Emulsifiers in poultry nutrition–molecular mechanism of lipid metabolism and energy utilization: A meta-analysis and KEGG mapping

Abstract

The feed energy derived from fat remains relatively low because young poultry have limited fat digestibility and inefficient lipid metabolism. This study aimed to evaluate the effectiveness of emulsifiers in regulating lipid metabolism and increasing lipase activity in poultry. A quantitative review of 57 studies (2015–2025) revealed that emulsifier supplementation significantly upregulated genes associated with lipogenesis, including diacylglycerol O-acyltransferase 2 (DGAT2), and increased lipolysis through the upregulation of perilipin 1 (PLN1). It also increased the activities of lipoprotein lipase and total lipase. Kyoto Encyclopaedia of Genes and Genomes (KEGG) mapping also suggested that fatty acid biosynthesis (gga00061), β-oxidation (gga00071), the PPAR signalling pathway (gga03320), cholesterol metabolism (hsa04979), and the regulation of lipolysis (gga04923) in adipocyte regulation pathways were synergistically activated. Taken together, these pathways suggest that emulsifiers improve lipid turnover, oxidative metabolism and lipoprotein formation in broiler chickens. In summary, emulsifiers modulate lipid metabolism at the molecular level, which offers mechanistic support for improving energy utilization and may also aid in developing precision nutrition strategies in poultry.

Introduction

In modern broiler chickens, feed energy intake and metabolism are critical factors for improving feed efficiency because both processes must be optimized to support rapid growth and production. Lipids are a dense energy source with a higher energy value than carbohydrates and proteins. Furthermore, since lipids do not significantly increase body heat, they are a primary energy source for livestock in hot climates such as the humid tropics (Kpomasse et al., 2021; Adli et al., 2025). However, poultry utilization of lipids as an energy source, particularly from fats and oils, remains limited. This is especially true for fast-growing broiler birds, which are young and have limited fat digestion capabilities.

Broiler chickens digest and absorb fat less efficiently, which limits their ability to utilize high-fat feeds. This is related to their relatively short digestive tract, limited bile secretion, and low lipase activity, all of which affect fat digestion (Ge et al., 2019; Fernandes et al., 2023; Priyatno et al., 2025). Consequently, the efficiency of converting feed fat into energy is reduced. Decreased energy conversion leads to reduced feed conversion and stunted growth (Marx et al., 2017; Wealleans et al., 2020). While broiler chickens are the primary focus of research, issues with lipid digestion and utilization are not limited to broiler chickens. Similar restrictions have been observed in other poultry species, including turkeys (Nemati et al., 2021), ducks (Zosangpuii et al., 2011; El-Katcha et al., 2021), and laying hens (Attia et al., 2009; Hu et al., 2022; Okasha et al., 2023), in their ability to digest fat and use more energy during high-production or early growth stages. Low bile secretion, inadequate digestive capacity, and variable lipase activity in these species can also limit lipid utilization and metabolic efficiency.

Therefore, the right feed must be formulated by combining it with emulsifier additives. Emulsifiers increase lipid dispersion, reduce surface tension, and accelerate triglyceride (TGL) hydrolysis by lipases (Siyal et al., 2017; Hu et al., 2019). Additionally, emulsifiers modulate molecular responses related to enzyme and gene activities involved in lipid metabolism (Zhang et al., 2022a).

Thus, research on emulsifiers in poultry nutrition is ongoing. However, the results of these studies are inconsistent. For example, Boontiam et al. (2017) reported a significant positive response to emulsifier administration in broiler chickens. In contrast, Arshad et al. (2020) reported that emulsifier administration had no significant effect on broiler growth. Similarly, Kwak et al. (2022b) confirmed that emulsifiers did not significantly increase broiler growth.

Molecular studies reinforce the idea that emulsifiers have the potential to improve fat digestion, lipid metabolism, and poultry productivity in general. However, physiological responses, particularly those related to lipid metabolism, show variable results at the molecular level, including gene expression and enzymatic activity. Several previous studies have confirmed significant increases in the gene expression of diacylglycerol O-acyltransferase 2 (DGAT2), stearoyl-CoA desaturase (SCD), carnitine palmitoyl transferase 1 (CPT1), and peroxisome proliferator-activated receptor alpha (PPARα). Conversely, other studies have reported inconsistent results (Huang et al., 2008; Mao et al., 2022; Li et al., 2023; Hu et al., 2024). Lipase activity also shows similar inconsistencies. Researchers attribute this variability to differences in emulsifier dosage, energy levels in feed, and species-specific responses. It was hypothesized that the activation or repression of specific genes serves as a key indicator of how effectively emulsifiers support lipid metabolism in poultry. Previous studies Ahmadipour et al. (2025) and Zhang et al. (2022a) have suggested that emulsifiers modulate molecular mechanisms regulating gene expression and enzymatic activity, thereby increasing performance.

However, owing to inconsistencies among reported findings, a quantitative synthesis was required to clarify their overall impact. Therefore, the present study performed a meta-analysis to evaluate the effects of emulsifiers on lipid metabolism by examining gene expression and enzymatic activity related to lipogenesis, lipolysis, and fatty acid oxidation. The analysis aimed to identify general trends, quantify effect sizes, and assess how moderator variables influence biological responses. This study presents a clear picture of the state of the art in understanding how emulsifiers modulate lipid metabolism in poultry. The results are expected to support the development of precision nutrition strategies in modern poultry production systems.

Materials and methods

Ethical approval

This meta-analysis did not involve direct experimentation on animals or the use of human data. Therefore, this study did not require ethical approval. Additionally, this meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021).

Study location and period

The study was conducted between January 8, 2025, and April 6, 2025. It relies on open-access repositories maintained by the National Research and Innovation Agency (BRIN). All the selected papers were digitally organized via Mendeley Desktop (version 1.19.8). The meta-analysis files were securely stored in a private GitHub repository.

The PICO framework

This study follows the PICO framework to define its research focus. The population (P) involves commercial poultry, whereas the intervention (I) includes dietary treatments involving the use of feed additives that contain emulsifiers. The comparison (C) consists of negative control groups that do not receive emulsifier additives. The outcomes (O) refer to markers of lipid digestibility, lipid profile, and lipid metabolism, including gene expression and lipase activity (Budiarto et al., 2025; Iswantari et al., 2025).

Keywords and search strategy

The primary sources of literature are three major scientific databases: PubMed, Scopus, and Web of Science. The search uses advanced keyword combinations organized into four thematic categories.

The first category includes emulsifiers and feed additives and contains terms such as emulsifier, dietary emulsifier, feed emulsifier, emulsifying agent, lecithin, lysolecithin, lysophosphatidylcholine, lysophospholipid, phosphatidylcholine, bile acid, bile salt, monoacylglycerol, monoglyceride, diglyceride, diacylglycerol, glyceryl polyethylene, glyceryl polyethylene glycol ricinoleate, polyglycerol polyricinoleate, stearoyl lactylate, sophorolipid, surfactant, surface-active agent, biosurfactant, and synthetic emulsifier.

The second category focuses on energy and fat digestibility. It also focuses on lipid profiles and genes involved in lipid metabolism. The relevant keywords included lipid digestion, fat digestibility, fat absorption, lipid absorption, lipid transport, fat transport, chylomicron transport, lipoprotein transport, serum triglycerides, blood triglycerides, plasma lipids, blood lipid profile, fatty acid metabolism, lipid metabolism, fatty acid synthesis, fatty acid oxidation, lipolysis, adipogenesis, lipogenesis, fat deposition, body fat accumulation, fat mobilization, and fat metabolism.

The third category is for identifying target species, which include poultry such as chickens, ducks, geese, and turkeys, as well as guinea fowl and quail.

