Methods

Top Methods Results Discussion Author & Article Info References

Participants

Twenty-six men and women (ages, 20 to 63 years) with moderate hypercholesterolemia were recruited from students and employees of the university and surrounding community. After obtaining informed consent, plasma samples were drawn for lipid analysis; persons with a total plasma cholesterol level exceeding 5.69 mmol/L but less than 7.24 mmol/L were asked to participate in the study. Participants were asked to complete a questionnaire relating to health status, provide a qualitative food frequency record, and answer questions regarding the intake of any drugs that would alter lipid-related measures and regarding the presence of any food allergies that would preclude eating the cheeses. Persons were excluded from the study if they had such health problems as diabetes or hypothyroidism, if they were taking drugs that alter lipid metabolism, if they had triglyceride levels exceeding 2.35 mmol/L, or if their diet records indicated that their diets differed markedly from the average American diet (that is, if the energy derived from fat was less than 20% or more than 50%, with the average American fat energy intake defined as approximately 38% [5]).

Protocol

The protocol was approved by the institutional review board at the University of California, Davis. Persons who successfully met the screening criteria were randomly assigned to two different groups of equal sizes. One group started on normal cheese and the other started on modified-fat cheese. Participants and nutritionists were blinded to the cheese type consumed at any stage.

Every 2 weeks, participants were required to complete a 3-day diet record and also to pick up the next 2 weeks' worth of cheese and meet with a nutritionist who would collect diet records, gauge their compliance, and reinforce the need to comply. Blood samples for lipid analysis were obtained at 1-month intervals.

Diets

Because ours was an outpatient, freely feeding crossover diet study, no effort was made to standardize the participants' diets and therefore the diets varied. The nutritionist assessed compliance (that is, adherence of each participant to their self-selected diet) by assessing weight, examining food intake records, and questioning participants about their experience during the previous 2-week period. The nutritionist repeatedly emphasized that participants should not modify their diets except to accommodate the intake of 100 g of cheese daily. In addition, participants were asked and given reminders to maintain a consistent pattern of exercise along with the unaltered food intake, so that their weights would remain stable.

Dietary Analysis

The nutrient composition of the diets, as shown on diet intake records, was analyzed using the Nutritionist III program (N Squared Computing; Silverton, Oregon). The composition of the modified-fat cheese product was added to the food database using information supplied by the manufacturer. The control cheese was a partial skim-milk mozzarella containing 6 g of fat per ounce, 9 g of protein per ounce, 90 calories per ounce, and 60% of calories as fat. The calorie content of the cheese product was identical to that of the control cheese; however, the cheese product had no cholesterol, whereas the control cheese had a cholesterol content of 0.6 mg/g.

Cheese Composition

The experimental, modified-fat cheese was manufactured by removing the animal fat and replacing it with vegetable oil as the fat source. The fatty-acid composition differed from that of the control cheese in that 45% of the fatty acids were polyunsaturated (Table 1).

Cheese Distribution

The cheeses were supplied in individually wrapped, 1-ounce (28-g) portions and were unmarked except for either a clear wrapping or a white backing to distinguish the two types. The cheese products were shipped frozen and stored at –20°C until distributed to participants. The cheese products were distributed in quantities sufficient for a 2-week period. The experimental group was instructed to increase consumption to four sticks of the cheese product daily (approximately 112 g/d), and the control group was asked to increase mozzarella consumption by 96 g/d.

Laboratory Analyses

Plasma samples were obtained between 0800 and 1000 h, after an overnight fast, by venipuncture of the antecubital vein; the drawing procedure was standardized respecting time and technique as suggested by the National Cholesterol Education Program [6]. Blood was drawn into 10-mL evacuated tubes containing 15 mg of sodium EDTA (ethylenediaminetetraacetic acid); the tubes were spun at 2000 revolutions per minute for 15 minutes, and plasma samples were transferred by pipette and stored at 4 °C until analysis, which was usually within 48 hours of drawing. Aliquots of plasma samples were stored frozen for batch analysis of plasma apolipoprotein levels.

Plasma total cholesterol and triglyceride levels were measured using a semi-automated clinical analyzer (Gilford Instruments, Ciba Corning, Oberlin, Ohio). The levels were determined using enzymatic reagents (Ciba Corning Diagnostics Corporation, Oberlin, Ohio). High-density lipoprotein cholesterol was quantified after precipitation with magnesium/dextran sulfate (molecular weight, 50 000 daltons) (Ciba Corning Diagnostics Corporation). Low-density lipoprotein cholesterol levels were calculated using the Friedewald equation. The laboratory is a participant in the National Heart, Lung, and Blood Institute-Centers for Disease Control and Prevention Lipid Standardization Program (LSP 206), and the interassay coefficients of variability of both cholesterol and triglyceride were less than 3%.

