IJE Advance Access originally published online on December 5, 2005
International Journal of Epidemiology 2006 35(2):307-314; doi:10.1093/ije/dyi270
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Article |
Micronutrient intake and the risk of herpes zoster: a case–control study
1 Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London, UK
2 Department of Public Health and Primary Care, Institute of Public Health, Cambridge University, Cambridge, UK
3 Present address: Pfizer Global Research and Development, UK.
* Corresponding author. Infectious Disease Epidemiology Unit, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK. E-mail: sara.thomas{at}lshtm.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
Background Herpes zoster can seriously impair quality of life and may also be a marker for age-related immune decline (immunosenescence). Diets low in micronutrients may increase the risk of zoster by temporarily compromising cell-mediated immune function or by hastening immunosenescence.
Methods Primary objectives were to examine the association between risk of zoster and (i) dietary intake of vitamins A, B6, C, E, folic acid, zinc, and iron, and (ii) fruit and vegetable consumption. We conducted a community-based case–control study. Cases were adults with incident zoster presenting to 22 general practices in London. Controls were individuals with no zoster history, matched to cases by age, sex, and general practice. Diet was ascertained for 243 cases and 483 controls using an interviewer-administered food-frequency questionnaire. We used conditional logistic regression to estimate odds ratios.
Results There was a strong graded association between lower fruit intake and increasing zoster risk; in adjusted analysis, individuals who ate less than one piece of fruit per week had more than three times the risk of zoster compared with individuals who ate more than three portions per day. None of the dietary intakes of the seven micronutrients examined had a statistically significant association with zoster risk when considered singly. However, amongst individuals aged >60 years, a measure of combined micronutrient intake and vegetable intake showed similar dose-related associations with zoster risk.
Conclusion A cocktail of nutrients such as those found in fruit and vegetables may act together, particularly in older individuals, to maintain immune health and prevent zoster.
Keywords Herpes zoster, cellular immunity, fruit, vegetables, micronutrients, aged
Accepted 24 October 2005
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
Herpes zoster occurs when latent varicella-zoster virus (VZV) reactivates. Zoster is common in older populations, with annual incidence ranging from 3.6 to 14.2/1000 amongst the oldest individuals in population-based studies.1 It causes significant acute morbidity and long-term sequelae including chronic pain that can seriously impair quality of life, even amongst those who receive appropriate antiviral treatment.2 Therefore, strategies to reduce the incidence of zoster are of public health importance in ageing populations. Zoster occurs following waning of VZV-specific cell-mediated immunity, and so the risk of zoster is increased amongst individuals with impaired cell-mediated immune function.3–5 However, we know little about the determinants of zoster in the majority of individuals who do not have underlying immunosuppression.1
The sharp increase in risk of zoster with age may be due to the generalized loss of immune competence (immunosenescence) that contributes to older individuals' increased susceptibility to cancer, autoimmune diseases, and infections.6 The extent of immunosenescence can vary considerably between individuals,7 but (like zoster) its determinants are largely unknown. If some exposures increase the risk of immunosenescence and this in turn increases the risk of zoster, then identification of such exposures may suggest interventions to minimize both zoster incidence and immune impairment with age.
Low micronutrient intake is a putative risk factor for zoster at any age and for immunosenescence amongst older individuals because immune system cells depend on micronutrients for functional integrity.8 Specific micronutrients may protect against immunosenescence by limiting free-radical formation, regulating cytokine production, and enabling continued lymphocyte maturation and proliferation. Many older individuals in the UK have micronutrient intakes that are lower than estimated average requirements,9 and micronutrient deficiencies are associated with diminished cell-mediated immunity and increased incidence of infections in elderly populations.10–12 We, therefore, investigated the association between low micronutrient intake and risk of zoster in a British population.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
This research was part of a community-based case–control study in 1997–98 investigating risk factors for zoster amongst adults. Our primary dietary hypotheses were that the risk of zoster is increased amongst individuals with low intake in the year before interview of (i) any of seven micronutrients important for immune functioning: vitamins A, B6, C, E, folic acid, zinc, and iron, and (ii) fruit and vegetables, good sources of some of the micronutrients of interest. Secondary hypotheses were that zoster risk is increased amongst individuals with (i) illnesses associated with deficiencies of these micronutrients, and (ii) low body mass index (BMI).