The search focused on the last 10 years, from 2015 to 2025. A total of 628 scientific articles were collected. The raw data from the articles were subsequently exported in bibliometric format and managed with Mendeley. Duplicate articles were automatically removed, and manual verification was performed to ensure the consistency and accuracy of the topic. This resulted in 434 articles, which were filtered again. To ensure relevance to the research topic, regrading was performed, resulting in 194 studies that met the criteria. These studies were evaluated using predetermined inclusion and exclusion criteria.

Types of emulsifiers and lipid parameters

The types and classification of emulsifiers can be found in Table 2. Emulsifiers are categorized by their source and chemical properties, encompassing bile acid–based, glyceride-based, phospholipid-based, non-ionic/synthetic surfactants, biosurfactants, and protein-based types. This categorization offers a systematic view of the range of emulsifiers used in poultry research and underscores their physiological and functional contributions to lipid emulsification.

Table 2.

Type and classification of emulsifiers used in the meta-analysis.

Type	Class
Bile acid-based emulsifiers	bile acid (BA) (including the primary bile acid chenodeoxycholic acid (CDCA), and bile salts (BS)
Glyceride-based emulsifiers	diacylglycerol (DAG), glycerine monostearate (GMS), and glycerol monolaurate (GML)
Phospholipid-based emulsifiers	lecithin (LEC), lysolecithin (LLC), lysophosphatidylcholine (LPC), lysophospholipid (LPLyso), and phosphatidylethanolamine (PE)
Non-ionic/synthetic surfactant	glyceryl polyethylene glycol ricinoleate (GPGR), polyoxymethylene (POE), and stearoyl lactylate (SSL)
Biosurfactants	rhamnolipid (RL) and sophorolipid (SL)
Protein-based emulsifiers	porcine globin (PG)

Energy and crude fat digestibility were expressed as apparent metabolizable energy (AME), gross energy digestibility (GED), and crude fat digestibility (CFD). Lipogenesis was evaluated via acetyl-CoA carboxylase (ACC), DGAT2, fatty acid synthase (FASN), sterol regulatory element binding transcription factor 1 (SREBF1), and SCD. Lipolysis is indicated by adipose triglyceride lipase (ATGL), lipoprotein lipase (LPL), perilipin 1 (PLIN1), lipoprotein lipase activity (LPL act), hepatic lipase activity (LIPC act), and total lipase activity (TL act). Fatty acid oxidation was assessed through acyl-CoA oxidase 1 (ACOX1), CPT1, PPARα, and peroxisome proliferator-activated receptor gamma (PPARγ). Fatty acid transport and binding involve fatty acid binding protein 1 (FABP1, liver-type), fatty acid binding protein 2 (FABP2, intestine-type), fatty acid binding protein 4 (FABP4, adipocyte-type), fatty acid translocase (FAT, also known as CD36), and fatty acid transport protein 4 (FATP4). Lipoprotein metabolism and transport are represented by apolipoprotein B (ApoB) and microsomal triglyceride transfer protein (MTTP). The serum lipoprotein profile included high-density lipoprotein (HDL), high-density lipoprotein cholesterol (HDLC), low-density lipoprotein (LDL), low-density lipoprotein cholesterol (LDLC), total cholesterol (CHOL), and TGL.

Inclusion criteria, exclusion criteria, and selection process

The inclusion criteria covered peer-reviewed scientific articles published in reputable journals and written in English. Eligible studies have examined emulsifiers or feed additives in relation to lipid metabolism in poultry. Only experimental studies with appropriate designs, clearly defined control groups, and complete quantitative data suitable for further analysis were eligible for inclusion.

The exclusion criteria included publications not accepted by the scientific community, those lacking valid publication identifiers, and those unavailable in full-text format. Studies that did not investigate poultry species, were unrelated to the main topic, applied poor research methods, or failed to provide sufficient data for quantitative analysis were also excluded (Fig. S1).

After removing duplicates, 194 unique articles remained. These included 136 studies on broiler chickens, 13 on ducks, 36 on laying hens, 8 on quail, and 1 on geese. The initial screening excluded 61 articles on the basis of the following criteria: lack of a DOI or inaccessibility (n = 13); publication in nonreputable or nonpeer-reviewed journals (n = 21); irrelevance to the study scope (n = 22); and classification as review articles, books, proceedings, book chapters, or other nonoriginal sources (n = 5). A total of 133 articles fulfilled the relevance criteria and underwent further assessment on the basis of research methodology and data availability. Articles were excluded for the following reasons: unsuitable parameters (n = 76), use of nonemulsifier additives or mixtures with other additives (n = 43), irrelevant outcome measures (n = 9), absence of a control group (n = 6), use of categorical data (n = 12), and missing essential values (sample size, mean, or standard error/deviation) (n = 6). As a result, 57 articles were selected for the final analysis (Table 1).

Table 1.

Selected articles on emulsifier addition in poultry feed (broilers, broiler breeders, ducks, geese, and laying hens) under various dietary fat and energy levels.

