Impact of Antioxidant Therapy on Symptoms of Anxiety and Depression. A Randomized Controlled Trial in Patients with Peripheral

Section: ORIGINAL RESEARCH
The aim of this study was to determine the effect of an antioxidant preparation on the progression of peripheral arterial disease, and to identify any effects on mood. A randomized, placebo-controlled trial of antioxidant was performed in 120 men and women with intermittent claudication and an ankle brachial pressure index (ABPI)

Keywords: randomized controlled trial, atherosclerosis, antioxidants, diet, anxiety, depression.

INTRODUCTION
Antioxidants have an essential role in the body's defence systems. Many naturally occurring antioxidants are vitamins, particularly C, E and beta-carotene, but other agents including zinc, selenium and pyridoxine are also important. Evidence suggests that a key role of antioxidants is in the protection of low density lipoprotein cholesterol from oxidation, thus preventing atherogenesis [ 1]. This is supported by epidemiological studies which suggest that a high dietary intake of antioxidant vitamins is associated with a reduced risk of heart disease [ 2, 3]. Protection against peripheral arterial disease is also likely, as patients have lower levels of vitamin E in skeletal muscle [ 4], and reduced vitamin C levels in both serum and diet compared with controls [ 5, 6].

Antioxidants also protect polyunsaturated fatty acids from peroxidation which, in addition to influencing atherogenesis, has also been associated with depression [ 7-9] and with cognitive impairment in elderly men [ 10]. Low levels of other antioxidant compounds, pyridoxine and zinc, have also been associated with depression [ 11-13] and decreased appetite [ 11, 14, 15]. There is also evidence from coronary prevention trials that low serum cholesterol concentrations might be associated with increased suicide rates [ 16].

A randomized controlled trial was therefore conducted to determine primarily whether dietary supplementation with a combination of antioxidants would have a favourable effect on patients with peripheral medal disease, but also to identify any effects on mood status. Walking distance, the ankle brachial pressure index (ABPI), lipid levels, haemostatic factors, cardiovascular events and death were recorded, but the outcome measures reported in this paper are anxiety and depression.

METHODS Study Design
A randomized controlled trial of an antioxidant combination was conducted in patients with peripheral arterial disease. Subjects received either antioxidant or identical placebo capsules using a minimization technique in blocks of four, taking into account age, sex, smoking habits and regular aspirin consumption. The randomization procedure was performed by Scotia Pharmaceuticals, and the code number was allocated by telephone. The physician, patients and all those involved in performing clinical and biochemical measurements were blind to the allocation code. All patients gave written informed consent that they were willing to participate in the trial, and were provided with a detailed information sheet. Approval for the study was obtained from the local Ethics Committee.

To achieve 80% power at the 5% significance level, approximately 60 patients were required to detect a > 1 mmol 1(n-1) change in serum cholesterol levels. One hundred and twenty patients were entered into the trial in anticipation of a relatively high withdrawal rate. The trial forms part of a larger study examining the effects of antioxidants and essential fatty acids in intermittent claudication.

Study Population
Subjects with stable intermittent claudication were recruited from the Peripheral Vascular Clinic at the Royal Infirmary of Edinburgh. Males and females of any age were eligible for the study if shown to have claudication on the Edinburgh Claudication Questionnaire [ 17] and an ABPI of

Medication
Treatment with antioxidants or placebo was for a two-year period. Compliance with the drug treatment was monitored by drug returns, counted by the Pharmacy Department at the Royal Infirmary of Edinburgh, and by direct questioning of patients. All medication and placebos were manufactured by Scotia Pharmaceuticals.

Antioxidant capsules contained 3 mg beta-carotene, 100 mg ascorbic acid, 25 mg pyridoxine hydrochloride, 10 mg zinc sulphate, 10 mg nicotinamide and 1 mg sodium selenite, and an identical placebo contained 255 mg coconut oil. This combination of antioxidants was chosen to provide a range of different antioxidants, broadly in line with the national minimum guidelines. Patients were advised to take one capsule of antioxidant or placebo daily, and not to take any other vitamin supplements during the trial period, or to alter their normal diet.

