`Apotex Inc. et al. v. Novartis AG
`IPR2017-00854
`
`
`
`346
`
`INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
`
`variations in its prevalence across continents and alteration in
`the risk of acquiring MS among children of immigrants to reflect
`the rates of the adoptive country3'4 suggest a role of lifestyle
`and local environmental factors.8 Increased risk has been asso-
`ciated with viral
`infections,9 higher socioeconomic status,10
`and dietary habits which characterize over-nutrition."ll Most
`reports on MS and nutritional factors are based on ecologic
`correlation studieslz'l‘F21 and suggest an association with some
`nutrients and foods such as elevated intakes of energy and
`animal fats, (particularly saturated fats), milk and oats. In a
`correlation analysis, Espar'za at a].12 using data from the food
`balance sheets of the United Nations Food and Agriculture
`Organization (FAQ) for the period 1979—1981, found signifimnt
`positive and independent correlations of MS mortality with
`saturated fatty acids, animal fat, animal minus fish fat, and lati-
`tude. In a case—control study conducted in Italy by T013 et al.'3
`cases had higher intakes of bread, pasta, butter and lard, legume
`soup, horse flesh, and caffeine-containing beverages such as
`coffee and tea, before the age of 15. During adulthood (after
`15 years of age), however, eggs (odds ratio [OR] = 5.95), wine
`(OR = 0.37) and mineral water (OR = 0.30) distinguished cases
`and controls. Age at exposure has also been reported to be of
`importance by flutter22 who suggested that
`the influence of
`sunlight and diet may be of greatest importance during the first
`two decades of life. There is very little in the literature on
`observational studies carried out in North America on diet and
`MS. We have carried out a case-control study of dental amalgam
`nutrition and sociodemographic factors and MS. The present
`article shows the results of this study which was conducted with
`validated dietary methods,
`to explore the role of nutritional
`factors in the aetiology of this disease.
`
`Methods
`Case ascertainment
`
`incident MS cases, resident in greater Montreal and diagnosed
`between January 1991 to December 1994, were identified with
`the collaboration of
`the MS Association of Montreal East.
`
`neurologists, and general physician referrals. Announcements
`were also placed regularly in city newspapers in order to reach
`other individuals diagnosed with MS during this period. These
`notices were issued six times during the course of the study, and
`similar announcements were run on some local radio stations.
`
`Eligible incident cases were contacted by phone, and upon
`lnforrned consent were visited at home and interviewed.
`A total of 353 MS mses were identified during the study period.
`Of these, 87 (24.6%) were not eligible because of incorrect
`diagnosis or because they were prevalent cases. The remaining
`266 (75.4%) eligible subjects were followed up. Of these, 11
`(4.1%) declined to participate and 52 (19.5%) were not inter-
`viewed because of poor health, loss of contact, or refusal of the
`physician to issue permission to contact the patient. Finally,
`six cases (2.3%) were excluded after the interview bemuse of
`incorrect diagnosis. We were therefore able to interview 197
`subjects (74.1%) of the eligible cases.
`
`Control ascertainment
`
`Controls were drawn at random from the general population,
`and frequency matched to the cases by age (5 years age group),
`sex and phone number. The controls were identified through
`
`random digit dialling (RDD), using the first three digits of the
`phone number of the cases.23 Controls were selected from
`the telephone directory in which the corresponding cases were
`listed (all patients studied had a listed telephone number: only
`1% of families in the Montreal region do not have a telephone).
`A page from the telephone directory was randomly selected
`from the sampling frame and the names and addresses of 10
`individuals with the same first three-digit telephone numbers as
`the cases were selected. These residences were then contacted
`
`by a letter explaining the aims of the study. Approximately one
`week later these residences were telephoned to see if they con.
`tained an individual who matched the cases for age and sex and
`who agreed to be interviewed. If so, an interview was arranged
`at the control’s home.
`If not, the procedure was repeated.
`if
`more than one eligible control was reached at a given number,
`this information was kept in a databank for further use. Non-
`residential numbers were discarded and a ’no answer‘ number
`
`was redialled up to eight times at various points in time, day
`and night, weekday and weekend, before being rejected. As with
`the cases, controls were also contacted by a letter and those who
`assented were interviewed at home. We contacted a total of 236
`
`population-based controls. Of these, 202 (85.6%) eligible con-
`trols were interviewed. The remaining 34 (14.4%) were not
`interviewed for the following reasons: wrong age group or area
`of
`residence, poor health,
`language problems or refusal
`to
`participate (only five subjects).
`
`Questionnaire
`Trained interviewers administered a modified version of the
`
`health and sociodemographic questionnaire developed in our
`Unit for numerous case-control studies of diet and chronic dis-
`ease. The forms have been designed to assess information (about
`150 items) on such variables as sociodemographim, anthro-
`pometrics, occupational and medical history, family history of
`several chronic diseases, tobacco and alcohol consumption as
`well as physical activity. Weight and height at diagnosis were
`self reponed.
`A 164vitem food frequency questionnaire (FFQ), designed
`by the Epidemiology Unit of the National Cancer Institute of
`Canada (NCIC), mainly for epidemiological studies of diet and
`mncer. was employed to determine usual dietary intakes during
`the year prior to the interview (prior to diagnosis). This ques-
`tionnaire was first evaluated by a self-administered method“’25
`and validated more recently against a 7-day food record and a
`lengthy diet history questionnaire.26 The questionnaire asks for
`the accurate frequency of consumption of various foods per day,
`week or month, in the year before diagnosis of cases, or one year
`before interview for controls. It contains three sets of pictures of
`foods in terms of small, medium and large portions of rice, meat
`and chicken to help visualize portions consumed.
