without active gastric lesions were enrolled. In addition, patients with mild symptoms were enrolled in the study, that patients requiring medical treatment such as acid suppressants or prokinetics for FD because of severe symptoms interferes with daily life were excluded. Exclusion criteria in detail are presented in Supplementary Table 1.

The registered subjects were randomly administered an oral dose of NOVAponin or placebo for 12 weeks. The efficacy and safety of the intervention were assessed through visits at 6 weeks and 12 weeks after administration. The random assignment of participants was performed using the permuted block randomized method, and the trial was designed as double-blinded to minimize bias. During the trial, use of medications that could affect the trial were prohibited, and the individual history of drug dosing was checked through questionnaires at each visit. To maintain double-blind nature of the trial, the details of the information on randomization for each group were sealed and kept undisclosed by the administrator. The code was only revealed for 1 patient due to the occurrence of pregnancy during the test period. The study protocol was approved by the Seoul National University Bundang Hospital Institutional Review Board (IRB No. B-2003-598-003). Additionally, the trial was registered with ClinicalTrials.gov (NCT04482478) and the Korean Clinical Research Information Service (KCT0005229). Written informed consent was obtained from all the participants.

Calculation of Number of Target Subjects

The results of a previous study that used the gastrointestinal symptom rating scale (GSRS) as the primary efficacy variable were referred. The average difference between the 2 groups was estimated to be 3.8, and the largest standard deviation (6.6) among the suggested GSRS upper abdominal symptom scores were used for calculating the required number of subjects.¹⁸ As a result, the number of subjects was decided to be 65 in each group, assuming Type I error of 5%, Type II error of 20%, and an elimination rate of 26%.

Furthermore, in the results of the oral administration of D. lablab L. extracts to rats, significant effects were observed at doses of 25 mg/kg/day and 50 mg/kg/day.¹⁷ Using a conversion coefficient of 0.1620, a dose of 50 mg/kg/day in rats was converted to 8 mg/kg/day in humans. Therefore, the daily intake of D. lablab L. extracts was set to 500 mg/day (715 mg/day as NOVAponin), by considering an adult weight of 60 kg.

Efficacy Measurement

After randomization, each subject was advised to take 2 tablets of NOVAponin (including 500 mg D. lablab L. extract powder) or a placebo once daily for 12 weeks. At baseline, symptoms were evaluated using the GSRS and functional dyspepsia-related quality of life (FD-QoL) questionnaires. Six weeks after the administration began, vital signs and compliance were checked, and symptoms were evaluated using the GSRS and gastrointestinal symptom (GIS) questionnaires. Twelve weeks after the administration began, physical examination and laboratory tests were conducted. The symptoms were evaluated using the GSRS, FD-QoL, and GIS questionnaires.

The primary efficacy measure in this study was based on the improvement in GSRS upper abdominal symptom scores after 12 weeks of administration compared to baseline. The GSRS is a symptom-specific instrument comprising 15 items combined into 5 symptom clusters depicting reflux, abdominal pain, indigestion, diarrhea, and constipation, which has a 4-point graded Likert-type scale.¹⁹ The second efficacy measure involved evaluating the improvement of GSRS upper abdominal symptom scores after 6 weeks of administration compared to baseline, the improvement of total GSRS scores, GIS scores, FD-QoL scores, and laboratory tests including erythrocyte sedimentation rate, C-reactive protein, IFN-γ, TNF-α, 8-hydroxy-2-deoxyguanosine (8-OHdG), and total antioxidation status (TAS) after 12 weeks of administration compared to baseline. The GIS questionnaire²⁰ and the FD-QoL questionnaire²¹ are validated tools used to assess the degree of dyspepsia, consist of 10 questions and 21 questions with 5-point graded Likert-type scale, respectively. We referred to several previous studies on drugs used for functional gastrointestinal diseases, which measured inflammation-related biomarker TNF-α,²² and antioxidation-related biomarkers 8-OHdG and TAS^23,24 for the purpose of confirming the effectiveness and discovering the mechanisms.

For each measure, subgroup analyses was performed according to the subtypes of FD and sex. The EPS score was calculated by summing the scores of upper abdominal pain, heartburn, acid reflux, and hunger pain in GSRS, and upper abdominal pain, abdominal cramps, retrosternal discomfort, and acid reflux/regurgitation in GIS. The EPS score was calculated by summing the scores of nausea, rumbling, bloating, and burping in GSRS, and feeling of fullness, early satiety, loss of appetite, sickness, nausea, and vomiting in GIS.

