Competitive season effects on polyunsaturated fatty acid content in erythrocyte membranes of female football players

ABSTRACT

Background

An optimal and correctly balanced metabolic status is essential to improve sports performance in athletes. Recent advances in omic tools, such as the lipid profile of the mature erythrocyte membranes (LPMEM), allow to have a comprehensive vision of the nutritional and metabolic status of these individuals to provide personalized recommendations for nutrients, specifically, the essential omega-3 and omega-6 fatty acids, individuating deficiencies/unbalances that can arise from both habitual diet and sportive activity. This work aimed to study the LPMEM in professional female football players during the football season for the first time and compare it with those defined as optimal values for the general population and a control group.

Methods

An observational study was carried out on female football players from the Athletic Club (Bilbao) playing in the first division of the Spanish league. Blood samples were collected at three points: at the beginning, mid-season, and end of the season for three consecutive seasons (2019–2020, 2020–2021, and 2021–2022), providing a total of 160 samples from 40 women. The LPMEM analysis was obtained by GC-FID by published method and correlated to other individual data, such as blood biochemical parameters, body composition, and age.

Results

We observed a significant increase in docosahexaenoic acid (DHA) (p 0.048) and total polyunsaturated fatty acid (PUFA) (p 0.021) in the first season. In the second season, we observed a buildup in the membrane arachidonic acid (AA) (p < .001) and PUFA (p < .001) contents when high training accumulated. In comparison with the benchmark of average population values, 69% of the football players showed lower levels of omega-6 dihomo-γ-linolenic acid (DGLA), whereas 88%, 44%, and 81% of the participants showed increased values of AA, eicosapentaenoic acid (EPA), and the ratio of saturated and monounsaturated fatty acids (SFA/MUFA), respectively. Regarding relationships between blood biochemical parameters, body composition, and age with LPMEM, we observed some mild negative correlations, such as AA and SFA/MUFA ratio with vitamin D levels (coefficient = -0.34 p = .0019 and coefficient = -.25 p = .042); DGLA with urea and cortisol (coefficient = -0.27 p < .006 and coefficient = .28 p < .0028) and AA with age (coefficient = -0.33 p < .001).

Conclusion

In conclusion, relevant variations in several fatty acids of the membrane fatty acid profile of elite female football players were observed during the competitive season and, in comparison with the general population, increased PUFA contents were confirmed, as reported in other sportive activities, together with the new aspect of DGLA diminution, an omega-6 involved in immune and anti-inflammatory responses. Our results highlight membrane lipidomics as a tool to ascertain the molecular profile of elite female football players with a potential application for future personalized nutritional strategies (diet and supplementation) to address unbalances created during the competitive season.

KEYWORDS:

• Membrane lipidome
• mature red blood cell
• female athletes
• inflammation
• exercise
• football

1. Introduction

The physiological demands are different across sports disciplines. In particular, football (soccer) is characterized as an intermittent activity with low-intensity exercise and dynamic high-intensity efforts [1]. An optimal diet maximizes physical and mental performance before, during, and after training sessions and match-plays, thus maintaining overall health through the long sportive season [2]. Specific nutritional recommendations for football players have been developed to maximize their performance and minimize injuries and illnesses [3]. Nutritional strategies, in general, are designed to tackle inflammatory and oxidative processes, being these some of the metabolic hallmarks of intense physical activity [4]. On the other hand, exercise always has a disruptive effect which implies an efficient repair of damaged cells [5] and, recently, the skeletal muscle mitochondria morphology has been related to distinct inflammatory signatures which impair physical performance [6]. Clinical and molecular research highlights the importance of inflammation control, especially in individuals at high-intensity exercise. The increased omega-6 fatty acid (FA) intake, compared to the omega-3 found in the Western diet, is a drawback in this scenario [7]. Indeed, lipids play determinant roles, no longer confined as mere energy suppliers but as fine regulators of cellular metabolism and signaling. In particular, due to their potential anti-inflammatory properties, omega-3 polyunsaturated fatty acids (PUFA) have become one of the most utilized food supplements in sports nutrition [8]. Long-chain omega-3 PUFA, eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) are key molecular factors in maintaining human health with multiple beneficial activities, such as (i) balancing the pro-inflammatory activity of the omega-6 arachidonic acid (AA); (ii) being precursors of protective molecules such as resolvins and neuroprotectins [9]; (iii) regulating the immune system [10]; (iv) maintaining muscle and increasing muscle protein synthesis (MPS) [11]; and (v) contributing to cardiovascular health [12]. In athletes, they also provide some benefits in (vi) skeletal muscle recovery [13]; (vii) cognition and mood [13]; (viii) attenuating pro-inflammatory cell responses; and (ix) increasing lipid peroxidation and nitric oxide post-exercise [13], so it is relevant to have an optimal omega-3 FA status for health and performance [14]. Humans obtain long-chain omega-3 FA, mainly EPA and DHA, from marine sources such as seafood and algae; however, before any supplementation is given, it is advisable to know their baseline levels. Alpha-linolenic acid, from plant origin, is the essential fatty acid (EFA) omega-3 precursor, but its conversion to EPA and DHA is poor in humans [15,16].

Recent advances in molecular technologies, such as metabolomics, genetics, and metagenomics, have provided new scientific evidence related to phenotypes in connection with health conditions, in response to diet and metabolism, which is very useful for developing more precise nutritional recommendations [17–19]. Amongst them, cell membrane lipidomics represents a comprehensive approach to evidence lipid status in a fundamental cellular compartment for life, which reflects nutritional needs as shown in different population groups, i.e. in cancer [20] and obesity [21], with a high metabolic significance in combination with human clinical evaluations [22,23]. There are several reasons why knowledge of the composition of cell membranes becomes important: (a) PUFAs are incorporated in membrane phospholipids and their release by the action of phospholipase A2 (PLA2) starts lipid signaling with the formation of eicosanoids and other lipid mediators (i.e. the mitogenic prostaglandin E2) [24] and (b) phospholipid FA residues are involved in the naturally occurring process of membrane remodeling, known as Lands' cycle [25]. In such step, the cellular FA pool, resulting from dietary and metabolic contributions, becomes a relevant determinant, inspiring “membrane lipid therapy” as a tool for controlling the membrane lipidome properties and functions during diseases and aging [26]. Erythrocyte membranes are considered an ideal site for FA evaluation, not only for their easy access but mainly for their significance resulting from genetic, metabolic, and dietary factors and well-developed analytical methodologies. The erythrocyte membrane FA composition gives information of all FA families (saturated, monounsaturated, and polyunsaturated fatty acids – SFA, MUFA, PUFA). It represents other body tissues, thus furnishing a comprehensive biomarker of the homeostatic status of an individual [27]. Several research groups have studied the effect of lipid differences in athletes for different disciplines and blood fractions, such as in plasma, skeletal muscle, and erythrocytes [28–34], to compare FA content with sedentary women, to determine if there is a need for nutritional intervention, to observe changes in the FA composition in response to an intensive exercise, etc.