Bird strain	Sex	Emulsifier	Administered level (mg/kg)	Fat source	Fat percentage (%)	Energy adjustment (kcal/kg)	Ref.
Broiler
Arbor Acres	Mixed (50:50)	Rhamnolipid	10 – 100	Soybean oil	2	0	(Cai et al., 2024)
Arbor Acres	Male	Bile acid	80	Lard	2 – 5.5	0 and 90	(Ge et al., 2019)
Arbor Acres	Male	Bile acid	8	Lard and Soybean oil	1.5 – 3	0	(Geng et al., 2022)
Arbor Acres	Male	Bile acid	10	Soybean oil	2.5 – 7	0 and 291	(Hu et al., 2024)
Arbor Acres	Male	Bile acid	40 – 80	Lard	1.5 – 3	0	(Lai et al., 2018)
Arbor Acres	Not specified	Bile acid	100	Soybean oil	1	0	(Liu et al., 2025b)
Arbor Acres	Male	Lecithin	1,000	Soybean oil	2.65 – 3.52	0	(Shen et al., 2021)
Arbor Acres	Not specified	Bile acid	60	Soybean oil	1.6 – 4.92	0	(Wang et al., 2024b)
Arbor Acres	Male	Bile acid	200	Soybean oil	4	0	(Yin et al., 2021)
COBB 500	Male	Chenodeoxycholic acid	1,000 – 9,000	Not Specified	-	0	(Piekarski et al., 2016)
Evian 48 and Ross 308	Unsexed	Lecithin	500 – 1,000	Soybean oil	3.52	0	(Okasha et al., 2023)
Not specified	Not specified	Lysolecithin	400	Soybean oil	3	−100	(Ahmad et al., 2023)
Ross	Mixed (50:50)	Lecithin	350	Soybean oil	5.91	0	(Zaazaa et al., 2023)
Ross 308	Male	Lysolecithin	1,000 – 2,000	Soybean oil	3.47	0	(Ahmadipour et al., 2025)
Ross 308	Male	Emulsifier blend	1,000 – 2,000	Soybean oil	3	0 and −100	(Ahmadi-Sefat et al., 2022)
Ross 308	Male	Glyceryl polyethylene glycol ricinoleate	1,000	Soybean oil	6.4	0	(Bontempo et al., 2018)
Ross 308	Male	Lysophosphatidylcholine	500 – 1,500	Corn oil	2.44	0	(Gholami et al., 2024)
Ross 308	Male	Bile acid	1,500	Not specified	-	0	(Hemati Matin et al., 2016)
Ross 308	Male	Stearoyl lactylate	300 – 500	Tallow	4.49	−50	(Hoque et al., 2022)
Ross 308	Male	Lysolecithin	1,000	Soybean oil	3.2	−100	(Hosseini et al., 2018)
Ross 308	Male	Lysolecithin	3,000	Lard and Soybean oil	5.3 – 5.81	0	(Jansen et al., 2015)
Ross 308	Not specified	Sophorolipid	250	Soybean oil	5.33	0	(Kwak et al., 2022a)
Ross 308	Male	Sophorolipid	5 – 10	Animal fat	5.41 – 5.47	0	(Kwak et al., 2022b)
Ross 308	Male	Lecithin	1,000	Tallow	3.8	0	(Liu et al., 2020b)
Ross 308	Male	Lecithin	1,000	Poultry fat and Soybean oil	4	0 and −100	(Majdolhosseini et al., 2019)
Ross 308	Male	Lecithin and Lysophosphatidylcholine	500	Vegetable oil	4.79	0	(Nutautaitė et al., 2021)
Ross 308	Not specified	Emulsifier blend	500	Tallow	3.5 – 4	0 and −100	(Oketch et al., 2022)
Ross 308	Not specified	Lysolecithin	300 – 500	Soybean oil	3.97	−65	(Papadopoulos et al., 2018)
Ross 308	Male	Lysolecithin	300 – 900	Tallow	3	100	(Park et al., 2018)
Ross 308	Male	Emulsifier blend	500	Soybean oil	1.5	−50	(Saleh et al., 2020)
Ross 308	Not specified	Bile acid, Lysolecithin, and Lysophospholipid	500	Poultry fat and Soybean oil	6.03 – 6.15	0	(Shoaib et al., 2021)
Ross 308	Male	Glyceryl polyethylene glycol ricinoleate	350	Soybean oil	1.62 – 2.18	0	(Tenório et al., 2022)
Ross 308	Male	Diacylglycerol	422 – 825	Tallow	3.9	−100	(Upadhaya et al., 2017)
Ross 308	Male	Emulsifier blend	500 – 1,000	Tallow	5.4	0	(Upadhaya et al., 2018)
Ross 308	Male	Stearoyl lactylate	500	Soybean oil	4.5	50	(Wang et al., 2016)
Ross 308	Not specified	Polyoxymethylene	1,000	Tallow	4.5	−100	(Wickramasuriya et al., 2020)
Ross 308	Not specified	Stearoyl lactylate	500	Vegetable oil	4	−100	(Wickramasuriya et al., 2022)
Ross 308	Mixed	Lysolecithin	500	Soybean oil	3.99 – 5.18	0 and 50	(Zhang et al., 2022b)
Ross 708	Mixed (50:50)	Porcine globin	500	Soybean oil	4.43	0	(Dabbou et al., 2019)
Sanhuang	Mixed (50:50)	Phosphatidylethanolamine	5 – 20	Not specified	-	0	(Liu et al., 2025a)
Yellow-Feathered	Male	Glycerol monolaurate	300 – 600	Not specified	-	0	(Liu et al., 2020a)
Yellow-Feathered	Male	Rhamnolipid	250 – 500	Soybean oil	2.2	0	(Ma et al., 2025)
Yellow-Feathered	Mixed (50:50)	Emulsifier blend	500	Soybean oil	5.8 – 6.8	0 and −51	(Wang et al., 2024c)
Yellow-Feathered	Male	Lysolecithin	500 – 1,000	Soybean oil	1.9 – 2.5	−60	(Xiang et al., 2024)
Yellow-Feathered	Male	Bile salt	1.5	Soybean oil	2	0	(Zhang et al., 2022c)
Broiler Breeder
Ross 308	Female	Lysophospholipid	500	Soybean oil	0.5 – 1.3	0 and −40	(Sedghi et al., 2024)
Duck
Cherry Valley	Mixed (50:50)	Glycerine monostearate	4	Poultry fat and Soybean oil	2.05 – 2.36	−50	(Hu et al., 2019)
Pekin	Unsexed	Lysolecithin	500	Vegetable oil	0.5 – 2.5	0	(El-Katcha et al., 2021)
Pekin	Male	Glyceryl polyethylene glycol ricinoleate	100	Poultry fat	5.4	0	(Zeng et al., 2023)
Zhijiang	Mixed (50:50)	Bile acid	250	Rapeseed oil	1.25	0	(Chen et al., 2025)
Geese
Holldobagy	Male	Bile acid	75	Soybean oil	2	0	(Li et al., 2023)
Laying hen
Hy-line Brown	Female	Bile acid	95 – 190	Not specified	-	0	(Wang et al., 2024a)
Hy-line Brown	Female	Bile acid	1,200	Not specified	-	0	(Xing et al., 2025)
Hy-line Gray	Female	Bile acid	30 – 1,200	Soybean oil	1	0	(Yang et al., 2022b)
Hy-line Gray	Female	Bile acid	30 – 90	Soybean oil	1	0	(Yang et al., 2022a)
Lohmann	Female	Lecithin	1,000	Vegetable oil	1.08 – 6.1	0 and 250	(Hu et al., 2022)
Sanhuang	Female	Bile acid	100 – 200	Soybean oil	10	672	(Li et al., 2024)

Risk of bias

The assessment of potential bias in the studies involved a risk of bias (RoB) evaluation based on the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE) tool, which uses five review domains (Hooijmans et al., 2014; Adli et al., 2025). The first domain, attrition, refers to incomplete reporting of study outcomes, which indicates a high risk of bias. The blinding domain addresses the prevention of measurement bias by applying blinding procedures to research subjects. The outcome domain evaluates the clarity of measurement procedures, including both the methods and the instruments used. The randomization domain concerns the reporting of random allocation, which must follow appropriate experimental design principles. The reporting domain highlights the risk of bias arising from selective outcome reporting, which reflects a high risk if present. The assessment employed a four-level scale: low, indicating a very low risk of bias with complete explanations across all domains; some concern, referring to a tolerable level of bias due to limited or partial reporting; high, indicating a high risk of bias when domain-related information is absent or unclear in the methods section; and inability to determine when available information is insufficient to conclude (Fig. S2).

Analysis of data, software, and bias

The random effects model estimated the effect size (ES) and the standard error of the ES (SE) for each study (Eqs. (1) and (2)). The interpretation of the ES followed these criteria: an ES of 0.2 indicated a small effect, an ES greater than 0.5 indicated a moderate effect, and an ES greater than 0.8 indicated a large effect (Marín-Martínez and Sánchez-Meca, 1999; Buck et al., 2022; Amirul et al., 2025).

$$ES=\left(\frac{\mathrm{x}{¯}_{treatment}-\mathrm{x}{¯}_{control}}{S{D}_{pooled}}\right)\times \left(1-\frac{3}{4\left(n-1\right)-9}\right)$$ (1)

$$SE=\sqrt{\frac{1}{\sum w_i}+va{r}^2}$$ (2)

In this context, x̄ represents the mean of the control and treatment groups, the pooled standard deviation (SD) indicates the pooled standard deviation, n denotes the number of replications, w represents the weight of each study, and var indicates the between-study variance.

$$Egger\text{'}s=\frac{{\widehat{\beta }}_0}{SE\left({\widehat{\beta }}_0\right)},\text{df}=k-2$$ (3)

The definitions are as follows: ${\widehat{\beta }}_0$ refers to the estimated intercept from Egger’s regression, SE (${\widehat{\beta }}_0$) denotes the standard error of ${\widehat{\beta }}_0$, k represents the number of studies, and df indicates the degrees of freedom in the regression model.

The percentage of variation in effect size was measured via the heterogeneity index (I²) (Huedo-Medina et al., 2006; Bowden et al., 2011), which was calculated via the following equation:

$$I^2=\frac{Q-df}{Q}\times 100\%$$ (4)

In Eq. (4), Q refers to Cochran’s Q statistic, which represents the total variation across studies, and df denotes the degrees of freedom, which is calculated as the number of studies minus one (k − 1).

