Original Article

European Journal of Clinical Nutrition (2009) 63, 1008–1015; doi:10.1038/ejcn.2009.19; published online 8 April 2009

Long-term effects of increased dietary polyunsaturated fat from walnuts on metabolic parameters in type II diabetes

L C Tapsell1, M J Batterham1, G Teuss1, S-Y Tan1, S Dalton1, C J Quick1, L J Gillen1 and K E Charlton1

1Smart Foods Centre, School of Health Sciences, University of Wollongong, Wollongong, New South Wales, Australia

Correspondence: Professor LC Tapsell, Smart Foods Centre, School of Health Sciences, University of Wollongong, Wollongong, New South Wales 2522, Australia. E-mail: ltapsell@uow.edu.au

Received 11 August 2008; Revised 20 November 2008; Accepted 12 January 2009; Published online 8 April 2009.

Top

Abstract

Background/Objectives:

 

Most dietary interventions have metabolic effects in the short term, but long-term effects may require dietary fat changes to influence body composition and insulin action. This study assessed the effect of sustained high polyunsaturated fatty acids (PUFA) intake through walnut consumption on metabolic outcomes in type II diabetes.

Subjects/Methods:

 

Fifty overweight adults with non-insulin-treated diabetes (mean age 54±8.7 years) were randomized to receive low-fat dietary advice ±30g per day walnuts targeting weight maintenance (around 2000kcal, 30% fat) for 1 year. Differences between groups were assessed by changes in anthropometric values (body weight, body fat, visceral adipose tissue) and clinical indicators of diabetes over treatment time using the general linear model.

Results:

 

The walnut group consumed significantly more PUFA than the control (P=0.035), an outcome attributed to walnut consumption (contributing 67% dietary PUFA at 12 months). Most of the effects were seen in the first 3 months. Despite being on weight maintenance diets, both groups sustained a 1–2kg weight loss, with no difference between groups (P=0.680). Both groups showed improvements in all clinical parameters with significant time effects (P<0.004), bar triacylglycerol levels, but these were just above normal to begin with. The walnut group produced significantly greater reductions in fasting insulin levels (P=0.046), an effect seen largely in the first 3 months.

Conclusions:

 

Dietary fat can be manipulated with whole foods such as walnuts, producing reductions in fasting insulin levels. Long-term effects are also apparent but subject to fluctuations in dietary intake if not of the disease process.

Keywords:

dietary fat, fasting insulin, walnuts

Top

Introduction

Dietary intervention can produce early metabolic outcomes but long-term effects remain a challenge (Wing and Phelan, 2005; Pirozzo et al., 2008). There is little long-term data on the effects of manipulating dietary fat. Fat (lipid) molecules are potent gene regulators, with polyunsaturated fatty acids (PUFA) having an important role in energy and fat metabolism as well as regulating adipocyte formation (Clarke, 2001; Wang et al., 2002). Although energy-deficit diets result in weight loss, mechanisms associated with adipocytes may contribute to recidivism. This could be observed through changes in leptin, an adipocyte secreted hormones, which signals energy availability to the hypothalamic–pituitary axis in energy-deficient states, and is involved in energy expenditure and appetite regulation (Chan and Mantzoros, 2005). Insulin levels would be another ‘metabolic indicator’ given the effects on central nervous system (CNS) receptor pathways that influence energy homeostasis (Morton, 2007). It may be possible to observe metabolic effects of fat manipulation through changes in these parameters.

Dietary fat, however, is delivered by food, a composite of nutrients and other molecules, the significance of which is yet to be fully appreciated (Jacobs and Tapsell, 2007). Manipulating dietary fat means targeting foods and gaining compliance with the recommended diet. In a previous shorter-term trial (Tapsell et al., 2004) we observed that including 30g per day walnuts in a low-fat diet significantly raised dietary PUFA and produced a trend to reduced body fat in subjects with type II diabetes mellitus (T2DM). Assuming that the effects of diet on body composition take time, we hypothesized that over 12 months this sustained high PUFA intake would significantly affect body fat, regulatory hormone levels (insulin, leptin) and other indicators of metabolic change (glucose, lipids, energy expenditure).