It is well known that women and men have different physiological demands and nutritional needs [35–37]. For instance, it is common to find deficiencies in women in some nutrients such as iron, which is caused by its loss during the menstruation period, or calcium and vitamin D, involving a loss of bone mass and an increased risk of fracture. There are a few nutritional studies involving female football players [36,38–40], especially regarding fat [38–40]. Therefore, investigation of molecular aspects connected to nutritional intakes and metabolic conversions in such women athletes can furnish important information to translate into more precise food recommendations, improving personalized strategies.

This study aims to characterize the LPMEM in professional female football players during the sportive season and compare it with those defined as optimal values for the general population and with a control group. This is the first time LPMEM is monitored during the competitive season, providing the molecular basis for future design of personalized nutritional strategies.

2. Material and methods

2.1. Subjects and study design

The observational study was carried out on 40 female football players from the Athletic Club (Bilbao, Spain), who competed in the first division of the Spanish league during three football seasons (2019–2020, 2020–2021, and 2021–2022). For each season, three samplings were carried out: at the beginning of the season between July and September (T1), in the middle of the season between December and March (T2) and at the end of May (T3), giving a total of 160 samples. We had three timings for season 20–21, whereas the other two had only two timings (T1 and T2). All subjects were healthy. We also had access to general data, such as age, playing position and anthropometric measurements, including body fat percentage (BF%). Recruitment was conducted through the medical services of the Athletic Club in Lezama, Spain. The inclusion criteria were females competing in the first division of the Spanish league that signed the informed consent to participate. Some had supplementation for iron, and vitamin D.

A control group including 44 Spanish women from a previously published study by our group [41] was used for comparisons. This control group was selected considering the following criteria: absence of disease and not taking drugs or supplements, no intense physical activity according to the IPAQ questionnaire [42], normal weight, and matched for age (19–35 years old). This data set included data on the LPMEM measured with the same methodology.

The study protocol was approved by the Euskadi Clinical Research Ethics Committee (TUE-ATH-2018-02) and accomplished according to the Helsinki Declaration in 1975, revised in 2013. Written informed consent was obtained from all athletes.

2.2. Lipid profile of the mature erythrocyte membranes analysis

After an overnight fast, venous blood samples (approximately 2 mL) were collected in vacutainer tubes containing ethylenediaminetetraacetic acid (EDTA). Samples were stored at 4°C until analysis. An automated protocol was performed for blood lipid extraction and lipid transesterification to fatty acid methyl esters (FAMEs), which included a selection of mature red blood cells, as reported previously [43–45]. Briefly, the EDTA blood was centrifuged (4000 rpm for 5 min at 4°C) and the cell fraction was isolated based on the high density of the aged cells [46] and controlled by the use of a cell counter (Scepter 2.0 with Scepter™ Software Pro, EMD Millipore, Darmstadt, Germany). All the subsequent steps were automated, including cell lysis, isolation of the membrane pellets, phospholipid extraction from pellets using the Bligh and Dyer method [47], transesterification to FAMEs by treatment with a potassium hydroxide/methyl alcohol solution (0.5 mol/L) for 10 min at room temperature and extraction using n-hexane (2 mL) as described in the previous work [48]. Finally, the FAMEs were analyzed using capillary column gas chromatography on the Agilent 6850 Network GC System (Agilent, USA), equipped with a fused silica capillary column Agilent DB23 (60 m × 0.25 mm × 0.25 μm) and a flame ionization detector. Optimal separation of all FA and their geometrical and positional isomers was achieved following published methods [44,45]. Identification was made by comparing them to commercially available standards and a library of trans isomers of MUFAs and PUFAs. The amount of each FA was calculated as a percentage of the total FA content (relative %), being more than 97% of the GC peaks recognized with appropriate standards.

2.3. Lipid profile of the mature erythrocyte membranes cluster

Twelve FAs were selected for analyses: for SFAs, palmitic acid (C16:0), and stearic acid (C18:0); for MUFAs, palmitoleic acid (C16:1; 9c), oleic acid (C18:1; 9c) and cis-vaccenic acid (C18:1; 11c); for omega-3 PUFAs, EPA (C20:5), DHA (C22:6); for omega-6 PUFAs, linoleic acid (LA) (C18:2), dihomo-gamma-linolenic acid (DGLA) (C20:3), and AA (C20:4); for trans-fatty acids (TFAs), elaidic acid (C18:1; 9t), and mono-trans arachidonic acid isomers (monotrans-C20:4; omega-6). Based on these FA, different indexes previously reported in the literature [20] were calculated: (%SFA/%MUFA) index related to membrane fluidity or saturation index (SI); Omega-3 index (DHA + EPA), an index suggested as a cardiovascular disease risk factor; inflammatory risk index (%omega-6)/(%omega-3); PUFA balance [(%EPA + %DHA)/total PUFA × 100]; free radical stress index (sum of trans-18:1 + summary (Σ) of monotrans 20:4 isomers); unsaturation Index (UI) [(%MUFA × 1) + (%LA × 2) + (%DGLA × 3) + (%AA × 4) + (% EPA × 5) + (%DHA × 6)] and peroxidation Index (PI) [(%MUFA × 0.025) + (%LA × 1) + (%DGLA × 2) + (%AA × 4) + (% EPA × 6) + (%DHA × 8)]. Additionally, the enzymatic indexes of elongase and desaturase enzymes, the two classes of enzymes of the MUFA and PUFA biosynthetic pathways, were estimated by calculating the product/precursor ratio of the FAs involved in these reactions. Optimal value intervals for each of the 12 FA were based on the literature [27].

