729

Plasma Lipids and Fatty Acids in Urbanized Bushmen, Hereros and Kavangos of Southern Africa (Namibia) H.Y. Tichelaar a,*, AoJ.S. Benad~, S.J.D. O'Keefe b, P.L. Jooste a, S.A. Swanevelder c and E. Van Staden a aNational Research Programme for Nutritionat tntervention, Tygerberg, 7505, Republic of South Africa, bDepartment of Gastroenterology and Internal Medicine, Mayo Clinic, Rochester, Minnesota 55905 and Clnstitute for Biostatistics of the Medical Research Council, Tygerberg, 7505, Republic of South Africa

Thirty-nine urbaniT.ed ethnic Namibian people comprising 21 Bushmen (semi-urbanlzed), 7 Hereros and 11 Kavangos were assessed for plasma lipids and fatty acid (FA) composition. ~btal cholesterol and triacylglycerol concentrations were measured by enzymatic methods, and neutral lipid FA composition by gas-liquid chromatography. The results demonstrated that while total cholesterol concentrations were not significantly different, significant differences in triacylglycerol concentrations (P < 0.05) were seen between Bushmen and Kavangos. By comparing Bushmen with Hereros and Kavangos, significant differences between Bushmen and Kavangos were also observed in plasma triacylglycerol FA compositions, pa~ ticularly 16~ {32.73% v& 25.05%), 16:1n-7 (7.00% v& 5.06%), 18:2n4} (9.30% v s . 22.25%) and 20:3n4~ (0.12% v& 0.48%), while Kavangos had higher 20:4n4} levels than Hereros (1.44% v& 2.00%). In plasma cholesteryl esters, Bushmen were significantly different from Kavangos in 16:1n-7 (8.85% v& 4.93%), 18:1n-9 (32.06% v& 23.07%) and 20:4n4~ (6.91% v& 10.00%). Significant differences were also o1~ served between Bushmen and Hereros in 18:0 (1.08% v& 1.29%) and 18:2n~ (35.68% v& 45.50%). The FA of Namibian groups were also compared with South African reference groups comprising urbanized whites and Xhosas and rural Vendas. The differences in blood lipid values can be explained primarily by excessive alcohol consumption. These results suggest that semi-urbanized Bushmen have changed their diets Ullde~ urbanized conditions which may increase their risk of coronary heart disease. Lipids 27, 729-732 (1992).

Western diet with a high fat, low polyunsaturated to sat~ urated fatty acids (I~S) ratio and low carbohydrate intak~ The urbanized Xhosas are rapidly adopting a Western life style which increases the risk of developing CHD. Rural Vendas consume a traditional diet of maize meal supplemented with mopanie worms.

