Energy, macronutrients and laryngeal cancer risk

C. Bosetti1,+, C. La Vecchia1,2, R. Talamini3, E. Negri1, F. Levi4, J. Fryzek5, J. K. McLaughlin5, W. Garavello6 and S. Franceschi7

1 Istituto di Ricerche Farmacologiche Mario Negri, Milan; 2 Istituto di Statistica Medica e Biometria, Università degli Studi di Milano, Milan; 3 Servizio di Epidemiologia, Centro di Riferimento Oncologico, Aviano (PN), Italy; 4 Registre Vaudois des Tumeurs, Institut Universitaire de Médicine Sociale et Préventive, Lausanne, Switzerland; 5 International Epidemiology Institute, Rockville, MD, USA; 6 Ospedale San Gerardo, Università di Milano Bicocca, Monza, Italy; 7 International Agency for Research on Cancer, Lyon Cedex, France

Received 22 August 2002; revised 22 January 2003; accepted 19 February 2003


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background:

A role for diet in laryngeal carcinogenesis has been suggested, but only a few studies have examined the potential relationship with a wide variety of macronutrients.

Patients and methods:

A case–control study was conducted between 1992 and 2000 in Italy and Switzerland, including 527 incident cases of laryngeal cancer, and 1297 controls hospitalized for acute, non-neoplastic conditions. The subjects’ usual diet was investigated through a validated food frequency questionnaire, including 78 foods and beverages. Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using unconditional multiple logistic regression models.

Results:

Cases reported higher energy intake than controls. The continuous OR for 100 kcal/day was 1.16 (95% CI 1.12–1.21) for alcohol energy, and 1.02 (95% CI 1.01–1.04) for non-alcohol energy. A significantly increased risk of laryngeal cancer was observed for animal protein (continuous OR = 1.21, 95% CI 1.03–1.41), polyunsaturated fats other than linoleic and linolenic fatty acids (OR = 1.43, 95% CI 1.19–1.70), and cholesterol intake (OR = 1.43, 95% CI 1.19–1.71). Laryngeal cancer risk was slightly reduced with increasing vegetable protein (OR = 0.75, 95% CI 0.62–0.91), sugar (OR = 0.84, 95% CI 0.71–1.00) and monounsaturated fatty acid intake (OR = 0.83, 95% CI 0.70–0.99).

Conclusions:

Laryngeal cancer cases have a higher energy intake than control subjects, and report a higher intake of animal protein and cholesterol.

Key words: case–control study, diet, laryngeal cancer, nutrients, risk factors


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Although tobacco smoking and alcohol drinking are the two major determinants of laryngeal cancer risk [1, 2], a role for diet has also been suggested [3]. Fruit and vegetable consumption, and consequently antioxidant nutrients, have been associated with a reduced risk of laryngeal cancer [3, 4]. There is, however, limited evidence for other dietary correlates of laryngeal cancer [3], and only a few studies collected dietary histories detailed enough to examine the potential relationship of laryngeal cancer with a wide variety of macronutrients.

Among these, a case–control study from New York on 250 cases of laryngeal cancer and 250 population controls found an increased risk for total fat [odds ratio (OR) = 2.6], but not for protein intake, after controlling for energy intake [5]. A large, multicenter case–control study conducted in six regions of southwestern Europe, including 1147 male cases and 3057 population controls, reported no significant associations with proteins, carbohydrates or total fat, although a reduced risk was found for a high polyunsaturated/monounsaturated ratio (OR = 0.5) [6]. In a case–control study from Uruguay on cancers of the upper aerodigestive tract, an increased risk of laryngeal cancer was found in relation to total fat (OR = 1.7), protein (OR = 1.6), and cholesterol intake (OR = 1.5) [7]. Another case–control study from the same country, including 140 laryngeal cancer cases and 420 hospitalized patients, reported increased risk of laryngeal cancer for total fat intake (OR = 3.1), and various types of fatty acids [8].

In order to provide further insight into the association between energy, various macronutrients and fatty acids and laryngeal cancer risk, we analyzed the data of a large case–control study from Italy and Switzerland, where diet was measured with an extensive and validated food frequency questionnaire (FFQ) [9, 10].


