Meta-analysis of the effect of the gluten-free diet on the lipid profile of patients with celiac disease

Abstract

Introduction

Celiac disease (CD) is an autoimmune disorder that affects the digestive system. Treatment for CD relies on a gluten-free diet, and previous studies have suggested alterations in the lipid profile. However, the literature shows no consensus regarding the specific lipid fraction affected, the magnitude of the change, or whether lipid parameters tend to increase or decrease. The aim of this study was to evaluate the effect of a gluten-free diet on HDL, LDL, total cholesterol, and triglyceride levels in patients with celiac disease reported in the scientific literature.

Methods

A literature review was conducted in four databases including studies published in English, Spanish, and Portuguese. Inclusion and exclusion criteria were applied. Heterogeneity was assessed using the I² statistic, publication bias was evaluated with Egger’s test, and meta-analyses were performed to estimate mean differences in lipid profile parameters using a random-effects model with the REML estimation method.

Results

Fifteen studies were included, most of them conducted in Italian populations, comprising a total of 1,820 patients with CD, of whom 706 were pediatric. In pediatric patients, the meta-analysis showed no significant change in total cholesterol (6.2 mg/dL; 95% CI − 7.1 to 19.6), a significant decrease in triglycerides (− 14.2 mg/dL; 95% CI − 22.6 to − 5.8), and an increase in HDL cholesterol (11.4 mg/dL; 95% CI 8.5 to 14.2). In contrast, among adults, there was a significant increase in total cholesterol (12.4 mg/dL; 95% CI 5.1 to 19.7), triglycerides (5.9 mg/dL; 95% CI 0.5 to 11.2), and HDL cholesterol (6.1 mg/dL; 95% CI 3.9 to 8.3).

Conclusion

The findings of this study indicate that patients with celiac disease experience significant changes in lipid profile following adherence to a gluten-free diet, with a differential pattern between pediatric and adult populations.

Introduction

Celiac disease (CD) is an autoimmune disorder that affects the digestive system, primarily the small intestine, in genetically predisposed individuals who ingest foods containing gluten, such as barley, rye, wheat and their derivatives. The development of disease symptoms depends on the interaction between immune complexes involving T cells and gluten proteins [1].

Patients with CD may present gastrointestinal symptoms such as vomiting, chronic diarrhea, anorexia, abdominal distension, and irritability. Beyond gastrointestinal manifestations, CD also leads to extraintestinal symptoms resulting from nutrient deficiencies, including growth retardation in children, muscle atrophy, iron-deficiency anemia, osteoporosis, arthritis, and arthralgia, among others [2].

The prevalence of CD in Latin America ranges between 0.46% and 0.64%, similar to that reported in Europe [3]. In Spain, it is a more frequent condition, being up to five times more active in children than in adults [4]. In the United States, CD affects approximately 0.5% to 1% of the total population and is more common in women than in men [5]. In contrast, prevalence estimates are 0.2% to 0.02% in India, 0.6% in Iran and Israel, 1.6% in Syria, and 1.2% in Turkey [6–8]. Furthermore, CD diagnosis is more frequent in Europe and North America due to greater awareness and access to diagnostic tools, while in developing countries the disease remains underdiagnosed [4].

The current treatment for patients with CD is the gluten-free diet (GFD), which consists of excluding foods containing gluten and replacing them with alternative products, thereby reducing clinical symptoms. However, it has been reported that adherence to this diet may affect patients’ lipid profiles [6].

The relationship between celiac disease and lipid metabolism has been explored from two complementary perspectives. On one hand, certain nutritional components of the GFD, characterized by higher fat and carbohydrate content, may induce adverse increases in serum lipid levels [9, 10]. On the other hand, active CD can cause severe intestinal inflammation and malabsorption, leading to significant reductions in total cholesterol, LDL, and particularly HDL cholesterol [11, 12]. This decrease in HDL has been interpreted both as an indirect marker of intestinal inflammation and as a consequence of altered intestinal secretion of apolipoprotein AI (Apo-AI), the main structural protein of HDL [13, 14]. On the other hand, the decrease in LDL is caused by a reduction in the absorption of cholesterol in the intestine, which is used to form circulating LDL, as well as by an increase in the rate of transport and elimination of LDL apo-B [15]. Consequently, the adoption of a GFD, by allowing recovery of the intestinal mucosa, contributes to the normalization of these lipid parameters.

