Intake of dietary plant sterols is inversely related to serum cholesterol concentration in men and women in the EPIC Norfolk population: a cross-sectional study


Objective: We examined the relation between intake of natural dietary plant sterols and serum lipid concentrations in a free-living population.

Design, setting and participants: Cross-sectional population-based study of 22 256 men and women aged 39–79 y resident in Norfolk, UK, participating in the European Prospective Investigation into Cancer (EPIC-Norfolk).

Main exposure and outcome measures: Plant sterol intake from foods and concentrations of blood lipids.

Results: Mean concentrations of total cholesterol and low-density lipoprotein cholesterol, adjusted for age, body mass index and total energy intake, decreased with increasing plant sterol intake in men and women. Mean total serum cholesterol concentration for men in the highest fifth of plant sterol intake (mean intake 463 mg daily) was 0.25 mmol/l lower and for low-density lipoprotein cholesterol 0.14 mmol/l lower than those in the lowest fifth of plant sterol consumption (mean intake 178 mg daily); the corresponding figures in women were 0.15 and 0.13 mmol/l. After adjusting for saturated fat and fibre intakes, the results for total cholesterol and low-density lipoprotein cholesterol were similar, although the strength of the association was slightly reduced.

Conclusions: In a free-living population, a high intake of plant sterols is inversely associated with lower concentrations of total and low-density lipoprotein serum cholesterol. The plant sterol content of foods may partly explain diet-related effects on serum cholesterol concentration.


Coronary heart disease (CHD) is a leading cause of death in industrialised societies. Raised serum cholesterol concentration is a well-established risk factor estimated to explain four-fifths of the geographic variation in CHD mortality (Law & Wald, 1994). A meta-analysis (Gould et al, 1995) has estimated that a reduction of serum cholesterol by 10% would reduce cardiovascular death by 13% and total mortality by 10%.

It is generally accepted that a high content of saturated fat in the diet increases serum cholesterol, while polyunsaturated fats reduce cholesterol. This has been shown in studies using controlled diets (Clarke et al, 1997), and to a lesser extent in free-living populations (Tang et al, 1998). A high content of soluble dietary fibre in the diet is also considered to have a cholesterol-reducing effect (Brown et al, 1999). Other factors in the diet apart from fat and fibre could also influence serum cholesterol. These may include the bioactive plant sterols.

Plant sterols, found in foods of plant origin, have a structural similarity to cholesterol and reduce the serum cholesterol concentration by inhibiting the absorption of both dietary and biliary cholesterol in the small bowel (Wilson & Rudel 1994). The serum cholesterol-lowering effect of esterified plant stanols or sterols in pharmaceutical doses has been reviewed (Moghadasian & Frohlich, 1999; Katan et al, 2003). Naturally-occurring plant sterols in a mixed experimental diet are inversely correlated with cholesterol absorption (Ellegard et al, 2000), but to date, no study has looked at the effect of naturally occurring plant sterols from the habitual diet on serum cholesterol concentrations in a free-living population (Jones and Raeini-Sarjaz, 2001).

We examined the relationship between intake of plant sterols from naturally occurring dietary sources and serum cholesterol concentrations in men and women living in the general community.


The study population

The European Prospective Investigation into Cancer (EPIC-Norfolk) prospective population study of approximately 25 000 men and women in Norfolk, UK, aged between 45 and 79 y at baseline, were recruited between 1993 and 1997(Day et al, 1999). All participants gave informed consent. At the baseline survey, participants completed a health and lifestyle questionnaire and took part in a health examination. The participants were recruited as part of a nine-country collaborative study (EPIC, European Prospective Investigation into Cancer and Nutrition) designed to investigate dietary and other determinants of cancer. Additional data were collected for the EPIC-Norfolk cohort to enable assessment of other end points including cardiovascular disease. The cohort is similar to UK population samples with respect to many characteristics including anthropometry, blood pressure and lipids, but there is a lower proportion of current smokers (Day et al, 1999).


