|

 

Rh37030450

British Journal of Rheumatology 1998;37:274–281
UNCOOKED, LACTOBACILLI-RICH, VEGAN FOOD AND RHEUMATOID
M. T. NENONEN, T. A. HELVE,* A.-L. RAUMA and O. O. HA
¨ NNINEN
Department of Physiology, University of Kuopio and *Kivela¨ Hospital, Helsinki, Finland We tested the effects of an uncooked vegan diet, rich in lactobacilli, in rheumatoid patients randomized into diet and controlgroups. The intervention group experienced subjective relief of rheumatic symptoms during intervention. A return to anomnivorous diet aggravated symptoms. Half of the patients experienced adverse effects (nausea, diarrhoea) during the diet andstopped the experiment prematurely. Indicators of rheumatic disease activity did not differ statistically between groups. Thepositive subjective effect experienced by the patients was not discernible in the more objective measures of disease activity( Health Assessment Questionnaire, duration of morning stiffness, pain at rest and pain on movement). However, a compositeindex showed a higher number of patients with 3–5 improved disease activity measures in the intervention group. Stepwiseregression analysis associated a decrease in the disease activity (measured as change in the Disease Activity Score, DAS ) withlactobacilli-rich and chlorophyll-rich drinks, increase in fibre intake, and no need for gold, methotrexate or steroid medication(R2 = 0.48, P = 0.02). The results showed that an uncooked vegan diet, rich in lactobacilli, decreased subjective symptoms ofrheumatoid arthritis. Large amounts of living lactobacilli consumed daily may also have positive effects on objective measuresof rheumatoid arthritis.
K : Rheumatoid arthritis, Vegan food, Activity index, Vitamin B12, Sodium excretion, Lactobacilli.
P with rheumatoid arthritis (RA) often claim 26 mg, n = 13), zinc (9 mg to 18 mg, n = 13) and that their symptoms are alleviated by a special diet or niacin (11 mg to 19 mg, n = 13) increased during the by simple elimination of certain constituents from intervention. There was also an increase in the intakes their free-choice diet. Foods most often linked with of vitamins C and E [10].
worsened symptoms are red meat, spices, flour prod- Kjeldsen-Kragh et al. [11] have reported positive ucts, citrus fruits, chocolate and alcohol [1]. Improving effects of fasting and 1 yr of a vegetarian diet in RA.
symptoms have been reported with vegetables, oils and Their patients were mainly in functional group II and fish [1]. True food allergy seems uncommon in patients had only mild medication. In their study, there was a with RA [2]. Fasting is an effective treatment of the high drop-out rate (35%), most (22%) due to flare-up symptoms of RA, but most patients relapse on reintro- of arthritis symptoms. Kjeldsen-Kragh [12] summar- duction of food [3, 4].
izes the studies of the group, stating that the beneficial Haugen et al. [5] have collected data from patients effect of the dietary treatment was perhaps not related suggesting that extreme vegan diets have alleviated to the diet per se, but was caused by alterations in the their rheumatic symptoms. ‘Living food' teachers and microflora secondary to changes in the diet.
consumers have also reported beneficial effects of the Alterations of intestinal bacterial flora may play a diet [6–8]. ‘Living food' is an uncooked vegan diet, role in RA. The composition of the intestinal flora of rich in lactobacilli, which contains no animal products, patients with RA seems to differ from that of healthy raffinated substances or added salt. A detailed descrip- subjects [13]. Uncooked vegan food increases the tion of the diet is presented by Ha¨nninen et al. [7].
counts of faecal lactobacilli in the healthy population The majority of food items are soaked and sprouted [14] and in rheumatoid patients [9]. Peltonen et al. (seeds and grains), and many are fermented. Some [15] have shown that a change in the faecal microflora items are blended and dehydrated (bread ). Fermented was connected with decreasing activity of RA in the products contain high amounts of various lactobacilli above-mentioned 1 yr intervention with vegan and [9]. Fermentation and mechanical processing distin- vegetarian diets [11].
guish this diet from other vegan diets.
