Massive Obesity in a Migrant Samoan Population by ucb13077


									             Massive Obesity in a Migrant Samoan Population
                                     IVAN G. PAWSON, PHD,            AND   CRAIG JANES, MA

     Abstract: Peoples of the Pacific Islands tend to be-            the participants' average height fell between the 25th
come overweight when they migrate or are exposed to                  and 50th percentile of the US population, about one-
modernization in situ. Recent evidence suggests that                 half our sample exceeded the 95th percentile for
Samoans are particularly susceptible, exhibiting a high              weight. The extreme overweight was accompanied by
prevalence of obesity and hypertension following mi-                 elevated blood pressure and, in females, by elevated
gration to Hawaii. We report the preliminary results of              FPG. The massive adult weight of migrants from the
a survey of height, weight, blood pressure, and fasting              Pacific Islands carries serious public health implica-
plasma glucose (FPG) among an urbanized Samoan                       tions for areas that support large migrant communities.
community in the San Francisco Bay Area. Although                    (Am J Public Health 1981; 71:508-513.)

Introduction                                                         blood pressure are linked to the degree of exposure to west-
                                                                     ern influences and lifestyles, then the Samoan population in
     One of the expected consequences of migration to a              California should exhibit greater contrasts to the native pop-
modem industrialized country from a rural, nonwestern en-            ulations in Samoa than the Samoan population in Hawaii.
vironment and of the accompanying alterations in traditional
dietary and exercise habits is increased body weight. Peo-
ples of the Pacific Islands seem particularly prone to becom-        Materials and Methods
ing overweight and to developing concomitant degenerative
and metabolic disorders when they migrate or are exposed to          Study Population
increased modernization. The trend has prompted several in-               The Samoans of our study reside in the southern part of
vestigations into the public health implications of resettle-        San Francisco and adjacent areas of San Mateo County. The
ment programs'-4 and the consequences of exploitation of             structure of their community centers around 27 churches in
island resources.5 -7                                                the Bay Area. The traditional Samoan hierarchy of chiefs
     Continued migration has led to the growth of sizable Pa-        functions in a semi-formal capacity, although the extent to
cific Islander communities in the mainland United States,            which it can exert social control over the community is open
but because these peoples are frequently classified in official      to question. The Office of Samoan Affairs, Inc., a national
records under the ethnic category '"Other," no formal esti-          council of chiefs and community leaders, promotes adult
mates of population size are available. In California, a major       education and other programs that ensure the maintenance
recipient state for migrants from the Pacific Islands, the size      of cultural identity and assists immigrants in integrating
of the Samoan community alone is estimated to be between             themselves as well as is possible into the host culture.
60,000 and 80,000 persons,* most of whom live in the metro-
politan areas of San Diego, Los Angeles, and San Francisco.          Sampling Procedures
Because previous studies have shown that Samoans livingin -               Because of the-virtual absence of ethnogra-phic, medi-
semi-urban conditions in Hawaii exhibit a substantially high-        cal, or other literature on Pacific Island peoples who have
er prevalence of obesity and hypertension than their coun-           migrated to the United States, there are no reliable estimates
terparts in Samoa,'0, we decided to undertake a pilot sur-           of population characteristics such as size, distribution, medi-
vey in San Francisco to determine the frequency of these             cal history, health status, or mortality. In pilot surveys such
conditions among Samoans who have moved to the main-                 as ours, lack of background data makes the selection of a
land. We hypothesized that if increased weight and elevated          representative sample difficult. We chose church congrega-
                                                                     tions as our sampling base because the great majority of Sa-
                                                                     moan families regularly attend church, rendering estimates
      From the Department of Epidemiology and International          of various biological parameters obtained from this source
Health, University of California, San Francisco. Address reprint     more likely to reflect true population characteristics. With
requests to Ivan G. Pawson, PhD, c/o Editor, Epidemiology and        the help of church ministers, who exert considerable influ-
International Health, 1699 HSW, University of California, San        ence in the Samoan community, we were able to examine
Francisco, CA 94143. This paper, submitted to the Journal July 30,
1980, was revised and accepted for publication December 17, 1980.    most members of two church congregations and about 50 per
*Information supplied by Department of Health and Human Serv-        cent of a third one. We also examined members of a Samoan
ices, Region IX, Office of Planning and Evaluation.                  adult education class on the recommendation of a minister,