Measurement of Biomarkers

Laboratory tests were conducted at screening and 12 weeks after administration. Blood samples were obtained from the subjects in a fasting state (for at least 8 hours). After collection, the blood samples were analyzed by an outsourced laboratory, and discarded in accordance with laboratory regulations.

Safety Issues

All treatment-emergent adverse events reported during the trial were carefully documented and coded according to MedDRA (version 3.0). Any abnormal reactions observed after the ingestion of the trial food were charted and evaluated.

Statistical Methods

The data obtained from the participants in this trial were analyzed in 3 main forms: safety, full analysis, and per protocol (PP). For evaluation variable, intragroup comparison of changes was analyzed using the paired t test. The comparison between the 2 groups was analyzed using the two-sample t test or Wilcoxon rank-sum test, depending on the normality of the data.

Study Participation and Demographics

One hundred and thirty-one subjects were randomly assigned, excluding 17 subjects who failed the screening process, finally 131 subjects were randomized (Fig. 1). Among the assigned subjects, 61 subjects in the test group and 51 subjects in placebo group were included in the PP set (Fig. 1). There were 42 EPS patients, 49 PDS patients, and 21 patients showed overlapping. In specific, there were 21 EPS patients, 24 PDS patients, and 16 showed overlapping in the test group, and there were 21 EPS patients, 25 PDS patients, and 5 showed overlapping in the control group (Supplementary Table 2). Nineteen subjects were excluded from the PP set due to consent withdrawal, adverse reactions, administration of prohibited concomitant medication, pregnancy, poor compliance, and violation of the selection/exclusion criteria. There were no significant differences between the groups except for sex, that a significant number of females were assigned to the test group compared to the control group (P = 0.006). One subject with a history of insomnia in the test group and another one with a history of anxiety disorder in the control group were identified. However, both were not undergoing medical treatment at the time of enrollment (Table 1).

Figure 1. Flow chart showing enrolled patients and dropouts from the study. Adverse reaction possibly related to study drug: diarrhea (n = 1). PP, per protocol; FA, full analysis.

Table 1 . Baseline Characteristics of Study Participants

Characteristics	NOVAponin(n = 61)	Placebo(n = 51) (%)	P-value
Sex
Male	8 (13.1)	18 (35.3)	0.006
Female	53 (86.9)	33 (64.7)
Age (yr)	52.28 ± 13.34	50.69 ± 13.52	0.533
Exercise
No	24 (39.3)	15 (29.4)	0.563
1-2 times/wk	17 (27.9)	14 (27.5)
3-4 times/wk	10 (16.4)	15 (29.4)
5-6 times/wk	4 (6.6)	3 (5.9)
Everyday	6 (9.8)	4 (7.8)
Smoking
Non-smoker	61 (100.0)	49 (96.0)	0.205
Ex-smoker	0 (0.0)	1 (2.0)
Smoker	0 (0.0)	1 (2.0)
Degree of stress
Nothing at all	5 (8.2)	5 (9.8)	0.633
A little bit	44 (72.1)	32 (62.8)
A lot of	10 (16.4)	10 (19.6)
Ton	2 (3.3)	4 (7.8)
Beverages containing Caffeine (U/day)	1.66 ± 1.00	1.39 ± 1.11	0.125
Chocolate (U/day)	0.10 ± 0.35	0.12 ± 0.38	0.987
Regularity of eating
Regular	40 (65.6)	34 (66.7)	0.903
Not regular	21 (34.4)	17 (33.3)
Time taken to eat one meal
Within 10 min	9 (14.8)	8 (15.7)	0.075
10 to 20 min	34 (55.7)	38 (74.5)
20 to 30 min	12 (19.7)	4 (7.8)
More than 30 min	6 (9.8)	1 (2.0)
Frequency of overeating
Less than 3 times/wk	49 (80.3)	42 (82.4)	0.785
3 or more times/wk	12 (19.7)	9 (17.6)
Alcohol drinking
No (never drink)	18 (29.5)	15 (29.4)	0.912
No (no drink)	9 (14.8)	9 (17.6)
Yes	34 (55.7)	27 (53.0)
A period of abstinence (mo)	30.67 ± 65.09	27.22 ± 25.38	0.339
Average alcohol consumption (U)	2.36 ± 2.05	2.44 ± 1.97	0.699

A P-value < 0.05 was considered statistically significant.