To elucidate the effect of the GFD in patients with CD, several studies have been conducted, yielding divergent results. One study analyzed the lipid profile of 132 patients before and after the GFD, showing a statistically significant increase (p < 0.001) in total cholesterol (mean difference 19.6 ± 30.2 mg/dL) and HDL cholesterol (mean difference 9.5 ± 11.6 mg/dL); however, no significant changes were observed in LDL or triglyceride levels [16].

In 2021, a cohort of patients was monitored before and one year after initiating the GFD. A significant increase in total cholesterol was observed, from 145.4 ± 38.7 mg/dL to 166.7 ± 39.0 mg/dL (p = 0.002) [17]. In the same study, triglyceride and LDL levels also increased significantly after one year of follow-up (p = 0.014 and p = 0.001, respectively), while HDL levels showed no significant change [18].

In another study conducted in 2009, a significant increase (p < 0.001) in total cholesterol and HDL was reported [19]. The LDL/HDL ratio decreased after one year on the diet (p < 0.0001), whereas LDL and triglycerides showed no significant changes [20].

A 2018 study compared the lipid profiles of pediatric patients with type 1 diabetes mellitus (T1DM) (Group 1) and those with both CD and T1DM (Group 2). Group 2 exhibited significantly lower HDL levels compared with Group 1 (p < 0.05), while LDL, total cholesterol, and triglycerides showed no significant differences [17]. Another investigation in 2016 found lower HDL levels in Group 2 before starting the GFD compared with Group 1. After adherence to the GFD, HDL levels significantly increased in Group 2 (p < 0.0001) [18].

Overall, the global scientific literature presents discrepancies regarding whether changes occur in the lipid profile before and after adherence to a GFD. Moreover, inconsistencies exist concerning which lipid parameters are affected, with no consensus on the direction or magnitude of these effects. Reaching a consensus on the impact of the GFD on the lipid profile could help identify potential weaknesses in the current treatment for CD and reduce associated cardiovascular risk factors linked to lipid metabolism [19]. Therefore, the aim of this study was to evaluate the effect of a gluten-free diet on HDL, LDL, total cholesterol, and triglyceride levels in patients with celiac disease reported in the scientific literature [20].

Methods

Study design

Systematic literature review and meta-analysis conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21].

PICO Question.

Population

Patients with biopsy-confirmed celiac disease [11, 13, 16, 22–29], those referred by hospitals with a confirmed diagnosis [30, 31], and those who met the Italian consensus criteria [32].

Intervention

Gluten-free diet (GFD).

Comparison

Patients with celiac disease before initiation of the GFD lasting at least 6 months to a year.

Outcome

Levels of total cholesterol, triglycerides, HDL cholesterol, and LDL cholesterolIdentification.

A systematic search was carried out in the PubMed, Scopus, Google Scholar, and SciELO databases. Articles published in English, Spanish, or Portuguese that included information on celiac disease, gluten-free diet, and alterations in lipid profile were retrieved.

The search terms used were “gluten free” OR “GFD” OR “gluten free diet”, “celiac disease” OR “CD” OR “gluten enteropathy”, and “lipid profile” OR “cholesterol” OR “triglycerides” OR “HDL” OR “LDL”, and “cardiovascular disease”.

Filters varied depending on the database: “Title/Abstract” in PubMed, “TITLE-ABS-KEY” in Scopus, “resumen” in SciELO, and “in the title of the article” in Google Scholar.

Examples of search formulas included: ((gluten free [Title/Abstract]) AND (celiac disease [Title/Abstract])) AND (lipid profile [Title/Abstract]); TITLE-ABS-KEY (gluten AND free AND celiac AND disease AND cholesterol); (ab: (gluten free)) AND (ab: (celiac disease)) AND (ab: (triglycerides)); allintitle: gluten free AND celiac disease AND cholesterol.