508                                                                                                   AJPH May 1981, Vol. 71, No. 5
                                                                                                      MASSIVE OBESITY IN MIGRANT SAMOANS

himself a member of the class. In all, 179 individuals were                              MALES

examined (88 males, 91 females-Table 1). Because of the
casual attitude with which most Samoans view overweight,
we deem it unlikely that an individual's weight had any bear-
ing on his or her decision to participate. Examinations took                 E
place after Sunday service or during scheduled church func-
tions.                                                                      ._L
     We report here data on height, weight, and triceps skin-
fold, measured according to standardized techniques."
Blood pressure was taken in the resting state, by the same
observer (IGP) using a mercury sphygmomanometer fitted
with the appropriate cuff size for individuals of large or nor-
mal size. Fasting glucose levels were determined using                                      18 -29         30-39               40-49                   50 & over
venous blood and Dextrostix reagent strips which were read                                                           Age in years
with the Ames reflectance meter.
                                                                                                                                    L      1: Western Samoa
Comparison Data
                                                                                                                                    i_ 2: American Samoa
     Comparison data were available for Samoan popu-                                    FEMALES
lations (Table 1) in American Samoa and Hawaii9 and in                                                                                     3: Hawaii
Western Samoa.'0 In each location, data were collected by                                                                           - 4: California
members of the Pennsylvania State University Migration
     The data from Western Samoa, described by Baker and
Hanna,'0 derive from two villages chosen for their remote-                   _-
ness from the capital, Apia, and comprise 318 adults. The
composition of comparison samples in American Samoa and
Hawaii are reported by McGarvey and Baker.9 A total of
2,219 adults (1,688 in American Samoa, 591 in Hawaii) were
examined from 1975 to 1977. In American Samoa, subjects
were recruited from the capital, Pago Pago, and the sur-                                    18 - 29        30 - 39             40 - 49                 50 & over
rounding areas. In Hawaii, they were recruited from rural                                                            Age in years
areas on the north and western shores of Oahu, as well as                         FIGURE 1-Height in Four Adult Samoan Subpopulations
from urban areas in Honolulu. In all cases, individuals were
recruited voluntarily and examined either in the home or in a
ceremonial house. Measurement techniques-for example,
the use of blood pressure cuff size appropriate to the sub-                Results
jects' arm circumference-were similar in all studies. How-
ever, in view of uncertainties over the background, composi-                Height
tion, and representativeness of these samples, and the fact                      Figure 1 depicts adult height of Samoans living in San
that different investigators collected the data, detailed com-              Francisco compared with populations in Western Samoa,
parisons between them are unwarranted. We present the                       American Samoa, and Hawaii. In general, the stature of
data here only to emphasize the unique physical character-                  adult Samoan males is close to the 25th percentile of height
istics of the California population, compared to those in Ha-               (171 cm) in the US population,'2 whereas females fall be-
waii and Samoa.                                                             tween the 25th (158 cm) and 50th (162 cm) percentiles. The

TABLE 1-Sample Sizes of Samoan Subpopulations
                       Western Samoa                       American Samoa                       Hawaii                                          California

                    Male             Female             Male             Female             Male               Female                    Male                Female

Age (years)    N           %     N            %   N            %     N            %     N             %    N            %           N            %           N      %

19-29          37           29   62        33     177           26   277           28 85            34    152         44            14           16          26     29
30-39          26           20   41        22     151           22   211           21  52           21     61         18            21           24          23     25
40-49          24           19    33       17     134           19   209           21  45           18     73         21            20           23          23     25
50+            41           32    54       28     231           33   298           30 67            27     56         16            33           37          19     21
TOTAL         128          100   190      100     693          100   995          100 249          100    342         99            88          100          91    100

AJPH May 1981, Vol. 71, No. 5                                                                                                                                      509

