ARTICLE
Cancer Incidence in Israeli Jewish Survivors of
World War II
Lital Keinan-Boker, Neomi Vin-Raviv, Irena Liphshitz, Shai Linn, Micha Barchana
Background
Israeli Jews of European origin have high incidence rates of all cancers, and many of them were exposed
to severe famine and stress during World War II. We assessed cancer incidence in Israeli Jewish survivors
of World War II.
Methods
Cancer rates were compared in a cohort of 315 544 Israeli Jews who were born in Europe and immigrated
to Israel before or during World War II (nonexposed group, n = 57 496) or after World War II and up to
1989 (the exposed group, ie, those potentially exposed to the Holocaust, n = 258 048). Because no individual data were available on actual Holocaust exposure, we based exposure on the immigration date for
European-born Israeli Jews and decided against use of the term “Holocaust survivors,” implying a known,
direct individual Holocaust exposure. Cancer incidences were obtained from the Israel National Cancer
Registry. Relative risk (RR) estimates and 95% confidence intervals (95% CIs) were calculated for all cancer
sites and for specific cancer sites, stratified by sex and birth cohort, and adjusted for time period.
Results
The nonexposed group contributed 908 436 person-years of follow-up, with 13 237 cancer diagnoses
(crude rate per 100 000 person-years = 1457.1). The exposed group contributed 4 011 264 person-years of
follow-up, with 56 060 cancer diagnoses (crude rate per 100 000 person-years = 1397.6). Exposure, compared with nonexposure, was associated with a statistically significantly increased risk for all-site cancer
for all birth cohorts and for both sexes. The strongest associations between exposure and all-site cancer
risk were observed in the youngest birth cohort of 1940–1945 (for men, RR = 3.50, 95% CI = 2.17 to 5.65;
for women, RR = 2.33, 95% CI = 1.69 to 3.21). Excess risk was pronounced for breast cancer in the 1940–
1945 birth cohort (RR = 2.44, 95% CI = 1.46 to 4.06) and for colorectal cancer in the 1935–1939 cohort (for
men, RR = 1.75, 95% CI = 1.19 to 2.59; for women, RR = 1.93, 95% CI = 1.25 to 3.00).
Conclusions
Incidence of all cancers, particularly breast and colorectal cancer, was higher among Israeli Jews who
were potentially exposed to the Holocaust than among those who were not.
J Natl Cancer Inst 2009;101:1489–1500
Jews of European or American origin have a higher incidence of
cancer than other Jewish or non-Jewish ethnic groups in Israel. In
2004, for example, the age-standardized rate of all cancers among
Jewish men who were born in Europe or America was 297.50 cancers per 100 000 person-years, compared with 255.79 and 286.52
cancers per 100 000 person-years, in Asian- and African-born men,
respectively. In 2004, the age-standardized rate for Jewish women
born in Europe or America was 272.90 cancers per 100 000 person-years, compared with 214.90 and 234.4 cancers per 100 000
person-years in Asian- and African-born women, respectively (1).
Many European-born Israeli Jews who survived World War II
were exposed to the Holocaust, which included severe starvation,
extreme mental stress, and exposure to various infectious agents
and to cold winter temperatures. Thus, a possible explanation for
the differences in cancer incidence observed among the various
Jewish ethnic groups may be differences in their specific exposure
to the traumas of the Holocaust during World War II.
The few studies that have investigated the association between
psychological stress and cancer incidence in various populations
have reached inconclusive results (2). Two studies (3,4) found a
jnci.oxfordjournals.org
positive relationship between self-reported stress levels and the
incidence of breast and prostate cancers, but their findings were
not supported by results of other studies (5,6). Also a systematic
review of stress and breast cancer (7) concluded that stress does not
seem to increase the incidence of breast cancer. In addition, one of
the most commonly used tools for assessing stress exposure, the
Affiliations of authors: School of Public Health, Faculty of Welfare and Health
Sciences, University of Haifa, Haifa, Israel (LK-B, NV-R, SL, MB); Israel Center
for Disease Control, Ministry of Health, Tel Hashomer, Ramat Gan, Israel (LK-B);
Israel National Cancer Registry, Ministry of Health, Jerusalem, Israel (IL, MB);
Unit of Clinical Epidemiology, Rambam Medical Center, Haifa, Israel (SL).
Correspondence to: Lital Keinan-Boker, MD, PhD, MPH, School of Public
Health, Faculty of Welfare and Health Sciences, University of Haifa, Mt
Carmel, Haifa 31905, Israel (e-mail: [email protected]; lital.keinan@
icdc.health.gov.il).
See “Funding” and “Notes” following “References.”
DOI: 10.1093/jnci/djp327
© The Author 2009. Published by Oxford University Press. All rights reserved.
For Permissions, please e-mail: [email protected].
Advance Access publication on October 26, 2009.
JNCI
|
Articles 1489
CONT E X T A N D C A VEAT S
Prior knowledge
High incidence rates of all cancers have been observed among
Israeli Jews of European origin, many of whom were exposed to
severe famine and stress during World War II.
Study design
Retrospective study in a cohort of Israeli Jews who were born in
Europe between 1920 and 1945. The exposed group (ie, those
potentially exposed to the Holocaust) immigrated to Israel before
or during World War II. The unexposed group immigrated after
World War II but before 1989.
Contribution
Exposure, compared with nonexposure, was associated with statistically significantly increased risk for all-site cancer for all birth
cohorts and for both sexes. This association was strongest in the
youngest birth cohort of 1940–1945.
Implications
Individuals exposed to severe famine and stress for prolonged
periods appear, especially those exposed at an early age, to be at
increased risk of all-site cancer and should be monitored accordingly. These findings warrant additional investigations that use
individual data to elucidate risk factors.
Limitations
Data on individual exposure to the Holocaust were not available, and
so a proxy variable for exposure was used that was based on immigration date. The study included only Jewish individuals living in
Israel. The birth cohort of 1940–1945 was the smallest cohort studied.
From the Editors
Schedule of Recent Experiences Questionnaire, which focuses on
traumatic life events in the recent past (8), was not designed to
capture stressful life events that occurred many years before and,
therefore, could not be used for World War II survivors. Thus, to
our knowledge, no studies exist on the relationship of World War
II and its associated mental stress to subsequent cancer incidence.
More is known about the relationship between caloric restriction and cancer in animals. This relationship has been used to
generate the hypothesis that caloric restriction decreases the risk of
cancer (9). The proposed mechanisms for this relationship include
increased DNA repair activity, reduced cell proliferation as a result
of the decreased availability of energy, and reduced oxidative stress
(9). Hormonal changes may also be involved (9,10).
The relationship between substantial—but transient—energy
restriction in humans and cancer risk has been explored in only a
few studies (11–19), with all being based on war-related extreme
situations that, unfortunately, created opportunities to address this
question. Four Norwegian studies (11–14) that focused on the
long-term effects of the calorie-restricted Norwegian diets during
World War II (average daily energy intake per person of 1240 kcal
in the winter of 1945 compared with a prewar value of 2500 kcal)
found positive associations between caloric intake and breast
cancer risk. The series of Dutch studies by Dirx et al. (15–17),
which were mostly based on ecological exposure data, detected no
association between caloric restriction (average daily energy intake
1490 Articles
|
JNCI
per person of approximately 800 kcal) and breast (15) or prostate
(16) cancer. However, a weak inverse relationship between energy
restriction early in life and subsequent colon carcinoma has been
reported for both men and women (17). Elias et al. (10) used individual exposure data to investigate the relationship between exposure to the Dutch famine from September 1944 through May 1945
and subsequent breast cancer risk. They reported an increased risk
for women who were severely exposed to this short but extreme
famine (ie, average daily energy intake per person of 800 kcal;
hazard ratio [HR] = 1.48, 95% confidence interval [CI] = 1.09 to
2.01) and a dose–response effect. In Guernsey, a British Channel
Island that was occupied by Germany between June 1940 and May
1945, the residents had an average daily energy intake per person
of 1200 kcal for almost a year (between June 1944 and May 1945).
However, breast cancer risk was not statistically significantly elevated among women who remained in the island under German
occupation and were aged 10–18 years in 1944 (HR = 1.28, 95%
CI = 0.61 to 2.75), compared with those who were evacuated (18).
