| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research Communications |
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: We examined whether pentadecanoic acid (15:0) in serum lipids can be used as a marker for intake of milk fat, the major dietary source of 15:0. We also investigated the relations between intake of milk fat and cardiovascular disease risk factors.
Design: Sixty-two 70-y-old men completed 7-d dietary records. The intake of milk products was studied in relation to the proportions of 15:0 in serum cholesterol esters and phospholipids, as well as to the clinical characteristics of these men, by using Spearman's rank correlation.
Results: The proportions of 15:0 in serum cholesterol esters were positively related to butter intake (r = 0.36, P = 0.004) and to the total amount of fat from milk products (r = 0.46, P < 0.0001); 15:0 in phospholipids was related to the amount of fat from milk and cream (r = 0.34, P = 0.008) and to the total amount of fat from milk products (r = 0.34, P = 0.008). Inverse associations were found between intake of milk products and body mass index, waist circumference, LDL-HDL ratio, HDL triacylglycerols, and fasting plasma glucose, whereas relations to HDL cholesterol and apolipoprotein A-I tended to be positive.
Conclusions: The results suggest that 15:0 in serum can be used as a marker for intake of milk fat. The explanation for the inverse associations between the intake of milk products and certain cardiovascular risk factors is not known. Am J Clin Nutr 1999;69:22–9.
Key Words: Pentadecanoic acid • milk fat • biomarker • fatty acid composition • diet • metabolic risk factors • serum • 15:0 • elderly men • Sweden
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The fatty acid compositions of serum and adipose tissue are known to partially reflect the relative fatty acid composition of the diet. This knowledge has been used in several ways: for estimations of the average composition of dietary fat (4, 5), as a control for adherence to a given diet (4, 6), as a complement to dietary surveys (7, 8), and in the evaluation and development of dietary assessment methods (9). The fatty acid composition of dietary fat is reflected by the fatty acids in different compartments of the body at different points in time. Thus, dietary fatty acid composition is reflected in the serum triacylglycerols a few hours after a meal; in erythrocyte membranes, serum phospholipids, and cholesterol esters weeks to months later; and in triacylglycerols of adipose tissue 2–3 y later (10–13). When analyzing the fatty acid composition of tissue samples to estimate the fatty acid composition of the diet, strong correlations have been found between essential unsaturated fatty acids and estimated dietary intake (6, 8, 11, 14, 15). Correlations for nonessential unsaturated and saturated fatty acids (SFAs) are weaker.
Human studies have shown that diets rich in milk and butter are hypercholesterolemic, in contrast with diets with a high content of polyunsaturated fatty acids (PUFAs) (16–19). All fatty acids in milk fat are not equal in their effects on serum lipoproteins and apolipoproteins (16, 18). The SFAs lauric (12:0), myristic (14:0), and palmitic (16:0) acids, which constitute 
40% of the fatty acids in milk fat, are known to be the main cholesterol-raising dietary fatty acids (16, 18). Few studies have addressed the effect of more moderate amounts of milk, butter, and fat from these products on serum cholesterol.
In the rumen of ruminants, the bacterial flora synthesize some fatty acids specific for ruminants, such as pentadecanoic (15:0) and heptadecanoic (17:0) acids (20). Because these fatty acids have an uneven number of carbon atoms, they cannot be synthesized in the human body. A recent study by Wolk et al (21) showed that the proportions of 15:0 and 17:0 in adipose tissue reflected milk fat consumption in women.
The purpose of this project was to study the relation between the estimated intake of fat of dairy origin and the proportion of 15:0 in serum lipids, with the aim of determining whether this fatty acid could be a biochemical marker for intake of milk products. In addition, the relation between the estimated intake of milk and clinical characteristics considered risk factors for ischemic heart disease was investigated.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. There was no significant difference in the cardiovascular risk profile between the 66 men and the whole cohort at the age of 50 y (data not shown). One of the 66 participants stated that he was restricted to a low-fat diet. The design of the study was approved by the Ethics Committee of the Faculty of Medicine of Uppsala University, and all participants gave their written, informed consent.
|
The amount of energy and the macronutrient and fatty acid compositions of the reported diets are shown in Table 2. All data regarding dietary intake used for correlation studies were adjusted for energy intake. The reported intake of milk products and the calculated amount of fat in these products are shown in Table 3.