All analyses used statistical software, including R version 4.4.3. A P value less than 0.05 indicated statistical significance. Publication bias was assessed via Egger’s test (Eq. (3)), where a significance level below 0.05 suggested the presence of bias (Lin and Chu, 2018; Ujilestari et al., 2025).

KEGG analysis and mapping

The molecular metabolism analysis from the meta-analysis data used the gene list included in Table S1, whose names were converted into the Kyoto Encyclopaedia of Genes and Genomes (KEGG) standard. The analysis was performed in R via the “KEGGREST” and “pathview” packages, version 1.46.0. The databases “org.Gg.eg.db” (gga; Gallus gallus) and “org.Hs.eg.db” (hsa; Homo sapiens) were used for KEGG pathway analysis (Kanehisa, 2000; Okuda et al., 2008; Kanehisa et al., 2025). The accession IDs included fatty acid biosynthesis (gga00061), fatty acid elongation (gga00062), fatty acid degradation (β-oxidation) (gga00071), the PPAR signalling pathway (gga03320), the adipocytokine signalling pathway (hsa04920), cholesterol metabolism (hsa04979), the insulin signalling pathway (gga04910), the regulation of lipolysis in adipocytes (gga04923), and fat digestion and absorption (hsa04975). The adipocytokine signalling (hsa04920), cholesterol metabolism (hsa04979), and fat digestion and absorption (hsa04975) pathways used human references because those pathways are not available for poultry. Pathway mapping was visualized via the Z score values of relative gene expression. The KEGG pathway construction was performed only for broiler chickens because of the availability of more complete gene data.

Results and discussion

Energy and fat digestibility

The digestibility of energy, including AME and GED, significantly increased following emulsifier treatment (p < 0.001; Table 3.1). However, AME showed a high level of bias (p = 0.05). In addition, fat digestibility, represented by CFD, also increased significantly (p < 0.001). AME demonstrated a large effect size, whereas GED and CFD had moderate effects.

Table 3.

Effect sizes of emulsifier additives in poultry on energy and fat digestibility, lipid profiles, and lipid metabolism indicators.

Observed parameters	k	ES	SE	Lower	Upper	P value	Egger	I² (%)
1. Energy and fat digestibility
AME	19	0.961	0.141	0.686	1.24	<0.001	0.005	51.1
GED	44	0.544	0.083	0.381	0.708	<0.001	0.082	71.6
CFD	59	0.742	0.073	0.598	0.885	<0.001	0.510	76.0
2. Serum lipoprotein profile
HDL	46	0.311	0.101	0.112	0.510	0.002	0.657	82.7
HDLC	61	0.367	0.073	0.224	0.510	<0.001	0.852	80.8
LDL	44	−0.196	0.099	−0.391	−0.002	0.048	0.292	74.5
LDLC	61	−0.429	0.074	−0.575	−0.284	<0.001	<0.001	80.3
CHOL	116	−0.064	0.055	−0.171	0.044	0.248	0.732	84.2
TGL	125	−0.187	0.053	−0.291	−0.082	<0.001	0.627	77.3
3. Lipogenesis
ACC	75	−0.206	0.098	−0.397	−0.015	0.035	0.656	86.3
DGAT2	15	0.525	0.168	0.195	0.855	0.002	0.679	85.9
FASN	92	−0.134	0.087	−0.305	0.037	0.124	0.757	87.6
SREBF1	73	0.145	0.096	−0.042	0.333	0.129	0.798	86.1
SCD	6	−0.799	0.272	−1.333	−0.266	0.003	0.054	95.9
4. Lipolysis
ATGL	8	−0.316	0.231	−0.769	0.138	0.173	0.350	82.7
LPL	18	0.796	0.180	0.443	1.15	<0.001	0.214	89.8
PLIN1	5	1.22	0.336	0.560	1.88	<0.001	0.013	93.3
LPL act	26	0.595	0.123	0.354	0.836	<0.001	<0.001	77.5
LIPC act	9	0.252	0.199	−0.138	0.642	0.205	0.133	78.1
TL act	36	0.566	0.103	0.364	0.768	<0.001	0.411	77.0
5. Fatty acid oxidation
ACOX1	5	−0.369	0.262	−0.882	0.144	0.159	0.386	84.2
CPT1	24	0.900	0.146	0.615	1.19	<0.001	0.170	90.2
PPARα	20	0.924	0.153	0.624	1.22	<0.001	0.162	89.3
PPARγ	10	0.045	0.172	−0.292	0.381	0.795	0.222	87.3
6. Fatty acid transport & binding
FABP1	15	0.389	0.178	0.040	0.738	0.029	0.411	86.4
FABP2	6	0.344	0.255	−0.155	0.843	0.177	0.899	84.4
FABP4	9	0.347	0.209	−0.064	0.757	0.098	0.938	88.7
FAT (CD36)	17	0.620	0.153	0.321	0.920	<0.001	0.166	80.3
FATP4	76	−0.118	0.102	−0.317	0.082	0.248	0.799	87.9
7. Lipoprotein metabolism and transport
ApoB	16	0.186	0.165	−0.138	0.510	0.261	0.397	90.5
MTTP	11	0.378	0.173	0.038	0.718	0.029	0.196	84.7

ACC, acetyl-CoA carboxylase; ACOX1, acyl-CoA oxidase 1; AME, apparent metabolizable energy; ApoB, apolipoprotein B; ATGL, adipose triglyceride lipase; CFD, crude fat digestibility; CHOL, cholesterol; CPT1, carnitine palmitoyl transferase 1; DGAT2, diacylglycerol O-acyltransferase 2; ES, effect size; FABP1, fatty acid binding protein 1 (liver-type); FABP2, fatty acid binding protein 2 (intestine-type); FABP4, fatty acid binding protein 4 (adipocyte-type); FASN, fatty acid synthase; FAT (CD36), fatty acid translocase; FATP4, fatty acid transport protein 4; GED, gross energy digestibility; HDL, high-density lipoprotein; HDLC, high-density lipoprotein cholesterol; k, observed data; LDL, low-density lipoprotein; LDLC, low-density lipoprotein cholesterol; LIPC act, hepatic lipase activity; LPL, lipoprotein lipase; LPL act, lipoprotein lipase activity; MTTP, microsomal triglyceride transfer protein; PLIN1, perilipin 1; PPARα, peroxisome proliferator-activated receptor alpha; PPARγ, peroxisome proliferator-activated receptor gamma; SCD, stearoyl-CoA desaturase; SE, standard error; SREBF1, sterol regulatory element binding transcription factor 1; TGL, triglycerides; TL act, total lipase activity; I², inconsistency index.

With respect to poultry species, only broiler data were available for AME, GED, and CFD, all of which significantly improved (ES > 0.5; p < 0.001; Fig. 1.i). Male poultry exhibited a relatively strong response across the three parameters (ES > 0.5; p < 0.001; Fig. 1.ii). No data were reported for female poultry, whereas the mixed-sex and unspecified groups presented an effect size close to 1 for CFD (p < 0.05). However, the mixed-sex group did not show a significant response to GED.

Fig. 1.

Effect sizes of bird species (i), sex (ii), emulsifier type (iii), and fat (iv) source on AME, GED, and CFD. Significance levels: P< 0.001 (***), P< 0.01 (**), P< 0.05 (*), and not significant (ns).

Among the emulsifier types, LLC had the greatest effect on AME, whereas DAG had the greatest effect on both GED and CFD (ES > 0.8; p < 0.001; Fig. 1.iii). DAG appeared to be more effective than other emulsifiers, likely because of its ability to actively form lipid globules in the digestive tract, which increased CFD digestibility.