Investigations
At recruitment, patients completed an interviewer-administered questionnaire including basic demographic data, medical and smoking histories and the Edinburgh Claudication Questionnaire [ 17]. At baseline and at the end of the trial, systolic and diastolic blood pressures were measured in both arms, and ankle systolic pressures were also measured in the dorsalis pedis and posterior tibial arteries of both legs, using a hand-held Doppler ultrasound probe (Oxford Instruments, UK). Patients performed a treadmill test at 1.5 mph at a 10% gradient, and walked until the onset of pain.

At recruitment and at the end of the trial, subjects completed the Hospital Anxiety and Depression Scale (HAD) [ 18]. This scale consists of 14 items, evenly distributed between the two constructs anxiety and depression. For each construct, a score below 8 is in the normal range, 8-10 is borderline and above 10 indicates a probable disorder of the relevant mood.

Samples of venous blood were taken without using a tourniquet at recruitment and after 24 months. Total cholesterol was analyzed on frozen plasma samples at the Department of Molecular and Cell Biology, University of Aberdeen. Cholesterol was measured using the Boehringer CHOD-PAD diagnostic kit, and Precinorm standard plasma (Boehringer) was included in each assay. Lipoprotein, rheological and haemostatic factors were also measured, results for which will be reported elsewhere [ 19].

Statistical Analysis
Data were entered on to a database (DBASE IV), and analysed on the University of Edinburgh mainframe computer using the SPSS-X SAS statistical packages. Throughout the analysis, a probability level of p

Scores for anxiety and depression were highly skewed, therefore the results were expressed as median and interquartile ranges. Baseline comparisons were performed using the Mann-Whitney test. The 2-year comparison between groups was determined using PROC GENMOD in SAS to perform analysis of covariance with the baseline value as covariate, to take into account baseline differences.

The ABPI was calculated for each limb by dividing the higher of the two ankle pressures by the higher of the two brachial systolic pressures; the lower ABPI for each patient was used in subsequent analyses. Walking distance was calculated by converting treadmill speed from miles per hour into metres per minute, and multiplied by time (in minutes) until onset of pain to give a distance in metres.

Analyses were performed on all patients, and also repeated for those who completed the trial only. Trends and statistical significances were similar in each case, therefore results for completers have not been presented in tabular form, but described in the text as appropriate.

RESULTS Patient Characteristics
One hundred and twenty patients were randomized into the trial, 55 to receive antioxidant and 65 to the placebo group. There were no statistically significant differences between the groups in age, sex, current smoking or aspirin therapy (Table 1). There was evidence of relatively severe lower limb disease in both treatment and control groups, with overall mean ABPIs of less than 0.7 and over 80% experiencing claudication on the level. In both groups, distances walked on the treadmill until onset of pain were less than 50 m, and plasma cholesterol levels were similar.

Thirty-three (60.0%) of those taking antioxidant and 42 (64.6%) of those on placebo completed the trial. Most subjects were withdrawn because of adverse events, but some subjects dropped out because they no longer wished to continue (10 subjects in the antioxidant group and seven of those on placebo). In those who completed the trial, compliance with the antioxidant preparation was good, with 91.5% taking the recommended dose.

Anxiety and Depression
Antioxidant and placebo groups had similar overall anxiety scores at baseline, which did not differ significantly throughout the trial (Table 2). In those who completed the trial, there was a non-significant trend of decreasing anxiety in the antioxidant group, with fewer having borderline or probable anxiety at the end of the 2 years, compared with an increase in those on placebo.

Depression scores at baseline also did not differ significantly, although the antioxidant group tended to be more depressed, having higher median scores and more borderline or probable depression (Table 2). After 2 years' treatment, however, the antioxidant group became significantly less depressed than those on placebo, representing a reversal of the baseline pattern (Fig. 1).

Lower Limb Disease and Lipid Levels
There was no significant correlation between anxiety or depression with either disease severity indicated by the ABPI, or with total cholesterol levels. The largest correlations were between anxiety and the ABPI in the antioxidant group, with a rank correlation coefficient of 0.19 at baseline and 0.18 at follow-up, but these did not reach statistical significance.