Screening

All records obtained from the search were imported into Zotero software, where duplicates were removed. Based on title and abstract review, inclusion criteria were applied as follows:

• (a) population diagnosed with celiac disease without cardiovascular comorbidities;

• (b) study reports at least one lipid profile parameter before and after adherence to the GFD;

• (c) original research article;

• (d) written in English, Spanish, or Portuguese.

Articles meeting these criteria were then subjected to full-text review and exclusion criteria were applied:

• (a) lipid profile results not expressed as mean and standard deviation;

• (b) lack of full-text access;

• (c) sample size smaller than 20 participants;

• (d) duplicated data from another study;

• (e) unclear sample size for comparison groups.

Selection

Articles that met both inclusion and exclusion criteria underwent qualitative and quantitative data extraction. Qualitative data included author, country, year of publication, sex distribution, and presence of comorbidities. Quantitative data included sample size, mean age, duration of GFD adherence, and lipid profile parameters (mean ± SD) before and after the GFD.

All lipid values were standardized to milligrams per deciliter (mg/dL).

Three researchers independently conducted the search and selection process. Any discrepancies were resolved by consensus or by consultation with a fourth researcher. Extracted data were consolidated using IBM SPSS Statistics software.

Methodological quality assessment

The methodological quality of the studies included in the meta-analysis was assessed using the Quality Assessment of Systematic Reviews and Meta-Analyses tool developed by the National Heart, Lung, and Blood Institute (NHLBI) [33].

Data analysis

Study characteristics were described using absolute and relative frequencies. Total cholesterol, triglycerides, HDL cholesterol, and LDL cholesterol values before and after the GFD were compared using meta-analyses of mean differences. The restricted maximum likelihood (REML) estimation method was applied.

Heterogeneity was assessed using the I² statistic, where I² < 25% indicates low heterogeneity, I² < 50% indicates moderate heterogeneity, and I² > 50% indicates high heterogeneity [34]. Due to the presence of heterogeneity in most parameters, random-effects models were employed. Publication bias was assessed using Egger’s test.

Additionally, subgroup analyses were performed to evaluate lipid profile effects separately for pediatric and adult patients.

The synthesis of results from each included study, along with corresponding confidence intervals, statistical significance, and contribution to the global meta-analysis, was visually represented using forest plots. All analyses were performed using SPSS version 30.0.

Results

During the identification process, a total of 235 articles were retrieved. After removing 57 duplicates and excluding 148 articles that did not meet the inclusion criteria, 28 articles remained for the application of exclusion criteria. Among these, two lacked full-text access, two analyzed patients with cardiovascular diseases, one did not clearly specify the number of patients included in the lipid profile comparison, three included fewer than 20 participants, one presented data identical to another already-included study, and five did not report results as mean and standard deviation. Consequently, 15 articles were included in the meta-analysis (Fig. 1).

Fig. 1.

PRISMA flow diagram showing the identification, screening, and inclusion of articles in the meta-analysis

The studies included originated from seven different countries. Italy accounted for eight records, India for two studies, and Egypt, Spain, the United Kingdom, Finland, and the United States each contributed one study (Fig. 2).

Fig. 2.

Geographic distribution of studies included in the meta-analysis. The color scale reflects the number of studies contributed by each country

Regarding patient characteristics, 65% of participants were female, most were pediatric patients, and the minimum duration of adherence to the gluten-free diet was six months. The studies by Pérez et al. [30], Ciampolini et al. [32], and Salardi et al. [23] included subgroup analyses; therefore, the data from these subpopulations were analyzed separately (Table 1).

Table 1.