Another recently published study (19) is of special interest. This
study focused on survivors of the Siege of Leningrad, which lasted
from September 1941 through January 1944, who were exposed to
severe starvation (average daily energy intake per person of 300
kcal that contained virtually no protein). The study results indicated that women exposed to the siege were at a higher risk of
dying from breast cancer (HR = 2.50, 95% CI = 0.92 to 6.80) than
those who were not exposed. Among those aged 10–18 years at the
peak of the siege in 1941–1942, these findings were statistically
significant (HR = 9.9, 95% CI = 1.1 to 86.5) (19).
Most of these studies (10–18) were conducted among nonJewish populations whose World War II experiences were considerably different, and probably less severe, than those of the
European Jewish population. The average daily energy intake
per person in Jewish ghettos and in concentration camps was,
roughly, 220–800 kcal (20,21), and the length of the exposure to
this daily diet of 220–800 kcal was substantial: Jews were interred
in concentration camps and ghettos at the beginning of the war
and remained there for the duration of the war. In fact, the longterm exposure of Jews to calorie-restricted diet often resulted in
malnutrition, which was clinically manifested as marasmus (severe
protein-energy malnutrition), kwashiorkor (severe protein deficiency), goiter, rickets, night blindness, anemia, and scurvy (21).
Furthermore, some Jews were very young when they were exposed
to severe conditions during the Holocaust, and an increasing body
of evidence supports the hypothesis that events occurring early in
a person’s life may strongly influence their subsequent health,
including their risk of cancer (22–28).
In this study, we investigated cancer incidence among a cohort
of European-born Israeli Jewish residents. We compared incidence
of cancer in the group who were potentially exposed to the
Holocaust with that in the nonexposed group to determine whether
the exposed group had a higher than expected cancer incidence.
Study Subjects and Methods
The Study Cohort
The study cohort included all Jewish people who were born in
European countries from 1920 through 1945 and who resided or
Vol. 101, Issue 21
|
November 4, 2009
had resided in Israel in 1983 and onward through December 31,
2004 (n = 315 544). The nonexposed group included Jews who
were born from 1920 through 1945 in Europe and who immigrated to the area that would become Israel before or during
World War II (ie, up to 1945). By use of these dates, we assumed
that the ability to immigrate during the war years reflected, in
itself, being nonexposed to the Holocaust. This nonexposed group
contained 57 496 subjects and so was small. The exposed group
included Jews who were born in Europe from 1920 through 1945
who immigrated to Israel (or to the area that would become Israel
in 1948) after the end of World War II (ie, after 1945) (n = 258
048). Exclusion criteria included immigration to Israel after 1989
because, in 1989 and the years that followed, Israel experienced a
massive wave of immigration from the former Soviet Union.
Because exposures of these immigrants between 1945 and 1989
may have been very different from exposures of those who immigrated to Israel between 1945 and 1989 and because information
regarding these potential exposures was mostly lacking, these
immigrants were excluded from the exposed group. According to
the Israeli Central Bureau of Statistics, 85% (approximately 268
212) of the study population immigrated to Israel before 1960 and
15% immigrated to Israel between 1960 and 1989 (29).
Exposure Status
We defined the exposure as living as a Jew in Europe directly or
indirectly under a Nazi regime during World War II (ie, from
1939 through 1945). Because no individual information regarding
actual exposure was available, we used a proxy variable that was
based on the date that European-born Jews immigrated to Israel:
after 1945 (ie, after potential exposure to the World War II
Holocaust) to define the exposed group and up to 1945 (ie, before
and during World War II) to define the nonexposed group. The
use of this proxy exposure variable is, in fact, the reason for our
deliberate decision not to use the term “Holocaust survivors,”
which implies a known, direct individual exposure to the horrors
of the Holocaust.
Assembling the Study Cohort
Data regarding Jewish immigrants from Europe before and after
World War II were retrieved from the Israeli Central Bureau of
Statistics. Two obstacles were encountered while assembling this
unique cohort. The first related to the country of origin of the
exposed group. The Israeli Central Bureau of Statistics categorizes
the Jewish Israeli population by birthplace, as Israel, Asia, Africa,
or Europe and America. We refined this categorization by birth
country to enable separation of European- and American-born
Jews. However, from 1983 through 1993, data on the specific
country of origin were not available, and the issue of what the
birthplace for those in the European or American group had to be
addressed for those in the exposed group of immigrates to Israel
from 1983 through 1989. The Israeli Central Bureau of Statistics
estimated that Jews born in America constituted 9%–11% of those
born in the total group of European- or American-born Jewish
immigrants from 1983 through 1993. From this information, the
size of the exposed group (n = 258 048) was uniformly reduced by
a factor of 10% across all age strata for all immigrants to Israel
from 1983 through 1989 to adjust for the number expected if only
jnci.oxfordjournals.org
European-born Jews were included (for a total of 232 244). The
size of the nonexposed group (n = 57 496) was not reduced,
because subcategorization by country of birth for this group,
which immigrated before 1945, was available.
The other obstacle related to determining the date of immigration is that the state of Israel was established in 1948. The Israeli
Central Bureau of Statistics does not have a specific immigration
date for Jews immigrating to Israel before 1947, but rather uses the
year 1947 as a cutoff point. Our inclusion criteria, however,
referred to the year 1945 as a cutoff point for the nonexposed
group. Indeed, most of the Jews who immigrated from Europe
before 1947 immigrated before 1939 and, therefore, were included
in the nonexposed group. From 1939 through 1945, immigration
was negligible but still existed. So, by use of these data, those
immigrating between 1946 and 1947 would be misclassified as
nonexposed. To reduce this expected misclassification, we estimated the size of this group. From 1946 through 1948, Israel was
under the British Mandate and was subjected to its policy of limited Jewish immigration into the country. Therefore, most of the
actual immigration into the country was, in fact, illegal. According
to historical data, a total of 67 100 World War II survivors aged
15–45 years immigrated illegally to Israel from 1946 through 1947
(30). Because only those aged 15–25 years are relevant to this study
(because the study cohort included people born in 1920 and later),
we estimated that roughly one-third of these illegal immigrants
should actually have been included in the exposed group. Therefore,
we transferred a total of 22 366 of these immigrants in 1946 and
1947 from the nonexposed group to the exposed group, equally
distributed among all relevant birth cohorts.
The study cohort is based on a dynamic population, and person-years were, therefore, used to describe the study groups. In
total, the study cohort included 4 919 700 person-years, 908 436
in the nonexposed group and 4 011 264 in the exposed group.
Ascertainment of Study Outcome
The main study outcome was the occurrence of a histologically
diagnosed malignant disease. The Israel National Cancer Registry
database was used to identify relevant study subjects with a cancer
diagnosis. The Israel National Cancer Registry was established in
1960, and since 1982 it has been compulsory that all newly diagnosed cancers in persons in Israel be reported to this registry. Data
collected by the Israel National Cancer Registry include demographic information (sex, date of birth, country of birth, date of
immigration to Israel if applicable, and date of death if applicable),
date and location of cancer diagnosis, histological type of the
malignant tumor, and disease stage at diagnosis. Completeness of
this registry is estimated at approximately 93% for solid tumors
(31).
We included all subjects diagnosed with a first primary malignant tumor, except for basal and squamous cell skin carcinomas.
Subjects with in situ melanoma or in situ breast or cervical cancers
were also included, as well as those with benign tumors of the brain
(because the prognosis of such tumors is often similar to that of
malignant tumors because of the specific anatomical site involved).
Second primary tumors were excluded. Definition and coding of
diagnoses of the various cancers were based on the International
Classification of Diseases–Oncology, Third Revision (32).
JNCI
|
Articles 1491
Statistical Analysis
Descriptive statistics were used for age, sex, and country of birth
distributions in both exposed and nonexposed groups and
expressed as means and SDs. The incidence rates were calculated
by using individual data in the numerators and aggregated data in
the denominators. Numerator data (ie, information on cancer
diagnoses in each group) were obtained from the Israel National
Cancer Registry. Denominator data were obtained for each group
from the Israel Central Bureau of Statistics for the following
5 years: 1983, 1988, 1993, 1998, and 2004. For each year, only
people still alive and living in Israel were included. These estimates were extrapolated to produce denominators for individual
calendar years, as needed. Thus, for example, cancer rates for
1988 in the exposed group were computed on the basis of the
exact number of incident cancers diagnosed for that year, which
was obtained from the Israel National Cancer Registry database
(ie, the numerator), and the population estimate for 1988, which
was obtained from the Israeli Central Bureau of Statistics (ie, the
denominator). Cancer rates for the nonexposed group for 2000
were computed on the basis of the exact number of incident
cancer diagnoses for that year, which was obtained from the Israel
National Cancer Registry database (ie, the numerator), and by
extrapolation from population estimates for 1998 and 2004, which
were obtained from the Israeli Central Bureau of Statistics (ie, the
denominator).
