Although the cardiovascular health of adults in Australia and other developed countries has improved greatly in the last decades, cardiovascular disease (CVD) is still a leading cause of premature death and morbidity [1
]. Poor diet is a modifiable behavior that can lead to risk factors for CVD, such as obesity, hypertension, hyperlipidemia, and type 2 diabetes. Dietary guidelines for the prevention of CVD recommend limiting fat intake, especially saturated fat, to reduce blood cholesterol concentration and body weight [2
Red meat is thought to be associated with the development of CVD due to the high proportion of saturated fat. However, after trimming visible fat, red meat such as beef and lamb contains up to 5% total fat [5
], which is no more than that of some white meats [6
]. Studies examining the association between red meat consumption and CVD risk have reported inconsistent results. The Nurses’ Health Study reported that red meat consumption was positively associated with an increased risk of nonfatal myocardial infarction or fatal coronary heart disease (CHD) [7
] whereas a meta-analysis in 2010 found no association between consumption of red meat and incidence of CHD or diabetes risk [8
Dairy products are also believed to increase CVD risk as the fat in dairy products is mainly saturated. However, different saturated fats have different effects on plasma lipids and lipoproteins and some may protect against CVD risk factors [9
]. The food matrix may also modify the effect of dairy fat on CVD outcomes [10
]. A recent review, funded by the dairy industry, found that most studies reported either an inverse association or no association between consumption of dairy products and CVD risk [12
]. Studies comparing the effect of full fat and reduced fat dairy on CVD risk are limited [12
Oils derived from plants have been promoted as being healthier than animal fat, with recommendations to replace animal fats with oil when cooking [13
]. The composition of monounsaturated and polyunsaturated fatty acids varies among different vegetable oils. Replacing saturated fatty acids with polyunsaturated fatty acids, found in plant-based oils, can result in an increased risk of death from CVD if the polyunsaturated fats are not omega-3s [14
]. This suggests that substitution of animal fat with vegetable oils may not necessarily reduce CVD risk.
Most studies have examined whether following advice for fat-related dietary behaviors is associated with lower risk of CVD among middle-aged to elderly individuals. It is important to determine whether the same associations exist among young adults. This study examined whether young adults who usually trimmed fat from meat, consumed low-fat products, and used oil for cooking had lower CVD risk than those who did not follow these practices. CVD risk was assessed using a continuous metabolic syndrome (cMetSyn) score.
In total, 2084 participants answered the food habits questionnaire and had data required to calculate the cMetSyn score. Participants were excluded if they were pregnant (n = 7) or used lipid lowing medication (n = 6), leaving 2071 participants for the analysis.
The characteristics of the participants are presented in Table 1
. Compared with the Australian population of 25–34-year-olds, the CDAH sample had a higher percentage of participants who were married or living as married (57% men, 64% women in the general population [27
]) and employed as professionals or managers (40% men, 38% women in the general population [28
]). The percentage of participants who were overweight or obese (BMI ≥ 25 kg/m2
) was comparable to the Australian population of similar age (58% men, 35% women [29
]). Mean (SD) cMetSyn score was −0.01 (−0.70) for men and −0.01 (−0.71) for women.
The analysis was stratified by weight status due to significant interactions between weight status and all fat-related behaviors. Most participants reported usually trimming fat from meat before or after cooking (Table 2
, 63% healthy weight, 78% overweight/obese). Participants who were overweight or obese tended to consume low-fat products more frequently than those who were healthy weight. Most participants usually used olive oil for cooking (62% healthy weight, 60% overweight/obese) whereas the proportion of participants who usually used canola oil (10% healthy weight, 10% overweight or obese) or butter (8% healthy weight, 7% overweight or obese) was low. Compared to healthy weight participants (3.3%), a higher percentage of those who were overweight or obese (8.2%) reported they were on a weight loss diet, whereas the percentage on a fat-modified diet was similar (11.9% healthy weight, 10.6% overweight/obese).
Among participants who were healthy weight, frequency of using low-fat oily dressing was the only fat-related behavior that was significantly associated with cMetSyn (Table 3
). Participants who never or rarely used low-fat oily dressing had lower cMetSyn scores than those who usually used it. The association remained significant after adjusting for education, daily fruit and vegetable consumption, weekly alcohol consumption, smoking status and intake of oil and creamy dressings. Further adjustment for diet quality, being on a weight loss or fat-modified diet did not attenuate the results.
Among participants who were overweight or obese, participants who usually used canola/sunflower oil for cooking had higher cMetSyn scores than those who never/rarely used it (Table 3
). The association remained after adjusting for smoking status and takeaway food consumption, and after adjusting for diet quality, being on a weight loss or fat-modified diet in the sensitivity analyses. A significant sex interaction was found for cooking with butter among those who were overweight or obese, with a trend for a higher score among males who more frequently used butter for cooking. Participants who were overweight or obese and usually consumed whole milk had higher cMetSyn scores than those who usually consumed skim or low/reduced fat milk. However, the association was attenuated and no longer statistically significant after adjusting for education and occupation.
This study examined whether fat-related eating behaviors (trimming fat from meat, consuming low-fat dairy, type of fat used for cooking) were associated with a cMetSyn score among young Australian adults. Healthy fat-related dietary behaviors were reported by a higher percentage of participants who were overweight or obese than those who were healthy weight. This may be a result of reducing fat intake to try to lose weight or greater pressure to report socially desirable answers. The cMetSyn score was higher among healthy weight participants who reported more frequent use of low-fat oily dressings and among participants who were overweight or obese and used canola oil more frequently for cooking. The other fat-related dietary behaviors were not associated with cMetSyn.