                  MALES                                                                                 relatively small size of the California sample makes it diffi-
                                                                                                        cult to judge whether the differences between the other three
                                                                                                        groups are biologically significant. Indeed, given our uncer-
                                                                                                        tainty about the background of migrant populations in Ha-
      oo-                                                                                   |           waii and California, and the difficulty in working with pub-
                                                                                                        lished variances and means, it is doubtful whether detailed
                                                                                                        comparisons are justified at this stage. However, to derive a
.c                                                                                                      measure of overall differences between the groups, we calcu-
     ~90                                                                                                lated pooled means and variances from the published age-
                                                                                                        specific means and variances for Hawaiian and Samoan sam-
                                                                                                        ples" and compared them with similar values from our own
                                                                                                        study. Values of the t statistic for comparison of means in
                                                                                                        the Califoria population with each of the other three groups
      70 -           | .s                                         s                         s           were insigificant once probability values were corrected for
                                                                                                        the number of groups in the comparisons (four), suggesting
                                                                                                        that sampling bias may have accounted for the higher heights
                                                                                                        seen in our sample. Increased stature among migrants would
                    18-29             30 -39               40 -49                 50 & over             be expected if migration takes place early in the life cycle;
                                               Age in years                                             however, all participants in our study were born in Samoa
                  FEMALES                                         KEY                                   and spent their early years there before moving to California.
                                                       L=l 1: Western Samoa                    Weight
     110                                                11 2: American Samoa                        Figure 2 shows weight among the same four sub-
                                                           3: Hawaii                           populations. The weight difference between the Western
                                                                  4: California                Samoa population and the populations in American Samoa,
     100                                                                                       Hawaii and California is striking. Among individuals in our
                                                                                               survey, 55 per cent of males and 46 per cent of females ex-
                                                                                               ceed the age specific 95th percentile for weight in the US
Z                                         : @
                                                          so 90
                                                          :@                             B
                                                                                             |population.'2 Pooled mean and variance estimates were cal-
                                                                                               culated for each sex from the published data for populations
                                                                                               in Samoa and Hawaii and compared with the California
      so1                  ;          :g                        @i                     g       sample by means of the t test. Even when corrected for the
                                                                                               number of groups being compared, the t statistic indicated
                                                                                              that males and females in the California subpopulation were
      70                                                                                       significantly heavier (.05 > p > .01) than samples from
                           @@.            i                   @ |                     |:=      Samoa or Hawaii. The average adult weights of individuals
                                                                                               in our samples, 103.2 kg among males and 93.05 kg among
                                                                                                        females, suggest that the California Samoan population may
                    18-29             30 -39                  40 -49              50& over              be the world's heaviest.
                                               Age in years
           FIGURE 2-Weight in Four Adult Samoan Subpopulations

     TABLE 2-Average Triceps Skinfold Measurements (mm) in Migrant and Native Samoans
                                                                                                      Age Groups

                                                    Males                                                                                  Females

                          20-29            30-39                      40-49                     50+                20-29           30-39             40-49           50+

                     N     sd     x   N       sd      x       N        sd         x     N       sd      x   N       sd     x   N    sd     x     N    sd     x   N   sd    k

     Amercan 134       12.8 151      17.9 134      18.6 231      18.5 235      27.4 211      38.0 209      39.4 298      37.3
     Samoa1        6.0          10.2          10.6          10.4          12.4          13.9          14.6          13.6
     Hawaii1    61     17.3 52       18.4 45       18.6 67       19.8 114      28.7 61       34.7 73       37.2 56       32.8
                   9.4           7.9           7.8           8.5          10.9          11.5          11.2          11.0
     California 13     20.5 20       17.2 18       19.8 31       17.7 25       25.2 21       31.8 20       32.7 20       31.3
                   9.9           6.7          10.9           8.2           8.7          10.9           7.9           8.3
             Data from   McGarvey and Baker, 1979

     510                                                                                                                                       AJPH May 1981, Vol. 71, No. 5
                                                                                                                   MASSIVE OBESITY IN MIGRANT SAMOANS

                                                                              *                p California
                                                                                             .A Hawaii

                                                                              +   -      -   -+ American Samoa
                                                                              o                0 Western Samoa