The study cohort was divided into five birth cohorts (1920–
1924, 1925–1929, 1930–1934, 1935–1939, and 1940–1945) to
represent the different ages at exposure. Person-years of follow-up
were calculated by sex and exposure status for each birth cohort, as
explained above.
Standardized incidence ratios and 95% confidence intervals
were computed by comparing the observed cancer rates in the
exposed subgroup with expected cancer rates in the general
European-born Jewish population of Israel, for which the rates
observed in the nonexposed group served as an approximation.
Because detailed data for the denominators required for these calculations were available for the study cohort from the Israeli
Bureau of Statistics only since 1983, standardized incidence ratios
were computed only for the period of 1983–2004. Standardized
incidence ratios were stratified by birth cohort and sex and adjusted
for time period (at 5-year intervals: 1983–1988, 1989–1993, 1994–
1999, and 2000–2004). Standardized incidence ratios and 95%
confidence intervals were computed for all cancer sites and for
selected cancer sites (colorectal, breast, prostate, and lung and
bronchial cancers) (33). For prostate cancer, standardized incidence ratios were stratified by time period into two categories,
1983–1990 and 1991–2004, by use of the date 1990—when the
prostate-specific antigen blood test for the early detection of prostate cancer was introduced in Israel—as the cut point. Prostate
cancer incidence increased by 36% from 1988 through 1991 in
Israel (1), a trend similar to that observed elsewhere. Because availability and accessibility of prostate-specific antigen testing may
have differed between the exposed and the nonexposed groups and,
hence, may affect outcomes for prostate cancer, we decided to
stratify these data by period.
To take into account the potential scatter within each birth
group, we have used exponential Poisson regression models to
1492 Articles
|
JNCI
compute relative risks (RRs) and 95% confidence intervals for all
cancer sites and for specific cancer sites, stratified by sex and birth
cohort, and adjusted for time period. Values of the relative risks and
standardized incidence ratios are similar and so only the relative
risks and 95% confidence intervals are presented in Tables 4–8.
The analyses were repeated for a subcohort of exposed participants who immigrated to Israel up to 1955 because, apart from the
large immigration wave from the former Soviet Union in 1990–
1991, a smaller one occurred in the 1970s and selection of an earlier cutoff point could have compensated for potential
misclassifications. The year 1955 was chosen because by then most
of World War II survivors who intended to immigrate to Israel
after the war had done so. Indeed, 268 212 (85%) of our eligible
study cohort of 315 544 had immigrated to Israel before 1960 (29).
This was done by assessing the expected number of deaths in the
exposed group in the time periods of 1955–1967 (calculated
through extrapolation) and 1968–2004 (given by the Israeli Central
Bureau of Statistics) and thus recalculating the corresponding
denominator.
The SAS statistical package (version 9.1) was used for statistical
analysis (34). The statistical significance level was set at a P value
of .05, and all tests were two-sided. Because this study involved no
direct linkage of existing databases and all analyses were carried
out in the Israel National Cancer Registry and the resulting data
contained no identifying details, the confidentiality of the study
participants was ensured. The study did not require an ethical
board approval and written informed consent but was approved by
the University of Haifa.
Results
General Description of the Study Population
Participants contributed a total of 4 919 700 person-years to the
follow-up: 908 436 person-years from the nonexposed group
(48.8% from men and 51.2% from women) and 4 011 264 personyears from the exposed group (43.7% from men and 56.3% from
women) (Table 1). Countries of birth with the highest numbers of
study participants were Poland (1 487 570 person-years, with 15%
of the total person-years in the nonexposed group and 34% in the
exposed group), the former Soviet Union (1 247 448 person-years,
with 14% of the total person-years in the nonexposed group and
28% in the exposed group), and Romania (892 644 person-years,
with 29% of the total person-years in the nonexposed group and
18% in the exposed group). Countries of origin with fewer study
participants included Germany and Austria (19% in the nonexposed group and 2% in the exposed group), Bulgaria (7% in the
nonexposed group and 6% in the exposed group), and
Czechoslovakia and Hungary (10% in the nonexposed group and
6% in the exposed group).
Cancer Incidence
Between January 1, 1960, and December 31, 2004, a total of 69
297 participants were diagnosed with cancer in the study cohort,
13 237 in the nonexposed group (6652 men and 6585 women) and
56 060 in the exposed group (24 773 men and 31 287 women)
(Table 2), reflecting crude incidence rates of 1457.1 and 1397.6
per 100 000 person-years in the nonexposed and the exposed
Vol. 101, Issue 21
|
November 4, 2009
Table 1. Distribution of person-years contributed by study participants by exposure status, sex, and birth cohort for the study period
1983–2004*
Nonexposed group, No. (%)
Group
Total
Birth cohort
1920–1924
1925–1929
1930–1934
1935–1939
1940–1945
Exposed group, No. (%)
Men
Women
Men
Women
444 484 (100.0)
463 952 (100.0)
1 804 978 (100.0)
2 206 286 (100.0)
158
127
102
39
17
161
132
112
40
17
052
525
040
129
738
(35.6)
(28.7)
(22.9)
(8.8)
(4.0)
125
929
007
244
647
(34.7)
(28.7)
(24.1)
(8.7)
(3.8)
397
342
326
338
399
834
575
174
512
883
(22.0)
(19.0)
(18.1)
(18.8)
(22.1)
519
492
387
377
429
343
743
166
733
301
(23.5)
(22.3)
(17.6)
(17.1)
(19.5)
* The nonexposed group included Israeli Jews who were born in Europe and immigrated to the area that would become Israel in 1948, before or during World War
II. The exposed group (ie, those potentially exposed to the Holocaust) included Israeli Jews who were born in Europe and immigrated to Israel (or the area that
would become Israel) after World War II and up to 1989.
groups, respectively. The mean age (±SD) at cancer diagnosis was
64.7 ± 10.7 years (66.4 ± 9.9 years for men and 63.0 ± 11.3 years
for women) in the nonexposed group and 62.9 ± 10.7 years (64.3 ±
10.2 years for men and 61.8 ± 11.0 years for women) in the exposed
group. For men in the birth cohort of 1920–1924, the mean age at
diagnosis in the nonexposed group was 67.3 ± 10.4 and in the
exposed groups was 67.3 ± 9.9 years. For men in the birth cohort
of 1940–1945, the mean age at diagnosis was 48.1 ± 11.4 years in
the nonexposed group and 50.7 ± 8.7 years in the exposed group.
Corresponding figures for women in the birth cohort of 1920–
1924 were 68.6 ± 9.5 years in the nonexposed group and 68.5 ± 9.1
years in the exposed group, and in the birth cohort of 1940–1045
these values were 47.1 ± 13.0 years in the nonexposed group and
50.9 ± 9.5 years in the exposed group.
The most common cancer in nonexposed men was prostate
cancer (16.8% of all malignant tumors) and the most common
cancer in exposed men was colorectal cancer (17.8% of all malignant tumors). The most common tumor in both nonexposed and
exposed women was breast cancer (30.4% and 31.4%, respectively,
of all malignant tumors) (Table 3).
Association between Exposure and Risk of Cancer
As explained earlier, only cancers diagnosed between January 1,
1983, and December 31, 2004, were used to compute risk esti-
mates for cancer. A total of 55 488 participants with cancer were
eligible (80.1% of all cancer patients), 10 282 in the nonexposed
group (5501 men and 4781 women) and 45 206 in the exposed
group (20 930 men and 24 276 women).
The risk of a cancer at any cancer site was statistically significantly higher in the exposed group than in the nonexposed group
across all birth cohorts and both sexes (ranging from RR = 1.17,
95% CI = 1.13 to 1.23, to RR = 3.50, 95% CI = 2.17 to 5.65). Risk
was considerably higher for both sexes among younger birth
cohorts than among older birth cohorts (for men born in 1940–
1945, RR = 3.50, 95% CI = 2.17 to 5.65; for women of the same
birth cohort, RR = 2.33, 95% CI = 1.69 to 3.21) (Table 4).