In terms of fat source, lard had the greatest effect on AME and CFD, whereas tallow had the greatest effect on GED (ES > 0.8; p < 0.01; Fig. 1.iv). These findings are noteworthy because, compared with plant-based fat sources, animal-derived fats tend to have stronger effects.

In addition, dietary fat and energy contents also control the physiological actions of emulsifiers. Under energy-restricted diets, emulsifiers increase the activity of digestion and lipid oxidation genes’ lipases, indicating a compensatory process for reduced caloric intake (Hu et al., 2019). At moderate and high-fat intakes, emulsifiers facilitate the redistribution of absorbed lipids and maintain well-balanced serum lipid profiles and evasion of lipogenesis above physiological requirements (El-Katcha et al., 2021; Kwak et al., 2022b). The close relationships observed between digestibility indices and enzymatic activities under these conditions confirm again that emulsifiers optimize lipid utilization in a wide range of nutritional environments. The modulatory effects observed across various lipid metabolism indicators were also influenced by dietary energy content. The response to emulsifiers appeared more pronounced in formulations with lower metabolizable energy, indicating that the emulsifier partially replaced energy needs by increasing lipid utilization efficiency (Hu et al., 2019). This finding aligns with the strong positive correlation between GED and TL, with an estimated correlation coefficient of approximately r ≈ 1 under negative energy conditions (Fig. S4.i.A). Similarly, a strong correlation emerged under low-fat conditions, where GED also showed a close association with TL (r ≈ 1; Fig. S4.ii.A). TL is known to act as an enzyme involved in fat digestion, fat absorption, and energy metabolism (Ahmadipour et al., 2025; Priyatno et al., 2025). On the basis of the type of emulsifier, the emulsifier blend presented a positive hub pattern between TL and GED, followed by an increase in blood TGL (Fig. S6). For lysolecithin, GED was strongly associated with improved CFD digestibility and lipogenesis, which was supported by FASN activity (Fig. S10). Similarly, FASN facilitates lipoic acid metabolism and fatty acid degradation, although its relative expression remains low in broiler chickens, laying hens, and geese (Figs. S7, S15, and S17). Moreover, a distinct pattern was observed for stearoyl lactylate, revealing a contradiction between GED and LDLC production and an increase in blood CHOL. This phenomenon likely indicated the conversion of CHOL to LDLC, which increased energy utilization and storage. Taken together, the KEGG-based pathway interpretation (Figs. S7–S13) corroborated the quantitatively obtained meta-analysis results, providing evidence that emulsifiers are involved in the orchestration of many lipid-related pathways, such as lipogenesis or lipolysis, and oxidation or transport through coordinated molecular regulation. These results provide mechanistic support for the use of emulsifiers as precision nutrients for lipid metabolism in poultry.

Lipid profile in blood serum

The administration of emulsifiers altered the serum lipid profile (Table 3.2). HDL and HDLC levels significantly increased, indicating a small effect size (ES ≈ 0.3; p < 0.01). In contrast, LDL and LDLC levels significantly decreased, with small effect sizes of 0.196 and −0.429, respectively (p < 0.05). TGL levels also declined significantly, with a small effect size (ES = −0.187; p < 0.001). However, CHOL did not significantly decrease.

An increase in emulsifier levels (mg/kg) significantly elevated HDL levels (R² = 0.83; p < 0.01), whereas HDLC significantly decreased HDL levels (Fig. 2.i). LDL and LDLC followed a nonsignificant quadratic pattern of decrease and increase, respectively. Similarly, CHOL and TGL displayed nonsignificant quadratic increases. Among the emulsifier types, LPC significantly increased HDL levels (p < 0.05; Fig. S3.i). Compared with the other emulsifiers, RL had the greatest effect on HDL, including LLC, LEC, and BA, in descending order of effect size (p < 0.05). LPC was the only emulsifier that significantly increased the effect size of LDL (p < 0.05). In contrast, LDLC levels significantly declined, primarily due to RL (p < 0.001), followed by LEC and BA (p < 0.05). The CHOL levels decreased, mainly due to POE, BA, and RL (p < 0.05), whereas LLC significantly increased CHOL (p < 0.001), with no significant effects observed for the other emulsifiers. Finally, TGL levels significantly decreased following the application of RL and BA, whereas LLC significantly increased the TGL effect size (p < 0.01).

Fig. 2.

Trends in effect size levels of emulsifier inclusion rate (%), energy difference (kcal/kg), and fat content (%) in relation to changes in the lipid profile. The trend lines indicate increases (blue) and decreases (red), with the data points marked by the significance level: P< 0.05 (green), 0.05 < P< 0.1 (orange), and P > 0.1 (brown).

Lipid mobilization in the bloodstream involves complex mechanisms. However, on the basis of the observed data, emulsifier supplementation generally improved the blood lipid profile. For example, LPC and RL increased HDL, HDLC, and LDL levels, suggesting that these emulsifiers enhanced lipoprotein conversion in the liver through transport-related mechanisms. When HDL and LDL levels change proportionally, maintaining a relatively constant ratio, the lipid accumulation process in the blood is considered physiologically normal. The increase in CHOL and TGL levels induced by BA likely resulted from increased intestinal absorption activity. These mechanisms indicated effective emulsification during fat digestion, which led to improved lipid absorption in the gastrointestinal tract. This finding was supported mainly by the performance of FAT (CD36) in the process of fatty acid absorption. Furthermore, in the small intestine (endoplasmic reticulum), fatty acids are converted into chylomicrons through the glycerolipid metabolism pathway, which is mediated by the activities of ApoB, FABP1, FABP2, and MTTP (Fig. 4).

Fig. 4.

Effect of emulsifiers on gene expression in the fat digestion and absorption pathway of broiler chickens.

A greater dietary energy difference (relative to the basal energy diet, kcal/kg) led to significant decreases in the LDLC, CHOL, and TGL concentrations, with determination coefficients of 0.793, 0.604, and 0.470, respectively (p < 0.05; Fig. 2.ii). Conversely, HDL, HDLC, and LDL levels did not significantly change, and the data were insufficient to establish robust regression models for these variables.

Increasing the dietary fat level (%) resulted in a significant reduction in the LDLC (R² = 0.319) and CHOL (R² = 0.253; p < 0.05; Fig. 2.iii). Although HDL, HDLC, and LDL levels also declined, the changes were not statistically significant. TGL levels showed a nonsignificant upwards trend. Vegetable fat sources, such as corn oil and soybean oil, significantly elevated HDL, HDLC, and LDL concentrations (p < 0.05; Fig. S3.ii). On the other hand, animal fat sources such as tallow significantly increased CHOL and TGL levels (p < 0.01), reduced LDLC, and increased HDLC levels (p < 0.01). These findings suggest that dietary fat inclusion enhances lipid intake and absorption. Vegetable-derived fats appear to be more readily transformed into HDL and LDL lipoproteins, whereas animal-derived fats tend to mobilize more slowly. This delayed mobilization is reflected in the initial accumulation of CHOL and TGL. These findings may provide a useful basis for feed modification, particularly under heat stress conditions, where slowly mobilized fat sources can reduce the thermal burden associated with feed metabolism.

Lipogenesis

The present meta-analysis demonstrated that emulsifiers significantly affect molecular markers associated with lipid metabolism. The mode of action begins with the influence of emulsifier additives on lipid digestion and metabolism through a concerted series of biochemical and physiological processes that go beyond their physical surfactant action. They enhance the dispersion of lipids within the intestine, improve the availability of triglycerides to digestive enzymes, and thus modulate gene expression and enzyme activity for lipogenesis, lipolysis, oxidation, and lipid transport. The combination of these effects synergistically improves the efficiency of nutrient utilization and energy conversion in poultry.