Recruitment and eligibility criteria for participants are described elsewhere.13 Briefly, cases were adults with incident zoster who presented to 22 general practices in London. Zoster was diagnosed via polymerase chain reaction analyses of vesicle fluid or a standardized clinical case definition. The latter comprised a dermatomal unilateral vesicular or maculopapular rash where (i) pain or rash extended to at least half the dermatome or (ii) pain lasted 1 month or more after rash onset, and where there was no history of a similar dermatomal rash at any site in the last 10 years. Two individually matched controls (with no history of zoster) were sought for each case, matched by age, sex, and general practice. Exclusion criteria included individuals with underlying immunosuppressive disorders or African ethnicity (a group at greater risk of undiagnosed HIV infection at this time), those incapable of responding to questions, or those temporarily registered with the practice. Potential participants were sent a letter inviting them to take part in the study—high response rates were obtained for both cases (94%) and controls (
83%). The study received ethical approval from five Research Ethics Committees. Participants gave written informed consent.
Data collection
One of us (S.L.T.) interviewed participants in their own homes. Diet in the last year was ascertained using the validated British version of the Nurses' Health Study food frequency questionnaire (FFQ), modified to capture seasonal consumption.14–16 The FFQ comprises a list of 139 food items with ‘open’ sections to allow recording of less usual foods and nine frequency options for each food item. Portion size information was collected as ‘units’ (e.g. pieces of fruit) or as ‘medium’ portions. For seasonal foods, we determined separately frequency of consumption ‘in season’ and at other times. We asked about specific brands of breakfast cereals and margarines consumed and the formulation of any micronutrient supplements taken in the last year. We also collected data on a wide range of other exposures that might confound the relationship between diet and risk of zoster. Participants' height and weight were measured wherever possible.
Sample size
A sample size of 244 cases and 488 controls for the study as a whole was derived from standard equations for matched case–control studies,17 estimating at least 10% exposure prevalence amongst controls, an odds ratio (OR) of 2.0, 5% significance (two-sided), 90% power, and a 20% increase to allow for multivariable analyses.
Statistical analysis
Portion sizes for ‘units’ of foods were obtained from British data sources.18 For other foods, age-specific and sex-specific ‘medium portion sizes’ were ascertained via finely stratified analyses of two national nutritional surveys of British adults.19,20 For broadly characterized FFQ items (e.g. beef), we identified the specific food that best represented the item for each sex/age/ethnic grouping from the national survey data and information collected at interview. The micronutrient and energy content of most foods were ascertained from standard British food composition tables (assessed primarily via IDA software)21; manufacturers' reported data were used for specific brands of cereals, margarines, and micronutrient supplements. We then calculated individuals' daily intakes of total energy and the seven micronutrients of interest by summing intakes from each food consumed, deriving the latter from daily frequency (adjusted for seasonal intake where necessary), nutrient content, and portion size.
Distributions of dietary micronutrient intakes were positively skewed and correlated with total energy intake. We adjusted for energy intake using the residuals method after log-transforming the data.22 Individuals were categorized into quintiles of energy-adjusted intake for each micronutrient, based on the intake's distribution amongst controls. The residual values used as cut-off points for each category were back-transformed for presentation.22 Fruit and vegetable intakes were categorized by pre-defining ‘low’ intake as less than once a week (fruits) or once a day (vegetables) and taking cut-offs for higher intakes to obtain approximately equal numbers of controls in each group. Categories for BMI were: ‘underweight’ (
20), ‘average’ (20.1–25, the reference category in regression models), ‘overweight’ (25.1–30), and ‘obese’ (>30).9
We used conditional logistic regression23 in STATA v.724 to investigate associations between the exposures of interest and risk of zoster. We examined the effects of intake of each of the seven micronutrients from foods alone and then from foods and supplements combined. The effects of micronutrient intake from food were re-examined after excluding individuals taking supplements, to check whether supplement intake was masking any effect of dietary micronutrient intake on zoster risk. We assessed the significance of associations between exposures and zoster using likelihood ratio tests of heterogeneity and of linear trend, for the latter scoring categories of exposure from 1 to 5.