MATERIALS AND METHODS

A total of 39 urbanized ethnic Namibian people, in the Ojinene district of east Hereroland, comprising 21 Bushmen, 7 Hereros and 11 Kavangos with average ages of 29, 33 and 41, respectively, were assessed for plasma lipids and FA compositions. The Bushmen studied were no longer purely nomadic; most of them live for variable times in association with Herero villages, but tend to migrate from one place to another depending upon work and food availability. They are accordingly very much "in transitiorL" Anthropometric measurements and dietary intake on the same study individuals have been reported previously by O'Keefe and Lavender (9). Approval to conduct a study on these ethnic groups was originally obtained from the local administrator or village headman. After an interview, informed consent was obtained from the subjects before blood samples were taken with the assistance of local district nurses. All participants were adults older than 18 years. All blood samples were taken in the early morning, in either fasting or semi-fasting (hght maize meal and water) conditions. Blood samples were taken in heparinized containers, and separated plasma samples were deepfrozen Very little information is available on the plasma lipids and until analysis could be carried out, which was always fatty acids (FA) in the different ethnic groups of Namibia" within two weeks. Total cholesterol (TC) and triacylglycerol especially on Bushmen {1,2). Various studies have demon- (TAG) concentrations were determined enzymatically strated that the plasma FA compositions, especially those (Boehringer Mannheim GmbH Diagnostica, Mannheim, of plasma cholesteryl esters (CE), reflect the dietary FA in- Germany; CHOD-PAP and GPO-PAP kits, respectively). take (3-6). Urbanization increases the risk of all groups to Neutral lipids were separated by thin-layer chromatogdevelop coronary heart disease (CHD), presumably due to raphy (TLC) and analyzed by gas-liquid chromatography changes in their dietary food intake (5-8), eg., increased (GLC) as described previously (10). The significance of alcohol consumption and a decreased dietary intake of vege these results was evaluated using the nonparametric table fat. An increased mortality rate from CHD may en- Kruskal-Wallis test to compare group means statistically at the 5% level. Although mean values were used, medanger the existence of Bushmen in modern society (1). We performed this study to examine the lipids and FA dian values may be a better estimate of the true location in semi-urbanized Bushmen per se, with comparison to of the mean. Robust standard deviations (SD) were used Hereros and Kavangos of the same area, and whit~ Xhosa as an improvement on the usual SD calculated for the and Venda South African reference groups in order to in- subgroup specifically, due to the small sample size of the vestigate the extent of these dietary changes White South Herero and Kavango groups. FA results are compared with Africans with a high incidence of C H D followed a typical urbanized whites and blacks (Xhosas), and rural blacks (Vendas) in South Africa (unpublished data). *To whom correspondenceshouldbe addressed at the National Re search Programme for Nutritional Intervention, P.O. Box 19070, RESULTS Tygerberg, 7505, Republic of South Africa~ Abbreviations: AA, arachidonicacid;CE, cholesterylesters;CHD, The findings of the trial are summarized in Tables 1, 2 coronary heart disease;DGLA, dihom~gammalinolenic acid;EPA, eicosapentaenoicacid;FA, fattyacids;GLC, gas-liquidchromatog- and 3. It should be noted that Table 1 includes data for raphy; LA, linoleicacid;OA, oleicacid;P/S,polyunsaturatedto sat- only 13 Bushmen, 4 Hereros and 10 Kavangos, compared urated fatty acid ratio;TAG, triacylglycerol;TC, totalcholesterol; with the data of 21 Bushmen, 7 Hereros and 11 Kevangos TLC, thin-layerchromatography. in Tables 2 and 3. This was due to insufficient sample size LIPIDS, VoI. 27, no. 9(1992)

730 H.Y. T I C H E L A A R E T AL. TABLE 1 P l a s m a Total Cholesterol and Triacylglycerol Concentrations in Urbanized E t h n i c Groups (in mmol/L) a Reference groups

Statistical

SAU

SAU

SAR

Whites

Xhosas

Vendas

Bushmen

Hereros

Kavangos

Lipid

(n = 13) mean (SD)

(n = 4) mean (SD*)

{n = 10) mean (SD*)

difference P < 0.05

(n = 50) mean (SD)

(n = 50) mean (SD)

(n = 61) mean (SD)

TC

3.71 (1.03)

4.21 (1.18)

3.61 (1.18)

KB H

6.24 (1.14)

4.08 (0.89)

4.23 (0.84)

TAG

2.01 (1.18)

1.35 (0.79)

1.06 (0.79)

K H B

2.01 (1.21)

1.19 (0.75)

1.26 (0.72)

aSD, s t a n d a r d deviation; SD*, robust s t a n d a r d deviation; SAU, South African urbanized; SAR, South African rural; non-significance is indicated by a connecting bar.

TABLE 2 P l a s m a Triacylglycerol Percentage F a t t y Acid Composition (weight percent) in Urbanized Ethnic Groups a Reference groups Statistical difference P < 0.05

SAU Whites (n -- 50) mean (SD)

SAU Xhosas (n = 50) mean (SD)

SAR Vendas (n = 61) mean (SD)

Fatty acid

Bushmen (n = 21) mean (SD)

Hereros (n = 7) mean (SD*)

Kavangos (n = 11) mean (SD*)

16:0

32.73 (5.23)

30.33 (6.06)

25.05 (6.06)

K H B

23.19 (3.69)

25.07 (4.35)

29.93 (5.50)

16:1

7.00 (2.19)

5.41 (2.36)

5.06 (2.36)

K H B

4.67 (1.21)

4.22 (1.65

5.70 (1.62)