Lipogenesis was increased by dietary emulsifier supplementation, as evidenced by a significant increase in DGAT2 mRNA expression (p = 0.02; Table 3.3). In contrast, ACC and SCD expression significantly decreased, with ES values of −0.206 and −0.799, respectively (p < 0.05). The bird species factor significantly reduced SCD expression in laying hens (ES ≥ 0.5; p < 0.05; Fig. 3.i.C). With respect to sex, the most prominent effect was observed for SCD expression in the mixed-sex group (50:50), which significantly increased (ES ≥ 0.5; p < 0.05; Fig. 3.ii.C). The emulsifier PE significantly increased SCD expression (ES ≥ 0.5; p < 0.05; Fig. 3.iii.C). Additionally, the use of animal fat-type lard significantly reduced SREBF1 expression (ES ≥ 0.5; p < 0.05; Fig. 3.iv.C). SCD serves as an indicator of lipogenesis and fat storage through the conversion of saturated fatty acids into unsaturated fatty acids. A decrease in SCD expression likely indicates increased lipid oxidation due to emulsifier supplementation, suggesting that fat was utilized directly (Fig. S9).

Fig. 3.

Volcano plot of effect sizes for bird species (i), sex (ii), emulsifier type (iii), and fat (iv) source on gene expression and lipase activity. Displayed effect sizes meet the criteria of |ES| > 0.5 and significance at P< 0.05; others are considered not significant.

At the lipogenic level, higher DGAT2 tends to improve the esterification of fatty acids into triglycerides for the assimilation of dietary fat. Conversely, decreased expression of SCD suggests a metabolic shift away from lipid storage through oxidation, indicating that fatty acids are directed toward energy formation rather than deposition (Zhou et al., 2025). Emulsifiers increase the availability of fatty acids and improve lipid digestion efficiency by enhancing SCD performance (Zhou et al., 2025). The conversion of saturated fatty acids (stearate and palmitate) into unsaturated fatty acids (oleate and palmitoleate) is catalysed by SCD itself. The DGAT2 enzyme then converts unsaturated fatty acids into TGL (Qi et al., 2020; Mao et al., 2022). DGAT2 combines a DAG with unsaturated fatty acids derived from SCD. Therefore, the functions of SCD and DGAT2 are closely related to TGL anabolism. TGLs are then stored in the form of oil coated with oleosin and deposited in the endoplasmic reticulum of poultry adipose tissue (Huang et al., 2008; Liu et al., 2012; Hosseini et al., 2018).

Lipolysis

With respect to lipolysis, the expression levels of the LPL and PLIN1 genes increased significantly, with ES values of 0.796 and 1.22, respectively (p < 0.001; Table 3.4). This upregulation was followed by a significant increase in the enzymatic activities of LPL act and TL act, with ES values of 0.596 and 0.566, respectively (p < 0.001). The highest LPL gene expression was detected in broiler breeders (ES ≥ 0.5; p < 0.05; Fig. 3.i.D). In contrast, PLIN1 expression was most prominent in the mixed (50:50) sex category (ES ≥ 0.5; p < 0.05; Fig. 3.ii.D). Among the emulsifier types, PE induced the strongest upregulation of PLIN1, whereas LPL expression decreased, likely due to the presence of LPLyso (ES ≥ 0.5; p < 0.05; Fig. 3.iii.D). Moreover, the use of tallow, an animal fat source, significantly increased TL activity (ES ≥ 0.5; p < 0.05; Fig. 3.iv.D). On the basis of the correlation results, dietary energy modification showed a notable pattern, particularly compared with the standard energy level (Fig. S4.i). Under conditions of a negative energy difference, LPL was strongly correlated (r ≈ 1) with CPT1, PPARα, and CHOL (Fig. S4.i.A). Under neutral energy conditions, LPL was also strongly correlated (r ≈ 1) with CPT1, whereas the correlations with PPARα and CHOL were moderate (r ≈ 0.5) (Fig. S4.i.B). In the case of a positive energy difference, LPL was moderately correlated (r ≈ 0.5) with CPT1 and PPARα but strongly correlated with ApoB and FAT (CD36) (r ≈ 1; Fig. S4.i.C). Under different dietary fat levels (Fig. S4.ii), LPL expression was correlated with various key metabolic markers. At low fat levels (≤ 2.5%), LPL was moderately correlated with CPT1, PPARα, and CHOL (r ≈ 0.5; Fig. S4.ii.A). At moderate fat levels, LPL was negatively correlated with CHOL, LDL, and TGL (r ≈ 0.5; Fig. S4.ii.B). At high fat contents (> 5%), LPL had a weak positive correlation with PLIN1 (r < 0.5) and a moderate negative correlation with PPARα and FABP2 (r ≈ 0.5; Fig. S4.ii.C).

In the lipolysis pathway, supplementation with emulsifiers increased the expression of LPL and PLIN1, as well as the activities of total and hepatic lipases (Kwak et al., 2022b). These reactions reveal that emulsifiers, in addition to physically breaking up fat droplets, also affect the intracellular mechanism controlling lipid mobilization (Olzmann and Carvalho, 2019). By increasing micelle formation in the intestinal lumen, emulsifiers increase substrate supplies for lipases and accelerate triglyceride hydrolysis into glycerol and free fatty acids. Concurrent stimulation with PLIN1 (Fig. S12). These findings suggest greater mobilization of stored lipids for metabolic use (Wang et al., 2013; Olzmann and Carvalho, 2019). These combined enzymatic and molecular effects explain the net increase in plasma free fatty acids and increase energy yield (Wang et al., 2013; Olzmann and Carvalho, 2019).

The energy stored in TGL is converted into free fatty acids (FFAs) and glycerol through the lipolysis pathway, which is carried out by the LPL act enzyme. LPL act activity increases FFA circulation in peripheral tissues (Kwak et al., 2022b). Consequently, this phenomenon elicits the mobilization of previously synthesized lipid energy reserves by TL act and LIPC act. This synergistic control indicates that emulsifiers control the precarious equilibrium between fat utilization and synthesis in a way that promotes optimum lipid turnover, leading to improved feed efficiency (Kwak et al., 2022b). This molecular response is consistent with what has been reported in the regulation of lipolysis in adipocytes (gga04923) pathway (Fig. S12), where PLIN1 activation stimulates the liberation of lipid droplets, which provide free fatty acids for β-oxidation. This pathway enrichment further supports the hypothesis that emulsifiers mediate not only extracellular but also, at least in part, intracellular lipid digestion.

An interesting observation from the meta-analysis was the reduction in ACC expression following emulsifier supplementation. Interestingly, ACC expression tended to decrease in response to emulsifier addition. These findings suggest the presence of an optimal dosage threshold, where excessive emulsifier administration induces a negative feedback mechanism in the lipogenic pathway (Ge et al., 2019; Wang et al., 2024c). This likely reflects a feedback response that limits excessive lipid synthesis once energy requirements have been met. This finding also suggests the presence of an optimal inclusion level of emulsifiers, above which lipogenic inhibition may occur (Ge et al., 2019). Such dosage sensitivity underscores the need for precision in emulsifier application, as excessive supplementation could disrupt metabolic equilibrium.

Fatty acid oxidation

The expression of genes involved in fatty acid oxidation, particularly CPT1 and PPARα, significantly increased, with a strong effect size (ES > 0.8; p < 0.001; Table 3.5). In contrast, ACOX1 and PPARγ did not have significant effects. This finding aligns with the observed increase in CPT1 expression in broilers and broiler breeders (ES ≥ 0.5; p < 0.05; Fig. 3.i.A). Additionally, CPT1 expression increased in male poultry, whereas PPARα expression increased in female poultry (ES ≥ 0.5; p < 0.05; Fig. 3.ii.A). Similar trends were observed with LPLyso-type emulsifiers, which increased CPT1 expression (ES ≥ 0.5; p < 0.05; Fig. 3.iii.A). Furthermore, lard induced a greater increase in CPT1 expression than did soybean oil (ES ≥ 0.5; p < 0.05; Fig. 3.iv.A). The upregulation of CPT1 suggests enhanced fatty acid transport, likely driven by increased dietary fat intake due to emulsifier supplementation. Consequently, the mitochondrial transport rate of fatty acids increases to support β-oxidation. This process corresponds to the role of PPARα and CPT1 as key transcription factors in lipid metabolism (Fig. S11). Furthermore, PPARα was highly expressed, and CPT1 was expressed at low levels in broiler chickens, whereas the opposite pattern occurred in ducks (Fig. S16).