Our previous studies [7, 8] have shown anecdotal Finnish RA patients often consume a deficient diet.
evidence of positive effects of the extreme, uncooked, Rauma et al. [10] have found that their calculated vegan diet (‘living food') in RA. The aim of the intakes for energy, iron, zinc and niacin were lower present study was to investigate subjective and than those in healthy persons. Shifting to a ‘living objective effects of this diet on chronic RA, and to food' diet in the present study significantly increased select possible therapeutic components of the diet for the daily intakes of energy (6.6 MJ to 8.9 MJ, n = further studies.
13). The calculated daily intakes of iron (13 mg to SUBJECTS AND METHODS Submitted 23 October 1996; revised version accepted 2 July 1997.
Forty-three consecutive adults with diagnosed (ARA Correspondence to: M. Nenonen, National Research and Development Centre for Welfare and Health, PO Box 220, criteria; [16 ]), chronic and active RA (Steinbrocker's FIN-00531 Helsinki, Finland.
functional class II–III; [17]) visiting the Rheumatic 1998 British Society for Rheumatology NENONEN ET AL.: VEGAN FOOD IN RHEUMATOID ARTHRITIS Outpatient Clinic at the Kivela¨ Hospital, Helsinki, daily urinary sodium excretion. Accurate use of the Finland, were selected for the study. There were no intervention diet [7] causes a decrease in urinary refusals. All selected patients had active joint symptoms sodium excretion to less than one-third part.
(more than three swollen or five tender joints) andelevated inflammatory parameters [erythrocyte sedi- mentation rate (ESR) >20 mm/h, or C-reactive pro- One rheumatologist ( TAH ) carried out the clinical tein (CRP) >10 mg/l ]. Patients were randomized into evaluation blindly. The patients filled in questionnaires two groups. The intervention group started the experi- recording their subjective experiences and gastrointest- mental diet and the control group continued their inal functions on the 0–10 scale at the beginning, in previous omnivorous diet. Three intervention patients the middle, and at the end of the dietary intervention.
could not eat all of their diet, and two of them refusedto continue in the study after a few weeks; one stopped This questionnaire has been modified from the ques- tionnaires used in our previous studies [7]. It is a ffered abdominal pains and distension.
One patient from the control group stopped the study modification of a visual analogue scale ( VAS ) with a for personal reasons, and one died of a heart attack numbered scale. Patients were interviewed after the just after the 3 month follow-up period had ended.
study, and their experiences were collected by a struc- Their data are used where adequate, no extrapolations tured discussion. Three months after the study period, their impressions about the effects were recorded again One intervention subject with high sodium excretion with the 0–10 scale.
(diet non-compliance) and one patient from the controlgroup (because of medication-induced hepatic toxicity) were excluded from analyses of interfering variables.
Fasting blood, daily urine and faecal samples were None were hospitalized and all patients continued their collected according to normal laboratory practice.
previous medication at the beginning. The medication Samples for the intervention period were collected: (1) was modified when necessary on clinical grounds (by before the dietary intervention (weeks −1 and 0); (2) rheumatologist TAH ). Caffeine-containing drinks, after the first month (weeks 4–5); (3) at the end of the chocolate, alcohol and tobacco smoking were prohib- intervention period (weeks 8–9 or 12–13); and (4) 3 ited in both groups.
months after the intervention period. Most of the The duration of the intervention was planned to be analyses were carried out immediately. The samples 3 months, but eight patients had to stop their inter- analysed later were stored at −20°C. The analyses vention diet after 2 months because of nausea, were carried out with normal laboratory methods used diarrhoea (n = 3) or difficulties with the taste of some in the hospital.
food items. The controls stopping the follow-up after2 months were selected to match these interventionsubjects by age, sex, disease activity and body mass index. There were no differences between the 2 and 3 The normality of interval variables was checked by month intervention groups, and the duration of the calculating the Shapiro–Wilk statistic W, and data intervention had no effects on the clinical outcome.
were scrutinized for outliers. The data for C-reactive The basic data of the subjects are presented in Table I.
