Advertisement

Bovine colostrum: Postpartum changes in fat globule size distribution and fatty acid profile

Open AccessPublished:March 02, 2022DOI:https://doi.org/10.3168/jds.2021-20420

      ABSTRACT

      Although “zero waste” valorization concepts are gaining increasing attention, colostrum, a byproduct of milk production, remains underused due to technological challenges. Information about the fat fraction and the size of fat globules is needed to address these challenges, but such information is currently lacking. This study aimed to fill this gap in the knowledge by measuring the size distribution of bovine colostrum fat globules (CFG) and analyzing its relationships with postpartum milkings, parity, and fatty acids (FA) profile. Four sequential postpartum colostrum samples were collected from 44 cows and analyzed for the abovementioned parameters. The results indicated that CFG size increases almost twice during postpartum milkings (from ∼5 to ∼10 µm), whereas lactation has little, if any, effect on CFG size. The FA profile analyses showed that the content of most FA in the fourth postpartum milking was different from the previous milkings. The correlation analyses between CFG size and FA profile also demonstrated that the fourth milking was clearly distinguishable from the first 3 postpartum milkings. For example, the saturated FA content from the first 3 milkings had a positive correlation with smaller CFG (and a negative correlation with larger CFG), whereas the fourth milking demonstrated no correlations. Based on these CFG size and FA profile analyses, the results of this study suggest that the first 3 postpartum milkings can be considered as colostrum, whereas the fourth milking represents transition milk. Information about CFG size distribution enables modification of the FA profile of colostrum products and the ability to create better valorization technologies for colostrum-based food and feed supplements.

      Key words

      INTRODUCTION

      Rapid human population growth and increasing food demands have made the “zero waste” and “full valorization” production concepts critical. Together with other essential nutrients, the need for animal-origin fatty acids (FA) is increasing. Due to improved breeding and farming practices, the individual milk production of cows has increased over the past centuries, whereas the consumption of milk by newborn heifers has remained the same. As a result, large-scale milk production inevitably leads to the production of considerable amounts (about 1% of annual milk production) of colostrum and transition milk. Colostrum composition differs significantly from milk (
      • Playford R.J.
      • Weiser M.J.
      Bovine colostrum: Its constituents and uses.
      ), containing higher levels of proteins (especially immunoglobulins;
      • Kehoe S.I.
      • Jayarao B.M.
      • Heinrichs A.J.
      A survey of bovine colostrum composition and colostrum management practices on Pennsylvania dairy farms.
      ;
      • Sats A.
      • Kaart T.
      • Poikalainen V.
      • Aare A.
      • Lepasalu L.
      • Andreson H.
      • Jõudu I.
      Bovine colostrum whey: Postpartum changes of particle size distribution and immunoglobulin G concentration at different filtration pore sizes.
      ), vitamins, minerals (
      • Kehoe S.I.
      • Jayarao B.M.
      • Heinrichs A.J.
      A survey of bovine colostrum composition and colostrum management practices on Pennsylvania dairy farms.
      ;
      • Duplessis M.
      • Mann S.
      • Nydam D.V.
      • Girard C.L.
      • Pellerin D.
      • Overton T.R.
      Folates and vitamin B12 in colostrum and milk from dairy cows fed different energy levels during the dry period.
      ), antioxidants (
      • Przybylska J.
      • Albera E.
      • Kankofer M.
      Antioxidants in bovine colostrum.
      ), and PUFA (
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      ). Due to the thermal sensitivity of proteins and FA profile-related cream separation issues (e.g., fat melting point), the utilization of colostrum using milk valorization technologies is complicated. As a result, traditional dairy industries do not collect colostrum and transition milk for further processing.
      Colostrum is a valuable source of proteinaceous bioactive factors and essential nutrients (
      • Playford R.J.
      • Weiser M.J.
      Bovine colostrum: Its constituents and uses.
      ) that are mostly marketed as protein products. At the same time, bovine colostrum is also a source of PUFA (
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      ), which are associated with cancer prevention (
      • Baum S.J.
      • Kris-Etherton P.M.
      • Willett W.C.
      • Lichtenstein A.H.
      • Rudel L.L.
      • Maki K.C.
      • Whelan J.
      • Ramsden C.E.
      • Block R.C.
      Fatty acids in cardiovascular health and disease: A comprehensive update.
      ). The FA profile of colostrum differs from that of mature milk (
      • Laakso P.
      • Manninen P.
      • Mäkinen J.
      • Kallio H.
      Postparturition changes in the triacylglycerols of cow colostrum.
      ;
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      ). However, despite an increasing number of studies addressing colostrum, its FA profile has gained little attention.
      The diameter of milk fat globules (MFG) determines the thermal and physiological properties of milk (
      • Michalski M.C.
      • Ollivon M.
      • Briard V.
      • Leconte N.
      • Lopez C.
      Native fat globules of different sizes selected from raw milk: Thermal and structural behavior.
      ) and its lipid composition (
      • Michalski M.C.
      • Briard V.
      • Michel F.
      • Tasson F.
      • Poulain P.
      Size distribution of fat globules in human colostrum, breast milk, and infant formula.
      ). Moreover, MFG size affects the physicochemical and functional properties of dairy products (
      • Logan A.
      • Auldist M.
      • Greenwood J.
      • Day L.
      Natural variation of bovine milk fat globule size within a herd.
      ;
      • Edén J.
      • Dejmek P.
      • Löfgren R.
      • Paulsson M.
      • Glantz M.
      Native milk fat globule size and its influence on whipping properties.
      ) as well as the digestion and bioavailability of FA (
      • Garcia C.
      • Antona C.
      • Robert B.
      • Lopez C.
      • Armand M.
      The size and interfacial composition of milk fat globules are key factors controlling triglycerides bioavailability in simulated human gastro-duodenal digestion.
      ;
      • Singh H.
      • Gallier S.
      Nature's complex emulsion: The fat globules of milk.
      ;
      • Uken K.L.
      • Schäff C.T.
      • Vogel L.
      • Gnott M.
      • Dannenberger D.
      • Görs S.
      • Tuchscherer A.
      • Tröscher A.
      • Liermann W.
      • Hammon H.M.
      Modulation of colostrum composition and fatty acid status in neonatal calves by maternal supplementation with essential fatty acids and conjugated linoleic acid starting in late lactation.
      ). The colostrum fat fraction also plays an important role from a herd reproduction point of view, as energy reserves of the newborn calves are scarce, and fat consumption increases their thermoregulation ability (
      • Murray C.F.
      • Leslie K.E.
      Newborn calf vitality: Risk factors, characteristics, assessment, resulting outcomes and strategies for improvement.
      ). In this regard, the FA composition and size distribution of fat globules are essential parameters that determine the effective metabolism of fat. Using “whole milk,” it has been shown that these 2 parameters, MFG size and composition, are related (
      • Martini M.
      • Salari F.
      • Altomonte I.
      The macrostructure of milk lipids: The fat globules.
      ;
      • Couvreur S.
      • Hurtaud C.
      Relationships between milks differentiated on native milk fat globule characteristics and fat, protein and calcium compositions.
      ). However, there is currently no information about the size distribution of bovine colostrum fat globules (CFG).
      For a more complete understanding about colostrum component size distribution, studies need to focus simultaneously on multiple parameters (
      • Sats A.
      • Kaart T.
      • Poikalainen V.
      • Aare A.
      • Lepasalu L.
      • Andreson H.
      • Jõudu I.
      Bovine colostrum whey: Postpartum changes of particle size distribution and immunoglobulin G concentration at different filtration pore sizes.
      ). In this study, the aim was to describe the bovine colostrum FA profile and CFG size. In addition, the relationships between FA profile, CFG size distribution, parity, and immediate postpartum milkings were analyzed.