Description of the characteristics of patients included in each study

First Author	Year of Publication	Sex F/M (n)	Mean Age (years)	Duration of Gluten-Free Diet	Sample Size (n)
El-Shaheed, A [22]	2018	17/33	8	≥ 1 year	50
Pérez, G [30]	2022	22/05	37.1	1 year	27
	2022	14/17	7.1	1 year	31
Salardi, S [23]	2016	ND/ND	Pediatric	1 year	201
Brar, P [11]	2006	87/45	44.4	≥ 6 months	132
Vuoristo, M [24]	1982	17/13	35	8 months	30
Forchielli, M [31]	2020	52/27	7.9	1 year	79
Zanini, B [25]	2013	504/211	35	1–5 years	715*
Capristo, E [13]	2009	17/09	ND	1 year	26
Riezzo, G [26]	2014	15/05	34	1 year	20
Siddha, L [27]	2016	20/30	< 16	6 months	50
Ciampolini, M [32]	1991	30/15	2	7 months	45
	1991	29/20	8.9	7 months	49
Lewis, N [28]	2009	ND/ND	51	1 year	100*
Salardi, S [23]	2017	ND/ND	< 6	1 year	90
	2017	ND/ND	> 6	1 year	111
De Marchi, S  [29]	2013	11/09	ND	6–8 months	20
Agarwal, A [16]	2021	26/18	29.5	1 year	44

ND No data

*In these studies, the sample size varied depending on the lipid profile parameter evaluated

Among adult patients, total cholesterol levels before the gluten-free diet ranged from 144.9 ± 27.8 mg/dL to 185.4 ± 37.8 mg/dL, while after dietary intervention they ranged from 141.5 ± 13.9 mg/dL to 204.4 ± 35.2 mg/dL. Triglyceride levels before the gluten-free diet varied between 83.2 ± 19.5 mg/dL and 136 ± 248.8 mg/dL, and after the diet ranged from 57 ± 0 mg/dL to 140.8 ± 30.1 mg/dL. HDL cholesterol levels before the gluten-free diet were between 38.9 ± 3.1 mg/dL and 52.6 ± 18.5 mg/dL, and after the diet ranged from 42.8 ± 8.7 mg/dL to 68.2 ± 17.4 mg/dL. Finally, LDL cholesterol levels before the intervention ranged from 84.3 ± 25.9 mg/dL to 130.3 ± 11.9 mg/dL, and after the gluten-free diet they ranged from 78.1 ± 13.9 mg/dL to 125.3 ± 9.3 mg/dL.

In pediatric patients, total cholesterol levels before the gluten-free diet ranged from 127.7 ± 25 mg/dL to 166.5 ± 36.2 mg/dL, and after the intervention they ranged from 119.5 ± 29.8 mg/dL to 168.5 ± 28.1 mg/dL. Triglyceride levels before the gluten-free diet varied between 64.3 ± 19.9 mg/dL and 120.4 ± 56.2 mg/dL, and after the diet ranged from 53.7 ± 22.3 mg/dL to 85.4 ± 21.2 mg/dL. HDL cholesterol levels before the gluten-free diet were between 28.5 ± 6.6 mg/dL and 64.5 ± 15.1 mg/dL, and after the diet ranged from 43.1 ± 13.1 mg/dL to 81.6 ± 19.3 mg/dL. Finally, LDL cholesterol levels before the diet ranged from 77.8 ± 30.7 mg/dL to 99.5 ± 23 mg/dL, and after the intervention ranged from 85.5 ± 21.7 mg/dL to 97.5 ± 25.6 mg/dL (Table 2).

Table 2.

Description of lipid profile values before and after the gluten-free diet reported in each study