Confounders considered in multivariable analyses included ethnicity, smoking, alcohol intake, socioeconomic status (house tenure and car ownership), recent illnesses, social and occupational child and varicella contacts, hours of outdoor physical leisure exercise in the last year, recent stressful events, time elapsed since acquiring varicella, and (for fruit and vegetable intake) total energy intake. We hypothesized that the effect of low micronutrient intake on zoster risk might be greater in older individuals (>60 years) because their immune function may be less efficient than that of younger individuals and may be further compromised if low micronutrient intake itself hastens immunosenescence. Therefore, effect modification by age was examined by comparing models with and without age interaction terms, using likelihood ratio tests.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
We obtained dietary data from 243 of the 244 enrolled cases and from their 483 matched controls. The median age of cases was 57.1 years (range: 16.5–91.2 years), and the mean age difference between cases and their matched controls was 4.1 days.
On univariable analysis, the only statistically significant association between food-derived energy-adjusted daily micronutrient intakes and zoster risk was a trend of increasing risk associated with decreasing vitamin C intake (Table 1). Apart from a weak trend associated with decreasing zinc intake, effect estimates remained non-significant after controlling for a range of confounders (listed below Table 1) and were not confounded by alcohol intake, socioeconomic status, or non-prescribed micronutrient supplements.
|
Individuals (n = 283) who reported taking non-prescribed supplements containing at least one of the seven micronutrients of interest were at similar risk of zoster compared with non-supplement takers [OR = 0.93; 95% confidence interval (95% CI): 0.67–1.30; P = 0.67]. On univariable analysis, none of the total (food plus supplement) micronutrient intakes were significantly associated with zoster risk. However, individuals (n = 24) with conditions associated with decreased micronutrient availability and/or increased requirement were at a 5-fold increased risk of zoster (adjusted OR = 5.16; 95% CI 1.73–15.44; P = 0.003)—conditions included dysphagia with food regurgitation (one case), ulcerative colitis (one case), eating disorders (one case, one control), pregnancy (one case, two controls), iron-deficiency or folate-deficiency anaemia (nine cases, five controls), and epilepsy with phenytoin medication (one case, two controls). Sixteen of these individuals and one other participant had been prescribed micronutrients and so had a high total micronutrient intake. To reduce potential confounding, we repeated analyses after excluding the 25 individuals with the above conditions and/or with prescribed micronutrients—effect estimates remained non-significant (available as Supplementary Data at IJE Online).
Estimates of BMI were available for 240 matched sets with data for at least one case and one matched control (677 individuals), after excluding individuals who were too unsteady to be weighed safely, were pregnant, or refused measurement. Overall, BMI was not associated with zoster risk (P = 0.52). The 281 individuals (98 cases, 183 controls) who were ‘underweight’ were not at increased risk of zoster compared with the 243 individuals with ‘average’ BMI (OR = 0.77; 95% CI 0.38–31.56; P = 0.47).
Unlike micronutrient intakes, we observed on univariable analysis a strong trend of increasing zoster risk associated with decreasing intakes of both fresh fruits and fresh/frozen vegetables (Table 2). When combined fruit/vegetable intake was considered, individuals consuming one or less portion a day were at nearly three times the risk of zoster compared with individuals with the highest intake. Similar trends were seen when fruit and vegetable intake were considered separately, with a larger and more statistically significant effect for fruit intake. Effect estimates for combined fruit/vegetable intake and fruit intake increased on multivariable analyses, but the protective effect of high vegetable intake disappeared (Table 2). Multivariable analyses included controlling for total energy intake and a range of other confounders (listed below Table 2); after these adjustments, effects were not further confounded by socioeconomic status, alcohol intake, non-prescribed micronutrient supplementation, or outdoor physical leisure activities.