Data are presented as n (%) or mean ± SD.

Validity Evaluation

Changes of gastrointestinal symptom rating scale scores

GSRS upper abdominal symptom scores at 12 weeks after administration showed significant improvement in both test groups (–5.30 ± 0.60, P < 0.001) and the control group (–2.35 ± 0.56, P < 0.001) (Fig. 2), a more pronounced change was observed in the test group compared to the control group (P < 0.001). Specifically, statistically significant differences between the 2 groups were observed in the improvement of upper abdominal pain (–0.70 ± 0.12 vs –0.22 ± 0.13, P = 0.005), heartburn (–0.56 ± 0.11 vs –0.20 ± 0.11, P = 0.018), bloating (–0.84 ± 0.14 vs –0.14 ± 0.15, P < 0.001), and burping (–0.61 ± 0.12 vs –0.20 ± 0.14, P = 0.033) (Table 2). In addition, at 6 weeks after administration, GSRS upper abdominal symptom scores showed significant improvement in both test groups (–5.13 ± 0.55, P < 0.001) and the control group (–1.92 ± 0.44, P < 0.001), but a more pronounced change was observed in the test group compared to the control group (P < 0.001) (Fig. 2). Specifically, statistically significant differences between the groups were observed in the improvement of upper abdominal pain (–0.79 ± 0.10 vs –0.14 ± 0.14, P < 0.001), heartburn –0.52 ± 0.11 vs –0.10 ± 0.12, P = 0.003), hunger pain (–0.36 ± 0.13 vs –0.04 ± 0.10, P = 0.021), rumbling (–0.64 ± 0.12 vs –0.25 ± 0.14, P = 0.025), bloating (–0.77 ± 0.12 vs –0.16 ± 0.13, P = 0.001), and burping (–0.66 ± 0.12 vs –0.25 ± 0.11, P = 0.019) (Table 2).

Figure 2. Gastrointestinal symptom rating scale (GSRS) upper abdominal symptom scores changes from baseline after 6 weeks and 12 weeks of administration.

Table 2 . Gastrointestinal Symptom Rating Scale Upper Abdominal Symptom Score Changes From Baseline After 6 Weeks and 12 Weeks of Administration

Variable	Time	NOVAponin	Placebo	P-value
Total Score	6 wk-baseline	–5.13 ± 0.55	–1.92 ± 0.44	< 0.001a
	12 wk-baseline	–5.30 ± 0.60	–2.35 ± 0.56	< 0.001a
Upper abdominal pain	6 wk-baseline	–0.79 ± 0.10	–0.14 ± 0.14	< 0.001a
	12 wk-baseline	–0.70 ± 0.12	–0.22 ± 0.13	0.005b
Heartburn	6 wk-baseline	–0.52 ± 0.11	–0.10 ± 0.12	0.003b
	12 wk-baseline	–0.56 ± 0.11	–0.20 ± 0.11	0.018c
Acid reflux	6 wk-baseline	–0.93 ± 0.12	–0.63 ± 0.14	0.214
	12 wk-baseline	–0.93 ± 0.13	–0.67 ± 0.14	0.185
Hunger pain	6 wk-baseline	–0.36 ± 0.13	–0.04 ± 0.10	0.021c
	12 wk-baseline	–0.52 ± 0.11	–0.25 ± 0.09	0.104
Nausea	6 wk-baseline	–0.46 ± 0.09	–0.35 ± 0.11	0.496
	12 wk-baseline	–0.48 ± 0.10	–0.37 ± 0.10	0.485
Rumbling	6 wk-baseline	–0.64 ± 0.12	–0.25 ± 0.14	0.025c
	12 wk-baseline	–0.66 ± 0.12	–0.31 ± 0.15	0.066
Bloating	6 wk-baseline	–0.77 ± 0.12	–0.16 ± 0.13	0.001b
	12 wk-baseline	–0.84 ± 0.14	–0.14 ± 0.15	< 0.001a
Burping	6 wk-baseline	–0.66 ± 0.12	–0.25 ± 0.11	0.019c
	12 wk-baseline	–0.61 ± 0.12	–0.20 ± 0.14	0.033c

^aP < 0.001, ^bP < 0.01, ^cP < 0.05; compared between groups: P-value for 2 sample t test or Wilcoxon rank sum test.

Data are presented as mean ± SE.