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
A case–control study of cancer of the larynx was conducted between 1992 and 2000 in two Italian areas and in the Swiss Canton of Vaud [11].

Cases were 527 patients (478 men and 49 women, median age 61 years, range 30–79) admitted to major teaching and general hospitals in the areas under study with incident, histologically confirmed squamous-cell carcinoma of the larynx, diagnosed no longer than 1 year before the interview. Laryngeal cancer cases included 271 glottis, 117 supraglottis, four subglottis, five laryngeal cartilage, 19 overlapping lesion of larynx and 111 unspecified laryngeal cancers.

Controls were 1297 subjects (1052 men and 245 women, median age 61 years, range 31–79) admitted to the same hospitals as cases for a wide spectrum of acute, non-neoplastic conditions, not related to smoking, alcohol consumption and long-term modification of diet, and frequency-matched with cases by 5-year age groups, sex and area of residence. To compensate for the rarity of laryngeal cancer in women, a control-to-case ratio of about five was chosen for females, as opposed to two for males. Twenty-seven per cent of the controls were admitted for traumas, 22% for other orthopedic disorders, 29% for acute surgical conditions, and 22% for miscellaneous other illnesses, including eye, nose, ear, skin or dental disorders.

All subjects were interviewed during their hospital stay using a structured questionnaire, administered by ad hoc trained interviewers. Information was collected on sociodemographic characteristics, anthropometric measures, lifestyle habits, including tobacco smoking and alcohol drinking, a personal medical history, and family history of cancer in first-degree relatives.

The subjects’ diet during the 2 years before cancer diagnosis or hospital admission (for controls) was investigated through an interview-administered FFQ, proven to be satisfactorily valid [10] and reproducible [9]. This included 78 foods and beverages, as well as a range of recipes common to the study areas, grouped into seven sections: (i) bread and cereal dishes (first courses); (ii) meat and other main dishes (second courses); (iii) vegetables (side dishes); (iv) fruit; (v) sweets, desserts and soft drinks; (vi) milk, hot beverages and sweeteners; (vii) alcoholic beverages. Subjects reported the average weekly frequency of consumption of each dietary item; intakes less than once a week, but at least once a month, were coded as 0.5/week. Several questions aimed at assessing fat intake pattern were also included in the questionnaire, and used to derive quantitative estimates of intake of fats used for seasoning and cooking. A food composition database was used to calculate the composition of diet in terms of total energy and various nutrient intake [12]. As a main source of data, a large, unpublished database of approximately 3000 items was provided by the National Nutrition Institute of Rome. The composition of food items included in that database was checked according to standard methods, and values for nutrients not included in the database were derived from other sources. The composition/100 g of 1991 food items with reference to total energy intake and 90 nutrients was provided by this database.

Statistical analysis
In order to examine the effect of various macronutrients independent of total caloric intake, ‘calorie-adjusted’ nutrient intakes were derived using the residual method suggested by Willett and Stampfer [13]. These ‘calorie-adjusted’ macronutrients were then categorized into quintiles based on the distribution of controls, and corresponding ORs and 95% confidence intervals (CIs) were estimated using unconditional multiple logistic regression models [14]. All models included terms for age (5-year groups), sex, center, years of education (<7, 7–11, ≥12), body mass index (quintiles), tobacco smoking (never, ex-smoker, current smoker of <15, 15–24, ≥25 cigarettes/day), alcohol drinking (<14, 14–27, 28–55, ≥56 drinks/week, plus a dummy variable for ex-drinkers), and non-alcohol energy intake (quintiles). In alternative models, ‘calorie-adjusted’ macronutrients were entered as continuous variables, with a measurement unit equal to the difference between the upper cut point of the fourth quintile and that of the first one. Tests for trend were based on the likelihood ratio test between models with and without a linear term for each macronutrient. ORs according to a fully partitioned model were also computed in order to allow for mutual confounding effects of all major macronutrients [15, 16].