|
|
Lipoprotein concentrations
VLDL and HDL cholesterol were assayed by a combination of preparative ultracentrifugation (29) and precipitation with a sodium phosphotungstate and magnesium chloride solution (30). Triacylglycerol and cholesterol concentrations in serum and in the isolated lipoprotein fractions were measured by enzymatic methods in a Monarch 2000 centrifugal analyzer (Instrumentation Laboratories, Lexington, MA). The CVs of the analyses of serum cholesterol and serum triacylglycerols were 8.1% and 2.7%, respectively. The concentrations of apolipoprotein A-I and B were measured by immunoturbidimetry with a Monarch apparatus. The CVs of the analyses of apolipoprotein A-I and B were 2.5% and 2.4%, respectively. Apolipoprotein(a) [apo(a)] was determined by an apo(a) radioimmunoassay method [Pharmacia (a) RIA, Pharmacia Diagnostics AB, Uppsala, Sweden], which is based on the direct sandwich technique with 2 monoclonal antibodies specifically directed against apo(a). The CV of the analysis of apo(a) is <5% according to the manufacturer.
Glucose metabolic studies
Insulin sensitivity was determined by the euglycemic hyperinsulinemic clamp procedure according to DeFronzo et al (31), as described in detail by Pollare et al (32). The insulin sensitivity index (M:I) was calculated by dividing the glucose disposal (M; mg
kg-1min-1) by the actual insulin concentration during the test (pmol/L; 1 pmol/L = 0.167 mU/L) multiplied by 100 (I). An oral-glucose-tolerance test was performed by giving 75 g glucose in 300 mL water. Plasma glucose was measured before, and 30, 60, 90, and 120 min after the ingestion of glucose. The results of the oral-glucose-tolerance tests were expressed as the area under the glucose curve.
Blood pressure measurements
Supine systolic and diastolic blood pressures were measured with a mercury manometer after subjects rested for 10 min. The value was recorded twice to the nearest even figure. The mean of these 2 figures was used.
Anthropometric measurements
Body weight was determined to the nearest 0.1 kg and height was measured in centimeters. Body mass index (BMI) was calculated as body weight in kilograms divided by the square of height in meters. Waist circumference was measured midway between the lowest rib and the iliac crest and the hip circumference was measured at the widest part of the hips, both in centimeters. The waist-to-hip ratio was calculated from these 2 measures. Physical activity was grouped into 4 categories on the basis of questions included in a medical questionnaire (33).
Statistical analyses
The results of this investigation were analyzed by using SAS and STATA (Stata Corporation, College Station, TX). Pairwise associations were examined by Pearson product-moment correlation analysis when the data were on a continuous scale, and by Spearman's rank correlation analysis when they were skewed (for example, the intake of most of the milk products) or ordinal (for example, the degree of physical activity). For comparison of the correlation coefficients of the associations with and without adjustments for physical activity, Spearman's rank correlation was used in both of the analyses.
To study whether intake of other groups of food items could possibly confound the investigated associations, we studied the relations of intake of other food items to milk products, to 15:0, and to clinical characteristics. From this analysis we concluded that intake of meat, vegetables, root crops, potatoes, and beer could be possible confounders, and we therefore analyzed the associations between milk products and 15:0, respectively, and clinical characteristics both with and without adjustment for intake of these foods. Correlations in the tables with P values < 0.05 are considered significant; all correlations with P values < 0.1 are included to show statistically significant relations as well as not fully significant trends, which may be of interest.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
).
|
. No significant associations were found between the proportion of 15:0 and intake of cheese or fat from cheese. Relations between the proportion of 15:0 in serum cholesterol esters and phospholipids, respectively, and the estimated dietary intake of 4:0–10:0 and 14:0 were positive (Table 6). Negative associations were found between the proportion of 15:0 in cholesterol esters and the percentage of energy intake as PUFAs. Adjustments for physical activity did not change the associations, except for the one between 15:0 in serum phospholipids and 14:0 in the diet, which was no longer significant.
|
|
). There were also positive relations between 15:0 in cholesterol esters and phospholipids and the long-chain n-3 fatty acids.