protein (S-CRP) and the duration of morning stiffnessdeviated intolerably from the normal distribution, andnormal scores calculated from the ranks [18] were used in statistical analysis. Analyses were performed with The 7 day dietary records were collected by a MANOVA for repeated measurements [19] using qualified dietitian three times: the first before theintervention, the second in the middle, and the third different combinations of independent variables and at the end of the intervention. Subjects in the dietary change in weight as a percentage as covariate.
intervention group received all the components of their The overall changes in the disease activity were daily diet from a specialized ‘living food' kitchen in analysed with a composite index described by Paulus packed form. The kitchen weighed the components, et al. [20]. The index was calculated from the changes and the subjects recorded items they did not consume in the following six variables: ESR, number of swollen and the amount of extra food. Subjects were super- joints, number of tender joints, rheumatic pains on a vised and tutored daily by the teacher of the ‘living VAS, HAQ and global patient estimate. A decrease or food' diet. The patients in the control group prepared increase of 20% or more was estimated as significant their omnivorous meals at their homes without and the number of significantly changed variables was calculated for each patient. Changes in ESR within The nutrient contents of the diets were calculated normal values and <5 mm/h were judged as non- with UNIDAP ( Unilever Dietary Analysis Program, significant. Zero values as divisor were replaced by Paasivaara Ltd, Finland ), and the results have been 0.01, and the clinical relevance and significance of the published separately [10]. The dietary compliance change in percentage was checked individually in these of the intervention group was followed by daily events. Separate indices were calculated for improved interviews, dietary records and by analysing their and deteriorated variables. The percentage of each BRITISH JOURNAL OF RHEUMATOLOGY VOL. 37 NO. 3 Test statistics (A), anthropometric data (B), history (C ) and medication (D) of rheumatic disease and compliance with the diet of theintervention and control groups (E ) in the 2 to 3 month intervention study with an extreme vegan diet, ‘living food'. Interval data are given as means and ..
Intervention group vs control A. Test statisticsRandomized for study Started the study Completed the study Follow-up after study Duration of test (2/3 months) B. Anthropometric dataSex: male/female P = 0.02† Body mass index (kg/m2) n.s. (P = 0.06)* C. History of rheumatic diseaseRheumatic disease (yr) Seropositive disease D. MedicationGold (i.m. or p.o.) Non-steroidal anti-inflammatory drugs E. Dietary complianceStrict dietary compliance Use of fermented wheat drink (500–1000 ml/day) (only 5 patients usedthe fermented wheatdrink daily and hada strict compliancewith other fooditems) *Fisher's exact test.
†t-test.
number of variables within the intervention and control groups was compared with Fisher's exact test.
The Disease Activity Score (DAS ) was calculated The randomization divided subjects into intervention and control groups with no difference in height, weight, Rheumatism ( EULAR; [21]). The stepwise regression body mass index, duration of rheumatic disease, sero- analysis was used to build models explaining the positivity and medication. Yet the intervention group changes in disease activity described by DAS.
was younger (49 vs 56 yr, P = 0.02) (Table I ). The Maximum R2 improvement, minimum R2 improve- dietary compliance in the intervention group was good, ment, stepwise and backward elimination methods as shown by the urinary excretion of sodium. Only were used. Medication, diet, disease history and one patient lacked a clear decrease. However, only six anthropometric data served as independent variables.
of the 19 subjects used all their fermented wheat The power of significance of the tests was calculated drink (Rejuvelac) daily (recommended amount 500– from the nomogram presented by Altman [22]. For a 1000 ml ). According to the daily records, nine patients 30% change in most variables, the statistical power had minor deviations from the diet. The diet of the was >0.5 at a significance level of 0.05. On pains at control group remained stable during the intervention rest, HAQ, morning stiffness, Ritchie indices, ESR, period. However, some started to use minor amounts alanine (ALAT ) and c-glutamyl (c-GT ) transferases, of ‘living food' items. The patients could keep the this power was achieved for 50% changes. The vari- secret and the clinical observer ( TAH ) remained ation in CRP was so high that the power to detect a blinded during the follow-up. All were on stable 50% change was only 0.25. The subjective estimates of medication and continued with least possible changes patients were analysed by Mann–Whitney U-test.