`interviews
`took place at the homes of the participants or at the Hotel-Dieu
`Hospital
`in Montreal. Cases and controls gave their written
`consent to being interviewed and permission to consult their
`neurologist to confirm diagnosis, and dentist for their dental
`amalgam record.
`
`Dietary analysis
`Food intakes were calculated by multiplying the reported
`serving size of each food consumed by its density and daily
`consumption frequency.
`Individual nutrient
`intakes were
`
`Downloaded from htzps;//academic.oup,com/ije/arcicleeabstracc/27/5/845/652592
`by guest
`on 03 April 2018
`
`
`
`calculated using a database based on Handbook No. 8 of the US
`Department of Agriculture, USDA}7 modified and expanded
`for Canadian foods. Dietary variables were restricted to 28
`nutrients and alcohol, as well as the separate contributions of 23
`individual foods or food groups. These nutrients represented
`major sources of energy and components of fat as a suggested
`risk factor, as well as vitamins and minerals. In this study the
`supplementary vitamins and minerals intake are not included in
`statistical analysis.
`
`Statistical methods
`
`(OR) were calculated from logistic regression
`Odds ratios
`models to estimate the risk of developing MS with each dietary
`variable and with physical measures of weight and height. Un-
`conditional rather than conditional logistic regression analyses
`were performed due to frequency matching of controls rather
`than individual matching. Body mass was computed by weight
`(kg)/height2 (m). In continuous variable analyses, the OR repres-
`ents the risk for each one 'unit' increase in the variable com-
`
`pared to a baseline risk of 1.00 for the same amount of variable.
`A 'unit’ refers to the amount of nutrient per day or the amount
`of body measure for which risk is estimated. For nutrients, one
`'unit’ was equivalent to the difference between the 75th per-
`centile and the 25th percentile level of intake. For other vari-
`ables, ttnit size is spedfied in the Tables. For foods and food groups.
`it was 100 g. Because of the skewed nature of dietary variables,
`in continuous variable analyses, the 'unit’ of variable was log-
`transforrned before perfomting regression analysis. Arithmetic
`means of daily nutrient intakes and some other variables are
`preferably included over geometric means because their stand-
`ard deviations are more meaningful. Both trends in risk with
`exposure (parameter esn'mate of slope), and relative odds by
`category (in quartiles) for selected nutrients were examined.
`All non energy-contributing nutrients were adjusted for ’total
`energy’ in regression models utilin'ng the standard method of
`energy adjustment.28 Energy-contributing nutrients were ad-
`justed by the partition method for ’other sources of energy“,
`that
`is
`the calculated difference between total energy and
`energy from the nutrient being examined.29 Body mass index
`(BMI) was included in all dietary models since it was associated
`with most dietary variables as well as with MS,
`Continuous variable analyses were performed for all subjects
`as well as for males and females separately. However, categor-
`ical analysis was undertaken only on all subjects due to the very
`small number of males in the study.
`
`Results
`
`The study included 197 cases and 202 controls matched for age
`and sex. Table 1 outlines the mean age, physical characteristics,
`and daily intakes of various nutrients for cases and controls.
`In total, 61 male and 136 female cases, and 64 male and 138
`female controls were evaluated. There were no statistically sig—
`nificant differences between the mean age, weight and height of
`cases and controls. The cases had a significantly lower BMI than
`the controls (P = 0.01).
`Table 2 shows the risk of developing multiple sclerosis per urtit
`of various nutrients (log-transformed). A higher energy intake
`(OR = 2.03; 95% C1:1.13—-3.67), and animal iat intake (OR =
`1.99: 95% CI: 1.12—3.54) are significantly positively associated
`
`NUTRITION AND MULTIPLE SCLEROSIS
`
`847
`
`with the risk in ’all' subjects. The trend is similar for males
`and females. Higher intakes of vegetable protein, dietary fibre
`especially fibre irom cereals, vitamin C,
`thiamirL riboflavin,
`calcium and potassium were negatively associated with the
`risk. When any of the three types of fats (saturated fat, animal
`fat, or total fat) were examined in multivariate analyses,
`in
`separate models along with dietary fibre, thiamin, riboflavin,
`vitamin C, calcium and potassium, and adjusted for BMI and
`’other energy’, each remained a significant risk factor. The OR
`for saturated fat was 3.94 (95% CI: 125—1246, P = 0.02), for
`animal fat it was 5.27 (95% C1: 2.10—13.22, P = 0.0004), and
`for total fat
`it was 3.37 (95% CI: l.01—ll.23, P = 0.05) in
`separate multivariate models in ”all subjects.
`In the same
`models. calcium and thiamin remained significant protective
`factors with 0R around 0.17 for thiarnin and 0.20 for calcium
`(P < 0.02).
`Similar findings were observed in the categorical analyses.
`Table 3 shows the risk of MS associated with the intake of
`
`certain nutrients by quartiles for all subjects combined. There
`was a negative linear trend between dietary fibre intake and the
`risk of MS (P for trend 0.004). The same pattern was also ob-
`served for vitamin E (P = 0.06), thiamin (P = 0.003), vitamin c
`(P = 0.001), riboflavin (P = 0.0005), potassium (P = 0.008) and
`calcium intake (P = 0.0003).
`Table 4 depicts the risk of MS per 100-g unit intakes of vari-
`ous foods or food groups. The meat group, consisting of pork.
`ham. luncheon meat, hot dogs, sausages or other processed meats,
`augmented risk in all subjects (OR = 1.24; 95% CI: 1.02—1.52).