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Table 1 gives the distribution of 527 laryngeal cancer cases and 1297 controls according to age, sex, and selected covariates. By design, the proportion of women was higher in controls than in cases, and the age distribution was similar in cases and controls. Cases did not differ from controls in terms of education, but reported a significantly higher tobacco (OR = 19.83 for current versus never smokers) and alcohol consumption (OR = 5.91 for drinkers of eight drinks/day versus never drinkers), and tended to have a lower body mass index (OR = 1.69 for subjects in the lower as compared with the higher quintile). The difference in body mass index between cases and controls was not restricted to the time of cancer diagnosis, but was also present in young and middle ages (data not shown).


View this table:
[in this window]
[in a new window]
 
Table 1.  Distribution of 527 laryngeal cancer cases and 1297 controls according to selected variables: Italy and Switzerland 1992–2000
 
In Table 2 the comparison of cases and controls according to their energy intake is presented. Cases reported higher energy intake than controls (OR = 4.35, 95% CI 2.75–6.88 for the highest versus the lowest quintile of intake). Alcohol energy, which accounts for 17% of total energy in cases and 11% in controls, showed a U-shaped relationship with laryngeal cancer risk, with an increased risk only for the highest quintile of intake (OR = 2.10, 95% CI 1.46–3.02). Non-alcohol energy was still higher in cases than in controls (OR = 2.47, 95% CI 1.60–3.80), although the trend in risk was weaker than for total energy. The continuous ORs for 100 kcal/day were 1.16 (95% CI 1.12–1.21) for alcohol energy, and 1.02 (95% CI 1.01–1.04) for non-alcohol energy.


View this table:
[in this window]
[in a new window]
 
Table 2.  Odds ratiosa (OR) of laryngeal cancer and corresponding 95% confidence intervals (CI) according to the daily energy intake: Italy and Switzerland 1992–2000
 
Table 3 shows the mean intake of various macronutrients, fatty acids and cholesterol among controls, and the corresponding OR. A significantly increased risk of laryngeal cancer was observed with increasing protein (OR = 1.62, 95% CI 1.08–2.43 for the highest quintile as compared with the lowest one), animal protein (OR = 1.71, 95% CI 1.16–2.52), polyunsaturated fats other than linoleic and linolenic fatty acids (OR = 2.18, 95% CI 1.48–3.21) and cholesterol intake (OR = 1.76, 95% CI 1.19–2.62). Laryngeal cancer risk was inversely related to vegetable protein intake (OR = 0.51, 95% CI 0.33–0.77). Reductions in risk, although not significant, were also found with increasing sugar (OR = 0.71, 95% CI 0.48–1.05), monounsaturated fatty acid (OR = 0.68, 95% CI 0.45–1.02) and oleic acid intake (OR = 0.72, 95% CI 0.48–1.08). The inverse relationship with monounsaturated fatty acid and oleic acid intake was reduced after including vegetables in the regression model, which are likely to be consumed with monounsaturated fats. Intake of starch, total fat, saturated fat and polyunsaturated fatty acids were unrelated to laryngeal cancer risk.


View this table:
[in this window]
[in a new window]
 
Table 3.  Odds ratiosa (OR) of laryngeal cancer and corresponding 95% confidence intervals (CI) according to the daily intake of selected macronutrients, fatty acids and cholesterol: Italy and Switzerland 1992–2000
 
The relationships between macronutrients found to be significantly associated with laryngeal cancer risk were further examined in Table 4 in separate strata of various covariates of interest. Risk patterns were consistent across strata of sex, age, alcohol drinking and tobacco smoking, although the OR tended to be stronger in men, in heavy alcohol drinkers (≥28 drinks/week) and in smokers or ex-smokers for less than 20 years. Similar ORs were observed in the two major subsites of laryngeal cancer, i.e. cancers of the supraglottis including epiglottis, false cords and laryngeal ventricles, and of the glottis, including true vocal cord.