                                                             Males                                                  Females
                        15o r                                                                 150 r

                        140   F                                      .                        1401-
                        1301-     A_--          +;+>   W-_   _---
                                                                             ,                130 1
                                  + --I
                        120 1                                                                 120p-

                   E                                                                          110 F
                   E 110 1

                        1001-                                                                 100    F

                         gol-                                                     A,            0o   F
                         80   F                                                                80o

                         701-                                                                  70 p

                         60                                                                    60

                                          I     I                        I         I
                                                                                                         a    l               I
                                  25            35             45        55       65                     25   35       45     55    65

                                                                                       AGE IN YEARS
                                              FIGURE 3-Blood Pressure in Four Adult Samoan Subpopulations

Skinfold Thickness                                                                           grant population has centered    on indicators of cardiovascu-
      Table 2 shows the measurements of the thickness of the                                 lar and metabolic risk. Figure 3 shows comparison of blood
triceps skinfold in migrant Samoan populations in Hawaii                                     pressure from the California sample with similar data from
and California, and in the native population of American                                     samples in Samoa and Hawaii. Although these data derive
Samoa. In all age groups, and in both sexes, mean skinfold                                   from measurements made by many different investigators, it
values exceed the 80th percentile for US whites whereas                                      appears that blood pressure variation parallels variation in
mean values for females in American Samoa over the age of                                    weight. However, partial correlation analysis, in which the
30 exceed the 95th percentile. The distribution of skinfold                                  relationship of body mass index (weight/height2) with blood
thicknesses among the four subpopulations is unexpected in                                   pressure was examined, while holding constant the effect of
view of the massive adult weights recorded for the migrants,                                 age, revealed a strong positive correlation in females, but a
particularly those in California. We calculated pooled mean                                  weak and statistically insignificant correlation in males. The
values for the triceps skinfold, in the manner described                                     preliminary findings suggest a different pattern of weight/
above. When these values were compared using the adjusted                                    blood pressure relationships in our samples than in those
t statistic, no significant differences were observed between                                studied by Hanna and Baker8 in Hawaii. In the Hawaiian
the group means, even though adult weight was significantly                                  sample, weight was strongly associated with blood pressure
greater in the California subpopulation. This suggests that                                  in males, whereas in females age was the strongest predictor.
the triceps skinfold may not be reliable as an indicator of                                       Table 3 illustrates the percentage of migrant and native
adiposity in Samoans. If the extra weight of the California                                  Samoan populations that exhibit definite hypertension. Al-
sample is carried as fat, as seems likely, the extra accumula-                               though the size of individual age groups (especially in our
tion is not reflected by skinfold thickness at the triceps.                                  sample) is too small to warrant detailed analysis, it is clear
However, compression measurement of skinfold thickness                                       that males as a group exhibit a higher prevalence of hyper-
on obese subjects is unreliable and may be inappropriate in a                                tension than the US norm, whereas Samoan females exhibit
                                                                                             a somewhat lower prevalence. While the distribution of hy-
population where obesity is so prevalent.
                                                                                             pertension among Hawaiian and Californian samples was
                                                                                             similar (x2 = 3.0 for males, x2 = 0.21 for females; .05 < p),
Blood Pressure                                                                               males and females in each sample differed significantly from
    To date, analysis of health outcomes in relation to the                                  each other (X2 = 6.12; .05 > p > .01 for Hawaiian Samoans,
massive adult weights encountered among the Samoan mi-                                       XI = 6.96; .01 > p > .001 for California Samoans). Strati-
AJPH May 1981, Vol. 71, No. 5                                                                                                                          511

TABLE 3-Proportion of Hypertensive' Individuals in Samoan Migrant and Native Populations

A. Migrant Populations in Hawaii and California
                                            Males                                                                              Females

                      18-34              35-54                  55+               All Ages          18-34            35-54             55+               All Ages

   Age            N           %     N            %          N          %         N           %        N        %       N       %           N    %        N          %

