Lisa B. Goode
Here it is:
Vitamin D intake and incidence of multiple sclerosis
To the Editor: We read with interest the article by Munger et al.1
A protective effect of sunlight on multiple sclerosis (MS) risk was
first suggested by Acheson et al.2 Vitamin D is a potential mediator
of this relationship. We are sympathetic to the hypothesis
being tested3 by Munger et al. but have the following concerns.
1) NHS studied women age _30, but more than half of female
patients with MS have onset below this age. Of those accrued,
some 50,000 were excluded from analysis.1 Was this done before
testing the vitamin D hypothesis? What were the characteristics,
when known, of exclusions for calculated vitamin D estimates
compared to those retained? Perhaps MS risk can be altered after
age 30, but earlier ages are implicated from migration studies.
2) 61 and then 130 questions were asked in the NHS and NHS
II questionnaires. Was correction made for multiple analyses?
Could the authors explain the assumptions and approach used to
calculate the “p trend” statistic that forms the basis of this report?
3) The apparent association of low MS risk with intake of _400
units of vitamin D from supplements per day seems at odds with
the recent report that those intakes of supplements have minimal
effects on 25(OH)D levels. Furthermore, young women who took
multivitamins were more likely to exercise outdoors. Multivitamin
use correlated better with summer 25(OH)D levels than winter.5
Vitamin D production in the skin requires UVB that is not intense
enough at latitudes >30 for at least 1 month each winter.
4) The association of MS with latitude seems unambiguous
from Kurtzke’s US Veterans’ studies and from Australia.6 The
lack of interaction with latitude in this study4 is surprising if
vitamin D intake in adulthood is causally related to MS risk, since
D levels and putative functional effects are dependent on latitude
5) We note that the NHSII cohort had more MS “cases/
person–y” (97/7.5 _ 105) compared to the NHS cohort (76/1.5 _
106). These data are difficult to compare. As age specific incidences
seem less in NHSII, is there evidence for a decreasing
incidence or prevalence in the areas surveyed?
6) How does the nurses’ D intake relate to that in the general
population? Vitamin intake could vary by ethnicity. Did the definition
of “white” include ethnic groups known to be resistant to
MS? Recall bias has been reported within weeks of illness associated
events. Has the accuracy of four yearly reports been
validated for the measures in this article?
George C. Ebers, MD, FRCPC, A.D. Sadovnick, PhD, Reinhold
Veith, PhD, Oxford, UK
Reply from the Authors: We thank Ebers et al. for their interest
in our study. We excluded women with an incomplete baseline
food frequency questionnaire. This exclusion was made a priori
following the same rules used in all previous dietary analyses in
these cohorts. Adjustment for multiple analyses in our study was
not necessary as only two measures of vitamin D intake were
considered: the combined amount of vitamin D intake from food
and from supplements. The p for trend was calculated using a
proportional hazards regression model using the median intake
values for each category of vitamin D from food or supplements as
a continuous variable. This method tests the overall null hypothesis
that vitamin D intake is unrelated to risk of MS without any
In contrast with the results of Vieth et al.5 quoted by Ebers et
al., in a substudy among 323 healthy women from the NHS cohort,
we found that vitamin D intake at levels below 400 IU does
increase 25(OH)D levels. Average winter plasma levels of
25(OH)D were positively correlated with levels of vitamin D intake:
40 nmol/L in the lowest quintile of intake (median _ 108
IU), 55 nmol/L in the third quintile (median _ 301 IU), and 70
nmol/L in the highest quintile (median _ 703 IU). Further, in our
cohorts, the association between vitamin D intake from supplements
and risk of MS was not materially altered by adjustment
for physical activity (unpublished data).
The women in the NHS cohort are older than those in the
NHSII cohort and because of modest overlap between the two
cohorts, age-specific incidence rates are difficult to compare. However,
age-specific incidence rates are slightly higher, not lower, in
NHS II (unpublished data). Further, the increasing use of MRI
may reduce the time between onset of MS and diagnosis, possibly
causing spurious changes in incidence rates.
The nurses’ intake of vitamin D is similar to that of women in
other US cohort studies.7 Race was self-reported and “White” did
not include ethnic groups with low risk of MS.
The lack of significant interaction between latitude and vitamin
D in our study may be explained by the insufficient power to
detect such an interaction.
As this was a prospective study, recall bias was not of concern
as none of the women had MS or MS symptoms when completing
the food frequency questionnaires.
Kassandra L. Munger, MSc, Eilis O’Reilly, MSc, Alberto
Ascherio, MD, DrPH, Boston, MA
Copyright © 2004 by AAN Enterprises, Inc.
1. Munger KL, Zhang SM, O’Reilly E, et al. Vitamin D intake and incidence
of multiple sclerosis. Neurology 2004;62:60–65.
2. Acheson ED, Bachrach CA, Wright FM. Some comments on the relationship
of the distribution of multiple sclerosis to latitude, solar radiation
and other variables. Acta Psychiatr (Scand) 1960;35 Suppl 147:132.
3. Steckley J, Dyment D, Sadovnick D, et al., and the Canadian Collaborative
Study Group. Genetic analysis of vitamin D related genes in multiple
sclerosis patients. Neurology 2000;54:729–732.
4. Hernan MA, Olek MJ, Ascherio A. Geographic variation of MS incidence
in two prospective studies of US women. Neurology 1999;53:1711–1718.
5. Vieth R, Cole DE, Hawker GA, Trang HM, Rubin LA. Wintertime vitamin
D insufficiency is common in young Canadian women, and their
vitamin D intake does not prevent it. Eur J Clin Nutr 2001;55:1901–
6. Hammond SR, English DR, McLeod JG. The age-range of risk of developing
multiple sclerosis: evidence from a migrant population in Australia.
7. Merlino LA, Curtis J, Mikuls TR, et al. Vitamin D intake is inversely
associated with rheumatoid arthritis: results from the Iowa Women’s
Health Study. Arthritis Rheum 2004;50:72–77.
Last edited by Nick
on Tue May 17, 2005 10:58 am, edited 1 time in total.