The amino acid content in the daily diet of seniors negatively correlates with the degree of platelet aggregation in a sex- and agonist-specific manner

Relationship between amino acid content in the diet and platelet reactivity in a group of men and women adjusted for age

In the first approach, Spearman’s rank correlation coefficients were calculated as the measure of associations between the estimated amounts of amino acids in food and platelet reactivity (AUC, A_max, AUC*A_max/1000) in the whole group of men and women after adjustment for age.

In this group, significant negative correlations were found between platelet aggregation and the amounts of ingested amino acids. This was true for all of the tested amino acids and for all three measured indices of platelet reactivity: AUC, A_max and (AUC*A_max)/1000. No preference regarding the used platelet agonist was noted, i.e., AA, COL or ADP; in all cases, negative correlations were found between the amounts of amino acids present in the diet and platelet reactivity (Table 1). Somewhat different patterns were noted for male and female subgroups.

Associations between amino acid content in the diet and platelet reactivity among women

Among women, significant negative associations were found between AA-dependent platelet reactivity and amino acid content, when Spearman’s rank correlations coefficients were adjusted for age (Table 2). For COL-induced aggregability, no significant association was observed (Table 2). For ADP-induced aggregability, a more heterogenous pattern of associations was found between the amounts of amino acids in the diet and the variables describing platelet reactivity; however, these associations always appeared as negative. It appears that older women show lower responsiveness of blood platelets to ADP-stimulation when food and beverages are enriched with Met, His, Ala, Asp, Arg and Gly (Table 2).

The female subgroup generally demonstrated less significant changes, where applicable, due to the lower sample size (sample size reduced from 246 to 122 individuals). To confirm the reduction in significance when dividing the sample into subgroups, and to see whether the sample size correction would reverse such a tendency, the subgroup data was analyzed using a bootstrap resampling procedure with an adjustment to the sample size of the overall group (n = 246). It was found that the amounts of all of the tested amino acids in the daily diet are negatively and significantly associated with all the hallmarks of blood platelet aggregability: AUC, A_max or the overall comprehensive measure of (AUC*A_max)/1000, when aggregation of female blood platelets is triggered by arachidonate (Table 3).

The same was found for collagen-stimulated blood platelets, for all amino acids but Arg (no statistical association with AUC, A_max or (AUC*A_max)/1000). Also Gly showed no significant correlations with the overall score of COL-induced platelet reactivity. Some amino acids (Met, Trp, His, Ala, Glu) only demonstrated insignificant negative associations with aggregability (Table 3). Thus, in the female group, Iso, Leu, Lys, Cys, Phe, Tyr, Thr, Val, Asp, Pro and Ser seem to remain the most effective inhibitors of collagen-induced platelet reactivity.

Lastly, in general, the ADP-dependent reactivity of the blood platelets was significantly and negatively associated with amino acid amounts with regard to AUC, A_max or (AUC*A_max/1000). The exceptions were the associations of A_max with Glu, Pro and Ser (p > 0.05). Some amino acids, including Leu, Tyr, Trp and Val, only demonstrated a tentative relationship with A_max (p > 0.05); with similar results being noted for Glu and Pro in the case of (AUC*A_max)/1000 (Table 3).

Associations between amino acid content in the diet and platelet reactivity in the male subgroup

In the male subgroup, no significant relationship was found between amino acid content in the daily diet and the aggregability of blood platelets induced with either AA, COL or ADP, when adjusted for age (Table 4).

As with the female group, the male subgroup was adjusted to the size of the total group (246 individuals) using bootstrap resampling. In the male subgroup, significant negative correlations, or tendencies, were noted between A_max, AUC and A_max*AUC/1000 and amino acids content but only for responsiveness to COL; no such relationships were noted for AA or ADP reactivity. Furthermore, in the case of COL-induced aggregation, the dietary amounts of Cys and Pro showed no relationship with the tested parameters of platelet reactivity (Table 5). Similarly to the female subgroup, these observations remained valid after bootstrap resampling; the procedure did not result in any changes in the observed associations.

Associations between amino acid content and comprehensive scores of platelet reactivity

In addition to the levels of single amino acids, they were grouped into four main sets for analysis: branched-chain amino acids (valine, leucine, isoleucine), sulfur amino acids (cysteine and methionine), exogenous amino acids (lysine, tryptophan, phenylalanine, threonine, valine, leucine, isoleucine, methionine, arginine and histidine) and endogenous amino acids (cysteine, tyrosine, alanine, aspartic acid, glutamic acid, glycine, proline, serine). The mean intakes of these groups were associated with the markers of platelet reactivity in the whole group of the studied individuals (n = 246), as well as in the separate subgroups of men and women.

The results indicate that all of the tested markers of platelet reactivity, i.e., AUC, A_max and (AUC*A_max/1000), were significantly and negatively associated with daily intakes of branched-chain, sulfur, exogenous and endogenous amino acids (Supplementary Table 1).