The risk for colorectal cancer was also statistically significantly higher in the exposed group than in the nonexposed
group across all birth cohorts and in both sexes (ranging from
RR = 1.31, 95% CI = 1.19 to 1.45, for men born in 1920–1924,
to RR = 1.75, 95% CI = 1.19 to 2.59, and RR = 1.93, 95% CI =
1.25 to 3.00, for men and women born in 1935–1939, respectively). Because of the small numbers of participants in the
youngest birth cohort and large scatter in their data (Table 5),
unstable risk estimates were obtained (data not presented). The
risk for breast cancer followed a similar pattern across all birth
cohorts (ranging from RR = 1.21, 95% CI = 1.10 to 1.33, for
those born in 1920–1924, to RR = 2.44, 95% CI = 1.46 to 4.06,
Table 2. Distribution of cancer diagnoses by exposure status, sex, and birth cohort, 1960–2004*
Nonexposed group, No.
Men
Group
No. of
diagnoses
Total
Birth cohort
1920–1924
1925–1929
1930–1934
1935–1939
1940–1945
Exposed group, No.
Women
Rate†
No. of
diagnoses
6652
1496.6
3726
1823
863
215
25
2357.5
1429.5
845.7
549.5
140.9
Men
Rate†
No. of
diagnoses
6585
1419.3
3248
1926
1105
260
46
2015.8
1448.9
986.5
646.1
260.7
Women
Rate†
No. of
diagnoses
Rate†
24 773
1372.5
31 287
1418.1
10 082
6554
4202
2448
1487
2534.2
1913.2
1288.3
723.2
371.9
11 692
8655
5141
3352
2447
2251.3
1756.5
1327.8
887.4
570.0
* The nonexposed group included Israeli Jews who were born in Europe and immigrated to the area that would become Israel in 1948, before or during World War
II. The exposed group (ie, those potentially exposed to the Holocaust) included Israeli Jews who were born in Europe and immigrated to Israel (or the area that
would become Israel) after World War II and up to 1989.
† Crude rate = number of diagnoses per 100 000 person-years.
jnci.oxfordjournals.org
JNCI
|
Articles 1493
Table 3. Distribution of cancer diagnoses by exposure status, sex, and cancer site*
Nonexposed group, No. (%)
Group
Total
Cancer site
Colorectal
Lung and bronchi
Urinary bladder
Lymphoma
Malignant melanoma
Central nervous system
Leukemia
Prostate
Breast
Corpus uteri
Ovaries
Kaposi sarcoma
Other
Exposed group, No. (%)
Men
Women
Men
Women
6652 (100.0)
6585 (100.0)
24 773 (100.0)
31 287 (100.0)
4422 (17.8)
3152 (12.7)
2330 (9.4)
1180 (4.8)
1003 (4.0)
925 (3.7)
698 (2.8)
3061 (12.4)
114 (0.5)
—
—
152 (0.6)
7736 (31.2)
4653 (14.9)
1752 (5.6)
629 (2.0)
1163 (3.7)
1205 (3.9)
1155 (3.7)
641 (2.0)
—
9817 (31.4)
1541 (4.9)
1491 (4.8)
77 (0.2)
7163 (22.9)
1099
624
629
336
450
236
189
1120
26
(16.5)
(9.4)
(9.5)
(5.1)
(6.8)
(3.5)
(2.8)
(16.8)
(0.4)
—
—
40 (0.6)
1903 (28.6)
920
384
130
283
431
252
128
(14.0)
(5.8)
(2.0)
(4.3)
(6.5)
(3.8)
(1.9)
—
2005 (30.4)
335 (5.1)
301 (4.6)
15 (2.3)
1401 (19.2)
* The nonexposed group included Israeli Jews who were born in Europe and immigrated to the area that would become Israel in 1948, before or during World War
II. The exposed group (ie, those potentially exposed to the Holocaust) included Israeli Jews who were born in Europe and immigrated to Israel (or the area that
would become Israel) after World War II and up to 1989.
for those born in 1940–1945) and was highest for those in the
youngest birth cohort (Table 6).
As previously explained, we stratified risk estimates for prostate
cancer by time period: 1983–1990 and 1991–2004 (Table 7). In the
first category 1983–1990, the risk for prostate cancer was not statistically significantly different between exposed and nonexposed
participants (except for the oldest birth cohort). However, in 1991
through 2004, the number of patients diagnosed with prostate
cancer was larger than expected in some of the birth cohorts in the
exposed group (especially in 1925–1929 and in 1930–1934) (Table 7).
The small numbers of patients in the youngest birth cohort and
the large scatter in data produced unstable estimates (data not
presented).
In the first four birth cohorts, risk for lung cancer was generally statistically significantly higher in the exposed group than in
the nonexposed group among men (ranging from RR = 1.59, 95%
CI = 1.39 to 1.83, for those born in 1920–1924, to RR = 2.27, 95%
CI = 1.89 to 2.72, for those born in 1925–1929) and among
women (ranging from RR = 1.05, 95% CI = 0.56 to 1.95, for those
born in 1940–1945, to RR = 1.93, 95% CI = 1.39 to 2.68, for those
born in 1935–1939). The small numbers of participants and large
scatter in the data in the youngest birth cohort produced unstable
estimates (data not presented). No clear trend in risk for lung
cancer was observed across birth cohorts (Table 8).
In a subanalysis, we extrapolated our data to restrict the exposed
group to people who had immigrated to Israel up to 1955. The
incidences calculated in this subanalysis showed the same trends,
including most of the dose–response trends with regard to age at
exposure, as those calculated for the total cohort, and were also all
statistically significant (data not shown).
Discussion
Assessment of cancer incidence in European-born Jews living in
Israel indicated a higher risk for all-site cancer and for specific
Table 4. Association between exposure and risk of all-site cancer,
diagnosed in 1983–2004, stratified by sex and birth cohort*
Table 5. Association between exposure and risk of colorectal
cancer, diagnosed in 1983–2004, stratified by sex and birth
cohort*
Birth cohort
Birth cohort
Men
1920–1924
1925–1929
1930–1934
1935–1939
1940–1945
Women
1920–1924
1925–1929
1930–1934
1935–1939
1940–1945
Observed No.
Expected No.
RR (95% CI)
8153
5591
3703
2197
1286
7155.00
3857.58
2320.65
1577.68
376.13
1.17
1.50
1.62
1.37
3.50
(1.13
(1.41
(1.50
(1.19
(2.17
to
to
to
to
to
1.23)
1.58)
1.76)
1.59)
5.65)
8474
6639
4164
2838
2161
6995.29
5239.30
2848.11
1835.50
933.36
1.30
1.33
1.48
1.55
2.33
(1.24
(1.26
(1.37
(1.34
(1.69
to
to
to
to
to
1.36)
1.41)
1.59)
1.79)
3.21)
* Analyses were adjusted for age and period. RR = relative risk; CI = confidence
interval.
1494 Articles
|
JNCI
Men†
1920–1924
1925–1929
1930–1934
1935–1939
Women†
1920–1924
1925–1929
1930–1934
1935–1939
Observed No.
Expected No.
RR (95% CI)
1682
1053
658
388
1341.40
705.20
359.38
213.38
1.31
1.56
1.84
1.75
(1.19
(1.37
(1.51
(1.19
to
to
to
to
1.45)
1.78)
2.24)
2.59)
1617
1151
638
374
1305.28
806.70
427.48
196.74
1.33
1.52
1.51
1.93
(1.19
(1.31
(1.25
(1.25
to
to
to
to
1.48)
1.75)
1.82)
3.00)
* This analysis was adjusted for age and period. RR = relative risk;
CI = confidence interval.
† Data for the birth cohort 1940–1945 are not presented because the small
number of cancers diagnosed did not allow calculation of the risk estimate.
Vol. 101, Issue 21
|
November 4, 2009
Table 6. Association between exposure and risk of breast cancer,
diagnosed in 1983–2004, among women, stratified by birth
cohort*
Birth cohort
Observed
No.
Expected
No.
1996
1761
1328
1124
896
1773.03
1538.04
828.51
683.48
373.60
1920–1924
1925–1929
1930–1934
1935–1939
1940–1945
RR (95% CI)
1.21
1.20
1.62
1.63
2.44
(1.10
(1.08
(1.41
(1.29
(1.46
to
to
to
to
to
1.33)
1.33)
1.86)
2.06)
4.06)
* Analyses were adjusted for age and period. RR = relative risk; CI = confidence interval.
cancers among Jewish survivors of World War II who were potentially exposed to the Holocaust than among those not exposed.
The excess risk was inversely associated with age at exposure and
was especially pronounced for breast and colorectal cancer.