2.4. Blood biochemical parameters

Fasting blood biomarkers were measured using standard laboratory assays on the same day venous blood was collected for LPMEM analysis. Serum concentrations of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), RBC, hemoglobin, and hematocrit, iron, ferritin, transferrin, white blood cells, aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma GT, uric acid, urea, cortisol, testosterone, creatine kinase (CK), creatinine, and vitamin D were considered.

2.5. Anthropometric measurements

Anthropometric measurements were performed before breakfast at Athletic Club training facilities. Body weight (Seca, Germany) and height (Seca, Germany) were assessed to the nearest 0.1 kg and 0.1 cm, following the manufacturer’s guidelines.

Six skinfolds (subscapular, triceps, supraspinal, abdominal, thigh, and calf) were measured using a skinfold caliper (Holtain, Crosswell, UK) to the nearest 0.2 mm, following the International Society of Anthropometric of Kinanthropometry (ISAK) protocol. Three measurements of each fold were performed using the mean value in mm. All measurements were performed by a qualified staff experienced tester. We used Yuhasz’s formula to calculate BF%.

2.6. Statistical analysis

Statistical analysis was carried out using RStudio (version 4.2.1). Before beginning the analysis, extreme outliers were removed for each FA variable and blood biochemical parameters. After that, normality was tested using the Shapiro-Wilks; since all variables were not normally distributed, results are expressed as means ± standard deviations (SD) and median and confidence intervals. Statistical comparisons of means were performed using Student’s t-test and one-way Analysis of Variance (ANOVA) for normally distributed variables, and Mann-Whitney U test and Kruskal Wallis for not normal variables. When statistical analyses revealed a significant effect, Bonferroni adjustments to the p-values were performed to identify differences between the three study times. The level of significance was set at p < .05. In addition, Spearman’s correlations were used to search for correlations between variables.

3. Results

3.1. Subject characteristics

Table 1 shows the demographic, playing positions, and anthropometric characteristics of the study population for each timing (T1, T2, T3) within the three competitive seasons. As most variables were not normally distributed, data are reported as median and confidence interval. The three competitive seasons had no differences in BMI and body fat composition. For the skinfolds, there were only statistically significant differences in season 2020–2021 in two out of six skinfolds, subscapular (increasing along the time point), and triceps (decreasing along the time point).

Table 1. Characteristics of the participants for each sampling point (beginning point T1, in the middle of the season T2, and at the end T3).

	SEASON 2019–2020			SEASON 2020–2021				SEASON 2021–2022
	T1 Oct 2019 n = 23	T2 Dec 2020 n = 22	p-Value	T1 Jul 2020 n = 23	T2 Dec 2020 n = 25	T3 May 2021 n = 23	p-Value	T1 Jul 2021 n = 20	T2 March 2022 n = 24	p-Value
Age	24.4 [21.5;28.3]	24.8 [21.6;28.5]	.633	24.7 [21.7;28.7]	24.7 [22.1;28.8]	23.8 [21.3;27.8]	.918	26.2 [21.5;29.9]	24.5 [21.6;30.2]	.741
Playing position
Forward	11	11		11	12	11		12	16
Defender	9	8		9	10	10		5	6
Goalkeepers	3	3		3	3	2		3	2
Body composition
Weight (kg) *	58.8 [55.3;62.0]	59.8 [55.8;63.1]	0.808	60.1 [57.1;61.8]	59.4 [56.4;62.0]	59.3 [57.2;63.8]	0.995	60.0 [57.1;64.5]	60.1 [58.6;65.0]	0.494
BMI (kg/m^2)	21.5 [20.1;22.0]	21.7 [20.1;22.3]	0.576	21.7 [20.5;22.7]	21.6 [20.7;22.2]	21.5 [20.2;22.3]	0.97	20.9 [20.2;23.3]	21.9 [21.0;23.2]	0.220
%BF *	12.7 [11.5;13.8]	12.6 [11.2;13.5]	0.780	13.9 [12.3;15.5]	14.0 [12.2;15.1]	12.9 [12.0;14.2]	0.421	13.0 [11.2;17.5]	13.3 [11.9;15.2]	0.925
Σ skinfolds (mm)	59.1 [50.9;66.3]	58.4 [49.0;64.4]	0.780	66.6 [56.7;77.0]	67.6 [55.4;74.6]	60.4 [54.7;68.5]	0.421	60.7 [49.5;89.8]	62.6 [53.7;75.3]	0.925
Subscapular *	8.2 [7.1;9.8]	7.2 [6.6;9.0]	0.165	8.5 [7.00;10.1]	9.4 [7.0;11.4]	10.2 [9.1;12.6]	0.038	8.4 [7.6;11.2]	7.1 [6.9;9.3]	0.193
Triceps *	9.3 [8.1;11.2]	11.2 [8.4;12.4]	0.401	10.0 [8.5;11.6]	9.8 [7.8;12.2]	7.2 [6.8;8.3]	0.004	9.9 [8.6;13.2]	11.4 [8.8;13.7]	0.350
Supraspinal *	6.8 [5.8;8.4]	6.6 [5.6;8.0]	0.627	6.8 [5.0;8.7]	6.6 [5.4;8.0]	6.8 [5.5;7.2]	0.644	7.0 [5.0;10.0]	6.2 [4.9;7.8]	0.202
Abdominal *	11.0 [7.9;12.5]	9.8 [7.8;11.4]	.0375	11.9 [9.1;14.8]	11.3 [9.8;15.2]	11.4 [8.9;13.2]	0.554	11.1 [8.6;15.8]	10.9 [7.9;13.8]	0.636
Thigh *	16.0 [14.0;18.4]	16.0 [13.8;17.4]	0.817	19.9 [15.8;24.2]	19.8 [15.2;22.0]	17.0 [14.9;19.6]	0.229	16.8 [14.1;23.8]	18.1 [15.0;23.0]	0.636
Calf *	5.9 [5.0;9.0]	5.4 [5.0;8.4]	0.913	8.2 [6.1;9.8]	8.6 [7.2;10.2]	8.4 [6.2;9.3]	0.604	7.9 [5.8;9.9]	7.8 [5.9;11.3]	0.542

Data are expressed with median and confidence intervals. *Not normally distributed variables.