18:0

5.94 (1.98)

5.85 (1.76)

3.68 (1.76)

K H B

5.03 (1.27)

6.71 (1.72)

3.82 (1.36)

18:1

43.38 (4.73)

47.60 (4.77)

41.34 (4.77}

K B H

39.02 (4.36)

39.63 (4.60)

40.96 (4.94) 17.79 (6.84)

18:2

9.30 (6.56)

9.37 (10.18)

22.25 (10.18)

B H K

24.43 (6.07)

20.70 (6.04)

20:3

0.12 (0.18)

0.18 (0.28)

0.48 (0.28)

B H K

0.48 (0.41)

0.42 (0.16)

20:4

1.44 (0.92)

1.07 (0.81)

2.00 (0.81)

H B K

2.38 (1.40)

2.59 (0.95)

1.72 (0.64)

20:5

0.07 (0.12)

0.19 (0.17)

0.14 (0.17)

B K H

0.81 (0.78)

0.66 (0.58)

0.64 (0.15)

ND

aSD, s t a n d a r d deviation; SD*, robust s t a n d a r d deviation; SAU, South African urbanized; SAR, South African rural; ND, not detected; non-significance is indicated by a connecting bar.

TABLE 3 P l a s m a Cholesteryl Ester Percentage Fatty Acid Composition (weight percent) in Urbanized E t h n i c Groups a Reference groups

Fatty acid

Bushmen (n = 21) mean (SD)

Hereros (n = 7) mean (SD*)

Kavangos (n = II) mean (SD*)

Statistical difference P < 0.05

SAU Whites (n = 50) m e a n (SD)

SAU Xhosas (n -- 50) mean (SD)

SAR Vendas (n = 61) m e a n (SD)

16:0

14.15 (4.42)

11.86 (2.54)

14.84 (2.54)

H B K

10.79 (1.42)

11.13 (1.62)

12.03 (1.78)

16:1

8.85 (3.93)

4.74 (3.40)

4.93 (3.40)

K H B

3.54 (1.26)

3.24 (1.80)

4.92 (2.19)

18:0

1.08 (0.67)

1.29 (0.37)

1.14 (0.37)

B K H

1.35 (0.32)

1.64 (0.50)

0.53 (0.32)

18:1

32.06 (6.88)

27.45 (7.35)

23.07 (7.35)

K H B

19.12 (2.22)

24.01 (4.79)

26.81 (4.02) 46.92 (5.11)

18:2

35.68 (13.08)

45.08 (11.77)

45.50 (11.77)

B K H

56.14 (5.11)

50.54 (6.75)

20:3

0.65 (0.64)

0.52 (0.49)

0.21 (0.49)

K H B

0.50 (0.29)

0.65 (0.22)

20:4

6.91 (2.99)

8.48 (2.63)

I0.00 (2.63)

B I~ K

6.63 (1.36)

7.68 (1.86)

8.63 (2.66)

20:5

0.62 (0.54)

0.57 (0.38)

0.31 (0.38)

K H B

1.92 (1.13)

1.11 (0.58)

0.17 (0.27}

ND

aSD,~standard deviation; SD*, robust s t a n d a r d deviation; SAU, South African urbanized; SAR, South African rural: ND, not detected; non-significance is indicated by a connecting bar.

LIPIDS, Vol. 27, no. 9 (1992)