The large cohort size (4 919 700 person-years) and the statistical significance of the results make the option of false results
rather unlikely. Selection bias may be inherent, however, because
the exposed subgroup was restricted to those surviving World
War II. Still, these survivors may, in fact, be more resilient than
those who did not survive World War II and so the true association might have been even stronger. Direct or indirect explanations for the increased cancer risks among the exposed group
include exposure to severe caloric restriction during World War
II, to prolonged psychological stress, and/or to long-standing
World War II–related posttraumatic stress disorder. Such exposures may have also contributed to the adoption of certain lifestyles or behaviors that have been associated with increased cancer
risks (eg, higher smoking and obesity rates). Past exposure during
World War II to various infectious diseases and poor hygienic
conditions could, at least in theory (35), also contribute to an
increased risk of cancer.
Table 7. Association between exposure and risk of prostate cancer, diagnosed in 1983–2004, among men, stratified by diagnosis
period and birth cohort*
Group
Observed
No.
Expected
No.
179
63
13
5
269.85
86.50
18.37
8.57
0.66
0.75
0.66
0.47
(0.51
(0.48
(0.23
(0.05
to
to
to
to
0.85)
1.18)
1.87)
4.18)
1050
757
529
308
1076.41
642.15
396.76
277.69
1.02
1.20
1.34
1.08
(0.91
(1.04
(1.11
(0.76
to
to
to
to
1.13)
1.37)
1.62)
1.53)
Diagnosis in
1983–1990†
Birth cohort
1920–1924
1925–1929
1930–1934
1935–1939
Diagnosis in
1991–2004†
Birth cohort
1920–1924
1925–1929
1930–1934
1935–1939
RR (95% CI)
* Analyses were adjusted for age and period. RR = relative risk; CI = confidence interval.
† Data for the birth cohort 1940–1945 are not presented because the small
number of cancers diagnosed did not allow calculation of risk estimates.
jnci.oxfordjournals.org
Table 8. Association between exposure and risk of lung and bronchial cancer, diagnosed in 1983–2004, stratified by sex and birth
cohort*
Birth cohort
Men†
1920–1924
1925–1929
1930–1934
1935–1939
Women†
1920–1924
1925–1929
1930–1934
1935–1939
Observed
No.
Expected
No.
944
755
452
263
605.79
355.18
226.69
160.79
1.59
2.27
2.04
1.66
(1.39
(1.89
(1.59
(1.04
to
to
to
to
1.83)
2.72)
2.61)
2.65)
545
447
261
121
492.91
357.34
135.80
117.16
1.15
1.35
1.93
1.05
(0.96
(1.09
(1.39
(0.56
to
to
to
to
1.37)
1.69)
2.68)
1.95)
RR (95% CI)
* Analyses were adjusted for age and period. RR = relative risk; CI = confidence interval.
† Data for the birth cohort 1940–1945 are not presented because the small
number of diagnoses did not allow calculation of risk estimates.
All-Site Cancer Incidence and Age at Exposure
Few studies (10–19) have investigated extreme calorie restriction
in humans and cancer incidence many years later. Their results are
contradictory. Three studies (10,18,19), however, indicated a
higher cancer risk among individuals exposed to severe but transient famine. The contradictory results may be partly explained by
different methodologies but may also reflect modification of the
outcome by factors such as age at exposure and the severity and
length of the exposure.
Jews who were exposed to the Holocaust during World War II
experienced especially high levels of psychological and mental
stress for extended periods (20,21,36). However, as mentioned
previously, the relationship between such stresses and subsequent
cancer incidence is still unclear (2,7).
Age at exposure modified the outcome in previous studies
(10,11,18,19) and also in this study. Among those who were
born in 1940–1945 and aged 0–5 years at exposure, all-site cancer risks were 3.50- and 2.33-fold higher than expected among
men and women, respectively. Early exposures, including antenatal ones, have previously been proposed (22,37–41) as modifiers of the individual susceptibility for future chronic morbidity.
The mechanisms involved possibly include long-term impact on
growth patterns, sensitivity of hormone receptors, basic hormonal levels, and behavioral responses that might alter longterm susceptibility to certain diseases ( 22,37–39). Age at
exposure is a known modifier of cancer incidence, as shown by
results of follow-up studies (40,41) in which associations have
been found between childhood exposure to therapeutic radiotherapy for tinea capitis, enlarged tonsils, or thymus gland and
an increased risk of cancers of the thyroid, salivary gland, central nervous system, skin, and breast, as well as leukemia.
Likewise, an excess risk of breast cancer has been reported
among scoliosis patients who had frequent diagnostic x-rays
during childhood and adolescence (40). Similar findings were
reported from a cohort of atomic bomb survivors, among whom
younger age at exposure was inversely associated with risk of a
solid malignant tumor (41).
JNCI
|
Articles 1495
Incidence of Breast Cancer
Compared with Jewish women in the nonexposed group, those in
the exposed group had statistically significantly higher risk for
breast cancer, with younger age at exposure being associated with
a statistically significantly higher risk. Known risk factors for
breast cancer include younger age at menarche, older age at menopause, infertility and low parity, higher age at first full-term pregnancy, no lactation, alcohol consumption, certain gene mutations,
and a positive family history of breast cancer (42). Women exposed
to the Holocaust, especially those exposed as children, had delayed
menarche or long periods of amenorrhea as a result of their living
conditions (43), two factors that are associated with decreased risk
of breast cancer. However, other factors may have contributed to
a higher risk for the disease, including weight changes. Weight
gain at adolescence and early adulthood has been previously proposed as a risk factor for postmenopausal breast cancer (44–48).
World War II survivors were prone to weight changes after World
War II, when they had access to more abundant food. Another
factor that may be influenced by caloric restriction is endogenous
hormone levels. Elias et al. (10) suggested that the higher susceptibility for breast cancer observed in women severely exposed to
the Dutch “hunger winter” may have been related to the abrupt
halt of the famine and the subsequent ready availability of an
unlimited food supply, which permanently and irreversibly affected
basic hormonal levels and modified the long-term risk of breast
cancer (10). Furthermore, Elias et al. (49) also showed that circulating levels of insulinlike growth factor, a hormone involved with
epithelial cells turnover and associated with higher risk for postmenopausal breast cancer (50), were statistically significantly elevated among those with the greatest exposure to the Dutch famine
(49). The amenorrhea experienced by many women who were
exposed to the War as adolescents and young adults could have
compromised their future parity and perhaps even have caused
infertility, two known risk factors for breast cancer. Finally, many
women in the exposed group who did have children had their first
full-term pregnancy at a relatively older age.
Increased alcohol consumption is associated with higher risk
for breast cancer (51). Alcohol, as is true with other addicting substances, is sometimes consumed by people suffering from posttraumatic stress disorder (52). World War II survivors are at high
risk for posttraumatic stress disorder (53), and the elevated breast
cancer risk observed in the exposed group may, at least partially, be
explained by higher alcohol consumption by some women in this
group.
Because both exposed and nonexposed groups contained only
European Jews, genetic factors, as well as specific mutations in
genes such as BRCA1 or BRCA2, cannot explain the observed
results. The question of a different gene expression that was caused
by extreme exposures, however, remains open.
Colorectal Cancer Incidence
Compared with the nonexposed group, the exposed group had a
statistically significantly higher risk for colorectal cancer, again
with younger age at exposure being associated with higher risk.
Known risk factors for colorectal cancer include older age, inflammatory bowel diseases, benign colorectal tumor, family history of
colorectal cancer, certain familial genetic diseases, obesity, and
1496 Articles
|
JNCI
nutritional factors, such as a diet lacking in fruit, vegetables, and
fiber but rich in red meat and saturated fat (54). Calorie restriction
was also suggested as a risk modifier. Dirx et al. (17), referring to
the Dutch famine, suggested that those exposed to more severe
caloric restriction during World War II were at a non-statistically
significantly lower risk for colorectal cancer. Similar findings were
reported from Norway, in which an unexpected drop in the incidence of colorectal cancer was noticed among the cohorts born
during or shortly after World War II (55), as well as in Estonia,
Sweden, and Denmark (56). Results of our study are contradictory.
Differences in the length and severity of exposures between Jewish
and non-Jewish populations during World War II, as discussed
above, may partially explain the contradiction.
Recently, lower calcium intake (57,58), lower folic acid consumption (59), and vitamin D deficiency (44,58) have been proposed as risk factors for colorectal adenomas and malignant
tumors. Inmates of ghettos and concentration camps suffered from
prolonged vitamin and mineral deficiencies because of their poorly
balanced and inadequate diet (20,21). These exposures may also
partially explain the contradictory results.