(decrease of medication for six patients in the interven- NENONEN ET AL.: VEGAN FOOD IN RHEUMATOID ARTHRITIS tion and for three patients in the control group, had a rising tendency at 3–6 months after intervention increase of medication for one control patient).
(P = 0.15) (Fig. 1). ESR increased during the inter- The intervention group lost weight, 9%, during vention period in both groups. This rise was seen the intervention, while the controls gained 1% of in methotrexate users (40%, P = 0.03), and it was weight. The difference in the weight changes between independent of the intervention diet.
the groups was significant (P = 0.0001) and was notexplained by medication when tested separately for Composite indices methotrexate, gold and steroids ( Fig. 1).
The composite index, after Paulus [20], for changes in disease activity showed in the intervention group Laboratory values an improvement of 20% or more in 2.9 variables ( Serum alkaline phosphatase (P = 0.01) and alanine 1.7). In the control group, only 2.0 variables improved aminotransferase (0.002) decreased in the interventiongroup. The change in weight as covariate abolished (.. 1.3, F = 3.79, P = 0.059, ANOVA). There were the significance of these changes. Serum protein values no differences in the mean amounts of deteriorated in the intervention group were significantly lower at variables (1.4 vs 1.6). The percentage of patients with the end of the intervention than before it (P = 0.025), four (P = 0.076) and five (P = The decrease in protein levels was seen in patients on 0.05) improved variables was greater in the interven- methotrexate treatment (P = 0.05) independently of tion group ( Fig. 2). This difference was nearly signi- the diet. Albumin levels decreased in both groups with ficant (P = 0.056–0.1) even when the data were no effect of diet (P = 0.005).
analysed according to the intention to treat principle.
Serum vitamin B12 levels decreased in the interven- The percentage of patients with one or two deterior- tion group (P = 0.0006). Values already differed after ated variables was not higher in the intervention 4–5 weeks (P = 0.02). This decrease was strongest in the most compliant diet subgroup (used the diet with A decrease in disease activity, as measured as a no aberrations and consumed all the fermented wheat change in the DAS [21], during the intervention was drink daily). Their values decreased from 308 to associated in a stepwise regression model (R2 = 0.48, 179 pmol/l (P for combined effect of accuracy of diet P = 0.02) with increasing daily amounts of wheat grass and use of drink = 0.05). Serum calcium (corrected drink and fermented wheat drink, increased intake of for serum protein) decreased in both groups (P = dietary fibre, and decreased intake of iron during the 0.0001). Serum sodium (P = 0.0001) levels decreased intervention, and no need for gold, methotrexate or in both groups in the first half of the intervention.
steroid medication at entry ( Table III ). In the interven- Daily excretion of sodium decreased (P = 0.0001) to tion group as a whole, the changes in DAS were not one-fourth of the pre-test level in the dietary interven- statistically significant (given as mean/95% confidence tion group, whereas excretion of potassium increased interval at the beginning, in the middle, at the end (P = 0.02).
and 3 months after: 3.26/2.88–3.63, 3.12/2.73–3.50,3.01/2.54–3.48 and 3.13/2.70–3.57 in the interven- Most patients in the dietary intervention group 3.46/3.02–3.90 and −3.56/3.18–3.94 in the control experienced positive subjective changes during the group, respectively; P = 0.7, MANOVA for repeated intervention diet in rheumatic pains, rheumatic joint swelling, morning stiffness and general impression( Table II ). Most of the control group experienced no change. The difference was statistically significant This study showed that an uncooked vegan diet rich (Mann–Whitney U-test, P < 0.03). When the interven- in lactobacilli, ‘living food', caused subjective improve- tion was over, the majority of intervention patients ment in the symptoms of RA. The objective measures reported either no change or a negative change in the of disease activity did not change when analysed above-mentioned parameters (P < 0.01), except the separately. Two indices describing the activity of RA ability to move. Correlation analysis showed thatthe subjective estimates of disease activity were mainly were used to analyse the results (composite index by dependent on the number of tender joints (Pearson Paulus, DAS by van der Heijde). Both indices showed fficients 0.41–0.6, P = 0.02–0.0001).
ffect of the dietary intervention with ‘living food' or an effect of some of its components on Activity measures of rheumatoid arthritis the symptoms of RA.