At 6 and 12 weeks after administration, GSRS total scores showed significant improvement in both the test group (6 weeks-baseline 7.02 ± 0.91, P < 0.001; 12 weeks-baseline –7.54 ± 0.94, P < 0.001) and the control group (6 weeks-baseline –3.33 ± 0.73, P < 0.001; 12 weeks-baseline –3.31 ± 0.85, P < 0.001), but more pronounced changes were observed in the test group compared to the control group (6 weeks-baseline P = 0.002; 12 weeks-baseline P < 0.001) (Fig. 3A). At 6 weeks and 12 weeks after administration, GSRS lower abdominal symptom scores showed significant improvement in both the test group (6 weeks-baseline –1.89 ± 0.46, P = 0.001; 12 weeks-baseline –2.25 ± 0.46, P < 0.001) and the control group (6 weeks-baseline –1.41 ± 0.41, P = 0.001; 12 weeks-baseline –0.96 ± 0.40, P = 0.022). A statistically significant difference between the groups was observed only at 12 weeks after administration (6 weeks-baseline P = 0.484; 12 weeks-baseline P = 0.030) (Supplementary Fig. 1 and Supplementary Table 3).

Figure 3. Gastrointestinal symptom rating scale (GSRS) upper total scores (A), gastrointestinal symptom (GIS) scores (B), and functional dyspepsia-related quality of life (FD-QoL) score (C) changes from baseline after 6 weeks and 12 weeks of administration.

Changes of gastrointestinal symptom scores

GIS scores at 6 weeks and 12 weeks after administration showed significant improvement in both the test group (6 weeks-baseline –11.21 ± 0.53, P < 0.001; 12 weeks-baseline –11.90 ± 0.52, P < 0.001) and the control group (6 weeks-baseline –6.65 ± 0.70, P < 0.001; 12 weeks-baseline –7.61 ± 0.73, P < 0.001), more pronounced changes were observed in the test group compared to the control group (6 weeks-baseline P < 0.001; 12 weeks-baseline P < 0.001) (Fig. 3B). Specifically, statistically significant differences between the groups were observed in the improvement of upper abdominal pain (6 weeks-baseline –1.66 ± 0.15 vs –0.88 ± 0.16, P = 0.001; 12 weeks-baseline –1.61 ± 0.16 vs –0.98 ± 0.17, P = 0.009), feeling of fullness (6 weeks-baseline –1.38 ± 0.14 vs –0.88 ± 0.20, P = 0.026; 12 weeks-baseline –1.70 ± 0.15 vs –1.04 ± 0.19, P = 0.012), early satiety at 6 weeks after administration (–1.38 ± 0.17 vs –0.92 ± 0.17, P = 0.043), sickness at 6 weeks after administration (–1.13 ± 0.18 vs –0.41 ± 0.19, P = 0.015), nausea (6 weeks-baseline –1.26 ± 0.14 vs –0.76 ± 0.16, P = 0.032; 12 weeks-baseline –1.25 ± 0.15 vs –0.76 ± 0.16, P = 0.024) and acid reflux/indigestion (6 weeks-baseline –2.10 ± 0.13 vs –1.29 ± 0.17, P < 0.001; 12 weeks-baseline –2.20 ± 0.14 vs –1.49 ± 0.17, P = 0.005) (Table 3).

Table 3 . Gastrointestinal Symptom Score Changes From Baseline After 6 Weeks and 12 Weeks of Administration