731

LIPIDS AND FATTY ACIDS IN URBANIZED BUSHMEN to analyze the lipids of all the 39 participants, as priority was given to fatty acid analyses. Statistically significant differences were found for TAG (P < 0.05) between Bushmen and Kavangos. Interestingly, Bushmen had the same TAG levels as the South African reference whites (Table 1). In neutral lipids the most important differences can be seen in plasma TAG percentage {TAG%) linoleic acid levels (LA; 18:2n-6}, which were significantly higher in Kavangos compared to Bushmen and Hereros (P < 0.05} with a concomitant reduction of stearic acid (18:0). Bushmen had significantly higher (P < 0.05) palmitic acid (16:0) and palmitoleic acid (16: ln-7) and lower dihomo-gammalinolenic acid (DGLA; 20:3n-6} levels than Kavangos, while Kavangos also had significantly higher arachidonic acid (AA; 20:4n-6) levels than Hereros. Compared to the South African reference groups comprising urbanized whites and blacks (Xhosas) and rural blacks (Vendas), Vendas appear to have plasma TAG% palmitic acid levels similar to those of Bushmen and Hereru~ while urban Xhosas and Kavangos have similar TAG% FA levels. Whites, however, have the highest LA levels of all the groups, with the Namibian groups all having higher oleic acid {OA. 18:1n-9) levels than any seen in the reference groups. Only Vendas have AA levels which are less than those of the Kavangos. In contrast, the reference groups have much higher eicosapentaenoic acid (EPA} levels than seen in the Namibian groups (Table 2). Bushmen had significantly higher (P < 0.05) palmitoleic acid and OA levels and lower AA levels than Kavangos in plasma percentage cholesteryl esters {CE%), where individual FA were expressed as percentages of the total area of all FA peaks Bushmen also had significantly lower (P < 0.05) stearic acid and LA levels than Hereros. Compared to the South African reference groups, Xhosas and Vendas have CE% composition palmitic acid levels which are similar to that of the Hereros, while whites seem to have the lowest levels of all groups. Palmitoleic acids are even lower in whites and Xhosas when compared with the levels of Hereros and Kavangos. Xhosas also have the highest stearic acid levels while Vendas have the lowest of all groups. Whites have the lowest OA levels of all groups, while Xhosas and Vendas have OA levels comparable to those of the Kavangos and Hereros, respectively. Both whites and Xhosas have higher LA levels than seen in all the Namibian groups, with whites having the highest. Although whites and Bushmen have similar AA levels, whites and Xhosas seem to have generally higher AA levels than seen in the other Narnibian groups (Table 3). It should be noted that the reference whites also have the highest eicosapentaenoic acid (EPA; 20:5n-3)/AA ratio (0.29}, compared to Xhosas (0.14), Bushmen (0.09), Hereros (0.07} and Kavangos (0.03}. Bushmen also have a lower P/S ratio in their CE% FA than whites, 5.57 vs. 7.23, respectively {Table 3). Unfortunately, no phospholipid levels were measured to support the CE% FA; this could have provided a more specific interpretation of the habitual dietary intake of each group. DISCUSSION

Bushmen, the original in_habitants of southern Africa, are scattered over large areas of northeast Namibia Their diet varies dramatically and depends primarily on availability, with the only consistent component being maize meal.

Home brewed beer consumption (2-4 liters per day) provides a daily energy intake from alcohol of 740-1480 kcal and contributes 156-312 grams of carbohydrate. Little or no vegetables and fruit are eaten. The Kavangos live in the far north and rely mainly on agriculture. Their staple diet consists of "African millet" (mahongo), a whole grain product, supplemented by fish from the river. Their diet is well balanced and reasonably constant. Hereros, the major black group, occupy the central areas of the territory. They mainly consume meat and dairy products with a staple diet of refined maize meal, 2-3 times daily. Hereros tend to consume more vegetables, fruits and fresh milk during the short rainy season. An interesting observation was made when comparing the data in Table 1. Bushmen had substantially higher mean TAG concentrations than Kavangos. While TAG concentrations for Kavangos and Hereros were similar to previous observations in rural groups (1,2), the values for Bushmen were nearly doubled. This could be ascribed, possibly, to excessive alcohol consumption which causes TAG accumulation in the liver (11). Whilst total caloric and animal protein intake is low (1,9), animal fat (lard) is used when available mainly for flavoring by mixing with maize meal. TC concentrations were similar in all groups, but the general tendency was lower than in the rural population (1) and lower compared to other groups(12-13). This communication compares for the first time the neutral lipid FA compositions in these southern African groups. Groups were not separated according to sex, because no sex difference could be detected in studies by Holman e t al. (14) and Manku e t al. (15). The effect of age should have little influence on the CE% FA composition (16); accordingly, population CE% FA differences are most likely to be due to dietary differences (6). The high LA content in the plasma CE of Hereros and Kavangos, and concomitant decreases in palmitoleic acid in these groups compared to Bushmen, suggests that these groups consume relatively less animal products than do the Bushmen, because plasma CE FA are known to have highly significant associations with dietary fat consumption due to the preference to plasma CE as indicators of dietary fat compositions (4-6). These findings are in accordance with a study of Melchert e t al. (17). The lower LA content in Bushmen may also indicate a reduced polyunsaturated fat intake by this group {18) and a relatively high saturated fat intak~ because vegetable and fruit consumption is little to none at all, which could be expected to increase the risk for CHD or atherosclerosis (6-7,19). Only few of the rural Bushmen, living from hunting and gathering, survive today. They had high LA TAG% FA values of 25.9% compared with the 9.3% as measured in the participating group (2,9). Because the plasma CE% FA of Bushmen are similar to those of normal Italian subjects, this may be indicative of a similar dietary P/S ratio (3). The FA of the Kavangos and Hereros are similar to values for groups reported by other investigators (19-21). However, South African whites and Xhosas have higher LA levels (unpublished data). Bushmen d~ however, have the highest OA levels in their plasma CE. Compared to the other groups, this could protect Bushmen because OA tends to resist oxidative modifications of FA while it simultaneously reduces the concentration of low density lipoprotein cholesterol. The net effect could be a reduction in the progression of