3.2. Blood biochemical parameters

Blood biochemical parameters determined in each season and timing are summarized in Supplementary Material 1. Each variable is reported as a median and confidence interval. In the 2019–2020 competitive season, we observed a significant decrease in mean corpuscular hemoglobin and mean corpuscular hemoglobin concentration levels, and an increase in hematocrit; however, these values are within the optimal values and this effect has no physiological relevance. In the 2020–2021 season, urea, LDL-C, mean corpuscular hemoglobin concentration, ALT, AST and CK significantly increased oversampling points. Nonetheless, hematocrit, iron, transferrin saturation index and vitamin D significantly decreased across sampling times. Finally, in the 2021–2022 season, we observed a significant decrease in uric acid and CK.

3.3. Lipid profile of mature erythrocyte membrane

The mature erythrocyte membrane FA profile of all participants expressed as the relative percentage of the cluster, the evolution through the different sampling points of each competitive season is reported in Table 2 Mature erythrocyte membrane fatty acid levels in professional female football players for each sampling point of each competitive season (beginning point T1, in the middle of the season T2 and at the end T3). We could observe that in the season 2019–2020, omega-3 DHA, total PUFA and unsaturation indexes showed significant increases over time. Nevertheless, stearic acid, vaccenic acid, Trans 18:1, total SFA, and total TFA showed significant decreases. In the second season, 2020–2021, there were significant increases in omega-6 AA, total PUFA, and peroxidation and unsaturation indexes when comparing the values along the sampling times. However, we observed significant decreases in palmitic acid, palmitoleic acid, and total SFA. In the last season, 2021–2022, only Trans 18:1 and total TFA experienced significant decreases.

Table 2. Mature erythrocyte membrane fatty acid levels in professional female football players for each sampling point of each competitive season (beginning point T1, in the middle of the season T2 and at the end T3).

	SEASON 2019–2020			SEASON 2020–2021				SEASON 2021–2022
	T1 Oct 2019 n = 23	T2 Dec 2020 n = 22	p-Value	T1 Jul 2020 n = 23	T2 Dec 2020 n = 25	T3 May 2021 n = 23	p-Value	T1 Jul 2021 n = 20	T2 March 2022 n = 24	p-Value
Palmitic acid (16:0)	22.4 ± 1.1	22.5 ± 1.3	0.639	23.1 ± 0.9	22.2 ± 1.2	21.4 ± 1.0	0<.001	21.9 ± 1.3	21.6 ± 1.0	0.294
Stearic acid (18:0) *	16.9 ± 1.2	16.1 ± 1.2	0.027	17.7 ± 1.1	17.5 ± 0.9	17.9 ± 0.9	0.404	17.5 ± 0.8	17.9 ± 1.1	0.229
Palmitoleic acid (16:1 n-7) *	0.34 ± 0.14	0.42 ± 0.15	0.095	0.39 ± 0.13	0.26 ± 0.12	0.27 ± 0.09	0<.001	0.28 ± 0.08	0.26 ± 0.05	0.604
Oleic acid (18:1 n-9) *	17.2 ± 1.3	17.4 ± 1.5	0.689	16.9 ± 1.0	16.7 ± 1.0	16.7 ± 1.3	0.629	17.2 ± 1.4	16.5 ± 0.8	0.103
cis-Vaccenic acid (18:1 n-11) *	1.38 ± 0.22	1.25 ± 0.18	0.027	1.48 ± 0.48	1.10 ± 0.14	1.20 ± 0.17	0<.001	1.13 ± 0.15	1.13 ± 0.13	0.869
LA (18:2 n-6)	14.0 ± 1.5	14.3 ± 1.5	0.459	13.6 ± 1.5	14.0 ± 1.3	13.6 ± 1.2	0.433	13.0 ± 1.1	13.8 ± 1.3	0.053
DGLA (20:3) *	1.64 ± 0.32	1.83 ± 0.3	0.097	1.73 ± 0.36	1.88 ± 0.38	1.79 ± 0.36	0.277	1.86 ± 0.34	1.73 ± 0.28	0.194
AA (20:4) *	18.1 ± 1.2	17.9 ± 1.3	0.621	18.0 ± 1.2	19.2 ± 1.3	19.5 ± 1.4	0<.001	19.1 ± 1.8	18.8 ± 1.2	0.561
EPA (20:5) *	0.83 ± 0.34	0.88 ± 0.33	0.644	0.96 ± 0.37	0.84 ± 0.37	0.95 ± 0.34	0.502	0.95 ± 0.25	0.93 ± 0.25	0.924
DHA (22:6)	6.5 ± 0.8	7.0 ± 0.9	0.048	5.9 ± 0.8	6.1 ± 1.0	6.5 ± 1.0	0.146	6.8 ± 1.0	7.2 ± 0.9	0.23
Trans 18:1 *	0.11 ± 0.04	0.08 ± 0.04	0.027	0.09 ± 0.07	0.10 ± 0.05	0.07 ± 0.02	0.404	0.09 ± 0.04	0.07 ± 0.01	0.034
Trans 20:4 *	0.14 ± 0.05	0.11 ± 0.05	0.051	0.11 ± 0.07	0.13 ± 0.05	0.11 ± 0.04	0.175	0.10 ± 0.03	0.08 ± 0.01	0.278
Total Trans *	0.26 ± 0.10	0.19 ± 0.08	0.008	0.22 ± 0.14	0.23 ± 0.08	0.18 ± 0.05	0.254	0.19 ± 0.06	0.15 ± 0.02	0.011
Total SFA	39.3 ± 1.0	38.6 ± 0.9	0.023	40.8 ± 1.1	39.7 ± 0.7	39.3 ± 0.8	0<.001	39.4 ± 1.2	39.5 ± 0.9	0.950
Total MUFA *	18.9 ± 1.4	19.1 ± 1.6	0.742	18.8 ± 1.1	18.1 ± 1.0	18.2 ± 1.3	0.052	18.6 ± 1.4	17.9 ± 0.9	0.065
Total PUFA *	41.2 ± 1.4	42.1 ± 1.1	0.021	40.2 ± 1.3	42.1 ± 0.9	42.3 ± 1.3	0<.001	41.8 ± 1.8	42.5 ± 1.0	0.111
SFA/MUFA	2.1 ± 0.2	2.0 ± 0.2	0.269	2.2 ± 0.2	2.2 ± 0.1	2.2 ± 0.2	0.91	2.1 ± 0.2	2.2 ± 0.1	0.081
ω-6/ω-3 *	4.7 ± 0.8	4.4 ± 0.8	0.211	4.9 ± 0.9	5.2 ± 1.1	4.9 ± 1.1	0.491	4.5 ± 0.9	4.3 ± 0.8	0.45
Omega-3 index	7.4 ± 1.2	8.0 ± 1.2	0.115	6.9 ± 1.0	7.0 ± 1.2	7.4 ± 1.2	0.232	7.8 ± 1.2	8.1 ± 1.1	0.326
PUFA Balance	18.0 ± 2.9	19.0 ± 2.9	0.267	17.2 ± 2.9	16.5 ± 2.8	17.6 ± 2.8	0.427	18.6 ± 2.6	19.1 ± 2.6	0.528
Peroxidation index	168 ± 4	172 ± 4	0.055	164 ± 3	169 ± 5	172 ± 6	0<.001	172 ± 7	174 ± 5	0.43
Unsaturation index	148 ± 7	152 ± 7	0.008	143 ± 5	149 ± 8	153 ± 9	0<.001	154 ± 10	156 ± 7	0.463
Δ6D+ELO 18:2/20:3	8.8 ± 1.9	8.1 ± 1.5	0.158	8.1 ± 1.7	7.7 ± 1.5	7.9 ± 1.7	0.684	7.3 ± 1.5	8.2 ± 1.4	0.056
Δ5D 20:4/20:3	11.4 ± 2.3	10.1 ± 1.8	0.039	10.8 ± 2.0	10.6 ± 2.3	11.2 ± 2.1	0.564	10.6 ± 1.7	11.1 ± 1.9	0.292
Δ9D 18:0/18:1 *	0.98 ± 0.14	0.93 ± 0.12	0.117	1.05 ± 0.09	1.05 ± 0.09	1.07 ± 0.11	0.761	1.02 ± 0.08	1.09 ± 0.09	0.017
Δ9D 16:0/16:1 *	72.6 ± 26.6	59.9 ± 21.0	0.110	66.4 ± 25.5	96.4 ± 30.4	88.5 ± 32.5	0<.001	86.4 ± 26.8	84.9 ± 13.4	0.777