Variable	Time	NOVAponin	Placebo	P-value
Total Score	6 wk-baseline	–11.21 ± 0.53	–6.65 ± 0.70	< 0.001a
	12 wk-baseline	–11.90 ± 0.52	–7.61 ± 0.73	< 0.001a
Upper abdominal pain	6 wk-baseline	–1.66 ± 0.15	–0.88 ± 0.16	0.001b
	12 wk-baseline	–1.61 ± 0.16	–0.98 ± 0.17	0.009b
Abdominal cramps	6 wk-baseline	–0.44 ± 0.12	–0.31 ± 0.12	0.450
	12 wk-baseline	–0.39 ± 0.13	–0.24 ± 0.14	0.386
Feeling of fullness	6 wk-baseline	–1.38 ± 0.14	–0.88 ± 0.20	0.026c
	12 wk-baseline	–1.70 ± 0.15	–1.04 ± 0.19	0.012c
Early satiety	6 wk-baseline	–1.38 ± 0.17	–0.92 ± 0.17	0.043c
	12 wk-baseline	–1.51 ± 0.16	–1.08 ± 0.16	0.051
Loss of appetite	6 wk-baseline	–0.77 ± 0.15	–0.63 ± 0.19	0.408
	12 wk-baseline	–0.79 ± 0.15	–0.71 ± 0.19	0.574
Sickness	6 wk-baseline	–1.13 ± 0.18	–0.41 ± 0.19	0.015c
	12 wk-baseline	–1.26 ± 0.19	–0.76 ± 0.18	0.089
Nausea	6 wk-baseline	–1.26 ± 0.14	–0.76 ± 0.16	0.032c
	12 wk-baseline	–1.25 ± 0.15	–0.76 ± 0.16	0.024c
Vomiting	6 wk-baseline	–0.41 ± 0.09	–0.27 ± 0.09	0.261
	12 wk-baseline	–0.46 ± 0.09	–0.24 ± 0.10	0.076
Retrosternal discomfort	6 wk-baseline	–0.69 ± 0.15	–0.27 ± 0.14	0.079
	12 wk-baseline	–0.74 ± 0.14	–0.31 ± 0.13	0.062
Acid reflux/indigestion	6 wk-baseline	–2.10 ± 0.13	–1.29±0.17	< 0.001a

^aP < 0.001, ^bP < 0.01, ^cP < 0.05; compared between groups: P-value for 2 sample t test or Wilcoxon rank sum test.

Data are presented as mean ± SE.

Changes of functional dyspepsia-related quality of life scores

At 12 weeks after administration, FD-QoL scores showed significant improvement in both the test group (–11.41 ± 1.75, P < 0.001) and the control group (–5.55 ± 1.20, P < 0.001), more pronounced changes were observed in the test group compared to the control group (P = 0.007) (Fig. 3C). Specifically, statistically significant differences between the groups were observed in the improvement of eating (–3.13 ± 0.49 vs –1.59 ± 0.42, P = 0.025), liveliness (–3.82 ± 0.60 vs –1.59 ± 0.40, P = 0.004). and role-functioning status (–2.03 ± 0.46 vs –0.86 ± 0.45, P = 0.020) (Table 4).

Table 4 . Functional Dyspepsia-related Quality of Life Score Changes From Baseline After 12 Weeks of Administration

Variable	NOVAponin	Placebo	P-value
Total Score	–11.41 ± 1.75	–5.55 ± 1.20	0.007a
Eating status	–3.13 ± 0.49	–1.59 ± 0.42	0.025b
Liveliness status	–3.82 ± 0.60	–1.59 ± 0.40	0.004a
Psychological status	–2.43 ± 0.60	–1.51 ± 0.49	0.170
Role-functioning status	–2.03 ± 0.46	–0.86 ± 0.45	0.020b

^aP < 0.01, ^bP < 0.05; compared between groups: P-value for 2 sample t test or Wilcoxon rank sum test.

Data are presented as mean ± SE.

Subgroup analysis by subtype of functional dyspepsia, sex, and Helicobacter pylori infection status

In the subgroup analysis for the EPS subtype, GSRS scores at 6 weeks and 12 weeks after administration showed more pronounced changes in the test group compared to the control group (6 weeks-baseline P = 0.009; 12 weeks-baseline P = 0.003). GIS scores at 6 weeks and 12 weeks after administration also showed more pronounced changes in the test group compared to the control group (6 weeks-baseline P = 0.009; 12 weeks-baseline P = 0.003). In the subgroup analysis for the PDS subtype, GSRS scores at 6 weeks and 12 weeks after administration showed more pronounced changes in the test group compared to the control group (6 weeks-baseline P = 0.040; 12 weeks-baseline P = 0.047). GIS scores at 6 weeks and 12 weeks after administration also showed more pronounced changes in the test group compared to the control group (6 weeks-baseline P < 0.001; 12 weeks-baseline P = 0.001). Statistically significant changes were observed in both groups, and the range of change in the EPS group seemed to be slightly larger than in the PDS subtype group (Table 5).