|
. After adjustment for physical activity, intake of fat from milk and cream remained significantly related to body weight, waist and hip circumferences, BMI, HDL triacylglycerols, and the results of the oral-glucose-tolerance test (data not shown). The relations between fat from cheese and apo(a), between fat from ice cream and the insulin sensitivity index, and between fat from butter and BMI, HDL cholesterol, LDL-HDL ratio, and apolipoprotein A-I also remained significant after adjustment for physical activity. When adjustments were made for intake of meat, beer, vegetables, and root crops, the intake of fat from milk and cream was still related to body weight, waist circumference, BMI, the waist-to-hip ratio, and the fasting plasma glucose concentration (data not shown). The relations between fat from ice cream and the insulin sensitivity index, and fat from butter and HDL cholesterol, LDL-HDL ratio, and apolipoprotein A-I, respectively, also remained significant. Adjustments for dietary calcium did not alter the relations (data not shown). Although the significance of the relations between intake of milk products and clinical variables were weakened after adjustment for physical activity and intake of other food items, the main pattern of the relations remained.
|
. After adjustment for intake of meat, beer, potatoes, root crops, and vegetables the associations were weakened, but significant inverse associations remained between 15:0 in serum cholesterol esters and body weight and BMI (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The associations between the reported intake of milk products and milk fat, respectively, and the proportion of 15:0 in serum lipids in this study are of the same magnitude as the relations between dietary fatty acids and the proportions of the corresponding fatty acids in serum and plasma shown in several earlier studies (5, 6, 13, 34). For example, Ma et al (5) showed that the Pearson's correlation coefficients between the proportion of 16:0 and 18:2 in plasma and the reported dietary intake of these fatty acids were 0.19 and 0.28 in cholesterol esters and 0.16 and 0.22 in phospholipids, respectively. Ma et al (5) have, as have several other groups, studied the relations between single fatty acids. In contrast, our study was based on intakes of groups of foods rather than on intakes of single fatty acids, which may have weakened the relations. In contrast with all other milk products, we did not find any associations between the proportion of 15:0 and the intake of cheese or fat from cheese. The reason for this is not clear, but during the study it became evident that some of the participants had difficulty adequately describing the fat content of cheese. There were also difficulties in estimating the size of cheese slices and subsequently estimating the amount of cheese eaten.
As mentioned earlier, 17:0 is also found in milk products, but in low concentrations (35). In this study we also investigated whether 17:0 could be used as a biochemical marker for intake of milk products and how it was related to clinical characteristics. However, the concentrations of 17:0 in serum cholesterol esters were too low to detect in 29 of the 66 subjects. The concentrations of 17:0 in serum phospholipids were detectable, but there were no significant relations to milk fat intake or to clinical characteristics in this study.
Fatty acid composition of the diet and serum lipids
The positive associations found between 15:0 and 16:0 shown in Table 7 could be because they are provided by some of the same dietary sources (35). This is certainly the explanation for the associations between 15:0 and 17:0, which both come mainly from milk products (35). This is also the explanation for the positive relations between the proportion of 15:0 in serum lipids and intake of short-chain fatty acids (4:0–10:0) and of 14:0 (Table 7) because these fatty acids also come mainly from milk products (35). According to this reasoning, 15:0 and 18:2 are probably provided by different dietary sources because they were negatively correlated in this study. This negative correlation may have been due to the choice of cooking fat and bread spread.
The positive correlations between 15:0 in serum and intakes of 20:5 and 22:6, the latter 2 of which are mainly from fatty fish or from endogenous synthesis from 18:3n-3, may have different explanations. One might be that persons who eat a lot of milk products also tend to eat a lot of fatty fish. Another possible explanation could be the enzymatic competition for desaturases and elongases during the endogenous synthesis of long-chain PUFAs from 18:1, 18:2, and 18:3n-3. The same enzymes are used in these reactions, but they have the highest affinity for the formation of 20:5 and 22:6 from 18:3n-3 (36, 37). Thus, a high consumption of 18:3n-3 or a low consumption of 18:2, as suggested by the negative correlation between the proportions of 15:0 and 18:2 shown in Table 7, may lead to relatively high proportions of 20:5 and 22:6 in the tissues. This may be true because of more efficient desaturation and elongation of 18:3n-3, even if the intake of these fatty acids has not been very high. The proportions of fatty acids are expressed as relative percentages of the fatty acids analyzed, and not as absolute amounts. Thus, one further possibility is that a result that appears to show an increase in a certain fatty acid is in fact secondary to a pronounced decrease in one or several other fatty acids present in high proportions in the serum.