CRP, ESR, B-haemoglobin, B-thrombocyte count, The following group of dietary factors was partially Ritchie index, HAQ, morning stiffness, VAS for pains (48%) responsible for the observed decrease in the at rest and on movement did not behave statistically disease activity index: fermented wheat drink, wheat differently in the intervention and control groups. The grass drink, dietary fibre and iron. These factors are steroid users in both groups had an 18% lower number indicators of compliance with the ‘living food' diet.
of tender joints (P = 0.02) with a decrease at the end Fermented wheat drink (Rejuvelac, [6 ]), the water of the study and 40% lower swollen joint numbers phase of germinated wheat seeds and water (1:3) (P = 0.002). The CRP values of the intervention group mixture (fermented for 48 h with freshly cut wheat BRITISH JOURNAL OF RHEUMATOLOGY VOL. 37 NO. 3 F. 1.—The results of the clinical and laboratory follow-up of the intervention and control groups in the 2 to 3 month intervention study inrheumatoid arthritis patients (N = 39) randomized to either intervention group starting an extreme vegan diet, ‘living food', for 2–3 monthsand omnivorous controls. Data were collected during and 3 months after the dietary intervention. Periods used in the figures are: Start, at thebeginning of the intervention; Middle, in the middle of the intervention; End, at the end of intervention (2 to 3 months); After, 3 months afterthe intervention. Data are given as means and 95% confidence interval (95% CI ). Intervention group, /; control group, --$--; regular useof fermented wheat drink, ···%···; no regular use of fermented drink, ···#···. Statistical significance (MANOVA for repeated measurements)is given as P values for the effects of time alone/effects of division to intervention vs control groups/effects of consuming fermented wheat drink.
grass), gives large amounts (2.4–4.5×1010/day) of effects on other bacteria in the gut and regulate their viable Lactobacillus plantarum and L. brevis strains number [23–25]. The role of lactobacilli is further [9], and it modifies the intestinal microflora. The supported by the findings of Peltonen et al. [15] and counts of faecal lactobacilli were higher in the interven- Eerola et al. [26 ]. They found that in both the study tion group [9] and their faecal b-glucuronidase activity of Kjeldsen-Kragh [11] and in this intervention, the decreased during the study (P = 0.04, MANOVA for positive clinical effect was associated with a change in repeated measurements). Lactobacilli have many the colonic microflora.
NENONEN ET AL.: VEGAN FOOD IN RHEUMATOID ARTHRITIS The subjective estimates of the rheumatoid patients about their disease symptoms during and after dietary intervention (uncooked, lactobacilli-rich, vegan food in the test group). Data were collected immediately and 3 months after completion of the study on 0–10 scales (5 = no change). Significant difference between groups was tested with the Mann–Whitney U-test (from the original 0–10 scale material ) Intervention group Swelling of joints Morning stiffness P = 0.0008 General impression Swelling of joints Morning stiffness General impression P = 0.07 n.s.
*Two patients stopped the diet after a few weeks, one later; some patients left some qustions unanswered.
†One patient stopped the study for personal reasons and one died just after the intervention period. Some patients left some questions Stepwise regression analysis of the disease activity score (DAS ) forthe change in the rheumatic disease activity during the 2 to 3 monthintervention period (maximum R2 improvement method, d.f. = 8)in a group (N = 39) of rheumatoid arthritis patients randomized to either an uncooked vegan or omnivorous diet R2 for the model Prob > F Prob > F Gold (p.o. or i.m.) Fermented wheat drink Wheat grass juice (ml/day) Change of fibre intake (%) Change of iron intake (%) Daily urinary sodium excretion at the start(mmol/24 h) The other drink, wheat grass drink, is consumed at 50–150 ml/day. It is pressed from fresh wheat grass F. 2.—The 2 to 3 month intervention study in rheumatoid arthritis and presumably contains high amounts of chlorophyll patients (N = 39) randomized to either intervention group starting and coumarin [27]. Its effects are unknown. The daily an extreme vegan diet, ‘living food', for 2–3 months and omnivorous intake of dietary fibre increased 2- to 4-fold when controls. Percentage of patients in the intervention and control shifting to the intervention diet [10]. This may have groups with at least 20% improvement or worsening in 0–6 of the contributed to the changes in colonic microflora [28] following variables: ESR, number of swollen and tender joints, by shortening the intestinal transit time [29]. In this rheumatic pains ( VAS), HAQ and global patient estimate. Statistical study, shifting to uncooked vegan food increased the significance is calculated with Fisher's exact test and the results aregiven as P values for all randomized patients (intention to treat calculated daily intake of iron by 98% [10]. The analysis of 42 patients) and for the patients completing the 2 to 3 haemoglobin levels remained unchanged, however. The month intervention (N = 39).