Plasma levels of apolipoproteins A-I and B were measured immuno-nephelometrically (Beckman, Brea, California) using reagents supplied by manufacturers. All samples were analyzed in the same assay run for each apolipoprotein. Intra-assay coefficients of variation for apolipoproteins A-I and B were 5% or less.

Statistical Analysis

The effect of cheese intake on study variables (plasma analytes, body weight) and percentage values (for example, dietary intake) were analyzed using a one-way, repeated-measures analysis of variance. Differences in any of the variables due to diet treatment were considered significant if the P value for the effect was less than 0.05 using the Fisher "least significant difference" test.

Results

Top Methods Results Discussion Author & Article Info References

Of the 30 persons who met the inclusion criteria and were initially entered into the study, 26 completed the study. Participants completing the study had a mean (±SE) plasma total cholesterol level of 6.34 ± 0.14 mmol/L. The mean total cholesterol level was 6.29 ± 0.10 mmol/L in men (n =17) and 6.31 ± 0.08 mmol/L in women (n = 9). The 4 persons who dropped out of the study showed no discernible pattern regarding the type of cheese consumed, and reasons for withdrawing from the study varied from the inability or unwillingness to comply with the study requirements to a reaction to the increased dairy products in the diet. The anthropometric data for the 26 participants who completed the study are shown in Table 2. The groups were similar in age and total cholesterol level; hypercholesterolemic men had a slightly higher body mass index. Participants changed neither their calorie intake nor their percentage energy intake as percentage of calories from protein (17.7% ± 0.6% compared with 18.3% ± 0.6%; P > 0.05), carbohydrate (46.5% ± 1.6% compared with 44.6% ± 1.3%; P > 0.05), or fat (34.9% ± 1.1% compared with 35.9% ± 0.9%; P > 0.05). Although the percentage of calories from fat did not change, analysis of diet records (data not shown) demonstrated that intake of polyunsaturated fat increased approximately 13% (P < 0.05). Another important measure of study compliance, body weight, in parallel with the data derived from diet intake analysis, did not change during the study (mean [±SE] starting weight of total study population, 78.5 ± 2.9 kg; mean midpoint weight, 78.9 ± 2.9 kg; mean end-point weight, 79.5 ± 3.0 kg; P > 0.05).

The effects on plasma lipoprotein levels of substituting the different cheese products into participants' diets are shown in Table 3. Overall, total cholesterol and LDL cholesterol levels decreased markedly during consumption of the modified-fat cheese product when compared with levels at both baseline and during consumption of the control cheese. Substitution of normal cheese product did not result in changes in lipid levels relative to levels observed at baseline. However, although men showed a statistically significant decrease in plasma cholesterol during consumption of the modified-fat cheese relative to levels observed either at baseline or during consumption of normal cheese, the decrease in women's total cholesterol values was not statistically significant compared to baseline total cholesterol values. In general, the results in women mirrored those in men, with decreases in LDL cholesterol occurring during modified-fat cheese consumption reaching statistical significance. Other plasma lipid levels did not change appreciably; no statistically significant changes occurred in plasma triglyceride, HDL cholesterol, or apolipoprotein B or A-1 levels (Table 3).

Discussion

Top Methods Results Discussion Author & Article Info References

The reduction of plasma cholesterol levels, particularly the LDL cholesterol level, through dietary changes has been one of the stated goals of both the Adult Treatment Panel of the National Cholesterol Education Program and the Surgeon General [1]. However, the best way to meet these goals remains unclear. Although the relation between a low-fat-content diet and a subsequent reduction in plasma cholesterol levels is clear, the effects of more moderate alterations in the amount of fat and the types of fatty acids present remain ill defined [2-4, 7-12]. Recent studies suggest that diets high in monounsaturated fats will lower plasma cholesterol levels, whereas diets particularly enriched in polyunsaturated fats result in a decline in not only total and LDL cholesterol levels but also in the HDL cholesterol level. The decline in HDL cholesterol has generated concern because an inverse relation between the HDL level and heart disease [13] has been shown.