`Sweets (candy, jam, jelly, chocolate) in females only (OR = 1.29:
`95% CI: 1.07—1.55) also appeared to favour an increased risk of
`MS. However, fruit juices among all participants (OR = 0.82;
`95% C1:O.74—O.92), margarine (OR = 0.73: 95% Cl : 0.55-0.96)
`intake in males, and cereal/bread in all subjects (OR = 0.62;
`95% CI:0.40~0.97), and in males alone (OR = 0.26; 95%
`CI : 0.10—0.70) were significantly protective.
`
`Discussion
`There is insufficient literature on North Amerimn observational
`
`studies looking at nutrition and MS to permit meaningful com-
`parisons of results with this study. This study was characterized
`by a low refusal rate by both cases and controls, and listed tele-
`phone numbers for the vast majority of the target population
`in the phone directory. which diminishes the possibility of a
`selection bias important enough to invalidate our results. The
`number of cases identified for this study was also close to the
`numbers expected in the study area for the period of the study.
`However there are some inherent limitations in the assessment
`of past diet. Not only are there inaccuracies in recall of past diet
`by any method but there are large lntra-individual variations in
`diet which make the process of characterizing typical diet for an
`individual rather difficult. In addition, the presence of disease
`before its diagnosis may have altered dietary preference among
`cases. Not much is known about the preferential food habits
`of MS patients at present, The study, however, employed a
`validated food frequency questionnaire and standard data
`collection and processing techniques used by us in several
`previous studies.
`The findings of a lower BMI among cases indicates perhaps
`individuals with lower BMI are at higher risk of developing MS.
`
`
`
`
`
`848
`
`INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
`
`Table 1 Means of age, physical characteristic and daily nutrient intakes for mses and controls, Montreal 1992—1995
`
`Maia
`Females
`
`Cases (11 = 61)
`Controls (11 = 64)
`Cases (:1 = 136)
`Controls (11 a 138)
`
`Variable
`Mean
`t SD
`Mean
`it SD
`Mean
`3: SD
`Mean
`3: SD
`
`I
`
`I
`
`II
`
`I
`
`H
`
`,
`
`.
`
`'
`
`I
`
`'
`
`I
`
`I
`
`I
`
`I
`
`.
`
`II
`
`.
`
`N
`
`..
`I.
`
`I
`
`.
`
`.
`
`.
`
`‘
`
`‘
`
`7
`
`.
`
`‘
`
`'
`
`-
`
`.
`
`10.0
`381 II
`9.5
`37.5
`10.1
`41.3 .
`9.7
`42.2
`AgIeIiyearsI
`12.5
`62.3
`11.2
`60.0 ’
`14.9 V
`79.3
`15.5 '
`77.5
`Weightkg ..
`I 5.8
`161.5
`6.4
`163.2. N
`6.6.
`'
`175.6
`5.4
`175.4
`Height;'anI
`4.7
`23.9
`4.1 .
`22.5 ‘
`'
`4.2
`25.7
`4.9
`25.2
`Body massindex'
`. 730
`2358
`918
`' 2550
`887
`2904'
`. 1491
`3312 l
`Energy,kml
`I
`31.2
`95.6
`f 36.3
`“98.6
`' 35.2.
`108.5
`75.8
`, 128.3
`Protein. g
`99.9
`277.6
`115.3II
`2 299.0
`106.3
`335.7
`150.9
`348.0
`Carbohydrates.gI
`33.7
`97.4
`43.5
`106.8 I.
`45.3
`120.2
`69.9
`142.0
`ITIOlalIfal.g
`13.3
`362
`17.7
`40.6 '
`18.0
`44.8
`27.8
`54.6
`ISaturaItcd fat,gI
`13,7
`38.5
`'17.0
`41.9
`18.3
`48.6
`28.3
`57.1
`Olcic acid g
`5.9
`12.1
`6.4
`12.7 H
`7.4
`14.7
`10.6 ‘
`15.6
`Linoleic acid. g
`143.2
`373.3
`193.6
`404.2
`V
`177.2
`450.1
`285.1
`560.6
`Cholesterol. rug
`250
`62.4 I
`32.6
`70.2 I
`34.2 .
`‘ 76.8
`' 48,5 .
`97.3
`Animal fat,IgI ..
`19,2
`34.4
`16.6
`36.0
`20.4
`43.1
`29.8
`44.5
`Vegetable 131,3 ,
`26.1
`67.2.
`31.0
`70.0
`31.3
`475.6
`68.0
`I944 '
`Animalerotein,IgI
`11.8
`28.1
`9.8
`27.8
`10.0
`32.6
`12.3
`335
`YegetableIprotIeIln, g
`9.8
`22.8
`8.8
`22.5
`9.2
`”24.7
`I100
`I242
`Dietary fibre,Ig
`3.8
`8.3
`3.7
`8.1
`4.3
`7.7
`4.7
`8.4
`Vegetable 811mg
`3.4
`5.1
`3.7»
`5.2
`3.8
`4.9
`4.6
`4.6
`FniiIt fibre. g
`42
`81
`4.3
`8.1
`6.2 f
`11.1
`4.7 -
`9.7
`ICIIeIreIaiI fibre, g
`4622.8
`9898.7
`' 4996.9
`9899.5”.
`4678.8 '
`9656.4.
`4450.9
`9730.0
`Vitamin A,fU
`1491.0
`1953.8
`1176.8
`‘ 1901.8
`1746.2 '
`2186.2
`1514.7
`2593.9
`ncdnot 1U
`3249.6
`6225;6 H
`”3763.9
`' 6285.3 .