Trained nurses measured height with a stadiometer to the nearest 0.1 cm. Weight was recorded to the nearest 0.1 kg. Body mass index (BMI) was calculated as weight in kilograms divided by height in metres squared. Participants were classified as current smokers or noncurrent smokers based on the health and lifestyle questionnaire. Physical activity level was categorised as (1) inactive, (2) moderately inactive, (3) moderately active and (4) active where ‘active’ was defined as a sedentary job with >1 h/day recreational activity or standing job with >0.5 h/day recreational activity or physical job with at least some recreational activity or heavy manual job.

Nonfasting blood samples were obtained by venepuncture. Serum concentrations of total cholesterol and high-density lipoprotein (HDL) cholesterol were measured with the RA 1000 Technicon analyser (Bayer Diagnostics, Basingstoke). Low-density lipoprotein (LDL) cholesterol concentration was calculated by using the Friedewald formula, except when the concentration of triglyceride exceeded 4 mmol/l (Friedewald et al, 1972).

Dietary plant sterol intake

A unique database that covers the major sources of plant sterols in the European diet, which includes more than 330 food items, including vegetables, fruits, cereals, bread, fats, nuts and beverages, was used to estimate plant sterol intake. The plant sterol data on vegetables, fruits and cereals have been published (Normén et al, 1999, 2002). Food items were analysed at the Department of Clinical Nutrition, Göteborg University, Sweden, using a modified gas–liquid chromatography procedure by (Jonker et al, 1985; Normén et al, 2001). Concentrations for the five most frequently occurring plant sterols were measured: the unsaturated plant sterols campesterol, stigmasterol, β-sitosterol, and the saturated plant stanols campestanol, and β-sitostanol. The sum of these five plant sterols constitutes ‘total plant sterol’.

Participants completed a 160 item food frequency questionnaire (FFQ) at recruitment (Bingham et al, 2001) to which plant sterol values were assigned using the database. Types of bread, pasta, breakfast cereals and types of fat used for frying, baking and spreading had been specified which assisted in the correct assignment of values. The plant sterol content of all products of pure animal origin was set at zero. Based on standard recipes, the plant sterol content of different baked products and mixed dishes was calculated. Average daily intake of total energy, total fat, saturated, monounsaturated and polyunsaturated fatty acids, dietary cholesterol and total fibre was also calculated from the FFQ as previously described (Bingham et al, 2001).

Statistical analysis

We used data on participants for whom total serum cholesterol concentration and LDL-cholesterol values were available, together with data from the FFQs. We excluded 45 participants (21 men and 24 women) with estimated total plant sterol intakes >750 mg/day. In all, 22 256 participants (10 016 men and 12 240 women) were included in the analysis.

Data were analysed separately for men and women using the STATA 7.0 software package (Stata Corporation, TX, USA, 2001). We calculated the quintile cut points for plant sterol intake using data for men and women combined. Dietary factors were examined in relation to the quintile groups of plant sterol intake. Cardiovascular risk factors were assessed for each quintile group adjusted for baseline age, BMI, total energy intake and menopausal status in women. We carried out tests for trend for continuous variables by fitting the quintile group number (1–5) as a variable in a linear regression and for categorical variables by using a score test for trend of odds. Linear regression was used to model blood lipids from plant sterol intake, in men and women separately, adjusting for age, BMI, total energy intake, menopausal status in women at baseline, saturated fat as percentage of energy intake, dietary fibre (g/MJ), physical activity level and smoking status.


Mean (s. d.) intakes of total plant sterol and the five individual plant sterols for men and women respectively were: total sterols— 310 (108), 303 mg/day (100); β-sitosterol—201 (70), 198 mg/day (66); campesterol— 69 (27), 66 mg/day (25); stigmasterol— 27 (10), 26 mg/day (9); sitostanol— 8 (4), 8 mg/day (4); campestanol— 5 (3), 5 mg/day (3). Mean (s.d.) intake of dietary cholesterol, sexes combined, was 260 mg/day (105). Higher intake of plant sterols was associated with higher total energy intake, total fat, saturated, mono- and polyunsaturated fatty acids, dietary cholesterol and dietary fibre (Table 1).