Top

Materials and methods

This was a 1-year parallel randomized controlled trial beginning with three monthly visits for dietary advice and measurements, followed by three monthly follow-ups and regular newsletters to encourage adherence. A sample size of 30 subjects per group, calculated from a similar trial with between-group differences in change in percentage body fat of 1.64±2.26 (Tapsell et al., 2004), provided a power of 0.967, allowing for 10 dropouts per group. Recruiting advertisements were placed in the local media between January and June 2005. After passing a screening questionnaire subjects underwent a diet history assessment and received an accelerometer and physical activity diary. Inclusion criteria were 35–75 years; previously diagnosed with T2DM, not insulin-treated, body mass index (BMI) >25 and <32kg/m2; waist circumference >102 for men and >94cm for women (NHMRC, 2003); and generally well. Exclusion criteria were major illnesses, food allergies or inhibitory habits, illiteracy and/or inadequate English. The study was approved by the Human Research and Ethics Committee of the University of Wollongong and the South Eastern Sydney and Illawara Area Health Service and was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12607000600448).

Randomization was conducted using a computerized random number generator, by a researcher independent of the subject interface (MB). Subjects, but not dietitians, were blinded to the type of overall diet (a prepackaged 30g snack portion of walnuts was given to the walnut group unbeknown to the controls), and advised not to take fish oil supplements. Of the 90 volunteers 50 met the inclusion criteria (mean age 54±8.7 years, range 33–70 years), and 24 and 26 subjects were randomized into the control and the walnut groups, respectively (Figure 1).


As per our previous study (Tapsell et al., 2004), both groups were given low-fat isocaloric dietary advice, targeting around 2000kcal (adjusted to individual requirements for weight maintenance). The control diet comprised 30% E fat (10% E saturated fatty acids (SFA), 15% E monounsaturated fatty acids (MUFA); 5% E PUFA, P/S ratio of 0.5), 20% E protein and 50% E CHO. The walnut diet included 30g per day walnuts and provided 10% MUFA, 10% E PUFA, and a P/S ratio of 1.0. Individuals received advice on the number of serves of core foods from a qualified dietitian. Dietary modeling ensured the advice matched macronutrient targets. A validated diet history interview (Martin et al., 2003) and 3-day food records were used to assess diet. Dietary data were analyzed using the FoodWorks software system (version 3, 2002; Xyris Software, Brisbane, Australia). Nutrient intake data were assessed using the AUSNUT fatty acid database (version 6, 2002) and Australian Fatty Acids Rev 6 2002 (RMIT, Melbourne, Australia). Participants were advised to walk briskly for 30min three times weekly. Physical activity was assessed by questionnaire (Baecke et al., 1982) and data were obtained from triaxial accelerometers (RT3, version 1.2; Stayhealthy Inc., Monrovia, CA, USA) worn by subjects for a consecutive 3-day period.

Body weight and percentage body fat was measured in an upright position in minimal clothing and without shoes using scales with a bioelectrical impedance component (Tanita TBF-622), which compares reasonably well with dual X-ray absorptiometry as a reference technique (Batterham et al., 2002). An abdominal CT scan was taken at the fourth lumbar vertebra by a trained radiographer at Southcoast X-ray (Wollongong, Australia). Subcutaneous and visceral adipose tissue areas (SAT and VAT) were measured at each of the levels using SIENET Sky software that runs the Siemens CT scanner.

Resting energy expenditure was measured using the BodyGem handheld indirect calorimeter (model 200-0001-01; HealtheTech, Golden, CO, USA), validated against other techniques (Melanson et al., 2004; Compher et al., 2005; Rubenbauer et al., 2006). Fasted blood samples were drawn by trained professionals and sent to a quality-assured pathology laboratory (Southern IML Pathology, Wollongong, Australia). Insulin sensitivity was assessed using the homeostasis model assessment (HOMA) method: glucose (mmol per 100ml) × insulin (μU/ml)/22.5 (Matthews et al., 1985).

Erythrocyte fatty acid composition concentrations were analyzed by gas chromatography using a Shimadzu GC-17A. Identification of fatty acids was based on the retention time of authentic fatty acid methyl ester standards (Sigma-Aldrich, Castle Hill, New South Wales, Australia). Fatty acid data were sorted before analysis to arrive at a total figure for n-3 fatty acids (notably α-linolenic acid (ALA), which is found primarily in walnuts, and the long chain n-3 metabolites, eicosapentaenoic acid (EPA) and docosahexanoic acid (DHA)).