`3179.6 I
`5333.7
`2824.5
`5477.4
`BetaItI:aroIIteIrI1c,1U
`I 8.7
`. 20.6
`10.3II'
`“21.8
`9.1
`424.0
`. 13.7.
`25.8
`VitaminIIlIS, mg
`126.8.
`I 4253.0
`121.5.
`. 230.1
`143.1
`232.1‘
`117.9
`210.5
`Vitamin c, mg
`06
`1.6
`0.6
`1.6
`0.7
`1.8
`0.7
`1.9
`Thiamin,Img
`I09
`24'
`0.9
`2.3 ‘
`1.0
`2,5
`1.2-
`2.7
`Riboflavinmg
`7.7
`”22.0
`‘ 85
`23.0
`7.9 '
`247
`21,4
`31.2
`Niacin, mg
`465.9
`1090.3
`498.1
`1041.8
`462.9
`1131.0
`577.5
`1142.4
`caIIIcIium. mg
`5.2
`18.2
`”5.4
`16.5
`6.1
`18.5
`7.9
`19.9
`lron.ImIgIIIIII
`1323.9
`4097.3
`1328.6
`4120.2
`I 1289.0
`43319
`III1815.2
`4614.4
`IPIotIasinum, rIng .
`Total 4091.5
`124.2
`66.4
`120.2
`51.5
`117.8
`54.7
`109.1
`I 45.8
`
`Alcohol,g
`23.1”
`37.6 ‘
`12.6
`17.2
`6.8
`15.4
`I
`l 5.6
`7.9
`- P = 0.01.
`
`I
`
`I
`
`I
`
`I
`
`.
`
`.
`
`.
`
`I
`.
`
`.
`
`I‘
`
`I
`
`.
`
`I
`
`_
`
`II
`
`I
`
`I
`
`f
`
`although the results do not reach statistical significance in men
`due to small sample size. To what extent this is a real association
`is not clear. 1t is possible that the presence of disease may have
`contributed to a hypermetabolic state and consequent lowering
`of body weight. However,
`it is the greater height and not the
`weight that appears to be the feature distinguishing cases from
`controls.
`-
`
`A high energy intake appears to be a risk factor for MS in this
`study, with animal fat contributing to most of this effect. Again.
`whether this is a real effect or a consequence of a hypermeta-
`bolic state, is difficult to say. The subclinicai phase of MS may
`have been present for several years and may have affected the
`intake of foods among cases in a selective manner. it is possible
`that high energy and high animal fat foods may have been
`
`differentially favoured by cases compared to controls. However,
`a concomitant increase in animal protein intake (if high energy
`animal foods were preferred) does not appear to be apparent
`among all subjects combined or among women. When looking
`for biological plausibility, attention has been directed specifically
`at fats since the myelin sheath, the target of disease, is mainly
`composed of lipids, and the imbalance in the proportion of
`saturated to unsaturated fatty acids is purported to affect sus-
`ceptibility to demyclinating agents.30 The protective effect of
`polyunsaturated fatty acids, particularly linoleic add. perhaps
`due to its immune-suppressive action. has prompted clinical
`trials on PUFA supplementation.31 A non-significant decreased
`risk was observed in this study with linoleic acid intake in the
`continuous variable analysis, but none in the categorical
`
`Downinadpd Evnw Hrrps //academic Cup com/ije/arcicle-abstract/27/5/845/552392
`by guest
`on 33 Ap:;l 20:8
`
`
`
`Table 2 Risk of multiple sclerosis per unit of age, physical characteristics and daily nutrient intakes, Montreal 1992—1995
`All subjects
`Males
`Females
`(197 cases/202 controls)
`(61 cases/64 controls)
`(136 cases/138 controls)
`
`Variable
`Unjt‘
`Odds ran‘ob 1c1)C
`P
`Odds rau‘ob 1C1)c
`P
`Odds ratiob 1C1)c
`P
`
`NUTRITION AND MULTIPLE SCLEROSIS
`
`849
`
`I
`
`.
`.
`II II
`
`I
`.
`
`'
`
`l
`
`'
`
`I
`
`I
`
`II
`
`..
`
`V
`
`'
`
`I
`
`I
`
`I
`002
`
`I
`
`0.002
`
`0.03
`
`I 0.951074—1211
`”1101072711531
`0991031421)
`I10
`Agedyears
`I0I.I3I4I(0.69I—1.I04)I
`0.92I'1I0.7I3II—I1.67I1
`0.911079—1031
`'10
`Welghld,
`15.111.012.021
`09610.53—173)
`1~1.5.(9-9.1i1:“l
`I10
`I
`‘
`Mug:
`941922422”.
`04819-91911
`.5. M<W -992
`" 'i
`Baum-min“
`I I1. 32 (0. 36—3. 381
`215910877773)
`I597
`20111.13;9.621
`0.02
`W .
`0.33 (0.10—1.14)
`1.5310.23I—10.11)I
`37
`0.52 (0.13—1.44)
`mums
`'
`V
`I
`l
`1.1110413-259)
`124100571071“ .