Table 1 Distribution of dietary components by quintile of total plant sterol intake in men and women aged 39–79 y, EPIC Norfolk 1993–1997 (values are mean (s.d.)).

Tables 2 (men) and 3 (women) show the distribution of cardiovascular risk variables by quintile of total plant sterol intake. There was a significant negative trend between blood lipids (total serum cholesterol and LDL cholesterol) and plant sterol intake in both men and women. Mean total serum cholesterol concentration for men consuming high amounts of plant sterols was 0.25 mmol/l lower and for LDL cholesterol 0.14 mmol/l lower than those with the lowest consumption; the difference in mean concentrations in women were 0.15 and 0.13 mmol/l for total and LDL cholesterols, respectively, after adjusting for age, BMI and total energy intake. HDL cholesterol was also negatively related to plant sterol intake in men but not in women. LDL/HDL ratio decreased significantly from lowest plant sterol intake to highest in both sexes, although this was more pronounced in women. Mean age and BMI decreased slightly with increasing plant sterol intake in men, but these associations were not significant in women.

Table 2 Distribution of cardiovascular risk indicators by quintile of total plant sterol intake in men aged 39–79 y, EPIC Norfolk 1993–1997 (values are mean (s.d.) unless otherwise indicated)
Table 3 Distribution of cardiovascular risk indicators by quintile of total plant sterol intake in women aged 39–78, EPIC Norfolk 1993–97 (values are mean (s.d.) unless otherwise indicated)

There was a significantly lower proportion of current smokers in the highest vs lowest groups of plant sterol consumers in both men and women (Tables 2 and 3). Adjusting for smoking status did not materially alter the significant association between plant sterol intake and blood lipid concentrations. The proportion of persons with self-reported high physical activity was significantly higher in the high vs low plant sterol consumer quintiles. There was no difference in the proportion of subjects taking cholesterol medications nor those with a history of hypercholesterolaemia between quintiles of plant sterol intake.

We undertook multiple regression analyses of lipid concentrations on dietary plant sterol intake after adjustment firstly for age, body mass index, total energy intake and menopausal status in women, then including, saturated fat and dietary fibre intake in the model, and thirdly including, in addition, mono- and polyunsaturated fat. We also ran additional models including physical activity level and smoking status. Table 4 shows the results of these models for total cholesterol, LDL and HDL cholesterol, with the regression coefficients representing the changes in lipid concentrations corresponding to a 200 mg/day increase in total plant sterol consumption. The models indicate that increased plant sterol consumption was associated with a significant decrease in both total and LDL cholesterol in men and women. After adjusting for saturated fat and fibre intakes, the results were similar, although the strength of the association was slightly reduced. Adjusting further for physical activity and smoking status made little difference to the results. Addition of mono- and polyunsaturated fat to the model made little difference in the men, but the relationship with cholesterol and LDL cholesterol was further attenuated in women.

Table 4 Multiple regression of lipid concentrations on dietary plant sterol intake in men (n=10 016) and women (n=12 240) aged 39–79 y in the EPIC-Norfolk population 1993–1997


Laboratory, animal and controlled human studies have provided evidence that pharmaceutical doses of plant sterols are effective in reducing serum cholesterol (Moghadasian & Frohlich, 1999; Law, 2000; Katan et al, 2003). However, there is a lack of data from free-living populations consuming naturally occurring plant sterols as part of a ‘normal’ diet. In the current study, in a free-living British population, a high intake of plant sterols from the diet is inversely and significantly associated with lower concentrations of total and LDL serum cholesterol.

This study has a number of limitations. Firstly, there is likely to be substantial measurement error in the assessment of dietary plant sterol intake, from measurement errors associated with FFQs, and limitations in nutrient databases for dietary plant sterols. Despite these large measurement errors, a significant relationship between dietary intake and cholesterol concentration was observed. However, random measurement error is likely to attenuate substantially any relationship making it likely that the association observed is real, and may, if anything be underestimated.