View this table:
[in this window]
[in a new window]
 
Table 4.  Odds ratiosa,b (OR) of laryngeal cancer and corresponding confidence intervals (CI) according to selected macronutrients, fatty acids and cholesterol, in strata of various covariates: Italy and Switzerland 1992–2000
 
A fully partitioned model including all major macronutrients as continuous variables produced results similar to individual macronutrients. The OR relative to 100 kcal/day was 1.30 (95% CI 1.04–1.62) for protein, 0.96 (95% CI 0.90–1.04) for sugars, 1.03 (95% CI 0.97–1.08) for starch, 0.97 (95% CI 0.85–1.04) for saturated fatty acids, 0.94 (95% CI 0.79–1.18) for monounsaturated fatty acids, and 0.93 (95% CI 0.79–1.10) for polyunsaturated fatty acids.


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
In this well-designed large study on laryngeal cancer, cases showed a significantly higher energy intake compared with controls. Consistent with the results of previous studies on laryngeal cancer [17, 18], as well as other upper digestive tract neoplasms [1921], we observed a tendency for cancer cases to report low body mass index, despite their high energy intake. A plausible explanation of this paradox is that the heavy alcohol intake, which characterizes these patients, could have influenced their diet, metabolism and energy balance. Alcohol, in fact, may reduce intake and bioavailability of a variety of essential nutrients [22, 23], and several other studies showed greater energy intake, but lower than expected body weight, in drinkers as compared to non-drinkers and in alcoholics [2426]. Tobacco smoking could also contribute with alcohol in reducing the weight of cancer cases [27, 28]. Cases may tend to report lower weight as a consequence of their weight loss after the onset and treatment for the disease [29], but this seems quite unlikely, since a tendency to leanness was often present many years before cancer diagnosis. A different recall of dietary intake by cases may also play a role, but it is also likely that cases increased their energy intake to maintain their body weight.

Although an increased energy intake was observed, no relevant associations with the major nutrients contributing to energy intake, i.e. carbohydrates and fats, were observed. Conversely, we found a slight protective effect of sugars on laryngeal cancer risk, which is not surprising, given that in this population they are derived mainly from fruit, found to be inversely related to laryngeal cancer risk [11].

An association between total fat and laryngeal cancer has been inconsistently reported in previous studies [4], although there are some indications of a more favorable effect of mono- and polyunsaturated fats than of saturated ones on laryngeal carcinogenesis [6]. Our study gives some support to this hypothesis, showing a reduced risk of laryngeal cancer with monounsaturated, mainly oleic acid, intake. This reduction was, however, moderate, and tended to disappear after taking into account vegetables, which are likely to be consumed with monounsaturated fats (from olive oil), although this may well represent an overadjustment. We also found an increased risk with the intake of polyunsaturated fats, other than linoleic and linolenic fatty acids. These, however, account for only about 3% of total polyunsaturated fats, and derive probably from mixed seed oils used mainly for frying.

The strongest association in our dataset was with protein intake from animal sources. A similarly increased risk in relation to protein intake has been reported in a study from Uruguay [7], although two other studies reported no relationship with protein intake [5, 6]. Only one study reported a protective effect for high protein intake: in that study, however, protein intake was measured as a score and high protein food intake was an indicator of a better/richer diet [30].

The significant association between animal protein intake and laryngeal cancer risk reflects the positive relationship found for red meat [11], which is one of the main sources of animal proteins in this population [31]. An association between red meat consumption and laryngeal cancer was also reported in a few other investigations [8, 32, 33]. However, the results on meat are inconclusive, and no significant relationships were found for other major protein foods, such as cheese, poultry or processed meat, in our or previous studies [3]. Thus it is difficult to interpret the association between animal protein intake and laryngeal cancer, and inferences on a possible causal effect must be made with caution.

As reported in another study [7], we found a significant association with cholesterol intake. Cholesterol, found mainly in meat and eggs, is, however, strongly correlated with protein intake, and, given our results for protein, it is not surprising that similar results were found for cholesterol.