Milk fat and clinical characteristics
The relations we found between the intake of milk products and several physiologic variables, such as body weight, BMI, waist circumference, and LDL-HDL ratio were negative, whereas the relations to HDL cholesterol and apolipoprotein A-I were positive (Table 8). If these relations reflect real, long-term causal associations, they are hard to reconcile with the opinion that high intakes of milk products, including the main cholesterolemic SFAs 12:0, 14:0, and 16:0 present in milk products, are connected with an increased risk of ischemic heart disease (16, 17, 19, 38). However, the effects of dietary SFAs on serum lipoproteins are complex and probably also depend on several other factors such as genetic disposition, concomitant intake of PUFAs and antioxidants, and other components of food (39). The positive relation between fat from butter and HDL cholesterol seen in this study could be due to the relatively high amounts of SFAs 12:0 and 14:0 in milk fat, fatty acids which have been concluded to increase both total serum and HDL cholesterol (40, 41). Hypocholesterolemic effects of milk products have been observed in some studies (42–45). It is not yet clear what substances in milk products might cause these effects, although several have been discussed, such as hydroxymethylglutaric acid, calcium, orotic acid, lactose, uric acid, or substances in the membrane of the fat particle (42). In several studies the possibly protective role of dietary calcium in the development of atherosclerotic cardiovascular disorders was discussed (46, 47). However, in this study, adjustments for calcium intake did not affect the relations between the intake of dairy fat and metabolic risk factors. This finding indicates that these relations were not explained by the intake of calcium in the dairy products.
It was suggested recently that conjugated linoleic acids, which are found in milk and meat products from ruminants, may have antiatherogenic and serum lipid-lowering effects in rabbits (48). In experimental studies, supplementation with conjugated linoleic acid significantly reduced whole-body fat and increased body protein in rats, mice, and chicks (49). A hypothesis to test could be that the inverse relations in this study between intake of milk fat and proportions of 15:0 in the serum on the one hand and risk factors for ischemic heart disease on the other are partially explained by the concomitant intake of conjugated linoleic acid.
Implication of lifestyle factors
When adjustments were made for physical activity, the relations between the proportion of serum 15:0 and intake of milk fat remained essentially unchanged, indicating that the degree of physical activity does not affect the functioning of 15:0 as a marker of milk fat in the diet. Although the main associations between the intake of milk fat and clinical characteristics also remained after adjustment for physical activity, some of the relations were weakened, indicating that variations in physical activity may confound the associations to some extent. After adjustment for intake of possibly confounding groups of food, such as meat, vegetables, potatoes, root crops, and beer, some of the relations between intake of milk products and serum 15:0 and clinical characteristics lost their significance. The alterations of the associations suggest that the intake of milk products is part of a more complex dietary pattern also characterized by, for example, the intake of meat, vegetables, root crops, and beer. It follows that the relations between the intake of milk products and clinical characteristics shown here should be interpreted with caution.
Study methods
There are some factors in this study that may have confounded the results. First, dietary records seldom show true dietary intake, and should always be interpreted cautiously. However, the imprecision of the dietary data would tend to diminish the strength of the investigated relations. Second, the intake of dairy foods among 70-y-old men may be associated with a health-conscious lifestyle also characterized by good eating habits, lower body weight, and higher physical activity. However, significant associations remained between intake of certain diary products or fat from dairy products and metabolic measures, even after adjustment for other food habits and physical activity. Third, the group of participants was not representative of the total population for age and sex, and other relations may be seen in other populations. A fourth possible confounder was the effect of selection of survivors in a cohort population that has been followed for several years.
Conclusion
From earlier studies it is known that fatty acids, mainly PUFAs, in serum and adipose tissue reflect the fatty acid quality of the habitual diet. The findings in this study of elderly Swedish men suggest that 15:0 in serum can be used as a marker for intake of milk fat at the group level. Although SFAs as a group raise cholesterol concentrations, the effects on the risk of ischemic heart disease of milk fat or dairy products as components of ordinary food need to be studied further.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Supported by The Swedish Dairy Council.