Table 5 . Score Changes According to Each Subtype From Baseline After 12 Weeks of Administration

Score	EPS						PDS
	Sex	Time	NOVAponin	Placebo	P-value		Sex	Time	NOVAponin	Placebo	P-value
GSRS	Total	6 wk-baseline	–2.61 ± 0.30	–0.90 ± 0.28	< 0.001a		Total	6 wk-baseline	–2.52 ± 0.32	–1.02 ± 0.33	0.001b
		12 wk-baseline	–2.72 ± 0.30	–1.33 ± 0.27	< 0.001a			12 wk-baseline	–2.57 ± 0.38	–1.02 ± 0.41	0.006b
	Male	6 wk-baseline	–2.00 ± 0.71	–1.11 ± 0.62	0.404		Male	6 wk-baseline	–1.75 ± 0.77	–1.17 ± 0.54	0.045c
		12 wk-baseline	–2.13 ± 0.64	–1.28 ± 0.43	0.281			12 wk-baseline	–1.75 ± 0.77	–0.61 ± 0.64	0.306
	Female	6 wk-baseline	–2.70 ± 0.33	–0.79 ± 0.27	< 0.001a		Female	6 wk-baseline	–2.64 ± 0.35	–0.94 ± 0.42	0.003b
		12 wk-baseline	–2.81 ± 0.34	–1.36 ± 0.35	0.003b			12 wk-baseline	–2.70 ± 0.42	–1.24 ± 0.53	0.035c
GIS	Total	6 wk-baseline	–4.89 ± 0.33	–2.76 ± 0.34	< 0.001a		Total	6 wk-baseline	–6.33 ± 0.43	–3.88 ± 0.54	< 0.001a
		12 wk-baseline	–4.93 ± 0.35	–3.02 ± 0.37	< 0.001a			12 wk-baseline	–6.97 ± 0.40	–4.59 ± 0.49	< 0.001a
	Male	6 wk-baseline	–3.50 ± 0.71	–2.67 ± 0.51	0.364		Male	6 wk-baseline	–5.88 ± 0.91	–4.61 ± 1.09	0.477
		12 wk-baseline	–4.38 ± 0.56	–2.89 ± 0.54	0.112			12 wk-baseline	–6.88 ± 2.53	–4.44 ± 3.48	0.090
	Female	6 wk-baseline	–5.09 ± 0.36	–2.82 ± 0.44	< 0.001a		Female	6 wk-baseline	–6.40 ± 0.47	–3.48 ± 0.60	< 0.001a
		12 wk-baseline	–5.02 ± 0.39	–3.09 ± 0.49	0.003b			12 wk-baseline	–6.98 ± 0.45	–4.67 ± 0.62	0.003b

EPS, epigastric pain syndrome; PDS, postprandial distress syndrome; GSRS, gastrointestinal symptom rating scale; GIS, gastrointestinal symptom.

^aP < 0.001, ^bP < 0.01, ^cP < 0.05; compared between groups: P-value for 2 sample t test or Wilcoxon rank sum test.

Data are presented as mean ± SE.

We performed an additional subgroup analysis according to sex, and the difference between the study group and the control group was pronounced in males than in females, that only GIS score at 12 weeks after administration in EPS patients showed significant change in the test group compared to the control group in males (P = 0.019), while all scores except GSRS total score at 12 weeks after administration in PDS patients showed significant changes in the test group compared to the control group in females.

In subgroup analyses according to H. pylori infection status, there was no statistically significant differences in symptom improvement between H. pylori-positive and negative patients (Supplementary Table 4).

Changes of the inflammatory and antioxidative markers

Erythrocyte sedimentation rate, C-reactive protein, IFN-γ and TNF-α, which are inflammatory biomarkers, as well as 8-OHdG and TAS, which are biomarkers for oxidative stress were measured. However, the levels of biomarkers after administration did not show significant changes compared to baseline (Supplementary Fig. 2 and Supplementary Table 5).

Safety

There were 9 adverse reactions in 5 subjects in the test group (5/66, 7.58%), and 5 cases in 4 subjects in the control group (5/65, 6.15%), without significant difference between the groups. The causality assessment between the adverse reactions and the test food showed that the test group had 1 case (diarrhea) as “possibly related,” 3 cases (vertigo, diarrhea, and dyspepsia) as “probably not related,” and 5 cases (hangover, pyrexia, folliculitis, breast pain, and oropharyngeal pain) as “definitely not related.” The control group had 5 cases (abdominal pain, back pain, Coronavirus infections, and urticaria) classified as “definitely not related.” Although all cases were mild in severity, the case classified as “possibly related” was dropped out (Supplementary Table 6). All changes in the analysis of vital signs, body weight, laboratory tests, and electrocardiograms were within the normal ranges and were considered to have no clinical implications. Overall, no significant adverse reactions occurred during the trial.