|
There was evidence that the effect of vegetable intake varied by age (Pinteraction = 0.021). There was no statistically significant association with zoster risk amongst individuals aged <60 years, but a statistically significant linear trend of increasing risk associated with decreasing intake amongst older individuals (Table 3). However, there was poor evidence that age modified the effect of fruit intake (Pinteraction = 0.59) or any of the individual dietary micronutrient intakes except for iron (Pinteraction = 0.076), where increasing risk was associated with decreasing dietary intake amongst older individuals (Table 3). We also explored whether the effect of combined micronutrient intake varied with age by creating a ‘total micronutrient score’ for each individual—we scored quintiles of intake for each of the seven micronutrients from one (lowest quintile) to five (highest quintile) and summed individuals' seven micronutrient scores to obtain their total micronutrient scores (range: 7–35). Total scores were then categorized into quintiles with similar numbers of controls in each group. Again, there was evidence that age modified the effect of combined micronutrient intake (Pinteraction = 0.033), with a statistically significant dose–response of increasing risk associated with decreasing combined intake only amongst older individuals (Table 3).
|
Additional analyses
Analyses were repeated after excluding (i) the 85 participants of non-white ethnicity, and (ii) the 283 individuals who took non-prescribed micronutrient supplements. In each revised analysis, effect estimates for dietary intakes were similar to those calculated using the whole dataset. Age-specific analyses were repeated for micronutrient scores derived from food and supplements combined—as with the individual micronutrient intakes, these scores had a less-strong association with zoster risk compared with scores derived from diet alone. To clarify whether the lower frequency of fruit and vegetable intake amongst cases was specific for these foods, analyses were repeated for other food groups, for example white bread and potatoes. We did not find generalized less frequent food consumption amongst cases, and cases aged >60 years ate potatoes more frequently compared with their matched controls (adjusted P = 0.016).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
In this population, we found a strong dose–response relationship between decreasing fruit intake and increasing zoster risk amongst both younger and older individuals and similar associations for vegetable intake and combined micronutrient intake amongst older individuals. People with conditions associated with micronutrient deficiency were at higher risk of zoster. These results suggest that a mix of nutrients such as those found in fruit and vegetables act together (particularly in older individuals) to maintain immune health. Our findings are consistent with studies of older populations that showed associations between micronutrient deficiencies and diminished cell-mediated immune function and between diets high in fruit and vegetables and lowered all-cause mortality.10–12,25–27 Low combined micronutrient intake may have a greater effect amongst older individuals because they have less robust immune function and/or because low intake itself hastens immunosenescence.
In contrast, most individual micronutrient intakes in our study were not associated with zoster risk, suggesting that single micronutrients in isolation have relatively weak effects on the immune system and on risk of zoster. Studies of other diseases have demonstrated a protective effect of nutrient-rich foods but no protection when constituent micronutrients are considered singly.28 Fruit and vegetables are a good source of some (although not necessarily all) of the seven micronutrients of interest, and may also include other biologically active substances that enhance immune function, such as polyphenols or other, as yet unidentified, factors.25 The weak association with vitamin C may reflect the effect of fruit intake, as vitamin C is mostly obtained from fruit and vegetables and plasma vitamin C levels are more strongly related to fruit consumption compared with other antioxidants.29 We were able to document the micronutrient content of all supplements consumed by participants. Non-prescribed supplements were not protective against zoster, perhaps because supplements do not replicate the full mix of nutrients in fruit and vegetables, or because some participants with poor immune function took micronutrients to compensate for poor diet or for feeling unwell.
A number of methodological issues must be considered when interpreting these data. First, non-differential misreporting of frequency of consumption could result in underestimation of effects for all dietary variables. In contrast, differential misreporting of fruit and vegetable intake is unlikely to have generated the strong dose–response effects seen for these intakes; the dietary questions formed only part of an extensive interview about a wide range of exposures, participants were unaware of specific dietary hypotheses, and additional analyses indicated that cases ate a different range of foods to controls.
Previous studies have inferred that age-matched and sex-matched individuals eat similar portion sizes of foods.30,31 However, any variation in food portion sizes between matched cases and controls could increase misclassification of intake from these foods. For example, if cases ate smaller portions of foods compared with controls, we could have overestimated their dietary micronutrient intake (calculated using falsely high age/sex-specific portion sizes) and thus underestimated the effect of micronutrient intake on risk of zoster. This was less likely to occur for fruit, which was mostly measured in ‘pieces’.