`115
`0.73I(0.36—l.48)
`Carbohydrates g
`15710554447) I
`2.2910 56—9. 371
`44
`171 1075-390)
`Total fat, g
`1130 10.614531) H
`I2. 51 I10. 59—1070)
`17
`1.83 (0.61—4.36)
`I
`Saturated (at. g
`099102943391
`145310131605)
`18:15
`1.12 (0.43—2.39)
`011:1Icac111 g
`0.39 (0.44—1.79)
`04010.12-1331
`6.64
`0.73 (0.40—1.33)
`Linoleic add. 3
`1.02 (0.40—2.64)
`3. 59 10.39—14.45)
`19
`I 1.55 (0.72—3.34)
`01016516101, mg
`lI.68 (0.03—3.42)
`I11111.09;2.I2I61
`I 32. 64I
`1294122121
`Animauau
`1.07 (0.59—1.92)
`0.43 10.19— 1 161
`22 33
`0.84 (0.52—1.36)
`Vegetable rat, g
`06310. 29—1591
`1.7710.5I1I—6.12I)
`21.6
`I 0.9310.47~1.351
`Animal protein, g
`023.111.120.221
`034100771961
`.143
`023.10% 10.02
`W162
`II 0.I62 (0.23—1.34)
`0.43 101471291
`12.19
`05110211001
`0.05
`W4
`I 0.66 10. 36—1211
`“1.15H10.53I—2I.52)
`4.75
`08410.53—1. 331
`Vegetable fibre, g
`I1 .00 10I..I7I5I—I1I34I1
`0.87 (0.64—1.18)
`I
`4.81
`0.93 (0.76—1.14)
`IFruiItlibre, g
`I0 77 10. 44—1. 34)
`02110021101
`5.57 mama I0.02
`mm
`03010.44—145)
`109102474253)
`5332
`03310.55—141)
`VitamiInAI. 111
`0.96 (0.03—I1I.4;6)
`1.61 (0.86—3.00)
`1491
`1:11 (0.79—1.57)
`Retinol, [U
`Q-BFill-Slzlsfiol
`63516422341
`3349
`0.36 1053—1291
`13m carotene, [U
`II
`09510.43—21 11
`0.5310113-158)
`13.24
`0.78 (0.41—1.43)
`vitamin E. Img
`0110mm I003
`‘ 0.5610.27—1I.151
`0.003
`W 162
`0211020202111
`022100302101
`002
`0.25 (0.04—1.53)
`WW
`I
`I
`0.66
`02110051621 0.005
`Maw 0.004
`I0.52I10.1I5—1.s1)
`11112011311111.2113
`1.3
`022102652631 0.003
`0. 75 1023—2.41)
`12.200.632.331
`W
`8.157
`1.6710.63-4.43)
`
`MW. 0.006
`0.461017—1231I
`0319011111;
`633 IIII
`029102251101
`130. 0751. 45)
`III
`0591009738111
`irorL mg I
`6.88
`0.4010 13-1 23)
`_ 0201002520321
`0.41100772541
`mm
`. 1670
`0.29.1020;0.361
`061 1025415010 3510 39—1 86)
`Total suga1'
`Ig
`60.35
`0. 75 1043—133)
`I 2. 59 I10 95—7 14)
`174 10 33—3I671
`1:666 1119131111;
`1075
`1.56 10.61390)
`
`AlcohoL g ‘ 10110.119-1151 10.4 1.0110.91—1.12) 1.03 10.37— 1.23)
`
`
`
`' Unit = for nutrients, units are the difference between upper and lower quartile cut points for all controls.
`b Each Line represents a separate model. Odds ratios for all non energy-contnbuting nutrients are adjusted for total energy and body mass index. Odds ratios
`for a” energy—contributing nutrients are adjusted for 'energy from other nutrients‘ and body mass index.
`C C] n 95% confidence intervals.
`
`.
`
`.
`
`I
`.
`
`V
`
`'
`
`I0.o3
`
`I
`
`I
`
`I
`
`0.003
`
`0.01
`
`.
`I0I.I02
`.091
`
`.
`
`H
`
`0.05
`‘
`
`0103
`'
`
`“1 No log uanslor‘mation.
`Note: items underlined are statistically significant.
`
`analysis. The increasing risk of MS with a high animal and
`saturated fat intake, as observed in this study, is consistent with
`findings from other studies.12 it is postulated that an excess of
`saturated fatty acids could be responsible for modifying the
`stability of
`the myelin sheath or
`for
`increasing platelet
`aggregation, with hypoxia in the rnicrocirculation in the central
`nervous system and subsequent perivascular demyelinimtion.32
`This study shows a protective effect of dietary patterns
`contributing to higher levels of dietary fibre, vegetable protein.
`
`vitamin C. vitamin E. thiamin. riboflavin. calcium. and potas-
`sium which are consistent with benefits of similar dietary
`patterns in reducing risk for cancer or cardiovascular diseases.
`These factors may be involved in the regulatory process of the
`nervous system or act as an antioxidant. The study generally
`supports a protective role for components commonly found in
`plants (fruit/vegetables and grains)and an increased risk with
`high energy and animal food intake. However, further invest-
`igations in similar populations are required to confirm which
`
`Downloaded from h::ps://acadevr1c.oup com/ije/article abstract/2 5/845/652592
`by guest
`on 03 Ap:;i 202a
`
`
`
`850
`
`INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
`
`Table 3 Risk of multiple sclerosis associated with the intake of selected nutrients by quartiles‘ for all subjects combined, Montreal 1992—1995
`Exposure category
`
` Nutrientb 01 02 03 Q4 P for trend
`
`
`
`
`
`
`Toralsmtxxakel...
`I
`I
`‘IbItalIiaLgII
`II
`_.
`I
`II
`Saw-WM.
`I
`I
`AnILrInalIIIaII,g
`
`I
`
`II
`
`,_
`.
`
`I
`T991.P’°“?m'.3..
`I
`IIIIII
`VegetableprotelngI
`I
`I
`II
`II
`pletaryIflbre,g
`II
`.