Lipid energy plays a role in the body’s energy production by activating the conversion of fatty acyl-CoA into acetyl-CoA. This process involves the degradation of fatty acyl-CoA into acetyl-CoA through β-oxidation, which is mediated by ACOX1, as shown in Fig. S8 (Lu et al., 2024). Acetyl-CoA then enters the Krebs cycle and the electron transport chain, where it produces ATP. This process is controlled by CPT1 (Tang et al., 2022). Inducing the expression of PPARα increases the activation of the fatty acid energy conversion pathway with acetyl-CoA. PPARα acts as the main transcription factor in lipid oxidative metabolism (Sedghi et al., 2024; Wang et al., 2024c).

An interesting part of the meta-analysis was the pervasive upregulation of CPT1 and PPARα, which play important roles in fatty acid oxidation in the mitochondria. First, CPT1 enables the entry of long-chain fatty acids into the mitochondrial matrix, and PPARα is a transcription factor that regulates β-oxidation genes (Schreurs et al., 2010; Tahri-Joutey et al., 2021). Second, emulsifiers stimulate oxidative metabolism, with increased conversion of fatty acids to acetyl-CoA and ATP (Li et al., 2022). This pathway accounts for the oxidation that allows broiler chickens to compensate for reduced dietary energy via the increased efficiency of energy generated from fat. Additionally, interspecies responses revealed relevant physiological differences. Broiler chickens, for example, tended to exhibit higher lipogenic gene expression than did ducks and geese, whereas broiler chickens presented increased expression of oxidative genes such as PPARα and CPT1. The variation indicates that emulsifier functionality depends on inherent metabolic characteristics, such as individual species, providing evidence for the principle of precision nutrition in poultry (El-Katcha et al., 2021; Kwak et al., 2022b), as supported by the results of the present meta-analysis (Fig. 3.ii.A). Moreover, male birds are likely to exhibit greater metabolism than females because they possess differences in the hormonal control of lipid turnover (Li et al., 2022). KEGG mapping (Figs. S6 and S7) revealed upregulated genes in the fatty acid degradation (β-oxidation; gga00071) and PPAR signalling (gga03320) pathways, which included CPT1 and PPARα, leading to increased mitochondrial fatty acid oxidation. Thus, emulsifiers may act as metabolic modulators that increase oxidative metabolism via PPAR-dependent transcriptional regulation.

Fatty acid transport and binding

Fatty acid transport and binding proteins, such as FABP1 and FAT (CD36), significantly increased, with ES values of 0.389 and 0.620, respectively (p < 0.05; Table 3.6). Consistently, FAT (CD36) expression in broiler chickens also increased significantly (ES ≥ 0.5; p < 0.05; Fig. 3.i.B). Both FABP1 and FAT (CD36) increased primarily in mixed-sex poultry (ES ≥ 0.5; p < 0.05; Fig. 3.ii.B). FABP4 exhibited the greatest increase under the PE treatment (ES ≥ 0.5; p < 0.05; Fig. 3.iii.B). Furthermore, FAT (CD36) levels significantly increased in response to animal fat and soybean-derived emulsifiers (ES ≥ 0.5; p < 0.05; Fig. 3.iv.B).

FABP1 and FABP4 strongly influence the distribution of β-oxidation, TGL formation, and lipid signal synthesis. FABP1 has been reported to play an important role in the intracellular transport of FFAs from the plasma membrane to mitochondria. Supported by FABP4, FABP1 distributes FFAs within adipose cells and regulates metabolic signals related to insulin sensitivity and energy balance (Wang et al., 2023). Specifically, FABP4 transports FFAs from lipid droplets to the β-oxidation pathway, connecting TGL to energy utilization in the mitochondria (Agellon, 2024). Moreover, FABP1 and FABP4 function as transport systems, and FAT supports ATP production and new lipid synthesis in the mitochondria (Huang et al., 2022). Supplementation with emulsifiers has been shown to increase the production of FABP1 and FABP4 (Fig. S9), indicating an effect on lipid metabolism due to increased energy availability from fats and oils (Zhang et al., 2022b). A similar pattern was observed in broiler commercial, broiler breeders, and laying hens, where increases in FABP1 and FABP4 were accompanied by increases in GED (Fig. S5). However, FABP1 expression, as well as FABP4 expression, was lower in broiler breeders, especially in laying hens (Figs. S14 and S15). The same trend was observed for blood lipids and lipoproteins, including CHOL, HDLC, LDLC, and TGL. The enzymatic activities of LPL act, LIPC act, and TL act also increased in ducks. Additionally, by upregulating several proteins, such as FABP1, FAB4, and FAT (CD36), emulsifiers can alter the way our bodies handle lipids rather than simply breaking down fats (Storch and Herr, 2001; Hotamisligil and Bernlohr, 2015). KEGG pathway enrichment (Fig. S10) suggested the participation of insulin signalling (gga04910) and adipocytokine signalling (hsa04920), which suggests that emulsifiers may regulate energy homeostasis via insulin sensitivity, as well as adipokine-mediated lipid transport. This modulation of PPARα/FABP1 and FABP4/PPARα interactions in these pathways indicates that there is crosstalk between lipid oxidation and energy cycle regulation.

The protein begins transporting fatty acids to the mitochondria from the cell surface where they are used as fuel for signalling pathways (Kastaniotis et al., 2017). A series of cholesterol metabolism processes were also found to be associated with the upregulation of the ApoB, FAT (CD36), and LPL genes due to emulsifier treatment (Fig. S13).

Lipoprotein metabolism and transport

The expression of the MTTP gene increased significantly in response to emulsifier treatment (p = 0.029; Table 3.7). Consistently, MTTP expression increased considerably in broilers (ES ≥ 0.5; p < 0.05; Fig. 3.i.E). A similar significant increase was detected in mixed-sex broilers (50:50 ratio) (ES ≥ 0.5; p < 0.05; Fig. 3.ii.E). The use of an emulsifier blend also significantly elevated MTTP expression (ES ≥ 0.5; p < 0.05; Fig. 3.iii.E). Furthermore, the use of soybean oil as the fat source induced a significant increase in MTTP expression (ES ≥ 0.5; p < 0.05; Fig. 3.i.E).

The emulsifiers can increase the release and production rates of VLDLs, which carry cholesterol from the liver to the body. All of these alterations operate to control lipid levels and provide cells with a steady fuel supply for growth and repair (Borén et al., 2024). Triglyceride levels in peripheral tissues increase due to very low-density lipoprotein (VLDL) and LDL (Borén et al., 2024). ApoB triggers the formation and secretion of VLDL and LDL (Fig. S13), which transport triglycerides from the liver to peripheral tissues (Borén et al., 2024). Furthermore, ApoB mobilizes TGL and cholesterol (Fig. S13) from their main sites of synthesis, the liver and enterocytes, to all tissues in the body (Borén et al., 2024; Sweeney et al., 2025). MTTP functions primarily in the assembly of VLDL and LDL by transferring TGL, phospholipids, and cholesterol esters into ApoB during the initial stages of lipoprotein assembly in the endoplasmic reticulum (Qin et al., 2021; Petrenko et al., 2023).Lipid phosphate phosphatase (LPP) assists in this process by acting as a catalyst in the hydrolysis of phosphatidic acid (PA), providing precursors for DAG synthesis and ultimately supporting TGL formation (Chandel, 2021). Finally, VLDL and LPL assembly requires TGL as the core lipid (Chandel, 2021). The emulsifiers in poultry feed can increase the expression of ApoB, MTTP, and LPP, which play instrumental roles in the transport system of TGL and cholesterol for energy production and distribution (Li et al., 2023; Hu et al., 2024). This finding was further supported by KEGG mapping of the cholesterol metabolism (hsa04979) pathway (Fig. S13), and the upregulation of ApoB and MTTP suggested improved VLDL assembly and lipid export. These types of molecular data support that emulsifiers increase systemic lipid trafficking, which is consistent with the better serum lipid profiles found in the meta-analysis.