At the end of the trial there were no significant differences in the ABPI between the two groups, and although pain-free walking distance tended to improve during the trial, particularly in the placebo group, there was again no significant difference between the groups. Both active and placebo groups also showed slight increases in total cholesterol levels during the trial, which were not statistically significant. These results will be reported in detail elsewhere [ 19].

DISCUSSION
A randomized controlled trial was conducted to determine the effect of antioxidant supplementation on lower limb atherosclerosis and to identify any effects on mood. Patients in the active and placebo groups were similar at baseline in terms of risk factor levels and extent of co-existent disease, confirming that the randomization method was adequate. As expected in subjects with intermittent claudication, over 95% were either current or ex-smokers. There were slightly more subjects in the placebo than the antioxidant group, a result of the complex stratified randomization process.

The prevalence of depression at baseline was slightly greater in the antioxidant group, although this difference was not statistically significant. This pattern was reversed at the end of the trial, however, with those in the placebo group having significantly more depression. One possible explanation for this reversal would be the selective withdrawal of depressed subjects from the antioxidant group during the trial, but an analysis of depression in completers only showed that this was clearly not the case. Alternatively, the changes in depression may have been prompted by changes in underlying lower limb disease, but this is unlikely as both the ABPI and walking distance remained similar in antioxidant and placebo groups at the end of the trial. There was also no significant correlation between depression and the ABPI or cholesterol levels.

Differences in anxiety levels were less marked, with no significant differences between placebo and antioxidant groups either at baseline or after 2 years. The same trend was present, however, of less anxiety in the antioxidant group compared with an increase in the placebo group at the end of the trial, particularly in completers only. The prevalence of anxiety in both groups was relatively high, possibly because of disease-induced stress, or because of anxiety related to the hospital attendance itself.

It is interesting that both anxiety and depression increased in the placebo group during the trial, possibly because less attention was paid to diet during the study period in the anticipation that they were receiving vitamin supplements. Unfortunately no blood levels for any of the antioxidant agents were obtained, so it has not been possible to determine the relative status of this population at baseline.

It is unlikely that deficiencies in the HAD questionnaire were responsible for the observed changes in mood, as it is a well-established tool shown to be valid in a variety of settings [ 20-23]. It is a self-administered rating scale designed to minimize the influence of associated physical illness by excluding items such as loss of appetite which may also be symptomatic of physical illness. One of the questions does ask if the patient feels "slowed down", however, which is a physical symptom of intermittent claudication. This question was therefore analyzed individually, but there was no evidence that it was responsible for overall changes in depression scores.

The combined antioxidant preparation therefore appears to have had a real effect on mood, producing less depression and possibly anxiety independent of any change in lower limb disease. A possible mechanism for this might involve the role of antioxidants in preserving polyunsaturated fatty acids, as it has been suggested that the ratio of certain fatty acids has an important effect on depression [ 9]. Alternatively, it may not be simply an antioxidant effect, as some components of the preparation including zinc, pyridoxine and selenium act as essential co-factors for many different enzyme systems [ 24-26]. Pyridoxine is a coenzyme involved in the biosynthetic pathway of serotonin [ 24], and serotonin has been implicated in the control of appetite [ 20]. Both pyridoxine and zinc in relatively high doses are known to improve eating habits in patients with anorexia nervosa, although this is not necessarily associated with depression [ 11, 14, 15]. Low levels of selenium in the diet have also been associated with increased anxiety and depression [ 27].

CONCLUSIONS
Treatment with a combined antioxidant preparation significantly reduced depression and lessened anxiety compared with untreated controls. This effect was independent of underlying changes in disease status or other risk factor levels. The mechanism for this effect is unclear, but warrants further investigation.

ACKNOWLEDGEMENTS
The authors are grateful to Ms Ruth Jepson who provided valuable assistance in collecting data for this trial, and to Drs Roy Skinner and Beverley Mowat for the measurement of cholesterol concentrations at the University of Aberdeen. The trial was funded by Scotia Pharmaceuticals Ltd.