      MATERIALS AND METHODS

      Sampling

      Colostrum samples were collected from 44 Holstein cows from the experimental farm of the Estonian University of Life Sciences between 2015 and 2017. The average (± SE) milk, fat, and protein production of previous lactation of sampling cows was 9,524 ± 159.2, 372 ± 6.2, and 313 ± 4.6 kg, respectively. The average lengths of the lactation and dry periods were 300 ± 1.1 and 60 d, respectively. All cows included in this study were in good health. To prevent suckling, calves were separated from dams. Cows were milked with a portable milking machine 30 to 120 min after calving and thereafter twice per day at regular milking times (at 12 ± 1 h intervals). Four sequential postpartum samples were collected from each cow (n = 4 × 44 = 176). After milking, the colostrum of each animal was stirred, and the sample (50 mL) was collected. Sampled cows were selected according to their lactation number (10 in the first, 9 in the second, 10 in the third, 8 in the fourth, 3 the in fifth, 2 the in sixth, and 2 in the seventh). Two cows were sampled in 2 yr (1 in its fourth and fifth and 1 in its fifth and sixth lactations). Samples were frozen at −20°C and stored. During the dry period and after calving, cows were fed ad libitum. Daily diets were based on hay and grass silage with the addition of a concentrate feed (barley flour, rapeseed cake) and Rindavital VK mineral feed (Schaumann Agri International GmbH). The ingredients, chemical composition, and nutritional value of the feed during the dry period and after calving are presented in Table 1. The farming conditions were constant in all lactations.
      Table 1The ingredients, chemical composition, and nutritional value in feed during the dry period and after calving
      ItemDry periodAfter calving
      Ingredient, g/kg
       Grass-clover silage687696
       Barley229232
       Heat treated rapeseed cake6263
       Mineral-vitamin feed99
       Anion mineral feed13
      Chemical composition, g/kg
       CP154154
       Crude fat3535
       Crude fiber211211
      Nutritional value
       ME, MJ10.310.3
       MP, g/kg9090

      Determination of FA Profile

      Extraction of FA was performed according to the boron trifluoride method described by
      • Yurchenko S.
      • Sats A.
      • Poikalainen V.
      • Karus A.
      Method for determination of fatty acids in bovine colostrum using GC-FID.
      . Fatty acid methyl esters were analyzed on a Varian 3900 gas chromatograph equipped with a flame ionization detector and autosampler CP-8400. Chromatographic separation of FAME was performed using a Supelco ionic liquid column SLB-IL111 (100m × 0.25 mm i.d., 0.20 μm film thickness). Tridecanoic acid methyl ester (C13:0) as internal standard and CRM47885 standard mixture as external standard were used. A more detailed method description has been described by
      • Yurchenko S.
      • Sats A.
      • Tatar V.
      • Kaart T.
      • Mootse H.
      • Jõudu I.
      Fatty acid profile of milk from Saanen and Swedish Landrace goats.
      .
      The following sums and indices were calculated: SFA, MUFA, PUFA, atherogenic index, desaturase index (DI), and ratio of n-6 and n-3 FA (n-6/n-3). The atherogenic index was calculated as has been proposed by
      • Ulbricht T.L.V.
      • Southgate D.A.T.
      Coronary heart disease: Seven dietary factors.
      as follows: atherogenic index = (C12:0 + 4 × C14:0 + C16:0)/(MUFA + n-6 + n-3), and n-6/n-3 as n-6/n-3 = (C18:2n-6 + C20:2n-6 + C18:3n-6)/C18:3n-3. The DI was defined as follows: product of desaturase/substrate of desaturase and calculated for 3 pairs of FA (DI14 = C14:1 cis-5/C14:0; DI16 = C16:1 cis-7/C16:0; DI18 = C18:1 cis-9/C18:0).

      Size Distribution of Fat Globules

      Size distribution was determined with a laser diffraction analyzer (Malvern Mastersizer 3000, Malvern Instruments Ltd.). Samples were frozen at −20°C after collection to store samples until more detailed measurements (below). Preliminary experiments in our laboratory have confirmed that freezing-thawing cycles have no significant effect on milk fat globule size distribution when measured by the Mastersizer 3000 (Table 2). Thawed colostrum samples were gently vortexed before transfer into an analyzer dispersion tank containing reverse osmosis water (grade 2, conductivity 0.055 µS/cm). The material (lipid, refractive index 1.6) and dispersant (water, refractive index 1.33) were selected as proposed by Mastersizer software, and the speed of the stirrer was 2,000 rpm. The size distribution of the fat globules was measured after sufficient amount of sample was added to obtain an obscuration of 5 to 8%. Size by volume distribution–based percentiles (Dv10 as the first decile, Dv50 as the median, Dv90 as the last decile), means (D[4,3], volume moment mean; D[3,2] surface moment mean), and mode were derived from Mastersizer software. D[4,3] and D[3,2] were defined by the following equation:
      D[k,z]=NidiNidi,


      where Ni is the number of globules in a size class of di, k = 4 and z = 3 for D[4,3], and k = 3 and z = 2 for D[3,2].
      Table 2Mean (SE) milk fat globule size (μm) first volume distribution decile (Dv 10), median (Dv 50), last decile (Dv 90), and D[4,3] (volume moment mean), according to number of freeze-thaw cycles
      Freeze-thaw cycles, countDv 10Dv 50Dv 90D[4,3]
      02.20 (0.03)4.21 (0.06)6.98 (0.15)4.35 (0.08)
      22.15 (0.02)4.09 (0.03)6.69 (0.06)4.38 (0.11)
      42.16 (0.02)4.10 (0.03)6.71 (0.06)4.58 (0.27)
      P-value
      P-value indicates the statistical significance of the effect of freeze-thaw cycles according to repeated-measures ANOVA (n = 6 × 3 = 18, 6 samples and 3 cycles).
      0.2130.0770.0740.157
      1 P-value indicates the statistical significance of the effect of freeze-thaw cycles according to repeated-measures ANOVA (n = 6 × 3 = 18, 6 samples and 3 cycles).

      Statistical Analyses

      For statistical analyses, the bovine CFG size distribution in the form of frequency tables obtained from Malvern Mastersizer 3000 software were used. In these tables, CFG sizes from 0.1 to 3,500 µm were divided into 100 intervals on a log10 scale, and the CFG volume percentage in each size interval was presented. Therefore, 100 relative frequencies of CFG size groups that summed 100% were obtained for each sample. The mean CFG size distributions for postpartum milkings and lactation numbers were obtained by calculating the mean frequencies of CFG size intervals by milkings and lactation numbers. Additionally the bovine CFG size mode, D[4,3], D[3,2], first decile, median, and last decile were evaluated for each sample from Malvern Mastersizer 3000 software.
      The postpartum milking and lactation numbers were considered as categorical factors, and their effects on bovine CFG size first decile, median, and last decile and on the colostrum FA content and FA sums and ratios were analyzed by fitting a general linear mixed model. In this model, fixed effects of the above factors, their interactions, and random effect of cow were considered. The marginal means (alias least squares means with 95% confidence interval) were estimated and compared by milkings and lactation numbers, applying the Tukey test for multiple testing. The magnitude of the cow effect was estimated using the intraclass correlation coefficient. Initially analyses were made with 5 parity groups. However, as there were no systematic differences between later lactations, especially in bovine colostrum fat globule size distribution, it was decided to consider all later lactations (>3) together.
      The relationships between CFG size and colostrum FA content were studied with Pearson correlation analysis. First, the correlation coefficients between median CFG size and FA content were calculated at each milking and lactation combination. Second, to study the dynamics of the relationships over the whole CFG size distribution, the correlation coefficients between FA content and CFG frequency in each size interval were calculated. To discover common CFG size distribution patterns, principal component analysis (PCA) was performed. To analyze to what extent these patterns reflected the differences between milkings and lactation numbers, general linear mixed models with principal components as dependent variables; milkings, lactation numbers, and their interactions as fixed effects; and cow as a random effect were fitted. The relationships between PCA patterns and colostrum FA content were studied with Pearson correlation analysis. In correlation analysis and PCA, only CFG size distribution in the interval of 1 to 100 µm was used.
      All statistical analyses (significance was declared at P ≤ 0.05) were performed, and all figures were constructed with R version 4.0.3 (https://www.r-project.org/). We used the packages ‘car', ‘lme4', ‘emmeans', and ‘multcompView' for modeling, and we used the package ‘ade4' for PCA.