Potential risks of oxidative stress and hepatic lipid redeposition with emulsifier use

Despite the fact that emulsifiers have an overall positive impact on lipid digestion, energy utilization, and metabolic efficiency, their excessive or inappropriate use may result in physiological risks. Higher levels of intestinal lipid absorption and enhanced maturation may lead to increased delivery of essential fatty acids to peripheral tissues (and potentially to the liver), exceeding the oxidative and metabolic capacity of hepatocytes (Ge et al., 2019; Zaefarian et al., 2019; Emami et al., 2020; Oketch et al., 2023). The accumulation of hepatic fats during this process can lead to lipid redeposition in the liver, particularly when the dietary energy density is high or emulsifier inclusion levels are not precisely controlled. This may occur under certain circumstances.

Additionally, increased lipid digestion and absorption may increase the susceptibility of lipids to peroxidation, thereby contributing to oxidative stress. During lipid metabolism, excessive generation of reactive oxygen species may interfere with redox homeostasis, disrupt cellular integrity and adversely affect liver function. This is due to several factors, such as the type and level of emulsifiers given, the total fat consumed by the birds, or nutrients in the feed; age and physiological status may influence these conditions (Ge et al., 2019; Wang et al., 2024c). Moreover, the consumption of high-energy food might influence oxidative stress responses and alterations in antioxidant defence mechanisms. Moreover, the potential beneficial effects of emulsifiers on feed efficiency and precision nutrition should be carefully monitored. The inclusion levels of the right emulsifier are crucial for optimizing metabolic outcomes while maintaining proper lipid metabolism and oxidative balance. Oxygen stress biomarkers, liver lipid deposition indices and long-term metabolism are also important for this condition and should be considered in future studies. Moreover, the use of emulsifiers, such as malondialdehyde, reactive oxygen species and several antioxidant enzymes, such as superoxide dismutase, catalase, or glutathione peroxidase, can provide insight into redox balance. Similarly, liver triglyceride content, histological lipid accumulation and the expression of genes related to upregulated fatty acids and storage can be associated with hepatic fat depletion.

Final assumptions, future suggestions, and study limitations

It is assumed that, beyond the molecular and biochemical mechanisms identified in this meta-analysis, the effects of emulsifier supplementation extend to physiological and productive responses in poultry. The apparent improvements in lipid turnover and oxidative metabolism imply that emulsifiers can increase nutrient utilization efficiency, leading to an improved feed conversion ratio, leaner carcass criteria and reduced abdominal fat deposition. These responses are particularly important in the case of modern, rapidly growing broiler chickens with high energy requirements, for which optimal lipid metabolism is a limiting factor to performance. From the point of view of feeding, adding emulsifiers offers a means to partially replace high-energy feedstuffs without affecting growth. Additionally, increased lipid metabolism may also help reduce excessive hepatic fat accumulation, which in turn may improve liver health and increase the likelihood of overcoming metabolic challenges in birds reared under intensive conditions.

Emulsifiers may also have systemic effects by modulating the gut‒liver axis. More recently, it has been reported that specific biosurfactant-based emulsifiers, such as rhamnolipids or bile acid derivatives, are able to modulate the gut microbial composition toward a better lipid profile, stimulate bile salt activities and modify fatty acid and TGL absorption and trafficking through host‒microbe interrelations. These studies indicate that emulsifiers are not only physiochemically perturbed but also metabolic disruptors of intestinal‒hepatic crosstalk. Despite the evidence on the molecules obtained from this meta-analysis, several information gaps remain. For example, data on nonchicken poultry species such as waterfowl, turkey, and ostrich remain limited. Additionally, future studies integrating multi-omics approaches such as transcriptomics, metabolomics, and lipidomic might reveal better metabolic pathways that are linked with emulsifiers for energy use and control of the environment of gut microbes.

The technical limitation of this study lies in the relatively high level of heterogeneity among the included studies. However, the applied random effects model provides valid effect estimates. Assessment of publication bias via Egger’s test revealed a low tendency for bias in most parameters, thereby strengthening the reliability of the conclusions. Furthermore, subgroup meta-analyses have confirmed that each factor of heterogeneity—including bird species, sex, type of emulsifier, and fat source—has reduced the level of heterogeneity and has also become a source of new findings and reviews (Budiarto et al., 2025; Michael, 2025).

Conclusion

A meta-analysis of 57 primary studies demonstrated that supplementation with emulsifiers consistently enhances the utilization of energy and fat in poultry, particularly in broiler chickens. Emulsifiers had a significant effect on both the serum fat profile and energy level by increasing HDL and HDLC and reducing LDL while also improving digestibility and decreasing crude fat digestibility.

Additionally, these compounds effectively modulated the serum lipid profile. The emulsifier type and the proportion of dietary fat sources affected the response, with animal fat origin and lysophospholipid-based alternatives generally being more effective. The modulation of genes and enzymes involved in lipolysis, fatty acid transport, and oxidation by emulsifiers had a significant impact on lipid metabolism at the molecular level. Molecular adjustments facilitate greater utilization of lipids as an energy source, with variations in research on the basis of both sex and diet. These results suggest that emulsifiers function through multiple physical and metabolic processes, including increased fat synthesis, improved intracellular fatty acid transport, lipid metabolism regulation, and the regulation of growth performance and efficiency. These findings provide a comprehensive overview of these findings. Defining the most effective concentrations of an emulsifier, testing long-term safety in terms of metabolism and liver function, and studying changes in species are essential for future research. Such efforts will also strengthen the use of emulsifiers as precision nutrition tools for sustainable poultry systems.

CRediT authorship contribution statement

Danung Nur Adli: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Funding acquisition. Sugiharto Sugiharto: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Funding acquisition. Tri Ujilestari: Writing – review & editing, Writing – original draft, Validation, Methodology, Data curation. Agung Irawan: Writing – original draft, Data curation, Formal analysis. Amirul Faiz Mohd Azmi: Writing – original draft, Formal analysis, Data curation. Fatim Illaningtyas: Writing – review & editing, Writing – original draft, Validation, Methodology, Data curation. Dimar Sari Wahyuni: Writing – original draft. Novia Qomariyah: Writing – original draft, Validation, Data curation. Pradita Iustitia Sitaresmi: Writing – original draft, Validation, Data curation. Rantan Krisnan: Writing – original draft, Validation, Data curation. Riris Delima Purba: Writing – original draft. Procula Rudlof Matitaputty: Writing – review & editing, Writing – original draft, Validation, Methodology, Data curation. Mochamad Dzaky Alifian: Writing – review & editing, Writing – original draft, Validation, Methodology, Data curation. Mohammad Miftakhus Sholikin: Writing – review & editing, Writing – original draft, Visualization, Supervision, Software, Methodology, Formal analysis, Data curation, Conceptualization, Project administration.

Disclosures

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Sugiharto Sugiharto reports financial support was provided by Diponegoro University. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.