TABLE 1. Characteristics of antioxidant and placebo groups at baseline
Legend for Chart:

A -
B - Antioxidant group (n = 55)
C - Placebo group (n = 65)

A
B C

Sex (% male(n))
65.5 (36) 70.8 (46)

Age (mean years (SE)[a])
66.2 (1) 65.3 (1)

Smoking history (%(n))

Current
38.2 (21) 40 (26)

Ex-smoker
54.5 (30) 60 (39)

Non-smoker
7.3 (4) 0 (0)

Aspirin therapy (%(n))
41.8 (23) 44.6 (29)

Ankle brachial pressure index(SE)
0.67 (0) 0.68 (0)

Pain-free walking distance (mean m (95% CI)[b])
45.4 (35.5, 58.2) 48.5 (39.0, 60.3)

Plasma cholesterol (mean mg dl-1 (SE))
231.5 (5.2) 228.5 (6)

[a] SE = standard error.

[b] Geometric mean and transformed 95% confidence intervals.
Between group differences were all non-significant (p > 0.05).
TABLE 2. Anxiety and depression in antioxidant and placebo groups
Legend for Chart:

A -
B - Baseline All subjects Antioxidant (n = 52)
C - Baseline All subjects Placebo (n = 65)
D - Baseline Completers only Antioxidant (n = 29)
E - Baseline Completers only Placebo (n = 39)
F - Two years Antioxidant (n = 29)
G - Two years Placebo (n = 39)

A B C D
E F G

Anxiety

Score (median (IQR)[a]) 4.5(3-7) 5 (2-9) 5 (2-8)
5 (3-9) 5 (3-7) 6 (2-10)

Category (%(n))

Normal 76.9(40) 69.2(45) 72.4(21)
69.2(27) 82.8(24) 61.5(24)

Borderline/anxious 23.1(12) 30.8(20) 27.6(8)
30.8(12) 17.2(5) 38.5(15)

Depression

Score (median (IQR)[a]) 4 (3-5) 3 (2 - 6) 4 (3-6)
3 (2-5) 3 (2-5) 4 (2-6)[*]

Category (%(n))

Normal 88.5(46) 92.3(60) 86.2(25)
94.9(37) 93.1(27) 92.3(36)

Borderline/depressed 11.5(6) 7.7(5) 13.8(4)
5.1(2) 6.9(2) 7.7(3)

[a] Interquartile range.

[*] p < 0.05 between group comparison, determined by analysis of
covariance.
S: FIG. 1. Depression scores in the antioxidant group at (a) baseline and (b) after 2 years.

REFERENCES
[1] Anonymous (editorial). Oxidised low density lipoproteins and atherogenesis. Br Heart J 1993; 69: 476-8.

[2] Stampfer M J, Hennekens CH, Manson JE, et al. Vitamin E consumption and the risk of coronary disease in women. N Engl J Med 1993; 328: 1144-9.

[3] Rimm EB, Stampfer M J, Ascherio A, et al. Vitamin E consumption and the risk of coronary heart disease in men. N Engl J Med 1993; 328: 1450-6.

[4] Larsson H, Haeger K. Plasma and muscle tocopherol contents during vitamin E therapy in arterial disease. Pharmacol Clin 1968; 1: 72.

[5] MacRury SM, Muir M, Hume R. Seasonal and climatic variation in cholesterol and vitamin C: effect of vitamin C supplementation. Scot Med J 1992; 37: 49-52.

[6] Leng GC, Horrobin DF, Fowkes FGR, et al. Plasma essential fatty acids, cigarette smoking, and dietary antioxidants in peripheral arterial disease: a population-based case-control study. Arterioscler Thromb 1994; 14: 471-8.

[7] Hibbeln JR, Salem N. Dietary polyunsaturated fatty acids and depression: when cholesterol does not satisfy. Am J Clin Nutr 1995; 62: 1-9.

[8] Maes M, Smith R, Christophe A, et al. Fatty acid composition in major depression: decreased Omega3 fractions in cholesteryl esters and increased C20:40 Omega 6/C20:50 Omega 3 ratio in cholesteryl esters and phospholipids. J Affect Disord 1996; 38: 35-46.