LIPIDS, Vol. 27, ho~ 9:(t992)

732

H.Y. TICHELAAR E T AL. atherosclerosis, which may explain in part the low incidence of CHD in Bushmen (22). The white control group had the lowest OA levels of all groups, which m a y be a contributory factor to the high incidence of CHD in this group (13). The white group also had higher CE I~S ratios than the Bushmen. Greenland Eskimos, well documented as having a reduced incidence of CHD (23), are characterized by having high OA levels and corresponding low CE I~S ratios (4,17). Lipid and FA comparisons of the different groups may oversimplify the risk elements of CHD and exclude the potential importance of genetic factors. However, very little information exists on the risk factors of the Namibian groups. The high palmitoleic acid and OA levels in the CE% FA of Bushmen with concomitant low LA levels have typical FA patterns of chronic alcoholics after acute alcohol abuse {24), and may indicate excessive alcohol consumption. B o t h palmitoleic acid and OA are known to correlate positively with alcohol consumption, while LA correlates negatively with these variables. Alcohol is known to alter lipid metabolism (8), probably by stimulating hepatic FA biosynthesis or inhibition of A6 desaturase activity in these individuals {25). Bushmen also had the lowest AA levels and the highest EPA/AA ratio of the Namibian groups. However, this is still considerably less t h a n the EPA/AA ratio of the reference whites, who are known to be at risk of CHD (12). Although low ratios may increase the risk of developing CHD, the group of Bushmen studied had no clinical signs of CHD. This could perhaps become a problem with economic improvements resulting in increased total caloric i n t a k e The change from their traditional lifestyle has resulted in a high incidence of malnutrition (9), especially reduced fat and protein stores (70% and 75%, respectively). The severe malnutrition is primarily responsible for the predominance of infectious disease in Bushmen, especially tuberculosis (9). One fact t h a t was extremely clear was the high incidence of alcoholism. Urbanization of Bushmen may thus have dire consequences for their continued survival as an independent ethnic group in modern society (9,26). ACKNOWLEDGMENTS The authors wish to thank Magdel H611and G.P. Engelbrecht for the data on the Xhosa reference group, T.J.V:W.Kotze for statistical guidance, M.J. Weight for his valuable editorial comments and Marita Kruger for dietary calculations.

LIPIDS, Vol. 27, no. 9 (1992)