We chose single foods to represent FFQ items for each age/sex/ethnicity group. The choice was narrow for fruit and vegetable FFQ items but wider for broadly characterized items such as ‘beef’.32,33 Misclassification of micronutrient intake could result from differences in micronutrient content between the foods chosen to represent FFQ items and the foods actually consumed by participants and by variations in processing and cooking of foods that (unlike fresh fruit) were not eaten raw. A more varied diet amongst younger subjects could increase misclassification and obscure any effect of vegetable and combined micronutrient intake.
We were able to assess many potential confounders of the protective effect of fruit and vegetable intake. In a recent British survey, fruit and vegetable consumption varied with age, sex, household income, smoking, alcohol consumption, children in the household, and BMI.34 All these variables were considered in the present analysis except for household income, which may not have been fully captured by household tenure and car ownership; however, it is unclear what unidentified proximate determinant of zoster this could represent. Physical activity affects immune functioning and could confound the effects of fruit and vegetable intake.35 We assessed only the effect of outdoor physical leisure activities, but we also adjusted for variables that are strongly correlated with overall physical activity (including age, sex, total energy intake and current illness). The effect of physical exercise on the immune system is probably non-linear,35 so residual confounding by physical exertion is unlikely to explain the strong linear associations with fruit and vegetable intake. Childhood socioeconomic status may be associated with adult vitamin C plasma levels, so if unknown childhood exposures could ‘programme’ the immune system to modify later zoster risk they could confound adult fruit and vegetable intakes.36,37 However, childhood determinants would then need to have strong dose–response relationships with adult intakes.
Daily intakes of nutrients estimated by the present FFQ are moderately correlated with intakes from weighed food records or plasma micronutrient levels (r = 0.43–0.55).14,15 We used the FFQ simply to rank individuals' micronutrient, fruit, and vegetable intakes, but some mis-categorization of intake into quintiles may have occurred, particularly for micronutrient intakes. Recent studies have suggested that measurement error with this FFQ can mask a disease association with quantiles of nutrient intake.16,38 However, the graded effect of combined micronutrient intake in our study (calculated from FFQ-derived individual micronutrient intakes) again suggests that a mix of micronutrients provides protection, not individual intakes.
We did not measure cell-mediated immune function in our participants and so could not investigate whether the effect of fruit and vegetable intake on zoster risk in older individuals was mediated via immunosenescence. Existing dietary cohort studies14,39 could measure immune function and zoster incidence to examine this hypothesis more directly and could utilize methods unsuited to our case–control methodology, such as food diaries and repeated biochemical measurements of micronutrients whose intakes are poorly estimated from food composition tables.40
The World Health Organisation has attributed nearly 4% of the overall disease burden in developed countries to low fruit and vegetable intake and has estimated that average fruit and vegetable intakes are lower than recommended levels in many areas, particularly amongst the elderly.41,42 Our findings provide a further reason to support existing recommendations to eat at least five portions of fruit or vegetables per day.43,44
|
Conflict of Interest |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
The authors have declared no conflicts of interest.
KEY MESSAGES
|
|
Acknowledgments |
|---|
The authors thank the 22 general practices, which collaborated on this study, Professor Judy Breuer for her help with the polymerase chain reaction analyses for zoster diagnosis, and Dr Mike Nelson for helpful advice on how to analyse the FFQ data. Material from the Dietary and Nutritional Survey of British Adults and the National Diet and Nutrition Survey of People aged 65 years and over was made available by the (then) Office of Population Censuses and Surveys, the Ministry of Agriculture, Fisheries and Food, the Department of Health, and the ESRC Data Archive, and has been used by permission of the Controller of H.M. Stationery Office. A.J.H. and S.L.T. designed the study, with input from J.G.W. S.L.T. ran the study, interviewed participants, and managed the data, and then (with critical contributions from A.J.H. and J.G.W.) carried out the analyses, interpreted the findings and wrote the paper. S.L.T. was funded first by a Medical Research Council (UK) research studentship, and then by the Research Foundation for Microbial Diseases, Osaka University (Japan).