`
`.
`I
`
`.
`
`.
`
`.
`
`I
`.
`
`I
`
`.
`
`I
`
`.
`
`II
`.
`
`I
`
`..
`
`.
`
`..
`
`II
`
`.
`
`..
`II
`
`I
`
`.
`
`_
`
`I
`I
`
`.
`.
`
`I
`I
`I
`......
`
`II
`II
`
`I
`
`..
`
`,.
`
`H
`
`. 0R.
`9556 of
`011
`95% c1
`01*
`9596 er
`OR
`95% or
`’95]
`.
`I' 050461
`on
`.
`95% c1
`on H I
`.
`.
`
`II
`.'
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`I
`
`I
`
`.
`
`1-00
`I}
`I
`I100
`.
`1.90
`
`1.00
`
`11-90
`
`I
`
`1.00
`
`I
`I100
`.
`I100 '
`
`1
`
`I
`
`'
`
`'
`'
`
`V
`
`'
`
`H
`
`I
`
`_
`
`I
`
`I
`
`M
`
`I
`
`.
`
`.
`
`I
`
`I
`
`.
`
`I
`
`.
`.
`I 0.00
`I
`I
`0.30 I
`
`0.25
`
`0.03
`
`0.12
`
`0.006
`
`.0000
`
`0.46
`
`0-8.9.
`(050—1591 .
`1.47
`(0.31—2.67)
`H.
`0.0%
`I
`(052—176)
`1.50 I
`(0.33—2.72)
`'
`0-7?
`(041-131)
`0.67II
`(oats—1.211
`082
`(007-000
`0.94
`(0.55—IlII67)I
`
`0-68.
`.
`(0.57—1.24)
`I
`0.89
`(0.47-1.71)
`f
`6-96,
`(0.51—1.32)
`1.03
`(0.55-1.95)
`[
`0-5'2‘9’
`'
`(016—0641 I
`0.I66III
`(0.36-1.23)
`'I
`0.61
`«00.00.01 .
`I
`1.03 I
`(0.5691041 I
`
`.
`
`I
`
`'
`
`'1
`
`_
`
`I
`
`.
`
`.
`
`..
`
`I
`
`..
`
`.
`
`.
`.
`
`..
`_
`
`I
`
`.
`
`I 0.76
`.
`.0-36
`.
`0.0.
`
`50.005
`
`0001
`I-
`0.0005
`
`I
`
`0.91
`
`0.00
`
`16.5
`(0.96—2.62)
`I.
`1.81
`(033-373)
`'
`'
`i010
`(0.77—5.33)
`i ”2.11.
`(1.094307)
`f
`'0570
`(0.32—1.53)
`0.29I-II'
`(0.13.063)
`I053}-
`~0:000:70
`I
`0.75
`013
`'
`I
`(0.40—1.40)
`959IsIIc1 I
`:
`00‘;
`I'(0.4's—1.60I)I
`(0.63—2.10)
`(0.30—1.05)
`950:. Ct
`'
`'I
`0.07
`1 072
`1.16
`QR
`mes-1:581
`194011305 .
`(0-69-04)
`9152451
`'0299'
`..
`vies.
`0170.
`0R
`«9129961
`’ 19991-19».
`10-39-1251.
`9526.61
`H
`O.36"’I
`H
`V 0.40"
`0.75
`OR
`WarnirtmgII
`(0.17—0.781'
`(021—0761
`(041-131)
`95% c1
`I
`II
`..
`0.57"?"
`0.55"”
`0.69”
`OR
`v1I1.IamlnIIc,Img_
`(0 20—069)
`(03141931
`(0.40—1.19)
`H9596 C1
`I
`I
`‘ 10.29"?
`0.44"
`(167‘
`I.
`"”60 1
`11110011017111.1115
`(0.14-0.61)
`(0.24—0.82)
`(0.38—1.18)
`7 9590a
`I
`II
`II
`'
`1.02
`H
`0.971
`1.15
`”1012'
`NlaclIn,ImgII
`II
`(0.40.217)
`10-51-1825
`10361-1911
`. #54051 i
`. ..
`033“
`0.44"
`0.32m
`1.00
`,
`{on
`13611155111111, mg
`(0.15-0.72)
`(02440.82)
`(0.10-0.61)
`II
`95561121
`I
`II
`I
`, 000::
`.‘00.0.+e:1..
`7700101.;
`‘ii 00
`j :00: :'
`.,
`'
`.
`00000000 ..... 'i‘
`VI 0.0003 95% c1 V H V (0.36-1.15) H ' (0.231.079) (0.15-0.59)
`
`
`
`
`
`
`
`
`‘ Cut points 1m quartiles: total energy 1952, 2427, 2848 katL' total [at 74.72. 99.37, 122.58 g: saturated [at 27.94, 36.15, 45.46 3; animal fat 46.59. 63.54. 30.79 3;
`total protein 76.9, 95.6, 114.2 g: vegetable protein 21.7, 27.4. 40.0 g; dletary fibre 16.73, 22.26, 28.80 g; vitamin A 6120, 8949, 12 124 IU; retina} 1067, 1916.
`2559 IU: beta airmen: 3589, 5-569. 7435 IU.’ vitamin 814.68, 20.34, 27.71 mg; Lhiamln 1.28, 1.60, 1.93 mg; vitamin C 156.38, 227.60, 309.87 mg; riboflavm
`1.75, 2.27, 3.00 mg; niacin 17.52, 21.71, 26.38 mg; potassium 3317. 3912. 4902 mg. calcium 7651,1047. 1353 mg.