[9] Adams PB, Lawson S, Sanigorski A, Sinclair AJ. Arachidonic acid to eicosapentaenoic acid ratio in blood correlates positively with clinical symptoms of depression. Lipids 1996; 31: S157-61.

[10] Kalmijn S, Feskens EJM, Launer LJ, Kromhout D. Polyunsaturated fatty acids, antioxidants, and cognitive function in very old men. Am J Epidemiol 1997; 145: 33-41.

[11] McLoughlin IJ, Hodge JS. Zinc in depressive disorder. Acta Psychiatr Scand 1990; 82: 451-3.

[12] Maes M, D'Haese PC, Scharpe S, et al., Hypozincaemia in depression. J Affect Disord 1994; 31: 135-40.

[13] Narang RL, Gupta KR, Narang APS, Singh R. Levels of copper and zinc in depression. Indian J Physiol Pharmacol 1991; 35: 272-4.

[14] Mira M, Stewart PM, Abraham SF. Vitamin and trace element status of women with disordered eating. Am J Clin Nutr 1989; 50: 940-4.

[15] McCain C, Stuart M, Kasarskis E, Humphries L. Zinc, appetite regulation and eating disorders. Prog Clin Biol Res 1993; 380: 47-64.

[16] Muldoon MF, Manuck SB, Matthews KA. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. BMJ 1990; 301: 309-14.

[17] Leng GC, Fowkes FGR. The Edinburgh Claudication Questionnaire: an improved version of the WHO/Rose questionnaire for use in epidemiological surveys. J Clin Epidemiol 1992; 45: 1101-9.

[18] Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983; 67: 361-70.

[19] Leng GC, Lee A J, Fowkes FGR, et al. Randomized controlled trial of antioxidants in intermittent claudication. Vasc Med 1997; 2: 279-85.

[20] Wurtman J J, Wurtman RJ. Impaired control of appetite for carbohydrates in some patients with eating disorders: treatment with pharmacologic agents. In: Perke KM, Ploog D, eds. The Psychobiology of Anorexia Nervosa. New York: Springer-Verlag, 1984: 12-21.

[21] Aylard PR, Gooding JH, McKenna PJ, Snaith RP. A validation study of three anxiety and depression scales. J Psychosom Res 1987; 31: 261-8.

[22] Jack TM, Walker VA, Morley SJ, et al. Depression, anxiety and chronic pain. Anaesthesia 1987; 42: 1235-6.

[23] Moorey S, Greer S, Watson M, et al. The factor structure and factor stability of the Hospital Anxiety and Depression Scale in patients with cancer. Br J Psychiatry 1991; 158: 255-9.

[24] Adams PW, Rose DP, Folkard J, et al. Effects of pyridoxine hydrochloride (vitamin B6) upon depression associated with oral contraception. Lancet 1973; 1: 899-904.

[25] Rayman MP. Dietary selenium: time to act. BMJ 1997; 314: 387-8.

[26] McClain C J, Kasarskis EJ, Allen JJ. Functional consequences of zinc deficiency. Prog Food Nutr Sci 1985; 9: 185-226.

[27] Benton D, Cook R. The impact of selenium supplementation on mood. Biol Psychiatry 1991; 29: 1092-8.

~~~~~~~~

By G. C. LENG, MD,( 1); A. J. LEE, PHD,( 1); F. G. R. FOWKES, FRCPE,( 1); I. J. DEARY, FRCPE,( 2) and D. HORROBIN

Adapted by MD,( 1) , PHD,( 1) , FRCPE,( 1) and FRCPE,( 2)

( 1)Wolfson Unit for Prevention of Peripheral Vascular Diseases, Department of Public Health Sciences, Edinburgh University, Teviot Place, Edinburgh EH8 9AG, UK; ( 2)Department of Psychology, Edinburgh University, George Square, Edinburgh EH8 9JZ, UK; ( 3)Scotia Pharmaceuticals Limited, Scotia House, Castle Business Park, Stirling FK9 4TZ, UK