Hawaii2         14/80     17.5    22/81      27.2          8/36       22.2     44/197     22.3 10/149         6.7    17/75   22.7      7/29    24.1   34/253 13.4
California       6/24     25.0    16/38      42.0          6/25       24.0     28/87      32.8 2/40           5.0     8/38   21.1      4/13    30.8   14/91  15.4
U.S.3                      7.0               22.9                     34.5                19.2                3.0            18.1              40.0          17.1

B. Native Populations in American Samoa and Western Samoa (All Ages)
                                       "Traditional"                                               "Intermediate"                              "Modem"
                                  Male                                Female            Male                        Female          Male                     Female
% American Samoa4                  9.0                                 18.0             20.0                         16.0           24.0                      21.0
% Western Samoa4
  (Both Sexes)                                       6.0

       Systolic Pressure 2 160 mm Hg or Diastolic Pressure 2 95 mm Hg.
      2Data from Hanna & Baker (1979).
      3Health and Nutrition Examination Survey (Roberts & Maurer, 1977).
      4P. T. Baker, Personal Communication.

fication of the American Samoan sample on the basis of                                       and related health risks in this population. Detailed dis-
proximity to Pago Pago (and presumed exposure to western                                     cussion of our data is unwarranted at this time because of
diet and lifestyles) illustrates a downward trend in the pro-                                uncertainties over the background of the Califomia Samoan
portion of hypertensive individuals as one moves away from                                   population, the chronology of weight gain among migrants,
the capital. Male Samoan nonmigrants designated as "mod-                                     and the general lack of demographic information. For ex-
em" exhibit a degree of hypertension similar to their coun-                                  ample, we are unsure of whether the obese Samoan adults
terparts in Hawaii, female nonmigrants exhibit a higher prev-                                we see so frequently are in some way a 'selected" group of
alence than either the Hawaiian or Californian sub-                                          Samoans, becoming fat before they migrate. At present, it is
population. These findings are in direct contrast to the                                     uncertain whether the obesity we have seen represents ex-
relatively low rate of hypertension in Western Samoa.                                        cess weight gained before migration, after migration, or dur-
                                                                                             ing both periods. In addition, the cultural beliefs that deter-
Fasting Blood Glucose                                                                        mine attitudes toward food and obesity are currently un-
     Blood glucose levels were examined in 68 males and 75                                   known, but may be of prime importance in understanding the
females. Although all individuals had been requested to at-                                  etiology of obesity in this population. If these problems can
tend the examination in a fasting state, interviews revealed                                 be resolved, the California Samoan population should prove
that only 40 of the male sample (59 per cent) and 43 of the                                  a useful model for the study of obesity in humans. Given the
female sample (57 per cent) could be considered truly fast-                                  present uncertainty over etiology, risks, morbidity, and mor-
ing, the remaining individuals having had sweetened coffee                                   tality that surrounds this condition, such a model could aid
or other sugar containing beverages immediately before the                                   greatly in understanding its natural history.
examination. We considered the fasting sample too small for
detailed analysis, but it was interesting to note that 18 per                                                       ACKNOWLEDGMENTS
cent of the male fasting sample, and 9 per cent of the female                                     Supported in part by National Institutes of Health Research Ca-
fasting sample had glucose levels that exceeded the 95th per-                                reer Development Award K04-AG00022 and by grants from the Uni-
centile for fasting blood glucose in a US white population.'3                                versity of California, San Francisco, School of Medicine.
When the effects of age were controlled, blood glucose cor-
related significantly with body mass index in females (r =                                                              REFERENCES
0.36, .05 > p > .01) but not in males (r = -0.04, .05 < p).                                    1. Prior IAM: Nutrition and epidemiology. J NZ Dietetic Assoc
                                                                                                  1966; 20:9-17.
                                                                                               2. Prior IAM: Cardiovascular epidemiology in New Zealand and
                                                                                                  the Pacific. NZ Med J 1974a; 80:345-352.
Discusssion                                                                                    3. Prior IAM, Stanhope J, Evand JS, Salmond CE: The Tokelau
                                                                                                  Island migrant study. Int J Epidemiol 1974b; 3:225-232.
                                                                                               4. Prior IAM, Beaglehole R, Davidson F, Salmond CE: The rela-
    The data presented here reflect the initial stages of what                                    tionship of diabetes, blood lipids and uric acid levels in Poly-
we expect to be a more extensive investigation of obesity                                         nesians. Adv Metab Disord 1978; 9:241-261.