In this trial, NOVAponin, D. lablab L. extract powder, significantly improved the GSRS upper abdominal symptom scores after 12 weeks of administration in the test group compared with the control group in mild FD. Furthermore, the GSRS upper abdominal symptom scores at 6 weeks after administration, GSRS total scores, GIS total scores, and FD-QoL total scores after administration also showed significant differences between the groups, without any significant adverse reactions. In addition, as a result of subgroup analysis, these differences were more pronounced in EPS subtype compared to PDS subtype, and more pronounced in female patients than in males.

As mentioned, current treatment strategies for FD is not satisfactory in all patients. Recent Korean guideline for FD suggested prokinetics for the PDS subtype and acid-suppressive drugs for the EPS subtype as first-line treatment.¹¹ If the treatment effect is insufficient, prokinetics for the EPS subtype and acid-suppressive drugs for the PDS subtype can be applied as opposite. In addition, other medications including fundic relaxants or tricyclic antidepressants can be considered. But, the overall rate of symptom relief with existing treatments is reported to be only around 50%.²⁵ Thus, it is reasonable to develop drugs derived from natural products that offer multiple modes of action. Biological drugs and functional foods may be superior to medicines that function with only one mechanism for treating diseases with complex mechanisms such as FD,²⁶ for instance, DA-9701, a prokinetic drug based on extracts of Corydalis tuber and Pharbitis seeds antagonizes the dopamine D2 receptor and boosts the 5-hydroxytryptamine 4receptor.²⁷ Rikkunshito²⁸ and Flos Lonicera extract²⁹ also showed improvement of symptoms, reduction of inflammation, and antioxidant effects in patients with FD.

The raw material of NOVAponin, the test food for this human trial, was obtained from white bean seeds, that have traditionally been widely used for many purposes; to stimulate the stomach, as an antidote for poisoning, and for the treatment of diarrhea.^30-32 Additionally, it is rich in physiologically active antioxidant substances, such as phenolic compounds and flavonoids as well as general nutrients,³³ and contains saponin-based substances such as Chikusetsusaponin IVa, Sandosaponin, Saponin D, Soyasaponin Bb, and Lablabosides.¹⁶ Furthermore, D.s lablab L. extract rescued stress-induced gastric mucosal disease in rats, that the levels of inflammatory mediators including IL-1β, TNF-α, and nuclear factor kappa-light-chain-enhancer of activated B cells were all significantly increased by stress, but pretreatment of D. lablab L. extracts significantly decreased the levels of inflammatory mediators, and showed gastric mucin protective effect.¹⁷

Taken together the suggested mechanism of action of D. lablab L. extract is as follows: First, D. lablab L. extract boosts nuclear factor erythroid-related factor-2-mediated mucoprotective, anti-inflammatory, and neuroprotective actions, as one of the main mechanisms of functional dyspepsia is the gastroduodenal inflammations following H. pylori infection and duodenal eosinophilia.^34,35 Additionally, it promotes the cytoprotective, anti-inflammatory, antioxidative, and anti-apoptotic activities of heme oxygenase-1 to heal gastric mucosal damage. As a result, it decreases the offense system (inflammatory mediators, apoptosis, and oxidative stress), increases the defense system (gastric protective mucins, MUC5A, and relieving hypoxia), and balances the regeneration, proliferation, and growth factors.¹⁷ Although our trial included subjects without definite organic diseases, D. lablab L. extract may have suppressed low-grade inflammation and improved epigastric symptoms. Second, a previous study demonstrated the anti-motility and membrane stability effects of D.s lablab L. extracts in a hemolysis of erythrocyte membranes animal model.³⁶ In addition, D. lablab L. extracts significantly suppressed zymosan-induced colonic inflammation in another experiment on mice, that decreased the expression of TNF-α, reduced the number of visceral pain-related behaviors and anxiety-like behaviors in mice.³⁷ The authors concluded that 200 mg/kg D. lablab L. extract was comparable to 30 mg/kg amitriptyline, a tricyclic antidepressant, and 30 mg/kg sulfasalazine, an anti-inflammatory agent.³⁷ As most FD patients have psychological factors, the anti-inflammatory and anti-anxiety effects of D. lablab L. extracts may have improved upper abdominal symptoms by regulating the brain-gut axis. In addition, D. lablab L. extract may act via multiple receptors similar to DA-9701, which improved gastrointestinal transit and lowered plasma adrenocorticotropic hormone levels via the central corticotropin-releasing factor pathway in an ileus animal model.^38,39 However, to date, there have been no research results or evidence, thus further studies on the mechanisms of D.lablab L. extract on the gut-brain axis in FD are needed in the future after proving its therapeutic efficacy for mild FD.