      RESULTS AND DISCUSSION

      Fat Globule Size

      CFG Size Distribution

      The colostrum size distribution consisted of 2 subdistributions, with the majority being around 10 µm and the minority (barely noticeable subdistribution) being <0.5 μm (Figure 1A). The CFG size recorded in this study was larger than the mean MFG size (3–6 µm) reported in previous studies (
      • Ménard O.
      • Ahmad S.
      • Rousseau F.
      • Briard-Bion V.
      • Gaucheron F.
      • Lopez C.
      Buffalo vs. cow milk fat globules: Size distribution, zeta-potential, compositions in total fatty acids and in polar lipids from the milk fat globule membrane.
      ;
      • Logan A.
      • Auldist M.
      • Greenwood J.
      • Day L.
      Natural variation of bovine milk fat globule size within a herd.
      ;
      • Jaakamo M.J.
      • Luukkonen T.J.
      • Kairenius P.K.
      • Bayat A.R.
      • Ahvenjärvi S.A.
      • Tupasela T.M.
      • Vilkki J.H.
      • Shingfield K.J.
      • Leskinen H.M.
      The effect of dietary forage to concentrate ratio and forage type on milk fatty acid composition and milk fat globule size of lactating cows.
      ). However, our findings were in line with those of
      • Michalski M.C.
      • Briard V.
      • Michel F.
      • Tasson F.
      • Poulain P.
      Size distribution of fat globules in human colostrum, breast milk, and infant formula.
      , who studied human breast milk within first postpartum days and reported large (∼9 µm) and small (∼0.1 µm) size populations similarly (Figure 1A). They suggested that both subdistributions represent fat globules. Unlike
      • Michalski M.C.
      • Briard V.
      • Michel F.
      • Tasson F.
      • Poulain P.
      Size distribution of fat globules in human colostrum, breast milk, and infant formula.
      , we did not dissociate the casein micelles. Therefore, in our case, the smaller subdistribution (<0.5 μm, Figure 1A) might have also represented casein micelles as the diameter of casein micelles is between 0.15 and 0.2 μm (
      • Mootse H.
      • Pisponen A.
      • Pajumägi S.
      • Polikarpus A.
      • Tatar V.
      • Sats A.
      • Poikalainen V.
      Investigation of casein micelle particle size distribution in raw milk of Estonian Holstein dairy cows.
      ). Nevertheless, because the <0.5 µm subdistribution is barely noticeable (Figure 1A) and far smaller than the first decile of size distribution (Table 3), we argue that it has no sizing effect.
      Figure thumbnail gr1
      Figure 1(A) Bovine colostrum fat globule size distribution at the first 4 postpartum milkings depending on lactation number (first lactation, n = 40; second lactation, n = 36; third and later lactations, n = 111). Each distribution shows the mean globule size distribution and globule size axes are presented on a log10 scale; red boxes denote the area of the middle 50% of globule sizes, strong black lines mark the median globule size by milkings, and dotted lines indicate the median globule size over 4 milkings. (B) Model-based means (with 95% CI) of bovine colostrum fat globule size first decile (Dv 10), median (Dv 50), last decile (Dv 90), D[4,3], D[3,2], and mode depending on milking and lactation. Significance of milking (pM), lactation (pL) and milking by lactation interaction (pM × L) effects are presented.
      Table 3Model-based means (SE) of fat globule size (μm) first decile (Dv 10), median (Dv 50), last decile (Dv 90), D[4,3], D[3,2], and mode depending on milking and lactation
      VariableMilkingLactationP-value
      P-values indicate milking (M), lactation (L) and milking by lactation interaction (M × L) effects.
      1234123–7MLM × L
      Dv 101.85 (0.23)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.13 (0.23)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.41 (0.23)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      3.56 (0.23)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.87 (0.27)1.98 (0.28)2.62 (0.18)<0.0010.0600.505
      Dv 505.84 (0.48)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      6.55 (0.48)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      8.50 (0.48)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      11.15 (0.48)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.00 (0.70)7.32 (0.74)7.72 (0.46)<0.0010.1940.465
      Dv 9016.21 (1.66)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      16.31 (1.66)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      22.74 (1.66)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      29.50 (1.66)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      26.85 (2.43)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      18.11 (2.56)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      18.61 (1.60)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.0110.736
      D[4,3]9.46 (1.53)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.59 (1.53)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      11.65 (1.53)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      16.08 (1.53)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      15.01 (1.63)9.30 (1.72)10.78 (1.06)<0.0010.0340.927
      D[3,2]3.48 (0.47)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      3.91 (0.47)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      4.45 (0.47)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      6.93 (0.47)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      5.57 (0.56)3.73 (0.59)4.78 (0.37)<0.0010.0770.590
      Mode6.08 (0.62)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      7.26 (0.62)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.99 (0.62)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      13.05 (0.62)
      Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.87 (0.90)8.78 (0.95)8.62 (0.59)<0.0010.4990.015
      a–c Means without common superscript letters in the same row among 4 milkings over 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1 P-values indicate milking (M), lactation (L) and milking by lactation interaction (M × L) effects.

      Postpartum Milking

      In samples of first postpartum milking, the CFG median size recorded in this study was in a similar range (4.5–5.5 µm) to the MFG (3–6 µm) reported in previous studies (
      • Lopez C.
      • Briard-Bion V.
      • Ménard O.
      • Beaucher E.
      • Rousseau F.
      • Fauquant J.
      • Leconte N.
      • Robert B.
      Fat globules selected from whole milk according to their size: Different compositions and structure of the biomembrane, revealing sphingomyelin-rich domains.
      ;
      • Fleming A.
      • Schenkel F.S.
      • Chen J.
      • Malchiodi F.
      • Ali R.A.
      • Mallard B.
      • Sargolzaei M.
      • Corredig M.
      • Miglior F.
      Variation in fat globule size in bovine milk and its prediction using mid-infrared spectroscopy.
      ;
      • Jaakamo M.J.
      • Luukkonen T.J.
      • Kairenius P.K.
      • Bayat A.R.
      • Ahvenjärvi S.A.
      • Tupasela T.M.
      • Vilkki J.H.
      • Shingfield K.J.
      • Leskinen H.M.
      The effect of dietary forage to concentrate ratio and forage type on milk fatty acid composition and milk fat globule size of lactating cows.
      ). During the subsequent 3 postpartum milkings, the CFG diameter almost doubled (Figure 1; Table 3), reaching 10 µm at the fourth postpartum milking. The effect of milking was significant (P < 0.001) in all size-describing parameters (Dv10, Dv50, Dv90, D[4,3], D[3,2] and mode, Table 3). There are no reports about bovine CFG size available. In a comprehensive review,
      • Martini M.
      • Salari F.
      • Altomonte I.
      The macrostructure of milk lipids: The fat globules.
      suggested that characteristics of MFG affect milk quality and digestive parameters, although available data regarding the effects of the MFG diameter on digestibility are conflicting. Our results suggested that smaller CFG diameter at the first postpartum milking can enable efficient fat and energy assimilation for calves. It can also relate to the unique ability of neonatal enterocytes to absorb protein macromolecules (
      • Sangild P.T.
      Uptake of colostral immunoglobulins by the compromised newborn farm animal.
      ). As colostrum fat content decreases over time (
      • Elfstrand L.
      • Lindmark-Mansson H.
      • Paulsson M.
      • Nyberg L.
      • Akesson B.
      Immunoglobulins, growth factors and growth hormone in bovine colostrum and the effects of processing.
      ), the increase of CFG size does not support previous reports that state positive relationships between daily fat secretion and average MFG diameter (
      • Wiking L.
      • Stagsted J.
      • Björck L.
      • Nielsen J.H.
      Milk fat globule size is affected by fat production in dairy cows.
      ;
      • Carroll S.M.
      • DePeters E.J.
      • Taylor S.J.
      • Rosenberg M.
      • Perez-Monti H.
      • Capps V.A.
      Milk composition of Holstein, Jersey, and Brown Swiss cows in response to increasing levels of dietary fat.
      ). However,
      • Argov N.
      • Lemay D.G.
      • German B.
      Milk fat globule structure and function: Nanoscience comes to milk production.
      hypothesized that as the membrane of the mammary epithelial cell is lost to envelop the globule, larger MFG may be secreted to reduce membrane loss. This might also be one of the explanations for the CFG size increase reported in this study. For ewes, the CFG in milk has been found to be larger immediately postpartum (10 h) than in subsequent days (
      • Martini M.
      • Altomonte I.
      • Salari F.
      The lipid component of Massese ewes' colostrum: Morphometric characteristics of milk fat globules and fatty acid profile.
      ), and a similar tendency has been reported for human colostrum (
      • Michalski M.C.
      • Briard V.
      • Michel F.
      • Tasson F.
      • Poulain P.
      Size distribution of fat globules in human colostrum, breast milk, and infant formula.
      ). The MFG mean diameter of cow milk has been found to decrease slightly during lactation, being close to 5.5 µm at the start of lactation (5–30 d) and ending up around 4.5 µm at >305 d (
      • Fleming A.
      • Schenkel F.S.
      • Chen J.
      • Malchiodi F.
      • Ali R.A.
      • Mallard B.
      • Sargolzaei M.
      • Corredig M.
      • Miglior F.
      Variation in fat globule size in bovine milk and its prediction using mid-infrared spectroscopy.
      ). These results, together with our measurements, suggest that the MFG mean diameter peaks shortly after postpartum (up to 40 h) and decreases relatively rapidly after this peak. Further studies are required to verify the rapid decrease of fat globule size in transition milk (2–5 d postpartum).