Genetic factors and specific genetic disorders should not
explain the results observed because the exposed and nonexposed
subgroups were European Jews. However, for colorectal cancer,
the question of a different gene expression caused by the extreme
exposures remains open.
Prostate Cancer Incidence
Before 1991, incidence of prostate cancer among Jewish men in
the exposed and nonexposed groups was essentially the same.
However, after 1990, the birth cohorts of 1925–1929 and 1930–
1935 in the exposed group had a higher-than-expected incidence
of prostate cancer. Known risk factors for prostate cancer include
older age and a positive family history of the disease. Nutrition and
obesity also may be risk factors for prostate cancer (60).
The higher risk for prostate cancer diagnosed in 1991–2004,
which was observed for some birth cohorts in the exposed group,
might be associated with nutritional deficiencies potentially experienced by this group during World War II. Caloric restriction has
been previously associated with a non-statistically significantly
higher risk for prostate cancer (RR = 1.30, 95% CI = 0.97 to 1.73)
(16). Lack of vitamins D and E as well as calcium and selenium
have also been reported as potential risk factors for prostate cancer
(61) and were potentially experienced by many in the exposed
group in this study. Furthermore, the modifying effect of age was
also evident.
High plasma levels of insulin-like growth factor I have been
positively associated with prostate cancer risk (62–64). In mice,
lifelong caloric restriction has been associated with lower circulating levels of insulin-like growth factor I and thus lower risk for
prostate cancer. In addition, compared with mice fed ad libitum
throughout life, calorie-restricted mice had higher plasma levels of
insulin-like growth factor I when refed (65), suggesting a permanent change in the hypothalamus–pituitary axis after the famine,
when food is again abundant. These findings are also supported by
those of Elias et al. (49). Thus, it is possible that this mechanism
would apply to men in the exposed group, especially those exposed
as adolescents.
Vol. 101, Issue 21
|
November 4, 2009
Incidence of Lung and Bronchial Cancer
We found that potential exposure to the Holocaust was associated
with a higher risk for lung and bronchial adenocarcinoma than
nonexposure. The best known risk factor for lung and bronchial
cancer is tobacco use (66). Unfortunately, we had no information
on the smoking status of the study participants, but some data support the hypothesis that smoking may be more prevalent in the
exposed group than in the nonexposed group. As mentioned earlier, persons who experienced the Holocaust are at high risk for
posttraumatic stress disorder (52,53,67). Tobacco smoking is often
used by people suffering from posttraumatic stress disorder (51).
Hapke et al. (68) found that smoking was statistically significantly
more prevalent among patients with posttraumatic stress disorder
than among those without the disorder (odds ratio [OR] = 1.28,
95% CI = 1.09 to 1.28) and that their addiction tendency was
higher (OR = 2.21, 95% CI = 1.16 to 3.90) (68). These findings
were more pronounced in men than in women (69). Interestingly,
in the study of the survivors of the Siege of Leningrad(19), higher
prevalence of past and current smoking was reported for those who
were exposed to the siege (80.9% and 20.7% in men and women,
respectively) than for those were not exposed (78.8% and 13.4% in
men and women, respectively). Thus, exposure to past traumas and
perhaps also subsequent posttraumatic stress disorder, which later
led to higher smoking rates, may have contributed to the higher
incidence of lung and bronchial cancer observed in the exposed
group.
Additionally, chronic lung diseases, such as chronic obstructive
lung disease, emphysema, and tuberculosis, have also been proposed as independent risk factors for lung cancer, after adjustment
for smoking (70). The exposed group potentially encountered
extreme living conditions and thus acquired acute and chronic lung
infections, including tuberculosis. Such infection often deteriorated into a severe and chronic condition, because neither proper
nutrition nor medical care was available, and thus may contribute
to an increased risk of lung cancer.
Advantages and Limitations
Although several other studies addressed the issue of calorie
restriction during World War II and its impact on subsequent
cancer incidence (10–19), this study is the first, to the best of our
knowledge, to refer specifically to the most exposed population at
that time—European Jews. Additionally, the study cohort is based
on all Israeli citizens who fulfill the study inclusion criteria, not
merely a sample of the reference population. The large cohort size
made the analyses of specific cancer sites possible and contributed
to the higher study power. Furthermore, follow-up was longer
than 40 years. In addition, data analyses took into account the
potential scatter in the birth cohorts by using exponential Poisson
regression models.
The study had several limitations and so caution is needed
when interpreting the findings. The study focused on World War
II survivors residing in Israel but did not include World War II
survivors living in other countries. Still, the Israeli population of
Jewish Holocaust survivors is currently the largest worldwide
(71,72).
The birth cohort of 1940–1945 was the smallest of all cohorts,
especially for the nonexposed group, and contained the few
jnci.oxfordjournals.org
people who managed to immigrate during the years of World
War II. Thus, the observed cancer incidence rates in this nonexposed group, which served as the “expected” rates for the standardized incidence ratio calculations, may have been unstable.
However, the trend observed across all birth cohorts and results
of previous publications (10,18,19,22–25) addressing the issue of
the modifying effect of age at exposure indicate that the
“expected” rates used for the youngest birth cohort were not
heavily biased.
The exposure in this study (ie, living as a Jew in Europe under
the Nazi regime during World War II) was not measured
directly but rather was assessed with a proxy variable that was
based on the date of immigration (before 1939 or after 1945).
The Jewish population in pre–World War II Europe included
more than 8 500 000 individuals, with more than 3 000 000 living in Poland alone (Appendix Table 1) (73). According to the
Molotov–Ribbentrop Pact (August 1939), a total of 1 200 000
Polish Jews became Soviet citizens after the Nazi occupation of
Poland in September 1939. Additionally, up to 300 000 Polish
Jews moved into the Soviet parts of Poland between September
1939 and February 1940. Some returned to Poland after a short
stay and others found themselves under Nazi occupation after
the German invasion of the former Soviet Union in June 1941
(73) but many stayed in the Soviet areas. Other European Jews
may have also been relatively protected during World War II
because of their successful escape from occupied territories to
nonoccupied countries in Europe, such as the United Kingdom,
Sweden, Spain, Portugal, or Switzerland. Because many of these
people immigrated to Israel when the War was over, some of
those defined as exposed may, in fact, have been misclassified
and should have been defined as nonexposed. However, because
this misclassification is not expected to be differential, the estimated standardized incidence ratios may be attenuated and thus
actually be larger.
The outcome in this study—cancer incidence—was ascertained
since 1960 when the Israel National Cancer Registry was established. Participants who were diagnosed with cancer before this
date are not included. This practice may have also caused a misclassification, especially for childhood cancer in the youngest birth
cohort. However, childhood cancers diagnosed among Israeli citizens account for no more than 1% of all cancers diagnosed each
year. Additionally, when the Israel National Cancer Registry was
established in 1960, the youngest birth cohort (1940–1945) would
have included individuals 15–20 years old, participants in that
cohort could still have been diagnosed with childhood cancer, and
their data could have been contributed to the Israel National
Cancer Registry. Finally, no individual data were available concerning other exposures that might have contributed to a higher
cancer incidence in World War II survivors, such as smoking or
alcohol consumption.
Conclusions
Jewish survivors of World War II who were potentially exposed to
the Holocaust were at a higher risk for cancer occurrence later on
in life than those not exposed. The risk appears to be modified by
age at exposure, with younger age at exposure being associated
with higher risk. These observations may have direct impact on the
JNCI
|
Articles 1497
health of World War II Jewish survivors and thus the care required
from their caregivers in Israel and elsewhere. Consequently, better
procedures for the early detection of all types of cancer are warranted. Furthermore, data from the exposed and nonexposed
groups in this study may be a valuable contribution to the etiologic
research of cancer.
These findings warrant further epidemiological studies (such as
case–control studies) of past and present risk factors that use individual data. Time is a key element in such research, because the
population of Jewish survivors of World War II is aging.
Appendix Table 1. Numbers of European Jews before and after
World War II (73)
Country
No. in 1939
The Baltic states
Belarus
Belgium
Bulgaria
The Czech Republic
Denmark
Finland
France
Germany and Austria
Greece
Hungary
Italy
Luxembourg
The Netherlands
Norway
Poland
Romania
Russia*
Slovakia
Ukraine*
Yugoslavia
Total
253 000
375 000
65 000
64 000
90 000
8000
2000
350 000
240 000
70 000
650 000
40 000
5 000
140 000
1800
3 300 000
600 000
975 000
90 000
1 500 000
43 000
8 861 800
No. killed or perished
during World War II (%)
228
245
40
14
80
000
000
000
000
000
—
—
90 000
210 000
54 000
450 000
8 000
1 000
105 000
900
3 000 000
300 000
107 000
75 000
900 000
26 000
5 933 900
(90)
(65)
(60)
(22)
(89)
(26)
(90)
(77)
(70)
(20)
(20)
(75)
(50)
(90)
(50)
(11)
(83)
(60)
(60)
(67)
* Not fully occupied by Nazi Germany during World War II.