The combination of the large study size and the collection of extensive dietary information using a validated FFQ render this study one of the best to date on laryngeal cancer and diet. Further strengths include the comparable catchment areas of study subjects, the almost complete participation for both cases and controls (>90%), the collection of data on other major risk factors for laryngeal cancer, and the use of an extensive FFQ, which allows for the assessment of a broad range of macronutrients. However, as in most case–control studies, some selection and recall biases are possible [14]. A recent cancer diagnosis may influence recall of diet for the cases, although awareness of dietary hypotheses in cancer etiology are still limited in the populations studied, and the comparability of dietary history between cases and controls are improved by interviewing subjects in the same hospital settings. Dietary habits of hospital controls could be different from those of the general population and could be associated with their reasons for hospitalization, but we paid great attention to select controls among diagnostic categories not related to diet modifications.

In conclusion, our study indicates that laryngeal cancer cases have a higher energy intake than control subjects. A positive association was found for protein from animal origin and cholesterol, whereas an inverse association was observed for sugar. No significant relationship was observed with total fat consumption, although there was an indication that monounsaturated fats—mainly oleic acid—could have a more favorable effect on laryngeal cancer than other types of fatty acids.


    Acknowledgements
 
The authors thank M. P. Bonifacino for editorial assistance. This work was conducted with contributions from the Italian Association for Cancer Research (AIRC), the Italian and Swiss Leagues Against Cancer and the Swiss Foundation for Research Against Cancer.


    Footnotes
 
+ Correspondence to: Dr C. Bosetti, Istituto di Ricerche Farmacologiche Mario Negri, Via Eritrea 62-20157 Milan, Italy. Tel: +39-02-39014526; Fax: +39-02-33200231; E-mail: bosetti{at}marionegri.it Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. International Agency for Research on Cancer. Tobacco smoking. IARC Monogr Eval Carcinog Risk Chem Hum 1986; 38: 35–394.[Medline]

2. International Agency for Research on Cancer. Alcohol drinking. IARC Monogr Eval Carcinog Risks Hum 1988; 44: 153–250.[Medline]

3. World Cancer Research Fund in association with the American Institute for Cancer Research. Food, Nutrition and the Prevention of Cancer: a Global Perspective. Washington, DC: World Cancer Research Fund 1997.

4. Riboli E, Kaaks R, Estève J. Nutrition and laryngeal cancer. Cancer Causes Control 1996; 7: 147–156.[ISI][Medline]

5. Freudenheim JL, Graham S, Byers TE et al. Diet, smoking, and alcohol in cancer of the larynx: a case–control study. Nutr Cancer 1992; 17: 33–45.[ISI][Medline]

6. Estève J, Riboli E, Péquignot G et al. Diet and cancers of the larynx and hypopharynx: the IARC multi-center study in southwestern Europe. Cancer Causes Control 1996; 7: 240–252.[ISI][Medline]

7. De Stefani E, Ronco A, Mendilaharsu M, Deneo-Pellegrini H. Diet and risk of cancer of the upper aerodigestive tract. II: Nutrients. Oral Oncol 1999; 35: 22–26.[CrossRef][ISI][Medline]

8. Oreggia F, De Stefani E, Boffetta P et al. Meat, fat and risk of laryngeal cancer: a case–control study in Uruguay. Oral Oncol 2001; 37: 141–145.[CrossRef][ISI][Medline]

9. Franceschi S, Barbone F, Negri E et al. Reproducibility of an Italian food frequency questionnaire for cancer studies. Results for specific nutrients. Ann Epidemiol 1995; 5: 69–75.[CrossRef][Medline]

10. Decarli A, Franceschi S, Ferraroni M et al. Validation of a food-frequency questionnaire to assess dietary intakes in cancer studies in Italy. Results for specific nutrients. Ann Epidemiol 1996; 6: 110–118.[CrossRef][ISI][Medline]

11. Bosetti C, La Vecchia C, Talamini R et al. Food groups and laryngeal cancer risk: a case–control study form Italy and Switzerland. Int J Cancer 2002; 100: 355–360.[CrossRef][ISI][Medline]

12. Salvini S, Gnagnarella P, Parpinel MT et al. The food composition database for an Italian food frequency questionnaire. J Food Compos Anal 1996; 9: 57–71.[CrossRef]

13. Willett WC, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol 1986; 124: 17–27.[Abstract]

14. Breslow NE, Day NE. Statistical Methods in Cancer Research, vol. I, The Analysis of Case–Control Studies; IARC Sci Publ 32. Lyon, France: IARC 1980.