`b Each line represents a separate model. odds ratios for all non energy-contributing nutrients are adjusted for total energy and body mass index. Odd ratios
`for all energy—contributing nutrients are adjusted for 'energy trom Other nutrients' and body mass index.
`‘ CI = 95% confidence intervals.
`" P < 0.05; " P < 0.01; ‘“ P < .001.
`
`Vttanttnaru
`I
`00000100
`I
`I
`I
`II
`Beta carotene,IU
`.
`.
`.,
`.
`“Emma,
`
`..
`
`.
`
`I
`
`H
`
`H
`
`I
`
`‘
`
`I
`
`.
`I
`
`1
`
`I
`
`.
`
`H
`
`'
`
`,
`
`.
`
`..
`.
`
`.
`
`.
`
`.
`
`V
`
`-
`
`‘
`
`.
`
`I
`
`.
`
`I
`
`II
`
`II
`
`I
`
`'
`
`1
`
`U
`
`,
`
`‘
`
`I
`
`I
`
`00
`.
`1.00
`.
`1.00
`
`1.00
`I
`1.00.
`
`1.00
`'
`1.00
`
`.
`I
`.
`
`.
`
`.
`
`I
`..
`
`‘
`
`.
`
`I
`
`1
`
`'
`
`.
`
`I
`
`factors consistently contribute to increased or decreased risk of
`developing MS. This study was somewhat exploratory in nature
`due to a lack oi sufficient published studies to generate a priori
`hypotheses. A large number of findings did not reach statistical
`significance due to small sample size, but it confirms a possible
`role of dietary factors in the causation of MS.
`
`Acknowledegments
`We wish to thank the Multiple Sclerosis Association 01 East
`Montreal for their finandal support. We also thank Ms Marie-
`Clalre Goulet,
`research assistant
`for
`this project, Ms Diana
`Sanderson for technical support, Mr Ovid Da Silva for his edit.
`orlal work, and Ms Helene Harnois for typing this manuscript.
`
`Downloaded frrw hhfins'//ar~ademic.oup.com/ije/artipleiabstrac:/27/S/845/552592
`by gues:
`on 03 April 2318
`
`
`
`Variable
`Milk
`YoIgunI
`Coiiee
`
`I
`
`.
`
`’
`
`1»
`
`.
`
`.
`
`004
`
`Table 4 Risk of multiple sclerosis per 100 grams of foods per day (log transformed), Montreal 1992—1995
`________~____________________________________________._______________________________________________________________,_________________._______________
`Males
`Females
`All subjects
`
`(61 asses/64 controls)
`(136 cases/138 controls)
`(197 cases/201 controls)
`Odds ratio“ (C1)b
`Odds ratio“ (C1)‘3
`Odds rado‘ (c1)b
`-.°-93 (Q.I83-1I. 05)
`099 (0134—113)
`.Oz".5 .(9157‘l395)
`097 (081 1.16)
`I0.99 (0. 84—I1I09)I
`0.96 (0.37—1.06)
`107 (0.9IEIlI-I.1Il'I7)
`1.013(093—195)
`1.07 (0.99—1.16)
`:19.212112945991121
`129292652951
`34(0 59—119)
`092 (017941.16)
`0.90 (0.74—103)
`use“;
`0.94 (0.65-1.33)
`IV'0I.23_1.I(I0'.63-1.'o4)
`035(069—104)
`Egg
`I 0.99 (0.53-1.69)
`H 1:.0I5 (0.81—1.36)
`103I(0.—3I21.30)I
`Pasta/pluné-HH
`0.86 (0.55-1.33)
`129(095—177)
`. 1.13 ((l..8&l44)
`Whitc'root veg.
`'
`0.91I(I0.6§—1.2(I5)
`095 (0.7W1. 19)
`0.94 (0.73—1.13)
`Red i661 veg.
`
`
`1.04I(I0.79_1.37)
`0.3910.7 £1.10)
`0.94 (0.30—1.12)
`011515605 veg.
`III1.01(0.7I6»1.35)
`"092 (075—1. 14)
`0.95.I(0.so—1.13I)
`Peas
`'
`‘
`I
`’
`0.95 (0791.13)
`Other veg.
`1.10 (073-167)
`090 (0.68-1 is) ‘
`.1.1.5.199%1-49>
`0861971 .1931
`0.96 (0.82—1.12)
`Butter.
`'
`m(w 0.02
`1.I09 (0. 90—1. 33)
`0.96 (032—112)
`W‘ I
`II 1.49 (0.73—1.67)
`o.30 (0I6IZ~l. 03)
`0.199(073-110)
`0115
`I
`1.10 (0.98-1.36)
`.
`.
`BeefI
`.. 1.IliI3I(t).I81I—l.7I3)
`II1III.I05(0. 89—137)
`I1I.2I9(0.93I—1.I60)
`.
`I
`1.21 (0.94-1 55)
`W gum 0.03
`Eish,
`'
`I
`091 (0.719-105)
`1.06 (0.34—1.40)
`muw II0.05
`12.211W 0.0013
`0.136 (050—1419)
`M 962(94952921
`9.77. (9.53.1101)
`99.5.1977:17.1 .
`Cakes
`0. 33 (0.75— 1. 04)
`0.85 (0.65—1.11)
`1.013 (013471.37)
`.
`I
`I
`Fruits
`0.97 (0.31- 1.15)
`0.83 (9.63-1.10)
`. 0.007 l
`51mm:
`1.12 (0.97—1.30)
`Chicken
`0.86 (0.67—1.10)
`0.74 (0.44—1.23)
`
`1.02 (0.81—1.29) ”0.92 (0.77-1.09)~
`Stew
`0.96 (0.83-1.l0)
`‘ Odds ratios for all foods are adjusted for total energy and body mass index.