Trimming fat from meat was not associated with the cMetSyn score. This findings is consistent with previous studies that have reported no significant relationship between intake of animal fat [30
] or saturated fat [33
] and the risk of CHD. In a meta-analysis of 17 cohort studies and three case-control studies, red meat intake was not associated with CHD or Type 2 diabetes [8
]. It is worth noting in that study the effects of lean and non-lean cuts of meat were not examined separately. Other studies that investigated lean red meat also found no negative effects on lipid levels, thrombotic risk factors, markers of oxidative stress or blood pressure [34
]. In contrast, a recent review of systematic reviews and meta-analyses recommended reducing meat-derived saturated fat for CVD health [4
Among healthy weight participants, more frequent use of low-fat oily dressing was associated with a higher cMetSyn score. Oily dressings are usually high in vegetable oils, which are recommended for cardiometabolic health. Although low-fat dressings usually have a lower energy content, the healthy oils have been replaced with sugar, starch and salt, components which are detrimental to cardiometabolic health [35
Our study found using olive oil for cooking was not associated with cMetSyn score. These results contrast those from studies of older participants that found higher olive oil to associate with better cardiometabolic health. In the PREvención con DIeta MEDiterránea (PREDIMED) Study (N
= 7216, 55–80 years old), the risk of developing CVD during the 4.8 year follow-up was 35% lower among participants with the highest baseline olive oil consumption, compared to those in the lower third [36
]. In a Spanish cohort of the European Prospective Investigation into Cancer and Nutrition (EPIC) study (N
= 40,142, 29–69 years old) [37
], olive oil, especially extra virgin olive oil, was related to a reduced risk of CHD events during a mean follow-up of 10.4 years. A meta-analysis of nine studies found an inverse relationship between olive oil consumption and stroke (and with stroke and CHD combined), but no association was found for CHD [38
]. In a cross-sectional study of Spanish adults (N
= 4572, 18 years and older), olive oil consumption was associated with a lower risk of obesity, impaired glucose regulation and hypertriglyceridemia than consumption of sunflower oil [39
]. The younger age of our study participants and assessed use of olive oil for cooking, not overall intake, could explain the observed differences with previous findings.
Among participants who were overweight or obese, higher consumption of canola/sunflower oil was associated with higher mean cMetSyn score. These findings are in contrast with a review, funded by the Canola industry [40
], in which most studies found canola oil was associated with beneficial effects on biomarkers of CVD risk (total cholesterol, LDL-cholesterol, and inflammatory biomarkers) compared with saturated fat. Sunflower oil appeared to have similar effects on plasma lipids as canola oil whereas olive oil was less protective [40
]. One potential reason for the conflicting findings is that most of the studies included in that review used raw oil which restricts the interpretation of these studies, because canola oil is mostly used for frying, and heat can cause the loss of some of the beneficial components such as α-linolenic acid [41
We found that using butter for cooking was not significantly associated with cMetSyn score. However, the number of participants who usually used butter for cooking was low and there was a trend for more frequent use of butter to be associated with a higher cMetSyn score among males who were overweight or obese. Previous studies have reported that butter consumption was not associated with Ischemic Heart Disease events [42
], incident CVD [43
] or stroke [44
Consumption of low-fat dairy products was not associated with cMetSyn scores after adjusting for potential confounders, which is consistent with recent studies [45
]. In a multicenter cross-sectional study of 4827 French adults aged over 35 years, no association was found between high-fat dairy intake and lipid profile or risk of mortality [45
]. In a Netherlands cohort (N
= 120,852, 55–69 years old), intake of full-fat or low-fat milk products did not predict the risk of Ischemic Heart Disease or stroke in men or women during the 10-year follow-up [46
]. A meta-analysis assessing six prospective cohort studies [47
] also indicated no association between total high-fat and low-fat dairy consumption and the risk of CHD. Among 2907 adults aged 65 years and older from the USA, serial measures of biomarkers for dairy fat were not significantly associated with incident CVD after 22 years of follow-up [48
]. Higher circulating heptadecanoic acid was associated with lower risk of CVD and stroke mortality [48
]. In the Nurses’ Health Study [49
] (80,082 women, aged 34–59 years), no association was found between high-fat or low-fat dairy consumption and risk of major CHD events during the 14-year follow-up; however, a higher ratio of high-fat to low-fat dairy consumption was associated with greater CHD risk. Whole-fat milk was also found to be positively associated with CHD risk in that study.
This study had limitations. Self-reported dietary data may be subject to error due to recall or reporting of socially desirable answers. The present study made no distinctions between canola and sunflower oil, or different varieties of olive oil. Data on participants use of oil was only collected in relation to cooking, not other uses such as dressings. The cross-sectional design means the direction of the association cannot be determined. It is possible that participants who are concerned about their metabolic health (e.g., have been informed they have high cholesterol) or have a family history of CVD may have recently changed their behavior to improve their cardiometabolic health [50
]. In addition, the study focused on components of the harmonized definition of the metabolic syndrome [20
] and did not include other risk factors such as non-alcoholic fatty liver disease, which is strongly associated with cardiometabolic health [51
]. Finally, data was only collected on the frequency of the eating behaviors rather than the amount consumed, therefore the overall fat composition of the diets, such as total, saturated and unsaturated fat intake is unknown. However, estimating portion sizes is difficult and prone to measurement error and the major variance of food intake is explained by frequency alone [52
Strengths of the study include studying a relatively healthy population, which avoids possible treatment and disease effects on the outcomes of interest. Most existing studies have focused on middle-aged and elderly populations whereas this study examined associations between fat-related eating behaviors and CVD risk among young adults. In addition, the CDAH study is a multicenter study with a large sample size from across Australia, with generalizability to healthy young Australian adults. The analyses were adjusted for various lifestyle and dietary factors.