Primer autor	Media ± DS del colesterol total (mg/dL)		Media ± DS de triglicéridos (mg/dL)		Media ± DS del HDL (mg/dL)		Media ± DS del LDL (mg/dL)
	Before	After	Before	After	Before	After	Before	After
El-Shaheed, A  [22]	153,7 ± 38.2	119.5 ± 29.8	111.7 ± 25.3	85.4 ± 21.2	64.5 ± 15.1	81.6 ± 19.3	ND	ND
Perez, G [30]	173.9 ± 35.2	154 ± 0	136.6 ± 248.8	57 ± 0	49.6 ± 9.9	ND	ND	ND
	155.6 ± 21.7	160.7 ± 46.4	68.1 ± 31.8	56 ± 23.5	50 ± 10.3	63.2 ± 17.5	ND	ND
Salardi, S [23]	ND	ND	ND	ND	51.3 ± 13.6	60.9 ± 13.7	ND	ND
Brar, P [11]	169.2 ± 35.3	188.8 ± 32.8	ND	ND	45.8 ± 13.3	55.1 ± 13.6	109.7 ± 26.6	114.5 ± 28.1
Vuoristo, M [24]	162.4 ± 38.6	201.1 ± 42.5	113.4 ± 58.5	92.9 ± 27.5	ND	ND	ND	ND
Forchielli, M [31]	152.4 ± 21.9	153.4 ± 25.7	66.4 ± 39.4	53.7 ± 22.3	53.9 ± 14.8	60.7 ± 13.4	88.4 ± 21.7	85.5 ± 21.7
Zanini, B [25]	171.2 ± 37.4	181.4 ± 35.1	87.9 ± 49.5	80.2 ± 42.8	48.1 ± 14.3	53.6 ± 14.2	102.5 ± 33.7	108.6 ± 34
Capristo, E [13]	172.1 ± 15.8	185.9 ± 10.8	135.5 ± 38.9	140.8 ± 30.11	38.9 ± 3.1	46.01 ± 4.2	130.3 ± 11.9	125.3 ± 9.3
Riezzo, G [26]	144.9 ± 27.8	141.5 ± 13.9	83.2 ± 19.5	83.2 ± 23.7	44.1 ± 5.1	46.8 ± 6.7	84.3 ± 25.9	78.1 ± 13.9
Siddha, L [27]	127.7 ± 25.0	159.2 ± 19.93	ND	ND	ND	ND	ND	ND
Ciampolini, M [32]	130 ± 32	155 ± 26	120.4 ± 56.2	80.2 ± 39.1	28.5 ± 6.6	43.1 ± 13.1	77.8 ± 30.7	94.1 ± 23.4
	148 ± 31	158 ± 27	80.4 ± 33.8	73.1 ± 38.3	36.1 ± 12.8	46.2 ± 14	96.3 ± 29.4	97.5 ± 25.6
Lewis, N [28]	187.1 ± 45.6	186.4 ± 47.5	109.8 ± 69.1	106.3 ± 61.9	52.6 ± 18.5	57.2 ± 22.4	ND	ND
Salardi, S [23]	158.3 ± 26.1	166.1 ± 30.6	64.3 ± 19.9	54.5 ± 17.1	ND	ND	99.5 ± 23	96.4 ± 27.5
	166.5 ± 32.6	168.5 ± 28.1	66.6 ± 31.9	65.5 ± 29.9	ND	ND	97.8 ± 30.9	93.4 ± 25.3
De Marchi, S [29]	185.4 ± 37.8	204.4 ± 35.2	85.9 ± 46.1	87.4 ± 45.9	51.4 ± 18.6	68.2 ± 17.4	110.7 ± 26.6	107.9 ± 22.8
Agarwal, A [16]	145.4 ± 38.7	158 ± 35.8	ND	ND	39.5 ± 10.3	42.8 ± 8.7	88.6 ± 28.4	97 ± 24

ND No data

Total cholesterol

When all patients were included in the meta-analysis of total cholesterol, substantial heterogeneity was observed among the included studies (I² = 90%), with no evidence of publication bias (Egger’s test p = 0.731). A significant increase in total cholesterol levels was found, with a mean difference of 9.6 mg/dL (95% CI 1.8 to 17.3) (Fig. 3).

Fig. 3.

Forest plot showing mean difference in total cholesterol levels before and after the gluten-free diet in patients with celiac disease

When analyzing only adult patients, the significant increase in total cholesterol persisted, with a mean difference of 5.9 mg/dL (95% CI 0.5 to 11.2). However, this difference disappeared when analyzing only studies including pediatric patients, for whom the mean difference was 6.25 mg/dL (95% CI − 7.1 to 19.6) (see Supplementary Material).

Triglycerides

Thirteen studies evaluating triglyceride levels in patients with celiac disease before and after initiation of a gluten-free diet were included. The meta-analysis of all patients showed substantial heterogeneity (I² = 72%), no evidence of publication bias (Egger’s test p = 0.802), and a significant decrease in triglyceride levels, with a pooled mean difference of − 10.7 mg/dL (95% CI − 16.8 to − 4.6) (Fig. 4).

Fig. 4.