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
1 Thomas SL, Hall AJ. What does epidemiology tell us about risk factors for herpes zoster? Lancet Infect Dis 2004;4:26–33.[CrossRef][ISI][Medline]
2 Johnson RW, Dworkin RH. Treatment of herpes zoster and postherpetic neuralgia. BMJ 2003;326:748–50.
3 Saibara T, Maeda T, Onishi S et al. Depressed immune functions in the early phase of varicella-zoster virus reactivation. J Med Virol 1993;39:242–45.[Medline]
4 Morens DM, Bregman DJ, West CM et al. An outbreak of varicella-zoster virus infection among cancer patients. Ann Intern Med 1980;93:414–19.[Medline]
5 Alliegro MB, Dorrucci M, Pezzotti P et al. Herpes zoster and progression to AIDS in a cohort of individuals who seroconverted to human immunodeficiency virus. Clin Infect Dis 1996;23:990–95.[Medline]
6 Castle SC. Clinical relevance of age-related immune dysfunction. Clin Infect Dis 2000;31:578–85.[CrossRef][ISI][Medline]
7 Barcellini W, Borghi MO, Sguotti C. Heterogeneity of immune responsiveness in healthy elderly subjects. Clin Immunol Immunopathol 1988;47:142–51.[Medline]
8 Field CJ, Johnson IR, Schley PD. Nutrients and their role in host resistance to infection. J Leukoc Biol 2002;71:16–32.
9 Finch S, Doyle W, Lowe C et al. National Diet and Nutrition Survey: People Aged 65 Years and Over. Report of the Diet and Nutrition Survey (Volume 1). London: The Stationery Office, 1998.
10 Payette H, Rola-Pleszczynski M, Ghadirian P. Nutrition factors in relation to cellular and regulatory immune variables in a free-living elderly population. Am J Clin Nutr 1990;52:927–32.
11 Penn ND, Purkins L, Kelleher J et al. The effect of dietary supplementation with vitamins A, C and E on cell-mediated immune function in elderly long-stay patients: a randomized controlled trial. Age Ageing 1991;20:169–74.
12 Chandra RK. Nutrition and immunity in the elderly: clinical significance. Nutr Rev 1995;53:S80–85.[Medline]
13 Thomas SL, Wheeler JG, Hall AJ. Contacts with varicella or with children and protection against herpes zoster in adults: a case–control study. Lancet 2002;360:678–82.[CrossRef][ISI][Medline]
14 Bingham SA, Gill C, Welch A et al. Validation of dietary assessment methods in the UK arm of EPIC using weighed records, and 24-hour urinary nitrogen and potassium and serum vitamin C and carotenoids as biomarkers. Int J Epidemiol 1997;26:S137–51.
15 McKeown NM, Day NE, Welch AA et al. Use of biological markers to validate self-reported dietary intake in a random sample of the European Prospective Investigation into Cancer United Kingdom Norfolk cohort. Am J Clin Nutr 2001;74:188–96.
16 Fletcher AE, Breeze E, Shetty PS. Antioxidant vitamins and mortality in older persons: findings from the nutrition add-on study to the Medical Research Council Trial of Assessment and Management of Older People in the Community. Am J Clin Nutr 2003;78:999–1010.
17 Schlesselman JJ. Sample size. In: Case–Control Studies: Design, Conduct, Analysis. New York: Oxford University Press, 1982, pp. 144–70.
18 Crawley H. Food Portion Sizes. London: HMSO, 1993.
19 Dietary and Nutritional Survey of British Adults, 1986–1987 (Study no. 2836). Colchester: ESCR Data Archive, 1991.
20 National Diet and Nutrition Survey of people aged 65 years and over (Study no. 4036). Colchester: ESCR Data Archive, 1999.
21 Nelson M. Integrated Dietary Analysis Software: Version 3. London: IDA Ltd, 1997.
22 Willett W, Stampfer M. Implications of total energy intake for epidemiologic analyses. In: Willett W (ed.). Nutritional Epidemiology. 2nd edn. Oxford: Oxford University Press, 1998, pp. 273–301.