REFERENCES 1. O'Keefe, S.J.D., Rund, J.E., Marot, N.R., Symmonds, K.L., and Berger, G.M.B. (1988) ~ A ~ Med, J. 73, 643-648. 2. 3u A.S., and Mann, J.I. (1972)Atherosclerosis 16, 15-29. 3. Dougherty, R.M., GallL C., Ferro-Luzzi, A., and Iacono, J.M. (1987) Am. J. Clin. Nutr. 45, 443-455. 4. Moilanen, T. (1987) Lipids 22, 250-252. 5. De Backer, G., De Craene, I., Roeseneu, M., Vercaemst, R., and Kornitzer, M. (1989) Atherosclerosis 78, 237-243. 6. Nikkari, T. (1986) Prog. Lipid Res. 25, 437-450. 7. Miettinen, T.A~,Nankkarinen, V., Huttunen, J.IL, Mattila, S~,and Kllmlin, T. (1982) Br. Med. J. 285, 993-996. 8. Baraona, E., and Lieber, C.S. (1979)J. Lipid Res. 20, 289-315. 9. O'Keefe, J.D., and Lavender, R. (1989) Lancet ii, 255-258. 10. Tichelaar, H.Y., Benad6, A.J.S., Daubitzer, A.K., and Kotze, T.J.V:W. (1989) Clin. Chim" Acta 183, 207-216. 11. Cunnane, S.C. (1987) Alcoholism 11, 25-31. 12. Steyn, K., Benad6, A.J.K, Langenhoven, M.L., Joubert, G., and Rossouw, J.E. (1987) ~ Aft:. Med. J. 71, 483-486. 13. Roesouw,J.E., Jooste, P.L., Steyn, K., and Benad6, A.J.S. (1985) S~ A ~ Med. J. 67, 533-538. 14. Holman, R.T., Smythe, L., and Johnson, S. (1979) Am. J. Clin. Nutr. 32, 2390-2399. 15. Manku, M.S~,Horrobin, D.F., Huang, Y-S.,and Morse, N. (1983) Lipids 18, 906-908.

16. Moilanen, T., Nikkafi, T., Riisimen,L., Viikari, J.,/~kerblom, H.K., Ahola, M., Dahl, M., Liihde, P-L., Pesonen, E., Pietik~nen, M., Sepp~inen, A., Suoninen, P., and Uhari, M. (1983)Atherosclerosis 48, 49-56. 17. Melchert, H-U., Limsathayourat, N., Mihajlovic, H., Eichberg, J., Thefeld,W., and Rottka, H. (1987)Atherosclerosis 65, 159-166. 18. Knuiman, J.T.,West, C.E., Hermus, R.J.J., and Hautvast, J.G.A.J. (1980) Atherosclerosis 37, 617-624. 19. Nikkari, T., Salo, M., Maatela, J., and Aromaa, A. (1983) Atherosclerosis 49, 139-148. 20. Navarro, X., Canete, J.D., Prat, J., and Segura, R. (1988)Revista Espanola De Fisiologia 44, 387-394. 21. Nikkari, T., Riis~Lrmn,L., Viikari, J.,/~kerblom, H.K., Vuori, I., Py6riil~i, K., Uhari, M., Dalfl, M., Liihde, P-L., Pesonen, E., and Suoninen, P. (1983) Am. J. Clin. Nutr. 37, 848-854. 22. Parthasarathy, S., Khoo, J.C., Miller, E., Barnett, J., Witztum, J.L., and Steinberg D. (1990)Prec. Natl. Aca& Sci. 87, 3894-3898. 23. Dyerberg, J., Bang, H.O., and Hjorne, N. (1975)Am. J. Cl/n~N u ~ 28, 958-966. 24. Alling, C., Dencker, S.J., Svennerholm, L., and Tichy, J. (1969) Acta Med. Scand. 185, 99-105. 25. Warnet, J-M., Cambien, F., Vernkier, V., Pecorar~ M., Flament, C., Ducimetiere, P., Jacqueson, A., Richard, J-L., and Claude, J-R. (1985) Br. Med. J. 290, 1859-1861. 26. Kondo, T., Ogawa, K., Satake, T., Kitazawa, M., Tald, K., Sugiyama, S., and Ozawa, T. (1986) Clin. CardioL 9, 413-416. [Received October 16, 1991, and in final revised form July 2, 1992; Revision accepted July 3, 1992]

Plasma lipids and fatty acids in urbanized Bushmen, Hereros and Kavangos of southern Africa (Namibia).

Thirty-nine urbanized ethnic Namibian people comprising 21 Bushmen (semi-urbanized), 7 Hereros and 11 Kavangos were assessed for plasma lipids and fat...
484KB Sizes 0 Downloads 0 Views