Potential confounding is another major issue. Persons eating diets high in plant sterols may well have other behaviours that influence cholesterol concentrations including different physical activity and smoking patterns; additionally, diets high in plant sterols are also different with respect to other components such as fibre and dietary fat quality that might confound the plantsterol and blood cholesterol relationship. Although men and women with higher plant sterol intakes also had higher fibre intake, higher physical activity levels and lower proportions of smokers, they also had higher intakes of total fat, saturated fatty acids and dietary cholesterol, known dietary factors related to raised serum cholesterol. The relation between plant sterol intake and blood lipid concentrations was apparent after adjusting for these possible confounding factors. Additionally, including monounsaturated and polyunsaturated fat in the models attenuated the relationship in women more than in men. These results suggest that higher plant sterol intake from natural food sources influence serum cholesterol independently of other factors although, given the measurement error in the dietary confounding variables, residual confounding remains a possibility. The weaker adjusted relationship in women may reflect additional important determinants of lipid levels in women such as hormonal status.


There are several mechanisms whereby dietary intake may influence serum cholesterol. Generally, these relate to the content and quality of fat and of dietary fibre. However, both high-fibre diets and diets high in vegetable fats also have high contents of plant sterols.

In studies using experimental diets in ileostomy subjects, foods and diets with a high plant sterol content and a known cholesterol lowering effect have shown higher ileal excretion of cholesterol compared to foods and diets with lower plant sterol contents (Andersson, 1996). Additionally, in ileostomy subjects, plant sterol intake from different experimental diets was inversely correlated with the absorption of cholesterol from the small bowel (Ellegard et al, 2000). The current study indicates that dietary plant sterols may explain some of the effects of different dietary patterns on cholesterol concentrations.

Dietary cholesterol intake in the current study was on average 260 mg/day. The daily endogenous production of cholesterol is estimated to be 4–5 times as high as the contribution from the diet to the micellar uptake of cholesterol (Grundy, 1983). As dietary plant sterols not only compete for uptake with the relatively small and variable intake of dietary cholesterol but also with endogenous cholesterol from the liver, it seems probable that plant sterols from the diet influence serum cholesterol concentrations to a greater degree than would dietary cholesterol.

Plant sterol intake

Plant sterols are generally consumed in an amount of 200–400 mg/day in Western diets (Jones et al, 1997), but diets based on a high intake of vegetable fats, whole grains, fruits and vegetables have higher plant sterol content. The effect of 1 g of processed plant sterol esters has been shown to reduce serum cholesterol by approximately 7% according to the most recent meta-analysis (Katan et al, 2003). The reduction of 2.8% as seen in men for an increased consumption of 200 mg is somewhat larger, which could be attributed to a relatively higher response from plant sterol in the low intake interval (a curvilinear response), as can be interpreted from the meta-analysis, and from what is known of the effects of dietary cholesterol on LDL. It could also be attributed to a higher efficacy of naturally occurring plant sterols. Lastly, this may reflect a long-term habitual exposure, compared to short-term randomised clinical trials.

Processed plant sterol esters have been incorporated into commercially available margarine spreads in the UK since the late 1990s (Law, 2000). In the current study, dietary intake was assessed before such products were in use and therefore should reflect effects on blood lipids from naturally occurring plant sterols only. Even with the comparatively lower plant sterol intake from the habitual diet compared to enriched products, a significant inverse association between plant sterol intake and serum cholesterol concentrations was observed.

The level of plant sterol intake found in this study (men— 310 mg/day; women— 303 mg/day) is remarkably similar to that recently found in a large Dutch population study, which also assigned values from the current plant sterol database to previously collected FFQs (Normén et al, 2001). The present findings are, however, considerably higher than that previously reported in a British population where mean intake of plant sterols (as total of β-sitosterol, campesterol and stigmasterol together with five other minor plant sterols) was 186 mg/day (Morton et al, 1995). Differences may be due to different methodologies, as the earlier report was based on chemical analyses of mixed samples of 20 food groups assigned from 119 food categories, as part of the UK Total Diet Study in the years 1987 and 1991. The current study utilises a more detailed approach using the FFQ method with chemical analyses of more than 330 different food items as well as standard recipes for calculating composite foods.