Plasma leptin was analyzed in batches by a single trained scientist following storage at −80°C, using the radioimmunoassay technique with I125 (HPLC-purified, specific activity 135μCi/μg) as the labeling isotope (Human Leptin RIA kit; Linco Research, St Charles, MO, USA). γ-Radiation emitted by the I125-labeled sample preparation was measured using the Wallac 1480 WIZARD (PerkinElmer Life Sciences Inc., Melbourne, Victoria, Australia) gamma counter. The standard curve based on the gamma counter output was generated using Prism version 4 (Macintosh) computer software by GraphPad Software Inc. (San Diego, CA, USA) Unknown sample concentrations in ng/ml were established using automated calculation of x-axis intercepts based on the standard curve equations.

Data were analyzed using SPSS (version 11.5.0, 2002; SPSS Inc., Chicago, IL, USA). The primary outcome measures of change in body weight, body fat and VAT were analyzed using the general linear model for repeated measures, and model assumptions were checked. The primary analysis was performed by examining the differences in the two study groups over the treatment time. Secondary analyses of differences in energy expenditure, energy intake, resting energy expenditure, leptin, insulin and glucose levels were performed using the general linear model for repeated measures with potential confounders (sex and body composition) included as covariates when appropriate. Correlations between reported dietary intake of fatty acids and erythrocyte cell membrane fatty acid concentrations were assessed using Spearman's correlation coefficient. Changes in erythrocyte fatty acids were assessed using repeated-measures analysis of variance (ANOVA) and the differences between baseline values and each measurement point were assessed by paired t-tests.

Top

Results

At baseline there were no significant between-group differences for the proportion of men and women, age, BMI or weight (mean=92.8±15.4kg, P=0.79). The walnut group had slightly higher body fat (P=0.04) and low-density lipoprotein (LDL) levels (P=0.04; Table 1). There were no significant differences between groups in reported physical activity, MedGem readings and dietary intake patterns (data not shown). The subjects consumed most of their energy from bread, meat-based dishes, dairy foods and foods later advised to avoid (biscuits, chips, pies, chocolate, ice-cream). Most of the fat in the diet came from meat-based dishes, spreads and oils, nuts and seeds, and those foods ‘later advised to avoid’. Nuts were already contributing around 10% fat in the diet for both groups (data not shown).


By 12 months both groups had reduced their energy intakes and increased the relative proportion of protein in their diets (time effects, P=0.008 for energy; P=0.035 for protein), but there were no differences in protein or energy intakes between groups (Table 2). There were no significant differences in alcohol intakes. From 3 months on, the walnut group reported almost double the amount of dietary PUFA than the control group (P=0.000). There was a concomitant significant group effect for changes in erythrocyte ALA levels (P=0.023) (Table 3). Post hoc analysis showed significant increases in erythrocyte ALA levels for the 0- to 3-month period only (P=0.000), with the walnut group showing a greater increase (1.36±0.97 to 3.24±1.38 vs 1.66±1.14 to 1.84±1.48% control). This difference remained throughout the study, but the value at 12 months (2.68±1.63%) was lower than that at 3 months. EPA levels also increased in both groups (Table 3). Although correlations between reported intakes and erythrocyte ALA levels were not significant, they were significant for EPA levels for the walnut group at all time points except 6 months and only for the control group at 6 months (Table 3).



At 12 months the nut food category contributed to 37% total dietary fat and 64% PUFA for the walnut group, compared to 11% total fat and 16% PUFA for the control group (data not shown).

Despite dietary advice targeting weight maintenance, both groups recorded sustained weight loss. The difference in mean weight between baseline and 12 months for the control group was 1.6kg, and for the walnut group was 2.34kg but there was no significant difference between groups (Table 4). The pattern of change however was different. The walnut group demonstrated a consistent drop in the first 6 months, which remained stable (and weight at 6 months was significantly different from baseline, P=0.028), but the control group began to regain weight at 6 months (Figure 2).

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Changes in body weight adjusted for baseline.