Forest plot showing the mean difference in triglyceride levels in patients with celiac disease before and after the gluten-free diet

In contrast, subgroup analysis revealed a significant increase in triglyceride levels among adult patients, with a mean difference of 5.9 mg/dL (95% CI 0.49 to 11.2), while in the pediatric population a marked and statistically significant decrease was observed, with a pooled mean difference of − 14.2 mg/dL (95% CI − 22.6 to − 5.8) (see Supplementary Material).

HDL cholesterol

Thirteen studies assessing HDL cholesterol levels in patients with celiac disease before and after initiation of a gluten-free diet were included. The meta-analysis including all patients showed substantial heterogeneity (I² = 75%), no evidence of publication bias (Egger’s test p = 0.102), and a significant increase in HDL levels, with a pooled mean difference of 8.6 mg/dL (95% CI 6.3 to 10.9) (Fig. 5).

Fig. 5.

Forest plot showing the mean difference in HDL cholesterol levels in patients with celiac disease before and after the gluten-free diet

This significant increase in HDL levels remained consistent in subgroup analyses, both in adults (6.1 mg/dL; 95% CI 3.9 to 8.3) and in pediatric patients (11.4 mg/dL; 95% CI 8.5 to 14.2) (see Supplementary Material).

LDL cholesterol

Ten studies evaluating LDL cholesterol levels in patients with celiac disease before and after initiation of a gluten-free diet were included. When all patients were pooled in the meta-analysis, heterogeneity was moderate (I² = 51%), with no evidence of publication bias (Egger’s test p = 0.270). The meta-analysis revealed no significant changes, with a pooled mean difference of 0.57 mg/dL (95% CI − 3.3 to 4.4) (Fig. 6).

Fig. 6.

Forest plot showing the mean difference in LDL cholesterol levels in patients with celiac disease before and after the gluten-free diet

Subgroup analyses showed that the gluten-free diet did not produce significant changes in LDL cholesterol levels in either adult patients (0.95 mg/dL; 95% CI − 4.3 to 6.22) or pediatric patients (0.5 mg/dL; 95% CI − 6.1 to 7.1) (see Supplementary Material).

Methodological quality assessment

The methodological quality assessment identified three critical items. The first was related to sample size, as most studies did not specify whether a sample size calculation had been performed. The second pertained to data analysis; although most studies met the general criterion, they did not perform statistical tests for repeated measures. Finally, only two studies reported multiple measurements before and after the intervention. The remaining criteria showed satisfactory compliance, exceeding 66% (Fig. 7).

Fig. 7.

Methodological quality assessment of the studies included in the meta-analysis

Discussion

The findings of this meta-analysis, which included 1,820 patients with celiac disease, most of them women, provide a more comprehensive understanding of the impact of a gluten-free diet (GFD) on the lipid profile. In the pediatric population, a decrease in triglycerides and an increase in HDL cholesterol were observed, suggesting an early metabolic improvement following the initiation of the diet. This pattern could be explained by the rapid reversal of mucosal damage at younger ages, which facilitates the recovery of intestinal absorption and the synthesis of apolipoproteins such as Apo-AI, a key structural component of HDL [11, 13, 28]. The decrease in triglycerides may also result from an improvement in dietary quality after implementing a gluten-free diet, as the reduction in the consumption of ultra-processed foods rich in simple sugars and trans fats lowers the glycemic index and consequently hepatic lipogenesis (triglyceride production) [35, 36]. In other words, maintaining an adequate GFD with a low glycemic index is associated with favorable triglyceride and HDL levels—an indirect indicator of small intestinal functional recovery and reduced cardiovascular risk [37].

In adults, a significant increase in total cholesterol, triglycerides, and HDL was observed after implementing the GFD. This pattern has been documented in studies reporting a progressive rise in lipid fractions once intestinal atrophy is reversed and lipid absorption restored [11, 13]. However, many current forms of the GFD, particularly among adults, include products with high fat and added sugar content [38, 39]. Such dietary patterns may contribute to elevations in lipid fractions beyond physiological limits. For example, the high caloric intake and imbalance between saturated and polyunsaturated fats in many commercial gluten-free products can negatively affect lipid homeostasis, offsetting the expected benefits of the GFD. Consequently, this may increase the risk of metabolic syndrome, dyslipidemia, and long-term cardiovascular disease [18]. Therefore, it is advisable to promote personalized dietary plans that emphasize not only gluten exclusion but also a nutritionally balanced and healthy profile—favoring natural foods such as fruits, vegetables, nuts, and lean meats, while avoiding ultra-processed gluten-free products.