Following this, the comprehensive scores (CS), a marker of total reactivity of blood platelets, showing ‘summarized’ or ‘global’ reactivity to all of the tested platelet agonists (arachidonate, collagen, ADP) in one common variable, were calculated. The results indicate that ‘global’ AUC, A_max and (AUC*A_max/1000) also demonstrate significant and negative correlations with daily intakes of branched-chain, sulfur, exogenous and endogenous amino acids in older men and women (Supplementary Table 1).

In the examined subgroup of older women, the daily intakes of branched-chain, sulfur, exogenous and endogenous amino acids remained negatively and significantly associated with ‘global’ platelet reactivity; however, the relationships with CS_{total_Amax} and CS_exo remained insignificant. The same associations were observed for the original subgroup of 122 women and when the data was adjusted to the size of full group (n = 246) (Supplementary Table 2). In contrast, no statistically significant associations were found between the hallmarks of the ‘global’ platelet reactivity and the daily intakes of any of the tested amino acid groups in the male subgroup, either for the original subgroup (n = 125) or after adjustment to n = 246. These relationships remained maintained (negative significant associations in women and no significant associations in men) also upon the adjustment of the discussed relationships to the set of confounders including several non-collinear blood morphology and plasma biochemistry variables of acceptable tolerance and showing significant differences between sexes: concentration of haemoglobin (HGB), plateletcrit (PCT), platelet-large cells ratio (P-LCR), white blood cell count (WBC), glucose, total cholesterol, high density lipoproteins (HDL), uric acid, animal protein, plant protein and the amount of energy derived of protein (Supplementary Table 2).

Interestingly, the comprehensive scores calculated for branched-chain, sulfur, exogenous and endogenous amino acids were not correlated with the age of the studied male and female probants (the mixed group of n = 246); however, covariance analysis found all to be positively and significantly associated with male sex (P = 0.00021, P = 0.00003, P = 0.0001 and P = 0.00006 for branched-chain, sulfur, exogenous and endogenous amino acids, respectively). While the ‘global’ intake of branched-chain, sulfur, exogenous or endogenous amino acids was found to be higher in men than women, the ‘global’ platelet reactivity was found to be higher in women than in men (Supplementary Table 3).

In the female subgroup, the total protein content in the diet significantly and negatively correlated with the reactivity of blood platelets to collagen or ADP, but not to arachidonate. However, no such relationship was noted for males. Some relationships were observed for amount of protein per kg of body mass and the content of animal protein in the diet. Neither the amount of plant protein in the diet, nor the amount of energy derived from protein, nor the protein density showed any statistically significant association with any of the tested hallmarks of blood platelet reactivity in all the tested sex groups (Supplementary Table 4).

Chemicals

Arachidonate, equine tendon collagen and ADP were purchased from Chrono-Log Corp. (Havertown, PA, USA). Physiological buffered saline (PBS) was from Avantor Performance Materials Poland S.A. (Gliwice, Poland). For blood sampling we used S-Monovette^® blood collection systems (Sarstedt, Nümbrecht, Germany).

Study design

This study presents the results obtained in subgroups of volunteers participating in the project entitled “The occurrence of oxidative stress and selected risk factors for cardiovascular risk and functional efficiency of older people in the context of workload”, funded by the Central Institute For Labour Protection – National Research Institute (Warsaw, Poland) and supervised by the Clinic of Geriatrics at the Medical University in Lodz (Poland). Participants responded to the announcements offering the participation in the project, which were published on local TV, radio and newspapers. Two basic inclusion criteria were age within the range of 60 to 65 years and the willingness to participate. The recruitment criteria have been described in more detail in an earlier report [45]. The research group included roughly 300 subjects (equal sex proportions), aged from 60 to 65 years (group-matched age distribution). After excluding volunteers currently taking antiplatelet drugs, we achieved a target population of 122 women and 124 men. The general characteristics of the studied group is presented in Supplementary Table 5.

Blood sampling, isolation of blood plasma, measurements of blood morphology and serum biochemistry

Blood was taken after overnight fasting and 15-minute rest directly before blood donation. Blood was collected by aspiration either to vacuum tubes (Sarstedt, Nümbrecht, Germany) supplemented with 0.105 mol/l buffered sodium citrate (citrate:blood ratio = 1:9, v/v) (for the measurements of platelet activation and reactivity) or to the tubes coated with EDTAK₃ in the case of samples taken for blood morphology analysis, or to tubes without any anticoagulant when serum was taken for further biochemical determinations. In all cases, blood was collected from a peripheral vein cannulated with an 18-gauge needle.

Blood morphology was measured with a 5-Diff Sysmex XS-1000i haematological analyser (Sysmex, Kobe, Japan), while a DIRUI CS 400 analyser (Dirui, Changchun, China) was used to evaluate serum biochemical parameters. In order to obtain blood serum, whole blood taken to the tubes without an anticoagulant was incubated for 30 minutes at 37°C and centrifuged (2000 × g/15 min./4°C). Supernatant (serum) was aspirated and used in further analyses (basic blood serum biochemistry).