512                                                                                                                                AJPH May 1981, Vol. 71, No. 5
                                                                                          MASSIVE OBESITY IN MIGRANT SAMOANS

 5. Zimmet P, Seluka A, Collins J, et al: Diabetes mellitus in an     10. Baker PT, Hanna JN: Modernization and biological fitness of
    urbanized isolated Polynesian population. Diabetes 1977;              Samoans: A progress report on a research program. IN: Pro-
    26:1101-1108.                                                         ceedings of the 1979 Seminar on Migration and Health, Epide-
 6. Zimmet P, Bjorntorp P: Adipose tissue cellularity in obese non-       miology Unit, Community Health Department, Wellington Hos-
    diabetic men in an urbanized Pacific Island (Polynesian) popu-        pital, Wellington, New Zealand.
    lation. Am J Clin Nutr 1979; 32:1788-1791.                        11. Weiner JS, Lourie JA: Human Biology: A Guide to Field Meth-
 7. Ringrose H, Zimmet P: Nutrient intakes in an urbanized Micro-         ods. Philadelphia: F. A. Davis, 1969.
    nesian population with a high diabetes prevalence. Am J Clin      12. Abraham S, Johnson CL, Najjar MR: Weight and height of
    Nutr 1979; 32:1134-1341.                                              adults 18-74 years of age: United States, 1971-74. DHEW Pub.
 8. Hanna JM, Baker PT: Biocultural correlates to the blood pres-         No. (PHS) 79-1659, National Center for Health Statistics,
    sure of Samoan migrants in Hawaii. Hum Biol 1979; 51:480-497.         Hyattsville, MD, 1979.
 9. McGarvey ST, Baker PT: The effects of modernization and mi-       13. Barrett-Connor E: Factors associated with the distribution of
    gration on Samoan blood pressures. Hum Biol 1979; 51:461-             fasting glucose in an adult community. Am J Epidemiol 1980;
    479.                                                                  112:518-523.

                  Kellogg Pharmaceutical Clinical Scientist Fellowships Available
                 Applications from qualified candidates are now being accepted for the Kellogg Pharmaceutical
            Clinical Scientist Fellowships offered by the University of Minnesota in conjunction with the St. Louis
            Park Medical Center Research Foundation, Minneapolis, MN. Ten fellowships remain to be awarded
            during the Fall 1981 academic term. Each three-year fellowship carries a stipend of $10,000 per year of
            study leading to the PhD degree in Social and Administrative Pharmacy.
                 Deadline for application for the fellowship is May 1, 1981; final decisions by the admissions
            committee will be made by May 15, 1981. Early application is encouraged.
                 The Program, funded by the W. K. Kellogg Foundation of Battle Creek, Michigan, represents an
            endeavor to fulfill a need identified by the 1975 Report of the (Millis) Study Commission on Pharmacy.
            During the course of study, experienced practicing pharmacists will fulfill university coursework and
            also participate in projects at various pharmacy practice sites. Degree credit will also be earned through
            these projects, the nature of the projects to be determined the goals and interests of the
            individual fellows.
                 Graduates of the Program are expected to become the major community resource for the
            application of pharmacy knowledge, to contribute significantly to the quality of drug-related services
            provided in their communities, and to assume leadership roles in drug-related research.
                 Criteria on which candidates will be judged include: a clinically-oriented pharmacy education, a
            minimum of two years clinical pharmacy practice experience relevant equivalent, professional
            licensure, superior prior academic performance, demonstrated leadership ability, and letters of
                 Specific information concerning the program and the formal application procedure may be
            obtained by writing to: Kellogg Pharmaceutical Clinical Scientist Program, Attention: Dr. Albert I.
            Wertheimer, College of Pharmacy, University of Minnesota, Minneapolis, MN 55455.

AJPH May 1981, Vol. 71, No. 5                                                                                                      513

To top