23 StataCorp. Stata Statistical Software: Release 7.0. College Station, TX: Stata Corporation. 1999.
24 Clayton D, Hills M. Conditional logistic regression. In: Statistical Methods in Epidemiology. Oxford: Oxford University Press, 1993, pp. 290–97.
25 Meydani, M. The Boyd Orr lecture. Nutrition interventions in aging and age-associated disease. Proc Nutr Soc 2002;61:165–71.[CrossRef][Medline]
26 Fortes C, Forastiere F, Farchi S, Rapiti E, Pastori G, Perucci CA. Diet and overall survival in a cohort of very elderly people. Epidemiology 2000;11:440–45.[CrossRef][ISI][Medline]
27 Seccareccia F, Alberti-Fidanza A, Fidanza F et al. Vegetable intake and long-term survival among middle-aged men in Italy. Ann Epidemiol 2003;13:424–30.[Medline]
28 Messina M, Lampe JW, Birt DF et al. Reductionism and the narrowing nutrition perspective: time for reevaluation and emphasis on food synergy. J Am Diet Assoc 2001;101:1416–19.[CrossRef][Medline]
29 Block G, Norkus E, Hudes M et al. Which plasma antioxidants are most related to fruit and vegetable consumption? Am J Epidemiol 2001;154:1113–18.
30 Heady JA. Diets of bank clerks. Development of a method of classifying the diets of individuals for use in epidemiologic studies. J R Stat Soc 1961;124:336–61.
31 Samet JM, Humble CG, Skipper BE. Alternatives in the collection and analysis of food frequency interview data. Am J Epidemiol 1984;120:572–81.
32 Holland B, Unwin ID, Buss DH. McCance and Widdowsons's The Composition of Foods: Fruit and Nuts. Cambridge: The Royal Society of Chemistry, 1992.
33 Chan W, Brown J, Lee SM. McCance and Widdowsons's The Composition of Foods: Meat, Poultry and Game. Cambridge: The Royal Society of Chemistry, 1995.
34 Bajekal M, Primatesta P, Prior G (eds). HSE 2001—Fruit and Vegetable Consumption. London: The Stationery Office, 2003.
35 Hoffman-Goetz L, Klarlund Pedersen B. Exercise and the immune system: a model of the stress response? Immunol Today 1994;15:382–87.[CrossRef][ISI][Medline]
36 Lawlor DA, Davey Smith G, Bruckdorfer KR, Kundu D, Ebrahim S. Those confounded vitamins: what can we learn from the differences between observational versus randomised trial evidence? Lancet 2004;363:1724–27.[CrossRef][ISI][Medline]
37 Hall AJ, Hall AJ, Yee LJ, Thomas, SL. Life course epidemiology and infectious diseases. Int J Epidemiol 2002;31:300–01.
38 Bingham SA, Luben R, Welch A et al. Are imprecise methods obscuring a relation between fat and breast cancer? Lancet 2003;362:212–14.[CrossRef][ISI][Medline]
39 Hu FB, Willett WC. Diet and coronary heart disease: findings from the Nurses' Health Study and Health Professionals' Follow-up Study. J Nutr Health Aging 2001;5:132–38.[Medline]
40 Hunter DJ, Morris JS, Chute CG et al. Predictors of selenium concentration in human toenails. Am J Epidemiol 1990;132:114–22.
41 World Health Organisation. The World Health Report 2002. Reducing Risks to Health, Promoting Healthy Life. Geneva: WHO, 2002.
42 Pomerleau J, Lock K, McKee M, Altmann DR. The challenge of measuring global fruit and vegetable intake. J Nutr 2004;134:1175–80.
43 World Health Organisation. Diet, nutrition and the prevention of chronic diseases. Technical Report Series 797. Geneva: WHO, 1990.
44 Department of Health. The NHS Plan. Improving Health and Reducing Inequality. Section 1320. London: The Stationery Office, 2001.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. A. Sheridan and M. A. Beck The Immune Response to Herpes Simplex Virus Encephalitis in Mice Is Modulated by Dietary Vitamin E J. Nutr., January 1, 2008; 138(1): 130 - 137. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||