On the other hand, LDL cholesterol levels did not show significant modifications after implementation of the diet, either in adults or in children. This finding is consistent with the available literature, which has reported similar results in follow-up studies lasting six months to one year [20, 28]. Because LDL is less influenced by acute intestinal inflammation, its stability may reflect a partial functional recovery of lipid metabolism without an immediate negative effect on cardiovascular risk. Nevertheless, from a clinical perspective, the persistence of normal LDL levels alongside elevated triglycerides—as observed in some adults—may indicate an atherogenic lipid profile. This pattern has been associated with metabolic syndrome, characterized by increased hepatic production of triglyceride-rich lipoproteins and reduced clearance of these particles, leading to the accumulation of remnant lipoproteins with high atherogenic potential [40]. Moreover, recent epidemiological studies have identified elevated triglycerides as an independent cardiovascular risk factor comparable to LDL, particularly when accompanied by low HDL levels [41]. Hence, LDL should not be monitored in isolation but rather as part of the overall lipid profile.

These results can be explained by two hypotheses mentioned earlier. First, changes in the lipid profile may result from improved intestinal absorption. In two of the studies analyzed [16, 32], HDL and total cholesterol levels were low prior to the diet but increased afterward, reflecting a return to clinically acceptable values following the resolution of intestinal inflammation. Second, the rise in triglycerides observed in adults may be a direct consequence of the nutritional quality of the GFD, particularly when it relies on industrially processed products—especially gluten-free baked goods, cookies, and pasta [42]—which often contain excessive carbohydrates and fats. The coexistence of these two mechanisms underscores the need for a critical evaluation of the GFD.

The substantial heterogeneity observed across studies may be explained by differences in age (pediatric vs. adult populations), variability in the duration of adherence to the gluten-free diet, and differences in dietary composition, particularly the consumption of ultra-processed gluten-free products. Moreover, variability in the composition of the gluten-free diet across studies may partly explain the heterogeneity observed in lipid outcomes.

The main limitations of this meta-analysis lie in the variability and, in many cases, the short duration of follow-up, which may limit the ability to capture long-term effects of the gluten-free diet on lipid profiles. This could lead to an underestimation or overestimation of the true metabolic impact of the diet over time.

In addition, most studies did not objectively verify dietary adherence or adjust for relevant confounding factors such as body mass index, dietary composition or physical activity. This lack of control may have introduced residual confounding, potentially influencing the direction and magnitude of the observed associations.

Finally, the concentration of studies in certain geographic regions—particularly Europe—may limit the generalizability of the findings to other populations. Therefore, the results should be interpreted with caution when extrapolating to more diverse settings.

Conclusion

The findings of this study suggest that a gluten-free diet is associated with changes in the lipid profile of patients with celiac disease, although the magnitude and direction of these effects vary across studies.

In pediatric patients, a trend toward improvement was observed, characterized by increased HDL cholesterol and decreased triglycerides, which may reflect recovery of intestinal function and metabolic status following dietary adherence. However, these findings should be interpreted with caution given the variability among included studies.

In adults, increases in HDL cholesterol, triglycerides, and total cholesterol were observed. These changes may be related both to improved lipid absorption and to differences in dietary composition, particularly the consumption of ultra-processed gluten-free products.

LDL cholesterol levels showed no consistent changes across populations

Overall, the substantial heterogeneity observed in the analyses highlights the need for cautious interpretation of these results. Further well-designed longitudinal studies, considering dietary quality and potential confounding factors, are required to better understand the metabolic impact of the gluten-free diet in patients with celiac disease.

Authors’ contributions

V.L.R., M.C.M.P., C.F.G. y L.F.H.G. contributed equally to all stages of the research process, including study conception and design, literature search and selection, data extraction, statistical analysis, interpretation of results, manuscript drafting, and critical revision of the final version. All authors read and approved the final manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript for the conduct of this research.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.