Full figure and legend (16K)


There were no significant interaction effects for changes in components of body fat distribution (and after adjustments for sex), but there were significant group effects for changes in VAT and SAT (P=0.006 and 0.005) (Table 4). The control group showed a tendency to lose body fat and VAT in the first 3 months (different from baseline, P=0.019 and 0.066) and gain SAT at 12 months (different from baseline, P=0.023), whereas the walnut group tended to lose SAT at 3 months (different from baseline, P=0.048) (Figure 3).

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Changes in percentage body fat adjusted for baseline.

Full figure and legend (15K)

There were no significant changes in resting energy expenditure adjusted for fat-free mass (Table 4). Leptin levels fluctuated in both groups (time effect, P=0.000), with both groups measuring an almost 50% drop by 12 months (Table 4).

All cholesterol measures fluctuated during the trial producing a significant time effect (P=0.000). For both the control and walnut groups, however, the high-density lipoprotein (HDL) levels were higher and LDL levels were lower at 12 months compared to baseline. The HDL level was significantly higher than baseline for the walnut group at 6, 9 and 12 months (P=0.055, 0.043 and 0.000, respectively), but only at 6 months for the controls (P=0.004). No significant effects were seen with triacylglycerol levels.

HBA1c levels also fluctuated in both groups producing a significant time effect (P=0.004), and lower 12-month readings compared to baseline. The levels at 6 months were significantly lower than baseline for the walnut group (P=0.023). The same pattern was observed for fasting glucose and insulin levels. However, an interaction effect was observed for the insulin values (P=0.046). This appears to be mostly accounted for with the significant drop from baseline to 3 months seen in the walnut group (P=0.002). The values remained significantly lower than baseline at 6 months (P=0.047) and 9 months (P=0.024). There was a significant difference between 9- and 12-month values for the control group (P=0.052). Post hoc analysis using the HOMA equation revealed a significant decrease at all time points compared with baseline for both groups (P=0.000; data not shown).

There were no major changes in physical activity (data not shown) or medications throughout the study at a group level. All but four subjects were on lipid-lowering and/or oral hypoglycemic agents throughout the study. Four subjects ceased the lipid-lowering medication and six subjects changed or slightly increased their dose of oral hypoglycemic agents after 6 months but these changes were distributed across groups and not considered statistically relevant.

Top

Discussion

This study confirmed that most of the effect of dietary intervention is seen in the first 3–6 months. Intervention itself produced favorable effects in both groups as participants improved their dietary patterns and inadvertently reduced energy intake. Differences between groups in type of fat consumed (P=0.000) were confirmed by measured increases in fatty acid biomarkers of intake. Levels of erythrocyte ALA (found in significant amounts in walnuts) increased and although correlations with reported intakes of ALA were not significant, they were significant for EPA, a metabolite of ALA (Cao et al., 2006). The biomarker data suggested compliance was best in the first 3 months and greater intensity of intervention in the latter stages might have improved the effect. The background diet for both groups was controlled, and the increased in EPA and DHA seen in both groups reflected the common advice to consume fish twice a week. Nevertheless, as the food with the greatest impact on dietary fat type, walnuts delivered three times the proportion of dietary fat and four times the proportion of PUFA in the walnut group (data not shown).

The intensity of treatment was the same for both groups but the weight loss effects appeared more sustainable in the walnut group where weight loss remained significantly different from baseline at 6 months (P=0.028). Changes in body fat were too difficult to discern between groups, and disappointingly, variability was quite large to begin with in this study. The drop in resting energy expenditure was expected under energy-deficit conditions (Chan et al., 2003) but no significant changes were noted. Likewise, the changed leptin concentrations with body weight and fat were consistent with the literature describing leptin as a mediator in the adaptation to starvation (Chan and Mantzoros, 2005).

The favorable changes in blood lipid levels were also expected given the improved diets, with higher PUFA intakes in the walnut group explaining the more favorable HDL results, as per our previous shorter-term study (Tapsell et al., 2004). The lack of change in triacylglycerol was unexpected but the baseline levels were not especially high. The significant interaction effect for insulin levels was likely due to the sharp decrease in the first 3 months in the walnut group that remained significant for the next 6 months. This was reflected in the significantly lower HbA1c levels observed for the walnut group at 6 months, confirming the early effects of higher PUFA intakes.