3 Address reprint requests to AEM Smedman, Department of Geriatrics, Unit for Clinical Nutrition Research, Uppsala University, PO Box 609, S-75125 Uppsala, Sweden. E-mail: annika.smedman{at}geriatrik.uu.se.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Saadatian-Elahi, N. Slimani, V. Chajes, M. Jenab, J. Goudable, C. Biessy, P. Ferrari, G. Byrnes, P. Autier, P. H. Peeters, et al. Plasma phospholipid fatty acid profiles and their association with food intakes: results from a cross-sectional study within the European Prospective Investigation into Cancer and Nutrition Am. J. Clinical Nutrition, January 1, 2009; 89(1): 331 - 346. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. L Crowe, N. E Allen, P. N Appleby, K. Overvad, I. V Aardestrup, N. F Johnsen, A. Tjonneland, J. Linseisen, R. Kaaks, H. Boeing, et al. Fatty acid composition of plasma phospholipids and risk of prostate cancer in a case-control analysis nested within the European Prospective Investigation into Cancer and Nutrition Am. J. Clinical Nutrition, November 1, 2008; 88(5): 1353 - 1363. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Sun, J. Ma, H. Campos, and F. B Hu Plasma and erythrocyte biomarkers of dairy fat intake and risk of ischemic heart disease Am. J. Clinical Nutrition, October 1, 2007; 86(4): 929 - 937. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M Hodge, D. R English, K. O'Dea, A. J Sinclair, M. Makrides, R. A Gibson, and G. G Giles Plasma phospholipid and dietary fatty acids as predictors of type 2 diabetes: interpreting the role of linoleic acid Am. J. Clinical Nutrition, July 1, 2007; 86(1): 189 - 197. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. L Crowe, C M. Skeaff, T. J Green, and A. R Gray Serum fatty acids as biomarkers of fat intake predict serum cholesterol concentrations in a population-based survey of New Zealand adolescents and adults. Am. J. Clinical Nutrition, April 1, 2006; 83(4): 887 - 894. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Skeaff, L. Hodson, and J. E. McKenzie Dietary-Induced Changes in Fatty Acid Composition of Human Plasma, Platelet, and Erythrocyte Lipids Follow a Similar Time Course J. Nutr., March 1, 2006; 136(3): 565 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Baylin, M. K. Kim, A. Donovan-Palmer, X. Siles, L. Dougherty, P. Tocco, and H. Campos Fasting Whole Blood as a Biomarker of Essential Fatty Acid Intake in Epidemiologic Studies: Comparison with Adipose Tissue and Plasma Am. J. Epidemiol., August 15, 2005; 162(4): 373 - 381. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Riserus, A. Smedman, S. Basu, and B. Vessby Metabolic effects of conjugated linoleic acid in humans: the Swedish experience Am. J. Clinical Nutrition, June 1, 2004; 79(6): 1146S - 1148S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Andersson, C. Nalsen, S. Tengblad, and B. Vessby Fatty acid composition of skeletal muscle reflects dietary fat composition in humans Am. J. Clinical Nutrition, December 1, 2002; 76(6): 1222 - 1229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Baylin, E. K Kabagambe, X. Siles, and H. Campos Adipose tissue biomarkers of fatty acid intake Am. J. Clinical Nutrition, October 1, 2002; 76(4): 750 - 757. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Wirfalt, B. Hedblad, B. Gullberg, I. Mattisson, C. Andren, U. Rosander, L. Janzon, and G. Berglund Food Patterns and Components of the Metabolic Syndrome in Men and Women: A Cross-sectional Study within the Malmo Diet and Cancer Cohort Am. J. Epidemiol., December 15, 2001; 154(12): 1150 - 1159. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Byberg, A. Smedman, B. Vessby, and H. Lithell Plasminogen Activator Inhibitor-1 and Relations to Fatty Acid Composition in the Diet and in Serum Cholesterol Esters Arterioscler. Thromb. Vasc. Biol., December 1, 2001; 21(12): 2086 - 2092. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. L. Ritzenthaler, M. K. McGuire, R. Falen, T. D. Shultz, N. Dasgupta, and M. A. McGuire Estimation of Conjugated Linoleic Acid Intake by Written Dietary Assessment Methodologies Underestimates Actual Intake Evaluated by Food Duplicate Methodology J. Nutr., May 1, 2001; 131(5): 1548 - 1554. [Abstract] [Full Text]
|
|
|
|
|
|
A. Wolk, M. Furuheim, and B. Vessby Fatty Acid Composition of Adipose Tissue and Serum Lipids Are Valid Biological Markers Of Dairy Fat Intake in Men J. Nutr., March 1, 2001; 131(3): 828 - 833. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