      Parity

      Lactation number had a significant effect on volume moment mean (D[4,3], P = 0.034) and larger fat globules (Dv 90, P = 0.011). It had a nonsignificant influence on median CFG size (Dv50, P = 0.194), smaller fat globules (Dv10, P = 0.06), and mode (P = 0.499; Table 3). However, mode was the only CFG size variable that demonstrated a significant milking by lactation interaction effect (P = 0.015). Previous findings about the lactation effects on cow MFG size are mixed;
      • Walter L.
      • Finch S.
      • Cullen B.
      • Fry R.
      • Logan A.
      • Leury B.J.
      The effect of physiological state, milk production traits and environmental conditions on milk fat globule size in cow's milk.
      reported smaller MFG size on first- and second-parity cows (140 individual cows from 1 herd), whereas
      • Fleming A.
      • Schenkel F.S.
      • Chen J.
      • Malchiodi F.
      • Ali R.A.
      • Mallard B.
      • Sargolzaei M.
      • Corredig M.
      • Miglior F.
      Variation in fat globule size in bovine milk and its prediction using mid-infrared spectroscopy.
      showed no significant effect of parity (399 individual cows from 44 herds).
      • Walter L.
      • Finch S.
      • Cullen B.
      • Fry R.
      • Logan A.
      • Leury B.J.
      The effect of physiological state, milk production traits and environmental conditions on milk fat globule size in cow's milk.
      suggested the effect of parity was relatively small compared with the between-herd and between-breed variations in the study by
      • Fleming A.
      • Schenkel F.S.
      • Chen J.
      • Malchiodi F.
      • Ali R.A.
      • Mallard B.
      • Sargolzaei M.
      • Corredig M.
      • Miglior F.
      Variation in fat globule size in bovine milk and its prediction using mid-infrared spectroscopy.
      , whereas following the same herd allowed capturing the independent effect of parity. Overall, it seems that lactation has little effect on CFG size.

      Fatty Acid Profile

      Altogether, 10 SFA, 4 MUFA, and 3 PUFA were identified from the colostrum samples used in this study, with SFA accounting for 64.9%, MUFA for 31.4%, and PUFA for 3.7% of the FA profile. Palmitic acid (C16:0), oleic acid (C18:1 cis-9), and linoleic acid (C18:2n-6) were the main SFA, MUFA, and PUFA, respectively (Figure 2; Table 4). These results are in line with FA profiles reported in earlier studies (
      • Laakso P.
      • Manninen P.
      • Mäkinen J.
      • Kallio H.
      Postparturition changes in the triacylglycerols of cow colostrum.
      ;
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      ;
      • O'Callaghan T.F.
      • O'Donovan M.
      • Murphy J.P.
      • Sugrue K.
      • Mannion D.
      • McCarthy W.P.
      • Timlin M.
      • Kilcawley K.N.
      • Hickey R.M.
      • Tobin J.T.
      Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
      ).
      Figure thumbnail gr2
      Figure 2Model-based means (with 95% CI) of colostrum fatty acids and fatty acid sums and ratios depending on milking and lactation number. Significance of milking (pM), lactation (pL), and milking by lactation interaction (pM × L) effects are presented. The following ratios were calculated: atherogenic index, desaturase index (DI), and ratio of n-6 and n-3 fatty acids (n-6/n-3).
      Table 4Model-based means (SE) of studied milk fatty acids and fatty acid sums (g/100 g) and ratios, depending on milking and lactation
      Variable
      The following ratios were calculated: atherogenic index, desaturase index (DI) and ratio of n-6 and n-3 fatty acids (n-6/n-3). DI was defined as follows: product of desaturase/substrate of desaturase and calculated for 3 pairs of fatty acids (DI14 = C14:1 cis-5/C14:0; DI16 = C16:1 cis-7/C16:0; DI18 = C18:1 cis-9/C18:0). Atherogenic index was calculated as proposed by Ulbricht and Southgate (1991): atherogenic index = C12:0 + 4 × C14:0 + C16:0)/(MUFA + n-6 + n-3 and n-6/n-3 as n-6/n-3 = (C18:2n-6 + C20:2n-6 + C18:3n-6)/C18:3n-3.
      MilkingLactationP-value
      P-values indicate the (M), lactation (L) and milking by lactation interaction (M × L) effects.
      1234123–7MLM × L
      C4:00.33 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.34 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.37 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.52 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.52 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.41 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.24 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      C6:00.35 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.37 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.38 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.51 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.49 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.40 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.31 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      C8:00.29 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.31 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.33 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.44 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.41 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.34 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.28 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      C10:00.88 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.92 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.95 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.30 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.03 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.18 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.82 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      C12:01.89 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.91 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.89 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.75 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.76 (0.09)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.20 (0.10)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.62 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.0010.001
      C14:011.16 (0.28)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      10.96 (0.28)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      10.82 (0.28)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.73 (0.28)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.78 (0.49)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      12.72 (0.52)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      9.51 (0.32)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.0010.085
      C15:00.66 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.63 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.64 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.53 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.61 (0.03)0.58 (0.03)0.65 (0.02)<0.0010.0830.545
      C14:1 cis-50.54 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.54 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.52 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.59 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.46 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.47 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.72 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      C16:037.64 (0.25)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      37.71 (0.25)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      37.58 (0.25)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      36.24 (0.25)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      36.85 (0.45)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      36.43 (0.48)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      38.60 (0.30)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.0010.355
      C16:1 cis-72.34 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.36 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.46 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.72 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.34 (0.07)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.30 (0.07)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.77 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.0010.003
      C18:012.65 (0.21)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      12.68 (0.21)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      12.78 (0.21)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      11.88 (0.21)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      13.15 (0.38)12.16 (0.40)12.18 (0.25)<0.0010.0790.035
      C18:1 cis-926.99 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      26.97 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      26.96 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      29.9 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      28.49 (0.64)26.63 (0.67)27.99 (0.42)<0.0010.113<0.001
      C20:00.28 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.28 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.29 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.23 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.28 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.25 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.27 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.0350.466
      C18:2n-62.31 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.32 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.33 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.15 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.19 (0.07)2.31 (0.07)2.34 (0.05)<0.0010.1860.002
      C20:1 cis-90.22 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.22 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.21 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.27 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.22 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.17 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.31 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      C18:3n-31.19 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.19 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.20 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.03 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.09 (0.03)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.18 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.19 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.0320.034
      C20:2n-60.28 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.29 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.29 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.24 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.34 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.27 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.21 (0.01)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      SFA66.12 (0.37)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      66.11 (0.37)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      66.02 (0.37)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      63.11 (0.37)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      64.88 (0.67)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      66.67 (0.70)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      64.47 (0.44)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.029<0.001
      MUFA30.09 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      30.1 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      30.15 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      33.48 (0.36)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      31.51 (0.64)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      29.57 (0.68)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      31.79 (0.42)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.018< 0.001
      PUFA3.79 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      3.8 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      3.82 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      3.41 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      3.61 (0.09)3.76 (0.10)3.74 (0.06)<0.0010.435<0.001
      Σ n-31.19 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.19 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.20 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.03 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.09 (0.03)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.18 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1.19 (0.02)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.0320.034
      Σ n-62.60 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.61 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.62 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.38 (0.04)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.52 (0.07)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.58 (0.08)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.55 (0.05)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.0010.859<0.001
      n-3/n-62.19 (0.16)2.21 (0.16)2.21 (0.16)2.48 (0.16)2.36 (0.16)2.20 (0.17)2.26 (0.10)0.1810.7980.953
      Atherogenic index2.52 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.50 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.46 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.12 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.23 (0.10)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.76 (0.10)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.21 (0.07)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.0010.011
      DI140.05 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.05 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.05 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.07 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.05 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.04 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.08 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.001<0.001
      DI160.06 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.06 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.07 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.08 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.06 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.06 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      0.07 (0.00)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      <0.001<0.0010.015
      DI182.17 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.16 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.15 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.55 (0.06)
      Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      2.21 (0.10)2.24 (0.11)2.31 (0.07)<0.0010.680<0.001
      a–c Means without common superscript letters in the same row among the 4 milkings or the 3 lactations are significantly different (P ≤ 0.05, Tukey post hoc test).
      1 The following ratios were calculated: atherogenic index, desaturase index (DI) and ratio of n-6 and n-3 fatty acids (n-6/n-3). DI was defined as follows: product of desaturase/substrate of desaturase and calculated for 3 pairs of fatty acids (DI14 = C14:1 cis-5/C14:0;DI16 = C16:1 cis-7/C16:0; DI18 = C18:1 cis-9/C18:0). Atherogenic index was calculated as proposed by
      • Ulbricht T.L.V.
      • Southgate D.A.T.
      Coronary heart disease: Seven dietary factors.
      : atherogenic index = C12:0 + 4 × C14:0 + C16:0)/(MUFA + n-6 + n-3 and n-6/n-3 as n-6/n-3 = (C18:2n-6 + C20:2n-6 + C18:3n-6)/C18:3n-3.
      2 P-values indicate the (M), lactation (L) and milking by lactation interaction (M × L) effects.

