healthy responders likely attenuated the associations, but this
should not have affected the relation between vitamin C and
stroke observed within the study participants; if anything, trun-
cation of the distribution is likely to reduce the power of any
associations. We used death certification and a hospital record
linkage system using ICD coding to identify stroke cases. Al-
though follow-up with the use of these methods is virtually com-
plete, this approach may underestimate incident nonfatal stroke
cases that are not admitted to the hospital. The use of self-
reported stroke to exclude prevalent cases may have missed some
prevalent strokes. We were not able to separately examine stroke
subtypes. Nevertheless, the primary focus of the study was to
assess the risk prediction of clinical stroke event severe enough
to lead to hospitalization or death regardless of stroke subtype. In
any case, the misclassification of strokes was likely to only at-
tenuate any associations.
Only single measurements of plasma vitamin C and other
covariates, such as cholesterol and BP, were made at baseline.
These measures as well as lifestyle behaviors, which may affect
vitamin C concentration, may have changed over the follow-up
period. Moreover, the blood sample taken was a nonfasting sam-
ple and was therefore less standardized for some of the variables
(eg, cholesterol concentration) than was a fasting blood sample.
Nevertheless, random measurement error was likely only to at-
tenuate any relations observed between plasma vitamin C and
Plasma vitamin C concentration was a good biomarker of plant
food, namely fruit and vegetable intakes, in our cohort. The
mol or 1-SD increase in plasma vitamin C concentration is
associated with approximately one additional serving of fruit and
vegetables daily (22). This agrees with more recently published
literature on the relation between higher dietary fruit and vege-
table intake and reduced stroke risk (23, 24). However, observa-
tional studies are not all consistent (25, 26), and antioxidant
supplementation including vitamin C did not produce a substan-
tial benefit in clinical trial settings in high risk individuals (6, 27).
The recent Women’s Health Initiative also reported no reduction
in cardiovascular disease in the group allocated to a low-fat and
higher fruit and vegetable target diet (28).
There are some plausible explanations why the discrepancy
exists between cohort studies and trials. Many of the supplemen-
tation trials, such as the Heart Protection Study and the Finnish
Alpha-Tocopherol Beta-Carotene Cancer Prevention Study,
were conducted in high-risk or highly selected rather than general
populations (27, 29, 30). Additionally, combinations of antioxi-
dants, some of them in pharmacologic doses, such as vitamin E,
may have unpredicted biological effects.
Moreover, the lack of benefit of vitamin C in clinical trials
could be explained by the relation between vitamin C dose and
plasma concentration. At doses 쏝100 mg/d, there is a large
change in plasma concentration for small changes in dose. Above
100 mg/d, there is little change in plasma concentration despite
large changes in dose. If the control group (ie, lowest quartile)
consumed 100 mg/d, then further increases in dose would be
predicted to cause little change in concentration. The outcome,
therefore, may not be affected at higher doses. This problem was
first pointed out by Levine et al in 1999 (31).
The association appeared to be independent of the most plau-
sible confounders (32), such as smoking, physical activity, and
social class, as we previously highlighted (22, 33). The mean
height of men and women by quartile of vitamin C concentrations
did not show material differences (Table 2); hence, life-course
factors are unlikely to be significant confounders. An intriguing
possibility is that the plasma vitamin C concentration is a good
marker of a wider range of health behaviors, such as fruit and
vegetable consumption, that may be protective against stroke.
Even then, it appears that the relation was independent of fruit
and vegetable consumption. It is also possible that the relation
could reflect measurement error related to the dietary instru-
ments. It is also plausible that vitamin C may biochemically
affect stroke risk. Given the current evidence, it unlikely that
long-term randomized controlled trials of isolated vitamin C
supplementation and cardiovascular disease endpoints will be
conducted. Nevertheless, the magnitude of the association be-
tween plasma vitamin C and subsequent stroke is substantial and
independent of known major risk factors for stroke.
We believe that these findings are of interest for several rea-
sons. First, the strong inverse association between plasma vita-
min C and stroke suggests that plasma vitamin C is likely to be a
good biomarker of whatever causal factors affect stroke risk,
most plausibly the dietary intake of plant foods. However, iden-
tification of the relevant factors may lead to better stroke pre-
vention. Second, irrespective of any causal associations, plasma
vitamin C appears to be a good predictive risk indicator of stroke,
independent of known risk factors such as age, BP, smoking,
lipids, diabetes, and BMI. Given that about half of the risk of
stroke is unexplained by conventional cardiovascular disease
risk factors (34) and that the predictive validity of traditional
cardiovascular disease risk factors appears to diminish with age
(35, 36), risk markers that may help to identify those persons at
greatest risk of stroke for targeted preventive interventions with
established therapies, such as BP reduction, may be of interest.
We thank the participants and general practitioners who took part in the
study, the staff of EPIC-Norfolk, and our funders.
The authors’ responsibilities were as follows—K-TK, SAB, and NJW:
principal investigators in the EPIC-Norfolk population study; RNL: respon-
sible for data management, computing, and data linkages; PKM: conducted
the analysis; and AAW: involved in the study design and in the writing of the
manuscript. All coauthors contributed to the writing of this article. KTK was
the guarantor. No conflicts of interest were reported by any of the authors.
1. Food Standard Agency. EVM review of vitamin C. Eat well, be well.
Helping you make healthier choices. Internet: http://www.food.gov.uk/
multimedia/pdfs/evm_c.pdf (accessed 15 October 2007).
2. Hickey S, Roberts H. Ascorbate: the science of vitamin C. Napa, CA:
Lulu Press, Inc, 2004. (ISBN 1-4116-0724-4.)
3. Hickey S, Roberts H. Misleading information on the properties of vita-
min C. PLoS Med 2005;2:e307; author reply e309.
4. Padayatty SJ, Katz A, Wang Y, et al. Vitamin C as an antioxidant:
evaluation of its role in disease prevention. J Am Coll Nutr 2003;22:
5. Yusuf S, Dagenais G, Pogue J, Bosch J, Sleight P. Vitamin E supple-
mentation and cardiovascular events in high-risk patients. The Heart
Outcomes Prevention Evaluation Study Investigators. N Engl J Med
2000;342:154 – 60.
6. Finnish ATBC Study. The effect of vitamin E and beta carotene on the
incidence of lung cancer and other cancers in male smokers. The Alpha-
Tocopherol, Beta Carotene Cancer Prevention Study Group. N Engl
J Med 1994;330:1029 –35.
7. Ness AR, Powles JW, Khaw KT. Vitamin C and cardiovascular disease:
a systematic review. J Cardiovasc Risk 1996;3:513–21.
8. Yokoyama T, Date C, Kokubo Y, Yoshiike N, Matsumura Y, Tanaka H.
Serum vitamin C concentration was inversely associated with subse-
quent 20-year incidence of stroke in a Japanese rural community. The
Shibata Study. Stroke 2000;31:2287–94.
68 MYINT ET AL
by guest on June 1, 2013ajcn.nutrition.orgDownloaded from