In general, increasing plant sterol intake by dietary means could be achieved by substitution of animal foods with plant food, and by substitution of plant foods with lower plant sterol concentration with alternatives richer in plant sterols. Thus, a 200 mg increase in daily dietary plant sterol intake can be achieved by a combination of several minor dietary changes, such as substitution of sunflower oil or olive oil with corn or rape seed (canola) oil, substitution of refined breakfast cereals with whole-grain alternatives, replacing white bread with whole-grain bread, substituting butter with polyunsaturated margarine, replacing apple juice with orange juice, and replacing green peas with broccoli.

The difference in cholesterol concentrations between top and bottom quintiles of plant sterol intake were not large, in the order of 0.15 mmol/l. Nevertheless, a difference of this magnitude, would be associated with, depending on age, a 5–12% difference in coronary heart disease risk (Law et al, 1994), so that even moderate changes in dietary intake within the usual feasible normal population range may be worthwhile.


This study is the first to show in a free-living population that consumption of plant sterols from the habitual diet are related to blood lipid concentrations. While it remains to be seen whether dietary plant sterols are also related to cardiovascular events and mortality, the findings add weight to the idea that there are many different components in foods that influence health. The demonstration of measurable effects even across the normal range of intake in the population may reinforce the relevance of current dietary recommendations to reduce heart disease in the whole community.


  1. Andersson H (1996): Diet and cholesterol metabolism in the gut- implications for coronary heart disease and large bowel cancer. Scand. J. Nutr. 40, 11–15.

    Google Scholar 

  2. Bingham SA, Welch AA, McTaggart A, Mulligan AA, Runswick SA, Luben R, Oakes S, Khaw KT, Wareham N & Day NE (2001): Nutritional methods in the European Prospective Investigation of Cancer in Norfolk. Publ. Health Nutr. 4, 847–858.

    CAS  Article  Google Scholar 

  3. Brown L, Rosner B, Willett WW & Sacks FM (1999): Cholesterol-lowering effects of dietary fiber: a meta-analysis. Am. J. Clin. Nutr. 69, 30–42.

    CAS  Article  Google Scholar 

  4. Clarke R, Frost C, Collins R, Appleby P & Peto R (1997): Dietary lipids and blood cholesterol: quantitative meta-analysis of metabolic ward studies. BMJ 314, 112–117.

    CAS  Article  Google Scholar 

  5. Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A & Wareham N (1999): EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br. J. Cancer 80 (Suppl 1), 95–103.

    Google Scholar 

  6. Ellegard L, Bosaeus I & Andersson H (2000): Will recommended changes in fat and fibre intake affect cholesterol absorption and sterol excretion? An ileostomy study. Eur. J. Clin. Nutr. 54, 306–313.

    CAS  Article  Google Scholar 

  7. Friedewald WT, Levy RI & Fredrickson DS (1972): Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without the use of the preparative ultracentrifuge. Clin. Chem. 18, 499–502.

    CAS  Google Scholar 

  8. Gould AL, Rossouw JE, Santanello NC, Heyse JF & Furberg CD (1995): Cholesterol reduction yields clinical benefit. A new look at old data. Circulation 91, 2274–2282.

    CAS  Article  Google Scholar 

  9. Grundy SM (1983): Absorption and metabolism of dietary cholesterol. Ann. Rev. Nutr. 3, 71–96.

    CAS  Article  Google Scholar 

  10. Jones PJ & Raeini-Sarjaz M (2001): Plant sterols and their derivatives: the current spread of results. Nutr. Rev. 59 (Part 1), 21–24.

    CAS  PubMed  Google Scholar 

  11. Jones PJ, MacDougall DE, Ntanios F & Vanstone CA (1997): Dietary phytosterols as cholesterol-lowering agents in humans. Can. J. Physiol. Pharmacol. 75, 217–227.

    CAS  Article  Google Scholar 

  12. Jonker D, van der Hoek GD, Glatz JFC, Homan C, Posthumus MA & Katan MB (1985): Combined determination of free esterified and glycosylated plant sterols in foods. Nutr. Rep. Int. 32, 943–951.