      Postpartum Milking

      In general, short-chain (C4–10) SFA concentrations increased with the number of postpartum milkings (Figure 3), whereas medium-chain (C11–15) and long-chain (C16–20) SFA decreased (except for second lactation C12:0). These results are in line with the dynamics of FA profile in colostrum and transition milk described by
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      and in a comprehensive study by
      • O'Callaghan T.F.
      • O'Donovan M.
      • Murphy J.P.
      • Sugrue K.
      • Mannion D.
      • McCarthy W.P.
      • Timlin M.
      • Kilcawley K.N.
      • Hickey R.M.
      • Tobin J.T.
      Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
      . It is also in agreement with the well-described positive association between MFG size (greater triglyceride to membrane ratio) and shorter-chain FA contents (
      • Couvreur S.
      • Hurtaud C.
      Relationships between milks differentiated on native milk fat globule characteristics and fat, protein and calcium compositions.
      ). There are some discrepancies between the findings of this study and the findings of
      • O'Callaghan T.F.
      • O'Donovan M.
      • Murphy J.P.
      • Sugrue K.
      • Mannion D.
      • McCarthy W.P.
      • Timlin M.
      • Kilcawley K.N.
      • Hickey R.M.
      • Tobin J.T.
      Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
      in long-chain FA (C18:0, C20:0) that can be explained by differences in housing and feeding conditions. Indeed, approximately half of long-chain FA (≥C18) are derived from the diet, whereas the majority of smaller (C4:0–C14:0) FA originate from de novo FA synthesis in the mammary gland (
      • Palmquist D.L.
      Milk fat: Origin of fatty acids and influence of nutritional factors there on.
      ).
      Figure thumbnail gr3
      Figure 3Pearson correlation coefficients between median fat globule size and colostrum fatty acids content and fatty acids' sums and ratios depending on milking and lactation number. The following ratios were calculated: atherogenic index, desaturase index (DI) and ratio of n-6 and n-3 fatty acids (n-6/n-3).
      Among medium-chain MUFA, only C14:1 cis-5 increased with milking number, with an exception for the second-lactation samples where it decreased. Long-chain MUFA also increased (except for second lactation C20:1 cis-9), whereas PUFA decreased. Overall, SFA and PUFA decreased, and MUFA increased. Similar dynamics in SFA, MUFA, and PUFA have been reported previously (
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      ;
      • O'Callaghan T.F.
      • O'Donovan M.
      • Murphy J.P.
      • Sugrue K.
      • Mannion D.
      • McCarthy W.P.
      • Timlin M.
      • Kilcawley K.N.
      • Hickey R.M.
      • Tobin J.T.
      Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
      ), with
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      reporting distinctive profiles in the longer term (72, 96, and 120 h postpartum) and not in the short term (24 and 48 h postpartum). The fact that PUFA are mainly associated with the MFG membrane (
      • Couvreur S.
      • Hurtaud C.
      Relationships between milks differentiated on native milk fat globule characteristics and fat, protein and calcium compositions.
      ), and that the content of the latter decreases when MFG size increases, may provide an explanation for the phenomenon of reduced PUFA over time. The atherogenic index decreased, and DI (except for second lactation DI14) increased with postpartum milking number in this study. The decreasing atherogenic index suggests that, in terms of cholesterol content (
      • Poppitt S.D.
      • Keogh G.F.
      • Mulvey T.B.
      • McArdle B.H.
      • MacGibbon A.K.
      • Cooper G.J.
      Lipid-lowering effects of a modified butter-fat: A controlled intervention trial in healthy man.
      ) and food production (
      • Bobe G.
      • Hammond E.
      • Freeman A.
      • Lindberg G.
      • Beitz D.
      Texture of butter from cows with different milk fatty acid compositions.
      ), transition milk can outperform colostrum.
      Differences in FA content are most evident at the fourth milking (Figure 2; Table 4). This supports the common understanding that colostrum production lasts around 24 h postpartum and that the fourth milking (36–40 h postpartum) should be considered transition milk. This knowledge is useful when considering nutritional and feeding aspects while developing colostrum and transition milk valorization technologies.