Data are expressed with means ± standard deviation. *Not normally distributed variables. ELO (Elongase).

The comparison between female football players and the control group is shown in Table 3 The mature erythrocyte membrane fatty acid profile of the participants compared to the control group and optimal values. In Supplementary Material 2 and Supplementary Material 3, the sample distribution for each FA with the optimal values is plotted as box plots. In relation to the difference between players and the control group, we observed a significant increase in LA, EPA, DHA, Trans 18:1 and 20:1, total PUFA, and omega-3 index. Female football players had significantly lower levels of stearic acid, DGLA, total SFA, and omega-6/omega-3 ratio. We also compared the beginning point, T1, with the control group, as the beginning point was taken after a rest where they did not have high-intensity training sessions. Players significantly increased oleic acid, vaccenic acid, LA, EPA, Trans 18:1, total MUFA and Omega-3 index. However, we found a significant decrease in DGLA in female players. Concerning the optimal ranges defined for the general population, we observed that the mean values in the football players for omega-6 DGLA, AA, Trans 20:4, EPA, and the ratio SFA/MUFA were out of the range. About 69% of the women had lower values for both DGLA and Trans 20:4; however, 88%, 44%, and 8% of the participants showed increased values in AA, EPA, and ratio SFA/MUFA, respectively. When comparing with the control group, we found statistically significant differences in football players for lower levels of DGLA and higher levels of LA, EPA, DHA, Trans 18:1, and Omega-3 index.

Table 3. The mature erythrocyte membrane fatty acid profile of the participants compared to the control group and optimal values.