This study showed that dietary fat can be manipulated with foods such as walnuts, producing reductions in fasting insulin levels. Weakening compliance or a threshold dietary effect could explain the flattening out of long-term effects.

Attention to diet is a cornerstone of diabetes management (Fowler, 2007), but with food delivering multiple compounds to multiple sites, dietary effects may be subtle and occur at multiple levels. Although weight loss is readily achievable, it is less likely to be sustained in the long term (Wing and Phelan, 2005) so the effect of diet on body composition may be of interest. Animal model research suggests that dietary PUFA may protect against obesity through mechanisms linked to fat deposition and hormonal regulatory systems (Wang et al., 2002). Human studies show that energy deficits alone do not change body fat distribution (Redman et al., 2007), but this is not the case with fat modification (Summers et al., 2002; Panigiagua et al., 2007). Dietary fat can influence the expression of genes associated with body weight regulatory systems (Summers et al., 2002; Wang et al., 2002). The effect of increased dietary PUFA on insulin action is another possible mode of action where hepatic fatty acid composition may influence insulin sensitivity (independently of cellular energy balance) (Matsuzaka et al., 2007), and insulin itself may influence hypothalamic control of energy homeostasis through CNS pathways.

In conclusion, this long-term study found that in T2DM, increasing dietary PUFA with walnuts in a low-fat diet reduced fasting insulin levels at least in the early stages of the intervention. Over 12 months, the pattern of dietary intake was aligned to patterns of body weight and body composition, which were reflected in clinical (glucose homeostasis and lipids) and metabolic (resting energy expenditure and leptin) variables that have implications for the long-term management of the condition. Although the focus was on PUFA, the intervention was food based and this subtle feature should not be underestimated. Whole foods deliver a matrix of molecules reflecting the original biological context of the nutrients (Jacobs and Tapsell, 2007). There are likely to be additional reasons why nuts, for example, have a place in weight management (Sabate, 2003). Likewise it is the total diet that has the effect, and here dietary modeling is important to control for confounding variables to match nutrient targets linked to mechanistic understandings (Gillen and Tapsell, 2005). This study exemplifies how food may deliver multiple compounds to multiple sites where the effects are subtle. In this case, walnuts in a healthy background diet delivered PUFA intakes that produced significant effects at least in the early stages of dietary change.