      Parity

      The short-chain SFA and C20:2n-6 decreased with increasing lactation number (Figure 2). Medium-chain and long-chain SFA, however, did not show clear change with increasing lactation number, as the second-lactation samples had the highest concentrations of C12:0 and C14:0, and lowest of C15:0, C16:0 and C20:0. Lactation number had a significant (P = 0.029) effect on overall SFA, with samples from cows in their second lactation having the highest values, especially at the fourth milking. The short-chain SFA content of the first milking of second-lactation samples was similar to lactation numbers 3 to 7, whereas the fourth milking resembled the first lactation. Thus, it seemed that although the colostrum of cows in the second lactation initially had similar quality to colostrum from cows in their third milking, this quality was not as persistent for cows that were in their third or later lactation. In the case of C14:1n-5 to C14:1 cis-5, C16:1n-7 to C16:1 cis-7, and C20:1n-9 to C20:1 cis-9, colostrum from cows in their third to seventh lactations had the highest values. The same trend was evident for MUFA in the first to third milkings.
      Our findings are in agreement with earlier studies showing that parity has the strongest effect on short-chain, de novo FA (
      • Mele M.
      • Macciotta N.P.P.
      • Cecchinato A.
      • Conte G.
      • Schiavon S.
      • Bittante G.
      Multivariate factor analysis of detailed milk fatty acid profile: Effects of dairy system, feeding, herd, parity, and stage of lactation.
      ). The difference in colostrum short-chain SFA profile is also evident when primiparous and multiparous colostrum samples are compared and lactation numbers are not distinguished, presumably reflecting individual responses of cows to increased energy requirements at the onset of lactation (
      • Contarini G.
      • Povolo M.
      • Pelizzola V.
      • Monti L.
      • Bruni A.
      • Passolungo L.
      • Abeni F.
      • Degano L.
      Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
      ).
      • O'Callaghan T.F.
      • O'Donovan M.
      • Murphy J.P.
      • Sugrue K.
      • Mannion D.
      • McCarthy W.P.
      • Timlin M.
      • Kilcawley K.N.
      • Hickey R.M.
      • Tobin J.T.
      Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
      reported nonsignificant effects of parity on colostrum short-chain SFA, which might stem from their small sample size (6 cows per lactation) and longer sampling period (0–5 d). However,
      • O'Callaghan T.F.
      • O'Donovan M.
      • Murphy J.P.
      • Sugrue K.
      • Mannion D.
      • McCarthy W.P.
      • Timlin M.
      • Kilcawley K.N.
      • Hickey R.M.
      • Tobin J.T.
      Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
      also demonstrated that parity had a significant effect on some medium and long-chain FA (C14:1, C15:0, C16:0, C18:0, C18:2n-6, C18:3n-3, C20:0, CLA, and C21:0), most of which are in agreement with our findings.
      For short-chain SFA content, second-lactation cows had similar values to the third- to seventh-lactation cows during the first milking, but increased with milking number, reaching a similar level as the first lactation cows by the fourth milking. As the majority of short-chain FA originate from de novo FA synthesis (
      • Palmquist D.L.
      Milk fat: Origin of fatty acids and influence of nutritional factors there on.
      ), this pattern can reflect the development and condition of the mammary gland. Highest atherogenic index values are in samples from the second lactation; therefore, from a consumer (
      • Poppitt S.D.
      • Keogh G.F.
      • Mulvey T.B.
      • McArdle B.H.
      • MacGibbon A.K.
      • Cooper G.J.
      Lipid-lowering effects of a modified butter-fat: A controlled intervention trial in healthy man.
      ) and food production aspect (
      • Bobe G.
      • Hammond E.
      • Freeman A.
      • Lindberg G.
      • Beitz D.
      Texture of butter from cows with different milk fatty acid compositions.
      ), second-lactation cow colostrum is least desirable.

      Correlations Between Fat Globule Size and FA Content

      CFG Size Median and FA

      The correlations between short-chain SFA content and CFG median size (Figure 3) indicated that samples from primiparous cows with larger CFG tended to have higher contents of C4:0 and C6:0 in the second, third, and fourth milkings, whereas no such positive correlation was found in samples collected from cows in the second to seventh lactations. Among medium-chain SFA, the clearest pattern was evident for C16:0 and C14:0. The samples from multiparous cows had a negative correlation between CFG size and C16:0 content, whereas samples from primiparous cows showed a negative correlation with C14:0. In first- and second-lactation cows, the CFG size was negatively correlated with overall SFA, and positively correlated with MUFA and DI14. In primiparous cows, the CFG size was negatively correlated with PUFA and omega FA contents, whereas these correlations in multiparous cows were positive.
      The correlations reported in this study agree with those reported by
      • Lopez C.
      • Briard-Bion V.
      • Ménard O.
      • Beaucher E.
      • Rousseau F.
      • Fauquant J.
      • Leconte N.
      • Robert B.
      Fat globules selected from whole milk according to their size: Different compositions and structure of the biomembrane, revealing sphingomyelin-rich domains.
      . However, here we also demonstrated the effect of parity, which was not addressed by
      • Lopez C.
      • Briard-Bion V.
      • Ménard O.
      • Beaucher E.
      • Rousseau F.
      • Fauquant J.
      • Leconte N.
      • Robert B.
      Fat globules selected from whole milk according to their size: Different compositions and structure of the biomembrane, revealing sphingomyelin-rich domains.
      . The negative correlation between CFG size and C14:0 was most likely driving the positive correlations between CFG size and DI14 value. The correlations between DI14, DI16, and DI18 and CFG size were not similar to each other (Figure 3). The most probable explanation for this is the different origin of these FA; <C14 FA originate from de novo FA synthesis, ≥C18 FA are derived from the diet, and C16 are derived from both (
      • Palmquist D.L.
      Milk fat: Origin of fatty acids and influence of nutritional factors there on.
      ). Interestingly, in the samples from the primiparous cows, the correlations (Figure 3) of the well-known relationships discussed above (larger MFG, more short-chain FA; smaller MFG, less MFG membrane, fewer long-chain FA) were more evident than in the samples from multiparous cows. The colostrum of primiparous cows has lower quality than that of multiparous cows (
      • Dunn A.
      • Ashfield A.
      • Earley B.
      • Welsh M.
      • Gordon A.
      • Morrison S.J.
      Evaluation of factors associated with immunoglobulin G, fat, protein, and lactose concentrations in bovine colostrum and colostrum management practices in grassland-based dairy systems in Northern Ireland.
      ). Thus, our results might suggest that primiparous and multiparous cows differ not only regarding their colostrum and milk composition, but also in their metabolism and milk syntheses mechanisms. The mechanisms behind these differences, which can lie for example in the metabolism, development, or condition of the mammary gland, need to be addressed in future studies.
      Higher values of DI and MUFA, accompanied by a decrease in SFA, have been connected to the prevention of the risks of cardiovascular diseases (
      • Reh W.A.
      • Maga E.A.
      • Collette N.M.B.
      • Moyer A.
      • Conrad-Brink J.S.
      • Taylor S.J.
      • DePeters E.J.
      • Oppenheim S.
      • Rowe J.D.
      • BonDurant R.H.
      • Anderson G.B.
      • Murray J.D.
      Hot topic: Using a stearoyl-CoA desaturase transgene to alter milk fatty acid composition.
      ), making the colostrum of first- and second-lactation cows especially desirable. Considering the atherogenic index, the colostrum with bigger fat globules, originating from cows in the first and second lactations (Figure 3), is highly valuable as a product with a lower atherogenic index can decrease both total and low-density lipoprotein cholesterol (
      • Poppitt S.D.
      • Keogh G.F.
      • Mulvey T.B.
      • McArdle B.H.
      • MacGibbon A.K.
      • Cooper G.J.
      Lipid-lowering effects of a modified butter-fat: A controlled intervention trial in healthy man.
      ). In addition, butter produced from milk with lower atherogenic index is more spreadable and less adhesive (
      • Bobe G.
      • Hammond E.
      • Freeman A.
      • Lindberg G.
      • Beitz D.
      Texture of butter from cows with different milk fatty acid compositions.
      ).

      CFG Size Distribution and FA

      The strengths and directions of correlations between FA and CFG size fluctuated between positive and negative values (Figure 4). This suggested that the FA concentration was dependent on the size of the CFG (or vice versa). In general, the first to third milkings showed similar correlations to each other, whereas milk from the fourth milking was clearly different for SFA, MUFA, and atherogenic index. This provided further support to our previous suggestion that the fourth milking should be considered transition milk. The associations between the CFG size and FA also have practical value as this information enables the modification of the FA profile of colostrum products by manipulating the size distribution. For example, colostrum with bigger fat globules should be preferred to achieve lower atherogenic index value and SFA content.
      Figure thumbnail gr4
      Figure 4Correlations between milk fatty acids content and fatty acids' indices (FA), and colostrum fat globule size distribution. The following ratios were calculated: atherogenic index, desaturase index (DI) and ratio of n-6 and n-3 fatty acids (n-6/n-3).