Fatty Acids (%)	Optimal values range ^#	Control group n = 44	Football players 160 samples	p-Value^+	Football player T1 66 samples	p-Value^++
Palmitic acid (16:0) *	17–27	22.1 ± 1.2	22.1 ± 1.2	0.928	22.5 ± 1.18	0.151
Stearic acid (18:0)	13–20	17.8 ± 1.6	17.4 ± 1.2	0.119	17.4 ± 1.13	0.160
Palmitoleic acid (16:1 n-7) *	0.2–.5	0.31 ± .17	0.32 ± .13	0.915	0.34 ± .13	0.414
Oleic acid (18:1 n-9) *	9–18	16.5 ± 1.4	16.9 ± 1.2	0.068	17.1 ± 1.24	0.028
cis-Vaccenic acid (18:1 n-11) *	0.7–1.3	1.23 ± .19	1.24 ± .27	0.754	1.34 ± .35	0.029
LA (18:2 n-6)	9–16	12.8 ± 1.1	13.8 ± 1.4	0<.001	13.6 ± 1.41	0.002
DGLA (20:3) *	1.9–2.4	1.98 ± .58	1.78 ± .35	0.036	1.74 ± .35	0.016
AA (20:4) *	13–17	19.0 ± 1.7	18.7 ± 1.4	0.324	18.4 ± 1.44	0.073
EPA (20:5) *	0.5–.9	0.76 ± .34	0.91 ± .32	0.018	0.91 ± .33	0.027
DHA (22:6)	5–7	5.9 ± 1.3	6.5 ± 1.0	0.006	6.41 ± .94	0.057
Trans 18:1 *	0.1–.3	0.09 ± .06	0.11 ± .05	0.030	0.12 ± .06	0.018
Trans 20:4 *	0.1–.4	0.10 ± .09	0.09 ± .04	0.486	0.10 ± .05	0.935
Total Trans *	0.1–.4	0.18 ± .13	0.20 ± .09	0.445	0.22 ± .11	0.120
Total SFA	34–45	4.1 ± 1.9	39.5 ± 1.1	0.061	39.9 ± 1.27	0.493
Total MUFA *	15–23	18.2 ± 1.5	18.5 ± 1.3	0.250	18.8 ± 1.29	0.045
Total PUFA *	30–43	41.5 ± 2.1	41.7 ± 1.4	0.433	41.0 ± 1.63	0.231
SFA/MUFA	1.7–2	2.2 ± .2	2.1 ± .2	0.121	2.14 ± .17	0.087
ω-6/ω-3 *	3.5–5.5	4.5 ± 1.4	4.7 ± .9	0.335	4.71 ± .88	0.359
Omega-3 index	0–4% high risk4–8% intermediate risk > 8% minimum risk 4–8% intermediate risk > 8% minimum risk	6.7 ± 1.5	7.5 ± 1.2	0.002	7.35 ± 1.17	0.017
PUFA Balance		16.1 ± 3.5	18.0 ± 2.9	0.003	17.9 ± 2.83	0.006
Peroxidation index		166 ± 7	170 ± 6	0<.001	148 ± 9.09	0.333
Unsaturation index		146 ± 9	151 ± 9	0.006	168 ± 6.49	0.120
Δ6D+ELO 18:2/20:3	<5 hyperactivity > 8 hypoactivity > 8 hypoactivity	6.8 ± 1.8	8.0 ± 1.7	0<.001	8.11 ± 1.84	0<.001
Δ5D 20:4/20:3	<6 hypoactivity > 9 hyperactivity > 9 hyperactivity	1.0 ± 2.3	1.8 ± 2.0	0.030	1.9 ± 2.03	0.030
Δ9D 18:0/18:1 *	<.7 hyperactivity > 1.3 hypoactivity > 1.3 hypoactivity	1.09 ± .14	1.03 ± .11	0.021	1.02 ± .11	0.011
Δ9D 16:0/16:1 *	<45 hyperactivity > 132 hypoactivity > 132 hypoactivity	85.6 ± 4.6	79.6 ± 28.2	0.362	74.7 ± 27.2	0.124

Data are expressed with means ± standard deviation. *Not normally distributed variables. # interval values obtained in general population [27]. ⁺p-Value is between female football players and control group. ⁺⁺p-Value is between beginning point of female football players and control group. ELO (Elongase).

Regarding the relationships between blood biochemical parameters, body composition, and age with LPMEM, AA, and SFA/MUFA ratio showed mild negative correlations with vitamin D levels (coefficient = −0.34 p = .0019 and coefficient = −0.25 p = .042, respectively). Moreover, DGLA also displayed mild correlations with urea and cortisol (coefficient = −0.27 p < .006 and coefficient = .28 p < .0028) (see Supplementary Material 4). On the other hand, we observed mild correlations between AA and age (coefficient = −0.33 p < .001) (see Supplementary Material 5).

3.4. Comparisons according to sampling points

We carried out three sampling points, T1 (at the beginning of the season), T2 (in the middle of the season), and T3 (at the end of the season), where we analyzed blood biochemical parameters, Supplementary Material 1 Biochemical values measured in plasma for each sampling point of each competitive season (beginning point T1, in the middle of the season T2 and at the end of T3). Data are expressed with median and confidence intervals. *Not normally distributed variables., body composition, Table 1 Characteristics of the participants for each sampling point (beginning point T1, in the middle of the season T2 and at the end T3), and LPMEM, Table 2 Mature erythrocyte membrane fatty acid levels in professional female football players for each sampling point of each competitive season (beginning point T1, in the middle of the season T2 and at the end T3).

In relation to body composition, there were no statistically significant differences between sampling points, except for the subscapular and triceps skinfolds. The subscapular fold had a statistically significant increase and the triceps significantly decreased. According to the LPMEM, we showed a significant increase in total PUFA during each sampling point, an increase, not statistically significant, of DHA and Omega-3 index, and a significant decrease in total SFA.

3.5. Comparisons according to playing position

When comparing any variable with the playing position, we removed goalkeepers as the sample size was too small. In the forward position, we included left and right-wingers, center-forward and second forward and in the defender position, we included attacking, central and defensive midfielders, full-backs, and central defenders. In Supplementary Material 6 Characteristics of the playing position for each competitive season. Data are expressed with median and confidence intervals. *Not normally distributed variables. We found significant differences in BF% and Σ skinfolds, with the lowest values found within the forwards compared to defenders in the three seasons are displayed. Regarding the differences in lipid profile, we found in the first season, 2019–2020, more statistically significant differences, where the forward group had higher values of vaccenic acid, total TFA, and trans 20:4 (Supplementary Material 7). Moreover, in terms of biochemical parameters, in season 2019–2020 forwards had statistically lower values for mean corpuscular hemoglobin, mean corpuscular volume, gamma GT, and vitamin D. In the second season, 2020–2021, there were statistically significant differences in mean corpuscular hemoglobin, ferritin, lymphocytes, neutrophils and testosterone, having defenders lower values in neutrophils (Supplementary Material 8 Biochemical values measured in plasma according to the playing position for each competitive season. Data are expressed with median and confidence intervals. *Not normally distributed variables.). In the last season, 2021–2022, the defenders’ group had statistically significant higher values in lymphocytes and CK; however, forwards had statistically significant higher values in neutrophils.

4. Discussion

This work aimed to characterize the lipid profile of the mature erythrocyte membranes (LPMEM) in elite female football players during the football season for the first time and compare it with those defined as optimal values for the general population and with a control group.