    CAS  Google Scholar 

  13. Katan MB, Grundy SM, Jones P, Law M, Miettinen T & Paoletti R (2003): Efficacy and safety of plant stanols and sterols in the management of blood cholesterol levels. Mayo Clin. Proc. 78, 965–978.

    CAS  Article  Google Scholar 

  14. Law M (2000): Plant sterol and stanol margarines and health. BMJ 320, 861–864.

    CAS  Article  Google Scholar 

  15. Law MR & Wald NJ (1994): An ecological study of serum cholesterol and ischaemic heart disease between 1950 and 1990. Eur. J. Clin. Nutr. 48, 305–325.

    CAS  PubMed  Google Scholar 

  16. Law MR, Wald NJ & Thompson SG (1994): By how much and how quickly does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ 308, 367–372.

    CAS  Article  Google Scholar 

  17. Moghadasian MH & Frohlich JJ (1999): Effects of dietary phytosterols on cholesterol metabolism and atherosclerosis: clinical and experimental evidence. Am. J. Med. 107, 588–594.

    CAS  Article  Google Scholar 

  18. Morton GM, Lee SM, Buss DH & Lawrance P (1995): Intakes and major dietary sources of cholesterol and phytosterols in the British diet. J. Hum. Nutr. Dietet. 8, 429–440.

    Article  Google Scholar 

  19. Norm L, Johnsson M, Andersson H, van Gameren Y & Dutta P (1999): Plant sterols in vegetables and fruits commonly consumed in Sweden. Eur. J. Nutr. 38, 84–89.

    Article  Google Scholar 

  20. Norm AL, Brants HA, Voorrips LE, Andersson HA, van den Brandt PA & Goldbohm RA (2001): Plant sterol intakes and colorectal cancer risk in the Netherlands Cohort Study on Diet and Cancer. Am. J. Clin. Nutr. 74, 141–148.

    Article  Google Scholar 

  21. Normén L, Bryngelsson S, Johnsson M, Evheden P, Ellegård L, Brants H, Andersson H & Dutta P (2002): The phytosterol content of some cereal foods commonly consumed in Sweden and in the Netherlands. J. Food Composition Anal. 15, 693–770.

    Article  Google Scholar 

  22. Tang JL, Armitage JM, Lancaster T, Silagy CA, Fowler GH & Neil HA (1998): Systematic review of dietary intervention trials to lower blood total cholesterol in free-living subjects. BMJ 316, 1213–1220.

    CAS  Article  Google Scholar 

  23. Wilson MD & Rudel LL (1994): Review of cholesterol absorption with emphasis on dietary and biliary cholesterol. J. Lipid Res. 35, 943–955.

    CAS  PubMed  Google Scholar 

Download references


We thank the participants and general practitioners who took part in EPIC-Norfolk. We also thank Mitra Ravand for technical assistance in the plant sterol analyses. EPIC-Norfolk is supported by programme grants from the Cancer Research Campaign and Medical Research Council with additional support from the Stroke Association, British Heart Foundation, Food Standards Agency, Department of Health and Wellcome Trust. The phytosterol analyses were supported by a grant from the Swedish government under the LUA agreement and the Swedish Cancer Foundation. There are no competing interests.

Author information




Guarantor: K-T Shaw.

Contributors: K-TK and SB are principal investigators in the EPIC-Norfolk population study. AW and SB are responsible for the nutritional analyses. SA developed the phytosterol nutrient databases with HA and LE. AM prepared the phytosterol data set in the EPIC cohort. JS conducted the statistical analyses with assistance from SA and K-TK. SA wrote the paper with contributions from coauthors.

Corresponding author

Correspondence to K-T Shaw.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Andersson, S., Skinner, J., Ellegård, L. et al. Intake of dietary plant sterols is inversely related to serum cholesterol concentration in men and women in the EPIC Norfolk population: a cross-sectional study. Eur J Clin Nutr 58, 1378–1385 (2004).

Download citation


  • phytosterols
  • cholesterol
  • lipids
  • diet
  • population

Further reading