15. Willett WC. Nutritional Epidemiology. New York, NY: Oxford University Press 1990.

16. Decarli A, Favero A, La Vecchia C et al. Macronutrients, energy intake, and breast cancer risk: implications from different models. Epidemiology 1997; 8: 425–428.[ISI][Medline]

17. Muscat JE, Wynder EL. Tobacco, alcohol, asbestos, and occupational risk factors for laryngeal cancer. Cancer 1992; 69: 2244–2251.[ISI][Medline]

18. D’Avanzo B, La Vecchia C, Talamini R, Franceschi S. Anthropometric measures and risk of cancers of the upper digestive and respiratory tract. Nutr Cancer 1996; 26: 219–227.[ISI][Medline]

19. Franceschi S, Levi F, Conti E et al. Energy intake and dietary pattern in cancer of the oral cavity and pharynx. Cancer Causes Control 1999; 10: 439–444.[CrossRef][ISI][Medline]

20. Franceschi S, Dal Maso L, Levi F et al. Leanness as early marker of cancer of the oral cavity and pharynx. Ann Oncol 2001; 12: 331–336.[Abstract]

21. Gallus S, La Vecchia C, Levi F et al. Leanness and squamous cell oesophageal cancer. Ann Oncol 2001; 12: 975–979.[Abstract]

22. Blot WJ. Alcohol and cancer. Cancer Res 1992; (Suppl 52): 2119s–2123s.

23. Blot WJ. Invited commentary: more evidence of increased risks of cancer among alcohol drinkers. Am J Epidemiol 1999; 150: 1138–1140.[ISI][Medline]

24. Jones BR, Barrett-Connor E, Criqui MH, Holdbrook MJ. A community study of calorie and nutrient intake in drinkers and nondrinkers of alcohol. Am J Clin Nutr 1982; 35: 135–139.[Abstract]

25. Hellerstedt WL, Jeffery RW, Murray DM. The association between alcohol intake and adiposity in the general population. Am J Epidemiol 1990; 132: 594–611.[ISI][Medline]

26. Colditz GA, Giovannucci E, Rimm EB et al. Alcohol intake in relation to diet and obesity in women and men. Am J Clin Nutr 1991; 54: 49–55.[Abstract]

27. Klesges RC, Meyers AW, Klesges LM, La Vasque ME. Smoking, body weight, and their effects on smoking behavior: a comprehensive review of the literature. Psychol Bull 1989; 106: 204–230.[CrossRef][ISI][Medline]

28. Perkins KA, Epstein LH, Marks BL et al. The effect of nicotine on energy expenditure during light physical activity. N Engl J Med 1989; 320: 898–903.[Abstract]

29. Collins MM, Wight RG, Partridge G. Nutritional consequences of radiotherapy in early laryngeal carcinoma. Ann R Coll Surg Engl 1999; 81: 376–381.[ISI][Medline]

30. Zatonski W, Becher H, Lissowska J, Wahrendorf J. Tobacco, alcohol, and diet in the etiology of laryngeal cancer: a population-based case–control study. Cancer Causes Control 1991; 2: 3–10.[CrossRef][ISI][Medline]

31. Favero A, Salvini S, Russo A et al. Sources of macro- and micronutrients in Italian women: results from a food frequency questionnaire for cancer studies. Eur J Cancer Prev 1997; 6: 277–287.[ISI][Medline]

32. Zheng W, Blot WJ, Shu X-O et al. Diet and other risk factors for laryngeal cancer in Shangai, China. Am J Epidemiol 1992; 136: 178–191.[Abstract]

33. De Stefani E, Deneo-Pellegrini H, Mendilaharsu M, Ronco A. Diet and risk of cancer of the upper aerodigestive tract. I: Foods. Oral Oncol 1999; 35: 17–21.[CrossRef][ISI][Medline]





This Article
Abstract
Full Text (PDF)
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Bosetti, C.
Articles by Franceschi, S.
PubMed
PubMed Citation
Articles by Bosetti, C.
Articles by Franceschi, S.