`5 Ci = 95% confidence intervals
`Note: items underlined are statistically s1gnlilcant.
`
`NUTRITION AND MULTIPLE SCLEROSIS
`
`851
`
`P
`
`0:006
`
`I
`
`I
`
`1»
`
`0.92
`
`I
`
`I
`
`I
`
`l
`
`
`
`References
`
`lSheUey E, Dean G. Multiple sclerosis. in: Mlliereral. (eds). Epidemiology
`ofDr‘seam. Oxiord: Blackwell Scientific Publishers. 1982. pp.347-S4.
`2 Confavreux C, Goudable B. Moreau T :1 al. Etiologle de la sdérose en
`plaques. Rev Pran'dm (Paris) i99l;41:1888—92.
`3Alter M, Yamoor M. Harshe M. Multiple sclerosis and nutrition.
`Arch Neural 1974;31:267—72.
`
`4'Hauser 51.. Multiple sclerosis and other demyelinating diseases. in:
`lsselbadte t! al. (eds). Han-Bonk Prina'pla aflnltrnal Medicine. 1301 Edn.
`New York: McGraw-Hlll Inc.. 1994, pp.2287—95.
`5Ernard JP, Thouez JP, Gauvreau D. Neurodegenerative diseases and
`risk (actors: a literature review. Sac Sa'Med 1995;40:847-58.
`
`6 lngalls TH. Endemic dustering oi multiple sclerosis in time and place,
`1934—1984. Confirmation of a hypothesis. Am J Farmn': Med Pall-ta!
`l986:7:3—8.
`
`7 laborde JM. Dando WA, Tedzen Ml. Climate, diflused solar radiation
`and multiple sclerosis (review). Soc Sa'Med 1988;27:231—38.
`511114: 1. Growing M. Klauber M :1 al. Clustering of residence of
`multiple sclerosis patients at age 13 to 20 years in Honialand, Norway.
`Am J Epumiol 1991;133:932-39.
`
`9 Dean G. Brady R. McLougghlin H. Motor neurone disease and mul-
`tiple sclerosis among immigrants to Britain. BrJ PrevSocMrd 1977:31:
`141—47.
`
`mCasetta I. Gram'eri E. Malagu 5 er 4!. Environmental risk factors and
`multiple sclerosis: a community-based. case-control study in the pro
`Vince of Ferrara. 1taly.Neuroepidemialogy i994;13:120—28.
`
`H Kriiger PG, Nyland ill. The role of mast cells and diet in the onset and
`maintenance of multiple sclerosis: a hypothesis. Med Hypothm 1995;
`44:66—69.
`
`‘2 Espam ML Sasaki S, Kesteloot H. Nutrition. latitude, and multiple
`sclerosis monality: an etiologic study. Am JEpidmtol 1995;10:733-37.
`”rota MR, Grartleri e, Malagu s a 41. Dietary habits and multiple
`sclerosis. A retrospective study in Ferrara.
`ltaly. Ana Neurol 1994:
`16:189—97.
`
`H lauer K. Dietary changes in temporal [elation [0 multiple sclerosis in
`the Faroe islands: an evaluation of literary sources. Namepidzmiology
`l989;B:200—06.
`
`lslauer K. Firnaba W. Epidemiologic aspects of multiple sclerosis.
`Vm‘tdierungmedizin 1992;44:125-30.
`l"Swank RL. Multiple sclerosis, a correlation of
`dietary fat. Am JMtdSa' 1950;220:421—30.
`
`its inddence with
`
`Downloaded from h:tp5://academ2c.cup.ccm/ige/articleVabstract/27/S/845/652592
`by guest
`an 01 April 2018
`
`
`
`852
`
`INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
`
`‘7 Swank KL Lersted O, Strom A (1 a1. Multiple sclerosis in rural Norway.
`N EnglJ Mtd 1952;146:721—28.
`
`)5 Bates D. Dietary lipids and multiple sclerosis. Ups J Med Sci 1990:
`4i(Suppl.):i73——87.
`
`19Muriell TG, Harbige LS, Robinson 1C. A review of the aetiology of
`multiple sclerosis: an ecological approach. AM Hum Biol 1991:18:
`95—112.
`
`20Malosse D, Perron H, Sasco A n a]. Correlation between milk
`and dairy product consumption and multiple sclerosis prevalence:
`a worldwide study. Nmraepidemt’ology 1992;11:304—12.
`in the
`21Ben-Shlomo Y, Davey~Smith G, Marmot MG. Dietary lat
`epidemiology 01 multiple sclerosis: has the situation been adequately
`assessed? Nturoepidemiology 1992;11:214—15.
`
`22 Hutter C. On the causes 01 multiple sclerosis Med Hypotheses 1993:41:
`93-96.
`
`23 Ghaditian P, Simard A, Balllargeon J n a]. Nutritional [actors and
`pancreatic cancer
`in the francophone community in Montreal,
`Canada. Int J Canar199l;47:l—6.
`
`2“ Jain MG, Harrison L Howe GR rial. Evaluation oi a sell-administered
`dietary questionnaire for use in a cohort study. Am J Clin ~qu 1982;
`36:931-35.
`
`25Jain MG. Howe GR, Rohan T. Dietary assessment in epidemiology:
`comparison of a food frequency and a diet history questionnaire with
`a 7-day food record. Am J Epidmiol 1996;10:953—60.
`
`“Shatenstein a, Ghadir