References
1. National Cancer Registry. Annual Reports: cancer incidence by gender
and place of birth among Jews: trend 1980–2002. Israel National Cancer
Registration. http://www.health.gov.il. Accessed August 2009.
2. Dalton SO, Boesen EH, Ross L, Schapiro IR, Johansen C. Mind and
cancer: do psychological factors cause cancer? Eur J Cancer. 2002;
38(10):1313–1323.
3. Metcalfe C, Davey Smith G, Macleod J, Hart C. The role of self-reported
stress in the development of breast and prostate cancer: a prospective
cohort study of employed males and females with 30 years of follow up.
Eur J Cancer. 2007;43(6):1060–1065.
4. Helgesson O, Cabrera C, Lapidus L, Bengtsson C, Lissner L. Selfreported stress levels predict subsequent breast cancer in a cohort of
Swedish women. Eur J Cancer Prev. 2003;12(5):377–381.
5. Nielsen NR, Kristensen TS, Zhang ZF, Strandberg-Larsen K, Schnohr P,
Gronbaek M. Sociodemographic status, stress and risk of prostate cancer.
A prospective cohort study. Ann Epidemiol. 2007;17(7):498–502.
6. Wakai K, Kojima M, Nishio K, et al. Psychological attitudes and risk of
breast cancer in Japan: a prospective study. Cancer Causes Control. 2007;
18(3):259–267.
7. Nielsen NR, Gronbaek M. Stress and breast cancer: a systematic update
on the current knowledge. Nat Clin Practice. 2006;3(11):612–620.
1498 Articles
|
JNCI
8. Holmes TH, Rahe RH. The Social Readjustment Rating Scale. J
Psychosom Res. 1967;11(2):213–218.
9. Hursting SD, Lavigne JA, Berrigan D, Perkins SN, Barrett JC. Calorie
restriction, aging, and cancer prevention: mechanisms of action and applicability to humans. Annu Rev Med. 2003;54:131–152.
10. Elias SG, Peeters PHM, Grobbee DE, van Noord PAH. Breast cancer risk
after caloric restriction during the 1944–1945 Dutch famine. J Natl Cancer
Inst. 2004;96(7):539–545.
11. Vatten LJ, Kvinnsland S. Body height and risk of breast cancer. A prospective study of 23,831 Norwegian women. Br J Cancer. 1990;61(6):
881–885.
12. Nilsen TI, Vatten LJ. Adult height and risk of breast cancer: a possible
effect of early nutrition. Br J Cancer. 2001;85(7):959–961.
13. Tretli S, Gaard M. Lifestyle changes during adolescence and risk of breast
cancer: an ecologic study of the effect of World War II in Norway. Cancer
Causes Control. 1996;7(5):507–512.
14. Robsahm TE, Tretli S. Breast cancer incidence in food vs non-food producing areas in Norway: possible beneficial effects of World War II. Br J
Cancer. 2002;86(3):362–366.
15. Dirx MJ, van den Brandt PA, Goldbohm RA, Lumey LH. Diet in adolescence and the risk of breast cancer: results of the Netherlands cohort
study. Cancer Causes Control. 1999;10(3):189–199.
16. Dirx MJ, van den Brandt PA, Goldbohm RA, Lumey LH. Energy restriction in childhood and adolescence and risk of prostate cancer: results from
the Netherlands Cohort Study. Am J Epidemiol. 2001;154(6):530–537.
17. Dirx MJ, van den Brandt PA, Goldbohm RA, Lumey LH. Energy restriction early in life and colon carcinoma risk: results of The Netherlands
Cohort Study after 7.3 years of follow-up. Cancer. 2003;97(1):46–55.
18. Fentiman IS, Allen DS, Ellison GTH. The impact of the occupation of
Guernsey 1940–1945 on breast cancer risk factors and incidence. Int J Clin
Pract. 2007;61(6):937–943.
19. Koupil I, Plavinskaja S, Parfenova N, Shestov DB, Danziger PD, Vagero
D. Cancer mortality in women and men who survived the siege of
Leningrad (1941–1944). Int J Cancer. 2009;124(6):1416–1421.
20. Shasha SM. Morbidity in the concentration camps of the Third Reich
[in Hebrew]. Harefuah. 2004;143(4):272–276,318.
21. Shasha SM. Morbidity in the ghettos during the Holocaust [in Hebrew].
Harefuah. 2002;141(4):364–368.
22. Okasha M, McCarron P, Gunnell D, Smith GD. Exposures in childhood,
adolescence and early adulthood and breast cancer risk: a systematic
review of the literature. Breast Cancer Res Treat. 2003;78(2):223–276.
23. Barton HA, Cogliano VJ, Flowers L, Valcovic L, Setzer RW, Woodruff
TJ. Assessing susceptibility form early-life exposure to carcinogens.
Environmental Health Perspect. 2005;113(9):1125–1133.
24. English DR, Milne E, Simpson JA. Ultraviolet radiation at places of residence and the development of melanocytic nevi in children (Australia).
Cancer Causes Control. 2006;17(1):103–107.
25. Hill DA, Preston-Martin S, Ross RK, Bernstien L. Medical radiation, family history of cancer, and benign breast disease in relation to breast cancer
risk in young women, USA. Cancer Causes Control. 2002;13(8):711–718.
26. Little JB. Ionizing radiation. In: Kufe DW, Pollock PE, Weichselbaum
RR, et al., eds. Cancer Medicine. 6th ed. London: BC Decker; 2003:
289–302.
27. Preston RJ. Children as sensitive subpopulation for the risk assessment
process. Toxicol Appl Pharmacol. 2004;199(2):132–141.
28. Whiteman DC, Whiteman CA, Green AC. Childhood sun exposure as a
risk factor for melanoma: a systematic review of epidemiologic studies.
Cancer Causes Control. 2001;12(1):69–82.
29. Yeshula V. Compensations and other payments Holocaust survivors in
Israel are entitled for. A report presented in the Knesset [Israeli
Parliament], May 2001, Israel [in Hebrew].
30. Gavzah-Braverman N, ed. History of the Jewish Settlement in Israel and the
Illegal Immigration in 1943–1948 [in Hebrew]. Jerusalem, Israel: National
Publication Service; 1984.
31. Fishler Y, Shetrit A, Barchana M, Modan B. Assessment of the Completeness
of the Israel Cancer Registry Database—Methods and Findings [in Hebrew].
Israel Center for Disease Control; 2003. www.health.gov.il. Accessed July
2008.
Vol. 101, Issue 21
|
November 4, 2009
32. Fritz A. ICD-O International Classification of Diseases for Oncology. 3rd ed.
Geneva: WHO; 2001.
33. Rothman KJ, Greenland S. Modern Epidemiology. 2nd ed. Philadelphia,
PA: Lippincott, Williams & Wilkins; 1998:63.
34. SAS Statistic software version 9.1. Heidelberg, Germany: SAS
International. http://www.sas.com.
35. Ziegler JL, Buonaguro FM. Infectious agents and human malignancies.
Front Biosci. 2009;14:3455–3464.
36. Trappler B, Cohen CI, Tulloo R. Impact of early lifetime trauma in later
life: depression among Holocaust survivors 60 years after the liberation of
Auschwitz. Am J Geriatr Psychiatry. 2007;15(1):79–83.
37. Lucas A. Programming by early nutrition: an experimental approach.
J Nutr. 1998;128(2):401S–406S.
38. Sallout B, Walker M. The fetal origin on adult diseases. J Obstet Gynaecol.
2003;23(5):555–560.
39. Ozanne SE, Fernandez-Twinn D, Hales CN. Fetal growth and adult diseases. Semin Perinatol. 2004;28(1):81–87.
40. Ron E. Cancer risks from medical radiation. Health Phys. 2003;85(1):
47–59.
41. Preston DL, Pierce DA, Shimizu Y, Ron E, Mabuchi K. Dose response
and temporal patterns of radiation-associated solid cancer risks. Health
Phys. 2003;85(1):43–46.