Top

References

  1. Baecke JA, Burema J, Frijters JE (1982). A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutr 36, 936–942. | PubMed | ISI | ChemPort |
  2. Batterham MJ, Tapsell LC, Jenkins AB (2002). A comparison of bioelectrical impedance and near infra-red interactance with dual energy X-ray absorptiometry for the determination of body fat. Nutr Diet 59, 120–126.
  3. Cao J, Schwichtenberg KA, Hanson NQ, Tsai MY (2006). Incorporation and clearance of omega-3 fatty acid in erythrocyte membranes and plasma phospholipids. Clin Chem 52, 2265–2272. | Article | PubMed | ChemPort |
  4. Chan JL, Heist K, De Paoli AM, Veldhuis D, Mantzoros CS (2003). The role of falling leptin levels in the neuroendocrine and metabolic adaptation to short-term starvation in healthy men. J Clin Invest 111, 1409–1421. | Article | PubMed | ChemPort |
  5. Chan Jl, Mantzoros CS (2005). Role of leptin in energy deprivation states: normal human physiology and clinical implications for hypothalamic amenorrhoea and anorexia nervosa. Lancet 366, 74–85. | Article | PubMed | ChemPort |
  6. Clarke S (2001). Polyunsaturated fatty acid regulation of gene transcription: a molecular mechanism to improve the metabolic syndrome. J Nutr 131, 1129–1132. | PubMed | ISI | ChemPort |
  7. Compher C, Hise M, Sternberg A, Kinosian BP (2005). Comparison between Medgem and Deltatrac resting metabolic rate measurements. Eur J Clin Nutr 59, 1136–1141. | Article | PubMed | ChemPort |
  8. Fowler MJ (2007). Diabetes treatment: part 1: diet and exercise. Clinical Diabetes 25, 105–110. | Article
  9. Gillen LS, Tapsell LC (2005). Structured dietary advice incorporating walnuts achieves optimal fat and energy balance in patients with type 2 diabetes mellitus. J Am Diet Assoc 105, 1087–1096. | Article | PubMed
  10. Jacobs DR, Tapsell LC (2007). Food the fundamental unit in nutrition. Nutr Rev 65, 439–450. | Article | PubMed
  11. Martin GS, Tapsell LC, Denmeade S, Batterham MJ (2003). Relative validity of a diet history interview in an intervention trial manipulating dietary fat in the management of type 2 diabetes mellitus. Prev Med 36, 420–428. | Article | PubMed
  12. Matsuzaka T, Shimano H, Yagahi N, Kato T, Atsumi A, Yamamoto T et al. (2007). Crucial role of a long chain fatty acid elongase, Elov16, in obesity-induced insulin resistance. Nat Med 13, 1193–1202. | Article | PubMed | ChemPort |
  13. Matthews D, Hosker J, Rudenski A, Naylor B, Treacher D, Turner R (1985). Homeostasis model assessment: insulin resistance and B-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28, 4412–4419. | Article
  14. Melanson EL, Coelho LB, Tran ZV, Haugen HA, Kearney JT, Hill JO (2004). Validation of the BodyGem hand-held calorimeter. Int J Obes Relat Metab Disord 28, 1479–1484. | Article | PubMed | ChemPort |
  15. Morton GJ (2007). Hypothalamic leptin regulation of energy homeostasis and glucose metabolism. J Physiol 583, 437–443. | Article | PubMed | ChemPort |
  16. National Health and Medical Research Council (2003). Clinical Guidelines for Weight Control and Obesity Management in Adults. Commonwealth of Australia: Canberra.
  17. Panigiagua J, Gallego de la Sacristana A, Romero I, Vidal-Puig A, Later J, Sanchez J et al. (2007). MUFA-rich diet prevents central fat distribution and decreases post-prandial adiponectin expression induced by a carbohydrate rich diet in insulin resistant subjects. Diabetes Care 30, 1717–1723. | Article | PubMed | ChemPort |
  18. Pirozzo S, Summerbell C, Cameron C, Glasziou P (2008). Advice on low-fat diets for obesity. Cochrane Database Syst Rev 2 Accession 12076496.
  19. Redman LM, Heilbronn LK, Martin CK, Alfonso A, Smith SR, Ravussin E (2007). Effect of calorie restriction with or without exercise on body composition and fat distribution. J Clin Endocrin Metab 92, 865–872. | Article | ChemPort |
  20. Rubenbauer JR, Johannsen DL, Baier SM, Litchfield R, Flakoll PJ (2006). The use of a handheld calorimetry unit to estimate energy expenditure during different physiological conditions. J Parenter Enteral Nutr 30, 246–250. | Article
  21. Sabate J (2003). Nut consumption and body weight. Am J Clin Nutr 78, 647S–650S. | PubMed | ChemPort |
  22. Summers LK, Fielding BA, Bradshaw HA, Ilic V, Beysen C, Clark ML et al. (2002). Substituting dietary saturated fat with polyunsaturated fat changes abdominal fat distribution and improves insulin sensitivity. Diabetologia 45, 369–377. | Article | PubMed | ISI | ChemPort |
  23. Tapsell LC, Gillen LJ, Patch CS, Batterham MJ, Owen A, Bare M et al. (2004). Including walnuts in a low fat/modified fat diet improves HDL: total-C in patients with type 2 diabetes mellitus. Diabetes Care 27, 2777–2783. | Article | PubMed
  24. Wang H, Storlien LH, Huang X-F (2002). Effects of dietary fat types on body fatness, leptin, and ARC leptin receptor, NPY, and AgRP mRNA expression. Am J Physiol Endocrinol Metab 282, E1352–E1359. | PubMed | ISI | ChemPort |
  25. Wing RR, Phelan S (2005). Long-term weight loss maintenance. Am J Clin Nutr 82 (Suppl), 222S–225S. | PubMed | ISI | ChemPort |
Top

Acknowledgements

Funding for this research was provided by the California Walnut Commission.

Extra navigation

.

natureevents

ADVERTISEMENT