There are some studies about the FA profile in the erythrocytes of athletes [28–34]. However, to our knowledge, this is the first time that the lipidomic profile has been applied in mature erythrocytes to study female football players and measured three times during the competitive football season. Our study showed that athletes differed in their lipid profiles comparing the sampling points within the same competitive season and that might be due to their exposure to high-intensity activity that might be accumulated along the competitive season, as Tepsic et al. suggested in their work [29]. It has been described that a rapid change in erythrocyte FA composition occurs after acute exercise [34,49], depending on the intensity and sport discipline [29]. In our study, overall, we observed a decrease in total SFA and an increase in total PUFA during the competitive season.

Regarding PUFAS, the main differences are due to changes in AA, DHA, and EPA.

Regarding AA, it is a promising biomarker for inflammation. According to Lands' cycle, inflammation processes start from the release of AA from the cell membrane phospholipids [50] to be a precursor for the formation of inflammation mediators such as leukotrienes and prostaglandins. In these female players, most of them (88%) had AA values (mean 18.7%) above the upper range for the optimal values in the general population (13–17%). This can indicate the existence of a cellular stimulation to produce proliferative mediators of inflammation or the consumption of a diet rich in omega-6 [51]. Although we do not have data on the individual food intake, we do know the general food patterns in the sports facilities provided by the nutritionist staff. They do not consume oils rich in omega-6 (such as sunflower oil), as they only use olive oil and adequate intake of other sources of AA, such as eggs and red meat. LA is an essential FA to be taken from foods, and its presence in erythrocyte membranes is a promising biomarker of such intake. Considering that it is a precursor of the omega-6 cascade and AA formation, and that LA levels in erythrocytes are within optimal ranges, we can say that the observed buildup of AA levels is connected to metabolic requests from inflammatory processes, not to dietary habits. In sportive activity, it is important to consider an inflammatory effect associated with tissue resistance. It is worth mentioning that AA, a pro-inflammatory FA, suffered a significant increase during the season (p = 0.007). Such an increase at the end of the season can be explained by the accumulation of high training loads and minutes playing matches and must be balanced with the role of omega-3. Total PUFA experienced a significant increase (p < .001) through sampling points and 37.5% of female players had higher values than the upper range for the optimal values in the general population (30–43%). Although our participants’ omega-6/omega-3 ratios were within range, it should be noted that values from Tepsic et al. [29,30] and Arsić et al. [28] were higher. However, a study with 24 athletes from the Olympic Training Centre of Sant Cugat del Vallés (Spain) [33] had similar values to ours. As this ratio is mainly associated with the diet and Tepsic et al. and Arsić et al. took place with Serbian athletes, their values could be explained due to their dietary habits, high intake of omega-6-rich oils, and low consumption of fish leading to a higher omega-6/omega-3 ratio [28].

Regarding omega-3 FA, EPA mean value is above the upper range for the optimal values, and 32.5% of female football players had greater values for DHA. Notwithstanding, our values are far away from other studies. For instance, basketball and football men players from Tepsic et al. obtained 3.72 ± 1.10 and 4.02 ± 0.76 for DHA and 0.34 ± 0.13 and 0.29 ± 0.18 for EPA [29]. Arsić et al. female water polo and football groups had DHA 3.54 ± 0.64 and 3.11 ± 0.83 and EPA 0.20 ± 0.05 and 0.32 ± 0.18 [28]. Likewise, Von Schacky et al. [32] showed 0.63 ± 0.23 for EPA and 4.34 ± 1.07 values for DHA. In a cross-sectional study where omega-3 FA in 404 National Collegiate Athletic Association (NCAA) was determined, football players obtained lower values, especially in DHA (2.30 ± 0.60) and EPA (0.38 ± 0.13) values [31]. These dissimilarities might be due to different analytical techniques for the FA measurement or the different dietary habits according to the country. PUFA in erythrocytes is a good reporter for long-term dietary habits and Stark et al. [52] showed substantial differences in the sum of EPA and DHA between the USA and Spain, with Spain having higher values. The control group showed good EPA and DHA levels, in accordance with the fish intake reported through FFQ [41]. Regarding the high EPA values found in the football players compared to the controls and the optimal values, considering that there was no difference in the fish intake in both groups, we can attribute this difference to metabolic issues. Intense exercise originates a significant series of inflammatory cascades which cause muscle damage and tissue injury [53,54]. EPA might be increased in response to that inflammatory response as it forms E-series resolvins with anti-inflammatory and pro-resolvins properties [9]. Moreover, EPA, instead of DHA, upregulates muscle protein synthesis (MPS) through the mechanistic target of rapamycin complex 1 (mTORC1), facilitating muscle protein turnover when athletes cannot consume, or they do not tolerate an adequate protein serving as recovery, post-exercise [55].

Regarding total SFA, the decrease during the competitive season is mainly related to stearic acid. As this FA acid level is within the optimal values, the general reduction in total SFA might be explained as a consequence of the increase of PUFA, as these FAs are measured as a percentage of the erythrocyte membrane.

When comparing to the optimal values for the general population, in our study population, palmitic acid, stearic acid, LA, total SFA, and total MUFA levels were within the optimal value range (Table 3 The mature erythrocyte membrane fatty acid profile of the participants compared to the control group and optimal values.). Nevertheless, we observed some variations in the levels of omega-6 FAs. Remarkably, 69% of the participants had low values of DGLA, while in the control group, 48% had a deficit in DGLA. In addition, these low values continue to be lower than the optimal range at T1, which came from a rest period. Von Schacky et al. [32] presented similar values (1.79 ± 0.34) for 106 German national elite winter endurance athletes. Whereas in another study with female football players, DGLA values in the erythrocyte membranes were even lower (1.46 ± 0.32) [28]. Despite the DGLA’s well-known metabolic functions, no previous study points out the low levels in athletes. DGLA is converted by cyclooxygenase and lipoxygenase enzymes into two oxidative metabolites which have been used as disease biomarkers involving inhibition of chronic inflammation, vasodilation, and decreasing blood pressure, and suppression of smooth muscle cell proliferation correlated to the development of atherosclerotic plaque [56,57]. Furthermore, it has been proved the reduced capacity to convert DGLA from LA that may be explained, among other causes, by premenstrual syndrome [57]. Considering the enzymatic activity (index Δ6D+ELO) above the optimal values measured indirectly by the ratio 18:2/20:3 (Table 3 The mature erythrocyte membrane fatty acid profile of the participants compared to the control group and optimal values.), we observed a hypoactivity in the conversion from LA to DGLA. Moreover, this deficit might be connected to an alteration of the immune system [58,59] and intestinal health, including intestinal permeability, expected to be disrupted when increasing training loads [60], which in turn is related to the inflammatory status [61] as explained previously, being DGLA a precursor of AA.