      Principal Component Analysis

      The PCA analyses indicated that most of the variation in CFG size distribution [the first principal component (PC1) explained 68.3% of the variation] was caused by the contrast between particle size frequencies below and above 10 µm. (Figure 5A). The PC1 considered mainly the differences between postpartum milkings (Figure 5C); at the first milkings, the frequency of smaller (10 µm) globules was higher. The random cow effect explained 53.9% of the variability of PC1, indicating that the particle size ranges related to PC1 were strongly animal-specific (Figure 5E). The PCA addressing the correlations between all measured FA simultaneously had stronger positive relationships (higher FA concentration, smaller CFG) with atherogenic index, medium- and long-chain SFA, omega FA, and PUFA, and negative relationships (higher FA concentration, larger CFG) with DI16, DI18, short-chain SFA, and MUFA (Figure 5B). These relationships are in line with the correlations between individual FA and CFG size discussed above and with well-established associations between MFG size, FA chain length, and MFG membrane (
      • Couvreur S.
      • Hurtaud C.
      Relationships between milks differentiated on native milk fat globule characteristics and fat, protein and calcium compositions.
      ). These relationships are in agreement with the results of the correlations between individual FA and CFG size discussed above. The second PCA (explaining 21.1% of the globule size variability) contrasted the frequency of particles with sizes around 10 µm and <4 or >20 µm (Figure 5A). It distinguished the samples according to lactation (Figure 5C); the samples from higher lactation numbers had a slightly higher frequency of fat globules of a size around 10 µm. This finding is in agreement with
      • Walter L.
      • Finch S.
      • Cullen B.
      • Fry R.
      • Logan A.
      • Leury B.J.
      The effect of physiological state, milk production traits and environmental conditions on milk fat globule size in cow's milk.
      , who reported smaller MFG size in first- and second-parity cows. However, no significant effect of parity has also been reported (
      • Fleming A.
      • Schenkel F.S.
      • Chen J.
      • Malchiodi F.
      • Ali R.A.
      • Mallard B.
      • Sargolzaei M.
      • Corredig M.
      • Miglior F.
      Variation in fat globule size in bovine milk and its prediction using mid-infrared spectroscopy.
      ), indicating that the lactation effect on MFG size is not yet fully clear.
      Figure thumbnail gr5
      Figure 5Results of principal component (PC) analysis of bovine colostrum fat globule size distributions in the interval 1 to 100 µm. (A) The weights (eigenvectors) of different fat globule size subintervals' frequencies in the first 2 principal components. (B) Correlations of the first 2 PC scores with milk fatty acids and fatty acid indices. (C–E) Location of samples according to their first 2 PC scores sorted by milking, lactation number, and cow. Each sample is marked with its group symbol, and samples from the same group are surrounded by a line; higher values in boxes mark the groups' centroids and P-values denote the significance of the factors according to the linear mixed model, considering fixed effects of milking and lactation number as well as random effect of cow. For the random cow effect, the proportion of variance of PC scores considered by the cow effect (intraclass correlation coefficient) is presented; for both PC, separate models were fitted.

      CONCLUSIONS

      This study provided the first evidence about bovine colostrum CFG size distribution. We showed that the CFG size almost doubled during the first 4 postpartum milkings, whereas lactation number had little effect on CFG size. The FA profile was similar in the first 3 milkings and became distinctive in the fourth milking. Correlation analyses between CFG size and FA showed medium strength relationships, and these also indicated that the fourth milking differed from previous milkings. Our results suggested that up to the third postpartum milking, milk can be considered colostrum; however, at the fourth milking, the milk should be considered transition milk. This knowledge enables more efficient and targeted production of colostrum-based foods and food supplements. Furthermore, our study suggested that size-based segregation of CFG allows for the modification of colostrum products FA profile.

      ACKNOWLEDGMENTS

      We are grateful to D. Arney (Estonian University of Life Sciences, Tartu) and K. Koorem (University of Tartu, Tartu) for comments on earlier versions of the manuscript. The publication of this work was supported by the European Union's Horizon 2020 research and innovation programme under grant agreement No. 810630 “ERA Chair for Food (By-) Products Valorization Technologies of the Estonian University of Life Sciences (VALORTECH)” and the Estonian University of Life Sciences research and development base financing (P170195VLTQ). The authors have not stated any conflicts of interest.