Whole blood impedance aggregometry

Platelet aggregability was determined with a MultiPlate Analyzer (Dynabayte, Munich, Germany), as described previously [46]. In brief, samples of citrated blood were left a 10-minute blood reposition at 37°C to avoid the interference of artefactual platelet activation caused by aspiration. The 300 μl aliquots of whole blood were mixed with an equal volume of PBS, gently mixed and left at 37°C for three minutes. Each sample was supplemented with the respective agonist: 0.5 mmol/l arachidonate, 1 μg/ml collagen or 10 μmol/l ADP in order to trigger platelet aggregation. The recording of platelet clumping started immediately thereafter and was tracked for 15 minutes. In this paper we employed three measures of blood platelet reactivity in response to the agonists (arachidonate, collagen, ADP). The area under the aggregation curve (AUC) and maximal aggregation (A_max) were used as two of three characteristic variables describing platelet aggregability. Those two variables give the optimal measure of the extent of platelet aggregation: how high it is when platelets respond maximally, and how long it lasts before disaggregation starts. These two measures together are much more reliable measures of the extent of platelet aggregation than just one. And it is more difficult to justify selecting only one of these variables for the final description of aggregation. The third one was a derivative variable, calculated according to the following formula: (AUC*A_max)/1000, combining the previous two to get the most complete description of platelet aggregation.

Amino acid intake

The daily amino acid intake was calculated based on a detailed analysis of the participants' menu. The intake of food and beverages during the preceding day was estimated by qualified investigators (dietitian and nutritionist) on the basis of a 24-h recall questionnaire using a portion size album. This interview was collected from three days, and then the menu that was most representative of the participant's typical diet was taken for further analysis.

The method of interview can be successfully used for dietary evaluation in the group of elderly people [47]. To decrease the risk of errors from dietary recall, participants were asked to prepare a list of food products (including snacks and beverages) they ate on the day before the appointment. Interviewers were instructed not to judge the diet of the respondents. A photo album was used to determine how many grams the eaten portion had. None of the study participants used protein supplements. The nutritional products declared by patients were analyzed with Dieta 5.0 software (The National Food and Nutrition Institute, Warsaw, Poland). This software platform allows calculation of energy intake and nutrients consumption. We estimated the amounts of the following amino acids: isoleucine (Ile), leucine (Leu), lysine (Lys), methionine (Met), cysteine (Cys), phenylalanine (Phe), tyrosine (Tyr), threonine (Thr), tryptophane (Trp), valine (Val), arginine (Arg), histidine (His), alanine (Ala), aspartic acid (Asp), glutamic acid (Glu), glycine (Gly), proline (Pro) and serine (Ser). Amount of the amino acids [mg] represents the level of amino acid consumption with the diet during the last 24 hours.

Statistical analysis

The continuous data were presented as either mean ± SD or median with interquartile range (from lower [25%] to upper [75%] quartile). In the case of categorical data, they are presented as percentages. Cumulative measures of amino acid intake were calculated based on the van der Waerden normal scores of individual amino acids for the following amino acid groups: branched-chain amino acids, sulfur containing amino acids, exogenous amino acids and endogenous amino acids. There were referred to as the relevant comprehensive scores: CS_BCAA, CS_sulfur, CS_exo and CS_endo, respectively. Likewise, the van der Waerden normal scores of blood platelet aggregation, cumulated through the three employed blood platelet agonists, collagen, ADP and arachidonate, were referred to as the overall scores of ‘cumulative blood platelet aggregation’, CS_{total_Amax}, CS_{total_AUC} and CS_{total_Amax*AUC}.

Data normality was tested with the Shapiro-Wilk W test and homogeneity of variances with Levene’s test. In the case when the assumptions of data normality and variance homogeneity were not violated, the groups were compared with a non-paired t-Student’s test, otherwise, the U-Mann-Whitney test was used. The associations between variables were calculated as either the partial correlation Pearson’s coefficients with adjustment for compounding variables (multiple regression, partial correlation) or as the Spearman’s rank correlation coefficients. For the purpose of comparisons between subgroups, in order to adjust the monitored variables for the confounders, like age, we used the analysis of covariance (ANCOVA). As a standard, due to relatively small sample sizes in the subpopulations studied and in order to ensure the sufficient statistical power of estimated inferences and associations, we employed the technique of resampling bootstrap (with 5,000 or 10,000 iterations) to minimize that the revealed differences/associations could be observed by pure chance. This approach was also used to estimate the associations in sexual subgroups of individuals upon the adjustment of subgroup sizes (n₁ = 122 for women and n₂ = 124 for men) to the sample size of the overall studied population (n = 246). In such situations, we refer in the description of the outcomes to the bootstrap-boosted test statistics instead of the classical approach. Statistical analyses were performed using Statistica v.13 (Dell Inc., Tulsa, OH, USA) and Resampling Stats Add-In for Excel v.4 (The Institute for Statistics Education, An Elder Research Company, Arlington, VA, USA).