42. Key TJ, Verkasalo PK, Banks E. Epidemiology of breast cancer. Lancet
Oncol. 2001;2(3):133–140.
43. Pasternak A, Brooks PG. The long-term effects of the holocaust on the
reproductive function of female survivors. J Minim Invasive Gynecol.
2007;14(2):211–217.
44. Clinton SK, Giovannucci EL, Miller EC. Nutrition in the etiology and
prevention of cancer. In: Kufe DW, Pollock RE, et al., eds. Cancer and
Medicine. London: DC Decker; 2003:397–412.
45. Eng SM, Gammon MD, Terry MB, et al. Body size changes in relation to
postmenopausal breast cancer among women on Long Island, New York.
Am J Epidemiol. 2005;162(3):229–237.
46. Radimer KL, Ballard-Barbash R, Miller JS, et al. Weight change and the
risk of late-onset breast cancer in the original Framingham cohort. Nutr
Cancer. 2004;49(1):7–13.
47. Lahmann PH, Schulz M, Hoffmann K, et al. Long-term weight change
and breast cancer risk: the European prospective investigation into cancer
and nutrition (EPIC). Br J Cancer. 2005;93(5):582–589.
48. Harvie M, Howell A, Vierkant RA, et al. Association of gain and loss of
weight before and after menopause with risk of postmenopausal breast
cancer in Iowa Women’s Health Study. Cancer Epidemiol Biomarkers Prev.
2005;14(3):656–661.
49. Elias SG, Keinan-Boker L, Peeters PHM, et al. Long-term consequences
of the 1944–1945 Dutch famine on the insulin-like growth factor axis. Int
J Cancer. 2004;108(4):628–630.
50. Rinaldi S, Peeters PH, Berrino F, et al. IGF-I, IGFBP-3 and breast cancer
risk in women: the European Prospective Investigation into Cancer and
Nutrition (EPIC). Endocr Relat Cancer. 2006;13(2):593–605.
51. Smith-Warner SA, Spiegelman D, Yaun SS, et al. Alcohol and breast
cancer in women: a pooled analysis of cohort studies. JAMA. 1998;
279(7):535–540.
52. Jacobsen LK, Southwick SM, Kosten TR. Substance use disorders in
patients with post traumatic stress disorder: a review of the literature. Am
J Psychiatry. 2001;158(8):1184–1190.
53. Joffe C, Brodaty H, Luscombe G, Ehrlich F. The Sydney Holocaust
study: posttraumatic stress disorder and other psychosocial morbidity in
an aged community sample. J Trauma Stress. 2003;16(1):39–47.
54. World Cancer Research Fund/American Institute for Cancer Research.
Colon and rectum. In: Marmot M, Atinmo T, Byers T, et al., eds. Food,
Nutrition, Physical Activity and the Prevention of Cancer: a Global Perspective.
Washington, DC: AICR; 2007:280–288.
55. Svensson E, Grotmol T, Hoff G, Langmark F, Norstein J, Tretli S. Trends
in colorectal cancer incidence in Norway by gender and anatomic site: an
age-period-cohort analysis. Eur J Cancer Prev. 2002;11(5):489–495.
56. Svensson E, Møller B, Tretli S, et al. Early life events and later risk of
colorectal cancer: age-period-cohort modelling in the Nordic countries
and Estonia. Cancer Causes Control. 2005;16(3):215–223.
jnci.oxfordjournals.org
57. Slattery ML, Neuhausen SL, Hoffman M, et al. Dietary calcium, vitamin
D, VDR genotypes and colorectal cancer. Int J Cancer. 2004;111(5):
750–756.
58. Hartman TJ, Albert PS, Snyder K, et al. The association of calcium and
vitamin D with the risk of colorectal adenomas. J Nutr. 2005;135(2):
252–259.
59. Giovannucci E. Epidemiologic studies of folate and colorectal neoplasia: a
review. J Nutr. 2002;132(8):2350S–2355S.
60. Hsing AW, Chokkalingam AP. Prostate cancer epidemiology. Front Biosci.
2006;11:1388–1413.
61. Sonn GA, Aronson W, Litwin MS. Impact of diet on prostate cancer: a
review. Prostate Cancer Prostatic Dis. 2005;8(4):304–310.
62. LeRoith D, Roberts CT. The insulin-like growth factor system and
cancer. Cancer Lett. 2003;195(2):127–137.
63. Stattin P, Rinaldi S, Biessy C, Stenman UH, Hallmans G. High levels of
circulating insulin-like growth factor-I increase prostate cancer risk: a
prospective study in a population-based nonscreened cohort. J Clin Oncol.
2004;22(15):3104–3112.
64. Platz EA. Energy imbalance and prostate cancer. J Nutr. 2002;132(11):
S3471–S3481.
65. Reed MJ, Penn PE, Li Y, et al. Enhanced cell proliferation and biosynthesis mediate improved wound repair in refed, caloric-restricted mice. Mech
Ageing Dev. 1996;89(1):21–43.
66. Khuder SA. Effect of cigarette smoking on major histological types of lung
cancer: a meta-analysis. Lung Cancer. 2001;31(2–3):139–148.
67. Collins C, Burazeri G, Gofin J, Kark JD. Health status and mortality in
Holocaust survivors living in Jerusalem 40–50 years later. J Trauma Stress.
2004;17(5):403–411.
68. Hapke U, Schumann A, Rumpf HJ, John U, Konerding U, Meyer C.
Association of smoking and nicotine dependence with trauma and posttraumatic stress disorder in a general population sample. J Nerv Ment Dis.
2005;193(12):843–846.
69. Thorndike FP, Wernicke R, Pearlman MY, Haaga DAF. Nicotine dependence, PTSD symptoms, and depression proneness among male and
female smokers. Addict Behav. 2006;31(2):223–231.
70. Littman AJ, Thornquist MD, White E, Jackson LA, Goodman GE,
Vaughan TL. Prior lung disease and risk of lung cancer in a large prospective study. Cancer Causes Control. 2004;15(8):819–827.
71. Ukeles Associates Inc. An Estimate of the Current Distribution of Jewish
Victims of Nazi Persecution. New York: International Commission on
Holocaust Era Insurance Claims; 2003.
72. DellaPergola S. Review of Relevant Demographic Information on World Jewry.
HESHVAN 5764. Jerusalem, Israel: International Commission on
Holocaust Era Insurance Claims; 2003.
73. Dawidowich LS. The War against the Jews 1933–1945 [in Hebrew]. Tel
Aviv, Israel: Zmora Bitan; 1982.
Funding
Israel Cancer Association (20050122, 2005).
Notes
L. Keinan-Boker wrote the manuscript. She assisted N. Vin-Raviv with the
analysis, writing, and submission of her MPH thesis, on which this article is
based. She with M. Barchana and S. Linn are currently supervising the PhD
dissertation of N. Vin-Raviv, which focuses on individual research of these
findings. N. Vin-Raviv was involved with data acquisition and verification
and assisted M. Barchana with formulating the main study idea and hypotheses. She carried out the work described in this article, including some of the
statistical analyses, as part of her MPH thesis. She assisted L. Keinan-Boker
with the writing of the manuscript. I. Liphshitz carried out the main statistical
analyses described in this study. S. Linn supervised N. Vin-Raviv on her MPH
thesis. S. Linn assisted L. Keinan-Boker with the writing of the manuscript.
M. Barchana formulated the main study idea and hypotheses and supervised the
MPH thesis of N. Vin-Raviv. M. Barchana, N. Vin-Raviv, and S. Linn were
involved with formulation of the main study idea and hypotheses and supervised
the MPH thesis of N. Vin-Raviv. M. Barchana assisted L. Keinan-Boker with
the writing of the manuscript.
JNCI
|
Articles 1499
The authors had full responsibility for the design of the study; the collection, analysis, and interpretation of the data; the decision to submit the manuscript for publication; and the writing of the manuscript.
The authors would like to thank Dr. Ilya Novikov from the Biostatistical
Unit in the Gertner Institute, the Sheba Medical Center, Israel, for his assistance with the final statistical analyses.
1500 Articles
|
JNCI
This work is based on the MPH thesis of Neomi Vin-Raviv, which
was submitted to the School of Public Health, the University of Haifa,
Haifa, Israel, as partial fulfillment of the requirements for her master’s
degree.
Manuscript received August 15, 2008; revised July 28, 2009; accepted
August 7, 2009.
Vol. 101, Issue 21
|
November 4, 2009