The correlation between medication compliance and drug effectiveness has been established,^40,41 and previous studies on FD have shown that proton pump inhibitors showed a faster effect than prokinetics or psychotropic drugs.^42,43 In our study, the trial period was set to 12 weeks, considering the possibility that the effect of NOVAponin is slower than that of proton pump inhibitor. However, the improvement in the GSRS and GIS scores at the 6-week time point was similar to that at the 12th week, suggesting that short-term treatment with NOVAponin could be similarly effective with long- term treatment. Another point is that we attempted to confirm the anti-inflammatory and antioxidative effects of D. lablab L. based on a previous study in rats;¹⁷ however, the laboratory test results did not confirm a difference between the groups. This may be because the aforementioned experiment was conducted after inducing gastric mucosal damage by stress. However, in the present study only patients with mild-to-moderate FD without organic diseases were targeted in this trial.

The sex difference in this study may be due to the difference in the number of participants according to sex with a smaller number of male participants, and it is not possible to conclude the effect of Novaponin was different according to sex based on the results of this study. Though, there is another possible explanation, the transient receptor potential cation channel subfamily V member 1 (TRPV1). The upregulation of TRPV1 plays an important role in functional GI diseases including IBS,⁴⁴ reflux diseases,^45,46 and FD,⁴⁷ in particular in coexistence with H. pylori infection.⁴⁸ The up-regulation of TRPV1 was observed even in FD patients lacking gastric mucosal inflammation in a recent study,⁴⁹ and recent animal study reported that there are sex differences in pain axis mediated by TRPV1.⁵⁰ Therefore, receptor expression such as TRPV1 may contribute to the sex difference of FD itself and the different responses according to the treatment. Further studies A follow-up study is needed to determine whether Novaponin is more effective in female and EPS subtype patients, and to figure out the mechanism in terms of sex differences, including TRPV1.

During the trial, nine adverse reactions were reported in 5 subjects (5/66, 7.58%) in the test group, while 5 adverse reactions occurred in 4 subjects (5/65, 6.15%) in the control group. All 14 cases were mild in severity, and no significant adverse reaction was occurred during the trial. All changes were within the normal range and were considered to have no clinical implications in the analysis of physical examinations and laboratory tests, that the safety of NOVAponin has been fully confirmed.

This study had several limitations. First, H. pylori status was analyzed in only half of the patients in this study. Although no statistically significant differences have been identified according to H. pylori infection status in our data, a larger study considering H. pylori infection status is needed in the future, based on the fact that H. pylori infection provokes gastrointestinal inflammation and may worsen the symptoms in FD patients.⁵¹ Second, this study was performed in a single institution, with a small number of subjects. We tried to analyze FD patients according to the subtypes and sex to find out the differences, but decisive conclusions were not obtained due to the small number of subjects and the difference in the number of male and female subjects between the 2 groups. Third, in this study, the amount of daily intake was calculated by referring to previous animal study results as mentioned in the methods section. However, the regimen was determined to take 2 tablets once a day in consideration of the compliance and the size of the tablet, and this may have affected the rate of absorption in the body and the degree of effectiveness. Further data on the pharmacokinetic aspect of Novaponin and follow-up study of different dosing intervals would be needed.

In conclusion, the findings of this trial prove evidence for the efficacy and safety of Novaponin in improving gastrointestinal symptoms in mild and moderate FD subjects, especially in EPS subtype and in females. Novaponin may be a good candidate for FD.

This work was supported by the Seoul National University Bundang Hospital Research fund (Grant No. 06-2020-0169). In addition, the clinical study was supported by NOVAWells.

None.

Yonghoon Choi analyzed the data and drafted the article; Nayoung Kim designed the study, collected the data, and supervised and edited the manuscript; and Dong Ho Lee supervised the writing of the manuscript and gave critical advises.

Note: To access the supplementary tables and figures mentioned in this article, visit the online version of Journal of Neurogastroenterology and Motility at http://www.jnmjournal.org/, and at https://doi.org/10.5056/jnm23180.