absorption of iron was probably impaired. There is BRITISH JOURNAL OF RHEUMATOLOGY VOL. 37 NO. 3 evidence about alterations in iron metabolism in RA subjective experiences mainly with the number of [30], and Haugen et al. [31] have shown a negative tender joints. Thus, their estimates may have been correlation between serum iron and ESR.
influenced by even small variation in their status and The calculated energy intake of this experimental strengthened by the great expectations caused by the diet was not hypocaloric, but it caused a decrease in strange and ‘mystic' diet.
body weight of 9%. This may be due to the low When the results of this study are compared with biological availability of the energy-yielding nutrients those of the minocycline intervention studies [33, 34], that was caused by the insufficient food processing and the most striking difference is the almost total lack of the high fibre content of the diet (42 g/day; [10]). It subjective improvement in the subjective variables could also result from the difficulties experienced with during the antibiotic treatment. However, both types the diet (nausea, etc.). The loss of weight could have of interventions probably caused changes in the intest- influenced the immune response and explained part of inal microflora. One explanation for this discrepancy the results. On the other hand, the multivariate analysis could be either less placebo effect in the double-blind of variance does not support this. The calculated antibiotic intervention or differences in the gastrointest- content of the experimental diet covered most of the inal effects of these studies (selection of bacteria?).
recommended daily requirements [10]. The calculated Many recorded parameters behaved differently in daily protein intakes in the intervention group patients with different medications (methotrexate, ster- increased from 58 to 80 g/day; in the control group, oids, etc.). Many of these effects probably have no real the intake was stable (57–59 g/day; [10]). These values significance because of the small number of patients in are near to those measured in long-term users of this each group. The effect of dietary manipulation might diet (48 g/day; [32]). The lowered serum albumin be better studied without drugs, but this was considered values were found in both groups and were not solely unethical. The association of decreasing disease activity associated with the intervention diet. Achieving this with no need for arthritis-specific drugs may denote a kind of daily intake demands that all food items are better effect of the diet in less active disease.
consumed in recommended amounts. This is not always This study showed that subjective relief of the symp- possible because of the unfamiliar taste and the strenu- toms of the RA could be achieved with a radical ous and time-consuming processing of some items. It dietary manipulation (uncooked extreme, lactobacilli- must also be remembered that the intervention diet rich vegan diet, ‘living food'). There were, however, corrected many deficiencies in the diet of the RA no significant effects on the separate objective disease patients studied by Rauma et al. [10]. This alone may markers (CRP, ESR, joint counts, etc.). The activity have had positive effects on the patients in the interven- index calculated from four disease activity indicators tion group. The results of regression analyses did not, did, however, find a statistically significant connection however, support this hypothesis. The stepwise regres- between compliance with the tested diet and the sion analyses did not include changes in the problem- decrease in disease activity.
atic nutrients in the models explaining positiveresponses.
Half of the patients experienced adverse effects The authors thank all their brave and patient (nausea, diarrhoea) during the diet and stopped the patients, Dr E. Leskinen ( Kivela¨ Hospital ), J. Laakso experiment prematurely, three during the first days or and I. Ruokonen (MILA Ltd ) for laboratory analyses, weeks and eight after 2 months. There were no severe Mrs M. Svennevig (Green-Way restaurant) for provid- side-effects caused by the diet, but the high premature ing the diet and tutoring the intervention group, Mrs cessation rate shows that extreme diets are not good L. Pajanne ( Kivela¨ Hospital ) for the practical arrange- for every patient. Caution and sound rationing is ments in the rheumatology out-patient clinic, Mrs A.
needed both from the patient and the doctor.