This 16-week crossover outpatient feeding study showed that total (6%) and LDL cholesterol (9%) levels decreased during consumption of a modified-fat cheese product (100 g/d) when compared with levels observed during consumption of normal cheese (100 g/d) (P = 0.003 and P = 0.007, respectively, for the group as a whole). Both men and women, when analyzed separately, showed similar effects (men: P = 0.04 and P = 0.02; women: P = 0.046 and P = 0.017, respectively). When analyzed relative to baseline levels, both total and LDL cholesterol levels decreased in the group as a whole (P = 0.02 and P = 0.0024, respectively) and in men (P = 0.04 and P = 0.03, respectively); however, in women, only the LDL cholesterol level showed a statistically significant decrease (P = 0.03). No changes in HDL levels relating to cheese consumption were found. No differences were found between baseline lipid levels and those observed when normal cheese was ingested. The reductions found in our study are similar to the decrease (8%) in total cholesterol reported for a group of hypercholesterolemic young men on a American Heart Association step I diet [4]. Another group in that same study was placed on an American Heart Association type I diet that was increased in the percentage of fat (monounsaturated) and showed a decline in total cholesterol of 10.4%.

The decline or reduction in plasma cholesterol levels does not appear to be attributable to differences in dietary cholesterol intake. The average cholesterol consumption for the diet including the control cheese was 246 ± 36 mg/d compared with 226 ± 23 mg/d for the diet including the modified-fat cheese product (P > 0.2). Such a decline in cholesterol intake should not appreciably affect plasma cholesterol levels because changes solely in dietary cholesterol content without concomitant changes in fatty-acid content result in only modest changes in plasma cholesterol levels [10, 11, 14]. Schaefer and colleagues [9] have reported that in normocholesterolemic persons, a diet low in cholesterol (250 to 300 mg), with a ratio of polyunsaturated fatty acid to saturated fatty acid of 0.2 and with fats making up 40% of the kcal intake (that is, a normal American diet), did not reduce plasma cholesterol levels. The ratio of polyunsaturated to saturated fatty acid for our sample's diet based on diet records was approximately 0.7 and did not change significantly during normal cheese and modified-fat cheese consumption.

Our study design—an outpatient crossover feeding study—presented difficulties in assuring compliance and properly controlling all study variables [15]. However, the validity of our findings is bolstered by the stability and congruence of the study records and results. Both men and women showed similar patterns of change in plasma lipid and lipoprotein levels. Both total calories and percentage of calories as fat, carbohydrate, and protein calculated from food record surveys remained unchanged. Further, the intake of polyunsaturated fatty acids increased only during consumption of modified-fat cheese. Body weight showed no significant changes, nor were any correlations found between body weight and plasma variables. Thus, the study findings cannot be readily attributed to changes in either body weight or caloric intake. Finally, because plasma cholesterol levels decreased only during consumption of modified-fat cheese and because participants were blinded to the type of cheese they were consuming, the possibility of selective compliance is unlikely; in addition, the exclusion of some other food does not appear to account for the changes observed. However, because the study findings may have resulted from a type II statistical error (that is, small sample size), they should be replicated in a larger population.

The 6% decline (group as a whole) in the total cholesterol level observed during consumption of the cheese product relative to the level observed during consumption of the control cheese is three times ( –0.4 mmol/L compared with –0.13 mmol/L) those reported in a meta-analysis of studies of oat bran as a hypocholesterolemic agent [16]. Given the documented relation between cholesterol lowering and risk for heart disease, this could translate into a 12% reduction, based on the decline in total cholesterol in coronary heart disease risk [17]. The decline in plasma cholesterol level occurred despite habits or genetic constitution predisposing to elevated plasma total and LDL cholesterol levels. No statistically significant alterations were observed in other variables such as body weight and calorie intake that could potentially confound changes or declines in plasma cholesterol levels [18]. However, it must also be emphasized that the reductions observed during consumption of modified-fat cheese did not lower participants' plasma cholesterol levels to National Cholesterol Education Program guideline levels.

Treatments ranging from major surgical interventions to modest lifestyle modifications have been advanced to reduce the risk for heart disease [19]. The relative safety, the low cost, and the reduced frequency of physician visits associated with dietary changes are reasons for emphasizing diet modification [20]. The modified-fat cheese product, other modified food products [21], and foodstuffs such as nuts, recently linked to reduced coronary heart disease [22], may be more effective because their palatability enhances patient compliance. Thus, modified-fat cheese products may be useful tools in lowering plasma cholesterol levels, both in specific patients and in the population as a whole.