Considering the high AA levels found, there is a clear benefit in having high levels in erythrocyte n-3 long-chain PUFAs, as it is correlated with membrane fluidity. As a consequence, there is a positive effect on oxygen diffusion, especially during exercise [62]. On the other hand, these high PUFA levels, from the omega-6 and omega-3 series, should be carefully considered to avoid undesirable oxidation of these long-chain FA in the membranes. We found that the football players had higher levels of total PUFA than the controls and that these levels increased during the competitive season. This should be considered when diet or supplementation with antioxidant properties is recommended.

Concerning the Omega-3 index or cardiovascular index, the sum of EPA and DHA, Stark et al. created a global map identifying which blood levels of omega-3, EPA, and DHA had countries and regions [52]. In that work, four ranges were classified, which correspond to the weight percent of EPA + DHA in the erythrocyte membrane. Spain was assigned to the range, from 4 to 6%, and Drobnic et al. work [33], which took place in Spain, obtained 4.9 ± 0.9. However, we observed higher values in the female players’ group (7.50 ± 1.22), where 34% of athletes had higher values than 8%. The control group’s values were lower than the athletes’ (6.69 ± 1.47) but higher than the Spain classification, which can be explained by the high fish consumption in the Basque Country population [63]. In other studies, notwithstanding, this index level was lower. For example, in German athletes was 4.97 ± 1.19 [32]; in NCAA football players it was 4.4 ± 0.8 [31]; in 31 NCAA Division I collegiate women soccer players, the mean of the index was 4.14% ± 0.74% [64]; and, in Martorell et al. study [34], where 11 healthy sportsmen were exposed to a calorie restriction program and acute exercise for 1 month, the basal measure was 3.41 ± 0.33. Those low and very low levels (<4%) can be connected with the incidence of cardiovascular diseases or an indicator of tissue impairment [65]. Moreover, it can also be related to a low intake of EPA and DHA food sources [64].

Although vitamin D levels decreased throughout the season associated with a decrease in sun exposure, they were always above the optimal ones (>30 ng/ml). This is an essential micronutrient for calcium homeostasis, bone health, inflammatory response, and well function of the immune system and skeletal muscle [66–69]. Deficiency levels of this vitamin depend on latitude, weather, reduction of ultraviolet radiation, and type of sports, mainly when training sessions are indoors. Furthermore, a deficiency is associated with an increased risk of autoimmune and chronic diseases [70,71]. Those low levels are estimated to affect 1 billion people, including athletes. Commonly, that decrement occurs in winter, as shown in a study with UK football players comparing summer and winter [72], and when vitamin D was analyzed during a competitive season of Division I female soccer players [73] and it is the same that we observed in our participants. This highlights the importance of monitoring this biomarker since supplementation in athletes has shown increased muscle strength and having the proper serum levels is linked to reducing injury risk and better sports performance [70].

A limitation of this study is the lack of measured dietary intakes for individual football players, and we only ascertained general dietary recommendation patterns. This is particularly challenging when interpreting the results regarding the origin of LPMEM changes.

In conclusion, the LPMEM gives us information related to lipid metabolism to form a cellular compartment crucial for life, which is influenced by the lipid pool formed with the contribution of dietary habits. A better knowledge of the nutritional status and needs of this particular population group will provide insights to define more specific personalized nutritional strategies. The findings in the present study confirmed our hypothesis that the optimal values of the lipidomic profile of the mature erythrocyte differed from elite female footballers, likely due to the professional sportive activity. Moreover, LPMEM showed the metabolic hallmarks of intense physical exercise, inflammation, and oxidative process throughout the competitive season. These facts should be considered when defining a personalized nutritional strategy (diet and supplementation) for female football players.

Author contribution

Conceptualization: I.T., J.A., T.V., J.Le.; methodology: J.A., A.L., J.La., J.Le., N.P., J.A., C.F.; formal analysis: N.P., J.A., G.M., C.F.; data curation: N.P., J.A., G.M.; writing – original draft preparation, N.P. and I.T.; writing – review and editing, N.P., I.T., J.A., J.M.O., C.F.; supervision, and funding acquisition: I.T. All authors read and agreed to the published version of the manuscript.

Ethics approval and consent to participate

The study was conducted according to the guidelines of the Declaration of Helsinki, updated at the World Medical Assembly in Fortaleza in 2013, and approved by the Euskadi Clinical Research Ethics Committee (permission number TUE-ATH-2018-02). Written informed consent was obtained from all athletes.

Acknowledgments

N.P. thanks the Department of Economic Development and Infrastructures of the Basque Government for receiving a PhD grant for young researchers in the scientific–technological and business environment of the Basque agricultural and food sector. We would like to express our gratitude to the female football players from Athletic Club that participated in this study.

Disclosure statement

C.F. is a cofounder of Lipinutragen srl, a spin-off of the CNR dedicated to membrane lipidomic analysis in human health. No potential conflict of interest was reported by the rest of the authors.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15502783.2023.2245386

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was partially supported by the Department of Environment: Territorial Planning: Agriculture and Fisheries of the Basque Country Government, ELKARTEK program from the Basque Country Government and the Centre for the Development of Industrial Technology (CDTI) of the Spanish Ministry of Science and Innovation under the grant agreement: TECNOMIFOOD project (CER-20191010). This is contribution number 1177 of AZTI, Food Research, Basque Research and Technology Alliance (BRTA).