Rokka for her help in the dietary analyses, and Dr H.
The indicators of disease activity behaved as Lenzner ( University of Tarto) for analysis of the faecal expected. The changes were not statistically significant bacteria. This study was supported by the Juho Vainio as in the Kjeldsen-Kragh et al. [11] study. Exclusion of the unfavourable patients in their study may havecaused this difference. The patients in the present study also had a more severe disease history and used more 1. Garrett SL, Kennedy LG, Calin A. Patients' perceptions medications. When the same kind of composite index of disease modulation by diet in inflammatory (rheumat- [20] was used in both studies, the proportions of oid arthritis/ankylosing spondylitis) and degenerativearthropathies. Br J Rheumatol 1993;32(suppl. 2):43.
patients with at least 20% improvement in at least five 2. van de Laar MAFJ, van der Korst JK. Rheumatoid variables were 30.6% [11, 15] and 21.1% (the present arthritis, food, and allergy. Semin Arthritis Rheum study). The difference in these percentages was not significant (difference 9.5%, 95% CI −35 to 16%).
3. Sko¨ldstam L, Larsson L, Lindstro¨m FD. Effects of The discrepancy between the subjective experiences fasting and lactovegetarian diet on rheumatoid arthritis.
of the patients and the more ‘objective' measures of Scand J Rheumatol 1979;8:249–55.
basically the same phenomena (HAQ, duration of 4. Hafstro¨m I, Ringertz B, Gyllenhammar H, Palmblad J, morning stiffness, pain at rest and pain on movement) Harms-Ringdahl M. Effects of fasting in disease activity, deserves a mention. The patients seemed to link their neutrophil function, fatty acid composition, and leukotri- NENONEN ET AL.: VEGAN FOOD IN RHEUMATOID ARTHRITIS ene biosynthesis in patients with rheumatoid arthritis.
Cooperative Systematic Studies of Rheumatic Diseases Arthritis Rheum 1988;31:585–92.
Group. Analysis of improvement in individual rheumat- 5. Haugen M, Kjeldsen-Kragh J, Nordva˚g BY, Fo¨rre O oid arthritis patients treated with disease-modifying anti- Diet and disease symptoms in rheumatic disease— rheumatic drugs, based on the findings in patients treated Results of a questionnaire based survey. Clin Rheumatol with placebo. Arthritis Rheum 1990;33:477–84.
21. Scott DL, van Riel PL, van der Heijde D, Studnicka 6. Wigmore A. Recipes for longer life, 1st edn. Wayne, NJ: Bence A (eds) Assessing disease activity in rheumatoid Avery Publishing Group Inc., 1980.
arthritis. The EULAR handbook of standard methods, 7. Ha¨nninen O, Nenonen M, Ling WH, Li DS, Sihvonen 1st edn. London: EULAR, 1993.
L. Effects of eating an uncooked vegetable diet for one 22. Altman DG. How large a sample? In: Altman GA, Gore week. Appetite 1992;19:243–54.
SM, eds. Statistics in practice, 1st edn. London: British 8. Nenonen M. Vegan diet, rich in lactobacilli (‘Living Medical Association, 1989:6–8.
food'): metabolic and subjective responses in healthy 23. Juven BJ, Meinersmann RJ, Stern NJ. Antagonistic subjects and in patients with rheumatoid arthritis.
effects of lactobacilli and pediococci to control intestinal Dissertation, University of Kuopio, Finland, 1995.
colonization by human enteropathogens in live poultry.
9. Ryha¨nen E-L, Mantere-Alhonen S, Nenonen M, J Appl Bacteriol 1991;70:95–103.
Ha¨nninen O. Modification of faecal flora in rheumatoid 24. Lindgren SE, Dobrogosz WJ. Antagonistic activities of lactobacilli rich lactic acid bacteria in food and feed fermentations.
FEMS Microbiol Rev 1990;87:149–64.