      REFERENCES

        • Argov N.
        • Lemay D.G.
        • German B.
        Milk fat globule structure and function: Nanoscience comes to milk production.
        Trends Food Sci. Technol. 2008; 19 (24363495): 617-623
        • Baum S.J.
        • Kris-Etherton P.M.
        • Willett W.C.
        • Lichtenstein A.H.
        • Rudel L.L.
        • Maki K.C.
        • Whelan J.
        • Ramsden C.E.
        • Block R.C.
        Fatty acids in cardiovascular health and disease: A comprehensive update.
        J. Clin. Lipidol. 2012; 6 (22658146): 216-234
        • Bobe G.
        • Hammond E.
        • Freeman A.
        • Lindberg G.
        • Beitz D.
        Texture of butter from cows with different milk fatty acid compositions.
        J. Dairy Sci. 2003; 86 (14594230): 3122-3127
        • Carroll S.M.
        • DePeters E.J.
        • Taylor S.J.
        • Rosenberg M.
        • Perez-Monti H.
        • Capps V.A.
        Milk composition of Holstein, Jersey, and Brown Swiss cows in response to increasing levels of dietary fat.
        Anim. Feed Sci. Technol. 2006; 131: 451-473
        • Contarini G.
        • Povolo M.
        • Pelizzola V.
        • Monti L.
        • Bruni A.
        • Passolungo L.
        • Abeni F.
        • Degano L.
        Bovine colostrum: Changes in lipid constituents in the first 5 days after parturition.
        J. Dairy Sci. 2014; 97 (24931528): 5065-5072
        • Couvreur S.
        • Hurtaud C.
        Relationships between milks differentiated on native milk fat globule characteristics and fat, protein and calcium compositions.
        Animal. 2017; 11 (27485694): 507-518
        • Dunn A.
        • Ashfield A.
        • Earley B.
        • Welsh M.
        • Gordon A.
        • Morrison S.J.
        Evaluation of factors associated with immunoglobulin G, fat, protein, and lactose concentrations in bovine colostrum and colostrum management practices in grassland-based dairy systems in Northern Ireland.
        J. Dairy Sci. 2017; 100 (28088405): 2068-2079
        • Duplessis M.
        • Mann S.
        • Nydam D.V.
        • Girard C.L.
        • Pellerin D.
        • Overton T.R.
        Folates and vitamin B12 in colostrum and milk from dairy cows fed different energy levels during the dry period.
        J. Dairy Sci. 2015; 98 (26094219): 5454-5459
        • Edén J.
        • Dejmek P.
        • Löfgren R.
        • Paulsson M.
        • Glantz M.
        Native milk fat globule size and its influence on whipping properties.
        Int. Dairy J. 2016; 61: 176-181
        • Elfstrand L.
        • Lindmark-Mansson H.
        • Paulsson M.
        • Nyberg L.
        • Akesson B.
        Immunoglobulins, growth factors and growth hormone in bovine colostrum and the effects of processing.
        Int. Dairy J. 2002; 12: 879-887
        • Fleming A.
        • Schenkel F.S.
        • Chen J.
        • Malchiodi F.
        • Ali R.A.
        • Mallard B.
        • Sargolzaei M.
        • Corredig M.
        • Miglior F.
        Variation in fat globule size in bovine milk and its prediction using mid-infrared spectroscopy.
        J. Dairy Sci. 2017; 100 (28088404): 1640-1649
        • Garcia C.
        • Antona C.
        • Robert B.
        • Lopez C.
        • Armand M.
        The size and interfacial composition of milk fat globules are key factors controlling triglycerides bioavailability in simulated human gastro-duodenal digestion.
        Food Hydrocoll. 2014; 35: 494-504
        • Jaakamo M.J.
        • Luukkonen T.J.
        • Kairenius P.K.
        • Bayat A.R.
        • Ahvenjärvi S.A.
        • Tupasela T.M.
        • Vilkki J.H.
        • Shingfield K.J.
        • Leskinen H.M.
        The effect of dietary forage to concentrate ratio and forage type on milk fatty acid composition and milk fat globule size of lactating cows.
        J. Dairy Sci. 2019; 102 (31421879): 8825-8838
        • Kehoe S.I.
        • Jayarao B.M.
        • Heinrichs A.J.
        A survey of bovine colostrum composition and colostrum management practices on Pennsylvania dairy farms.
        J. Dairy Sci. 2007; 90 (17699028): 4108-4116
        • Laakso P.
        • Manninen P.
        • Mäkinen J.
        • Kallio H.
        Postparturition changes in the triacylglycerols of cow colostrum.
        Lipids. 1996; 31 (8882972): 937-943
        • Logan A.
        • Auldist M.
        • Greenwood J.
        • Day L.
        Natural variation of bovine milk fat globule size within a herd.
        J. Dairy Sci. 2014; 97 (24835967): 4072-4082
        • Lopez C.
        • Briard-Bion V.
        • Ménard O.
        • Beaucher E.
        • Rousseau F.
        • Fauquant J.
        • Leconte N.
        • Robert B.
        Fat globules selected from whole milk according to their size: Different compositions and structure of the biomembrane, revealing sphingomyelin-rich domains.
        Food Chem. 2011; 125: 355-368
        • Martini M.
        • Altomonte I.
        • Salari F.
        The lipid component of Massese ewes' colostrum: Morphometric characteristics of milk fat globules and fatty acid profile.
        Int. Dairy J. 2012; 24: 93-96
        • Martini M.
        • Salari F.
        • Altomonte I.
        The macrostructure of milk lipids: The fat globules.
        Crit. Rev. Food Sci. Nutr. 2016; 56 (24915408): 1209-1221
        • Mele M.
        • Macciotta N.P.P.
        • Cecchinato A.
        • Conte G.
        • Schiavon S.
        • Bittante G.
        Multivariate factor analysis of detailed milk fatty acid profile: Effects of dairy system, feeding, herd, parity, and stage of lactation.
        J. Dairy Sci. 2016; 99 (27665132): 9820-9833
        • Ménard O.
        • Ahmad S.
        • Rousseau F.
        • Briard-Bion V.
        • Gaucheron F.
        • Lopez C.
        Buffalo vs. cow milk fat globules: Size distribution, zeta-potential, compositions in total fatty acids and in polar lipids from the milk fat globule membrane.
        Food Chem. 2010; 120: 544-551
        • Michalski M.C.
        • Briard V.
        • Michel F.
        • Tasson F.
        • Poulain P.
        Size distribution of fat globules in human colostrum, breast milk, and infant formula.
        J. Dairy Sci. 2005; 88 (15905422): 1927-1940
        • Michalski M.C.
        • Ollivon M.
        • Briard V.
        • Leconte N.
        • Lopez C.
        Native fat globules of different sizes selected from raw milk: Thermal and structural behavior.
        Chem. Phys. Lipids. 2004; 132 (15555609): 247-261
        • Mootse H.
        • Pisponen A.
        • Pajumägi S.
        • Polikarpus A.
        • Tatar V.
        • Sats A.
        • Poikalainen V.
        Investigation of casein micelle particle size distribution in raw milk of Estonian Holstein dairy cows.
        Agron. Res. (Tartu). 2014; 12: 153-158
        • Murray C.F.
        • Leslie K.E.
        Newborn calf vitality: Risk factors, characteristics, assessment, resulting outcomes and strategies for improvement.
        Vet. J. 2013; 198: 322-328
        • O'Callaghan T.F.
        • O'Donovan M.
        • Murphy J.P.
        • Sugrue K.
        • Mannion D.
        • McCarthy W.P.
        • Timlin M.
        • Kilcawley K.N.
        • Hickey R.M.
        • Tobin J.T.
        Evolution of the bovine milk fatty acid profile – From colostrum to milk five days post parturition.
        Int. Dairy J. 2020; 104104655
        • Palmquist D.L.
        Milk fat: Origin of fatty acids and influence of nutritional factors there on.
        in: Fox P.F. McSweeney P.L.H. Advanced Dairy Chemistry, Volume 2: Lipids. Springer, 2006: 43-92
        • Playford R.J.
        • Weiser M.J.
        Bovine colostrum: Its constituents and uses.
        Nutrients. 2021; 13 (33477653): 265
        • Poppitt S.D.
        • Keogh G.F.
        • Mulvey T.B.
        • McArdle B.H.
        • MacGibbon A.K.
        • Cooper G.J.
        Lipid-lowering effects of a modified butter-fat: A controlled intervention trial in healthy man.
        Eur. J. Clin. Nutr. 2002; 56 (11840182): 64-71
        • Przybylska J.
        • Albera E.
        • Kankofer M.
        Antioxidants in bovine colostrum.
        Reprod. Domest. Anim. 2007; 42 (17635778): 402-409
        • Reh W.A.
        • Maga E.A.
        • Collette N.M.B.
        • Moyer A.
        • Conrad-Brink J.S.
        • Taylor S.J.
        • DePeters E.J.
        • Oppenheim S.
        • Rowe J.D.
        • BonDurant R.H.
        • Anderson G.B.
        • Murray J.D.
        Hot topic: Using a stearoyl-CoA desaturase transgene to alter milk fatty acid composition.
        J. Dairy Sci. 2004; 87 (15377629): 3510-3514
        • Sangild P.T.
        Uptake of colostral immunoglobulins by the compromised newborn farm animal.
        Acta Vet. Scand. Suppl. 2003; 98 (15259783): 105-122
        • Sats A.
        • Kaart T.
        • Poikalainen V.
        • Aare A.
        • Lepasalu L.
        • Andreson H.
        • Jõudu I.
        Bovine colostrum whey: Postpartum changes of particle size distribution and immunoglobulin G concentration at different filtration pore sizes.
        J. Dairy Sci. 2020; 103 (32534920): 6810-6819
        • Singh H.
        • Gallier S.
        Nature's complex emulsion: The fat globules of milk.
        Food Hydrocoll. 2017; 68: 81-89
        • Uken K.L.
        • Schäff C.T.
        • Vogel L.
        • Gnott M.
        • Dannenberger D.
        • Görs S.
        • Tuchscherer A.
        • Tröscher A.
        • Liermann W.
        • Hammon H.M.
        Modulation of colostrum composition and fatty acid status in neonatal calves by maternal supplementation with essential fatty acids and conjugated linoleic acid starting in late lactation.
        J. Dairy Sci. 2021; 104: 4950-4969
        • Ulbricht T.L.V.
        • Southgate D.A.T.
        Coronary heart disease: Seven dietary factors.
        Lancet. 1991; 338 (1681350): 985-992
        • Walter L.
        • Finch S.
        • Cullen B.
        • Fry R.
        • Logan A.
        • Leury B.J.
        The effect of physiological state, milk production traits and environmental conditions on milk fat globule size in cow's milk.
        J. Dairy Res. 2019; 86 (31722760): 454-460
        • Wiking L.
        • Stagsted J.
        • Björck L.
        • Nielsen J.H.
        Milk fat globule size is affected by fat production in dairy cows.
        Int. Dairy J. 2004; 14: 909-913
        • Yurchenko S.
        • Sats A.
        • Poikalainen V.
        • Karus A.
        Method for determination of fatty acids in bovine colostrum using GC-FID.
        Food Chem. 2016; 212 (27374514): 117-122
        • Yurchenko S.
        • Sats A.
        • Tatar V.
        • Kaart T.
        • Mootse H.
        • Jõudu I.
        Fatty acid profile of milk from Saanen and Swedish Landrace goats.
        Food Chem. 2018; 254 (29548460): 326-332