10. Rauma A-L, Nenonen M, Helve T, Ha¨nninen O. Effect 25. Mehta AM, Patel KA, Dave PJ. Purification and proper- of a strict vegan diet on energy and nutrient intakes ties of the inhibitory protein isolated from Lactobacillus by Finnish rheumatoid patients. Eur J Clin Nutr acidophilus AC1. Microbios 1983;38:73–81.
26. Eerola E, Peltonen R, Nenonen M, Helve T, Ha¨nninen 11. Kjeldsen-Kragh J, Haugen M, Borchgrevink CF et al. O, Toivanen P. Intestinal flora and disease activity in Controlled trial of fasting and one-year vegetarian rheumatoid arthritis during vegan diet. Abstract submit- diet in rheumatoid arthritis [see comments]. Lancet ted for XIIIth European Congress of Rheumatology, Amsterdam, The Netherlands, 18–23 June 1995.
12. Kjeldsen-Kragh J. Dietary treatment of rheumatoid arth- 27. Rauma AL. Nutrition and biotransformation in strict ritis. Dissertation, Department of General Practice, vegans. Dissertation, University of Kuopio, Finland, University of Oslo, Norway, 1995, 29 pp.
1996, 114 pp. ( Kuopio University Publications D.
13. Severijnen AJ, Kool A, Swaak AJG, Hazenberg MP.
Medical Sciences 102).
Intestinal flora of patients with rheumatoid arthritis: 28. Benno Y, Endo K, Mizutani T, Namba Y, Komori T, Induction of chronic arthritis in rats by cell wall frag- Mitsuoka T. Comparison of fecal microflora of elderly ments from isolated Eubacterium aerofaciens strains. Br persons in rural and urban areas of Japan. Appl Environ J Rheumatol 1990;29:433–9.
14. Peltonen R, Ling WH, Ha¨nninen O, Eerola E. An 29. Davies GJ, Crowder M, Reid B, Dickerson JW. Bowel uncooked vegan diet shifts the profile of human fecal function measurements of individuals with different microflora: computerized analysis of direct stool sample eating patterns. Gut 1986;27:164–9.
gas-liquid chromatography profiles of bacterial cellular 30. de Jong G, van Noort WL, Feelders RA, de Jeu Jaspars fatty acids. Appl Environ Microbiol 1992;58:3660–6.
CM, van Eijk HG. Adaptation of transferrin protein 15. Peltonen R, Kjeldsen-Kragh J, Haugen M et al. Changes and glycan synthesis. Clin Chim Acta 1992;212:27–45.
of faecal flora in rheumatoid arthritis during fasting 31. Haugen MA, Ho¨yeraal HM, Larsen S, Gilboe I-M, and one-year vegetarian diet. Br J Rheumatol 1994; Trygg K. Nutrient intake and nutritional status in chil- dren with juvenile chronic arthritis. Scand J Rheumatol 16. Arnett FC, Edworthy SM, Bloch DA et al. The American Rheumatism Association 1987 revised criteria for the ˚ gren JJ, Ha¨nninen OO et al. Univalent classification of rheumatoid arthritis. Arthritis Rheum cation fluxes in human erythrocytes from individuals with low or normal sodium intake. J Cardiovasc Risk 17. WHO Workgroup. International classification of impair- ments, disabilities and handicaps. Geneva: WHO, 1980.
33. Kloppenburg M, Breedveld FC, Terweil JP, Mallee C, 18. Blom G. Statistical estimates and transformed beta vari- Dijkmans BAC. Minocycline in active rheumatoid arth- ables. New York: John Wiley & Sons, 1958.
ritis. A double-blind, placebo-controlled trial. Arthritis 19. SAS Institute Inc. The GLM procedure. In: SAS user's guide: statistics, Version 5, 1st edn. Cary, NC: SAS 34. Tilley BC, Alarcon GS, Heyse SP et al. Minocycline in Institute Inc., 1985:433–506.
rheumatoid arthritis. A 48-week, double-blind, placebo- 20. Paulus HE, Egger MJ, Ward JR, Williams HJ, controlled trial. Ann Intern Med 1995;122:81–9.

Source: http://www.frozenwheatgrass.co.uk/downloads/research/raw-food.pdf