Abstract
The association of dietary fat intake with ovarian cancer risk has been inconsistent across populations. We examined dietary fat intake, overall and by type, and ovarian cancer risk in two prospective cohort studies. We assessed long-term dietary fat intake among Nurses’ Health Study (NHS) and NHSII participants using food frequency questionnaires administered every 2–4 years beginning in 1984 and 1991, respectively. We examined cumulative energy-adjusted intake of total fat, specific types of fat (animal, vegetable, saturated, monounsaturated, polyunsaturated, trans fat), and cholesterol. We identified 700 ovarian cancer cases in NHS and 196 in NHSII with dietary information. Cox proportional hazards regression was used to estimate associations between intake and ovarian cancer risk. Dietary fat intake changed over time in both cohorts and was lower in NHS than NHSII. Higher cumulative average intakes of animal fat and cholesterol were significantly positively associated with risk of ovarian cancer in NHS (relative risk [RR] comparing extreme quartiles=1.57, 95%CI: 1.20,2.06 and 1.35, 95%CI: 1.08,1.69, respectively), but not in NHSII. Other dietary fat sources were not clearly associated with risk in either population. We did not observe clear associations between dietary fat and ovarian cancer risk in two large prospective cohort studies.
Keywords: Dietary fat, ovarian cancer, prospective cohort
Introduction
In 2017, approximately 22,440 cases of ovarian cancer were diagnosed in the U.S. and over 14,000 women died of the disease.1 It is the most fatal gynecologic malignancy due to clinical presentation at advanced stages and poor prognosis following diagnosis.1 There are few known preventive factors for ovarian cancer; of those that have been identified, many cannot be recommended on a population level. Due to its high fatality as well as lack of prevention recommendations and early detection, identifying modifiable risk factors, such as diet, for ovarian cancer is critical for reducing incidence and mortality.
Specifically, dietary fat is hypothesized to alter ovarian cancer risk by increasing circulating estrogen levels, which increases cell proliferation.2 Further, animal studies have observed that female offspring of mice fed high-fat diets in pregnancy had more reproductive tumors than the offspring of mice fed low-fat diets.3 In the Women’s Health Initiative randomized controlled trial, a low-fat/high fruit, vegetable and grain dietary intervention compared with usual diet was associated with a statistically significant reduction in ovarian cancer risk in the last 4 years of follow-up (HR=0.60; 95%CI 0.38–0.96).4 However, results from observational studies have been inconsistent. While several case-control studies reported an increased risk of ovarian cancer with higher dietary intakes of animal, trans and saturated fat5–8, others have reported no association.9–11 In one case-control study there was a reported decrease risk in ovarian cancer with higher intake of polyunsaturated fatty acids.8 In a meta-analysis of over 6,000 women from 8 studies (7 case-control studies, 1 cohort study), there was an increased risk of ovarian cancer for high vs. low intake of total (RR=1.24, 95%CI=1.07–1.43), saturated (RR=1.20, 95%CI 1.04–1.39), and animal fat (RR=1.70, 95%CI 1.43–2.03).12 As diet was assessed after diagnosis in the case-control studies, there is concern that results may have been influenced by recall bias.
While several prospective cohort studies, in which diet is assessed prior to diagnosis, have examined dietary fat and ovarian cancer risk, results from these studies have been mixed.13–18 In a pooled analysis of 12 cohorts, including the Nurses’ Health Study (NHS) and NHSII, there was no association between baseline dietary fat intake and risk, although there were suggestive positive associations for animal and saturated fat across extreme deciles of intake.15 In a more recent analysis in Europe, researchers observed a positive association between baseline polyunsaturated fat and invasive ovarian cancer risk (HR comparing extreme quartiles = 1.22, 95% CI = 1.02–1.48, p-trend = 0.02), but no association with other types of fat.16 A limitation of prior cohort analyses was the use of only a baseline food frequency questionnaire (FFQ), which may not represent long-term intake given temporal changes in diet. Therefore, utilizing repeated FFQs, we examined the associations between long-term dietary fat intake and ovarian cancer risk among participants in two large prospective cohorts, the NHS and NHSII.
Materials & Methods
Study population
The NHS was initiated in 1976 when 121,700 female registered nurses, 30–55 years of age and residing in 11 U.S. states, completed an initial questionnaire. NHS women have completed biennially mailed questionnaires to update information on various reproductive, lifestyle, and anthropometric factors as well as incident diseases. Deaths usually are reported by family or postal authorities. In addition, non-responders are identified via the National Death Index, which is sensitive and specific.19 Extensive information has been collected on diet as well as known and putative ovarian cancer risk factors. The NHSII began in 1989, with 116,429 female registered nurses, aged 25–42, from 14 states. The questionnaires and follow-up are similar to those in the NHS. Response rates in the two cohorts have been between 85–90% at each cycle. The Institutional Review Boards at the Brigham and Women’s Hospital and Harvard T.H. Chan School of Public Health approved this study.
Diet assessment
Adult diet intake was measured by a self-administered, semi-quantitative, ~130-item FFQ. In the NHS, this version of the FFQ was first asked in 1984, 1986, and every four years afterwards. In NHSII, the FFQ was first administered in 1991 and every 4 years thereafter. For each food item on the FFQ, women were asked to choose from 9 possible intake frequencies for a given portion size, from never to ≥6 servings per day. Nutrient intakes were calculated by multiplying the frequency by the nutrient composition of the specified portion size for each food, and then summing across all foods. Further, women provided information on types of fat used for cooking and baking (i.e., frequency of eating fried or sautéed food at home, kind of fat usually used for frying and sautéing, kind of fat usually used for baking, and brand and type of cooking oil usually used at home), and the brand and type of margarine or spread usually used; this information was incorporated into nutrient intake calculations. Due to food composition and fortification changes over time, the food composition database was updated over follow-up. All measures were adjusted for total energy intake. The validity of our FFQ was evaluated by comparison with diet records in 632 women from the Nurses’ Health Studies; the correlation for deattenuated, energy-adjusted total fat was 0.67.20
Ovarian cancer assessment
On the biennial questionnaires, NHS and NHSII women indicated if they had been diagnosed with incident ovarian cancer. For each participant who reported a new ovarian cancer diagnosis, or for diagnoses identified through death certificates, we requested medical records and pathology reports from the participant or next of kin. A gynecologic pathologist, blinded to diet and other risk factors, reviewed the records to confirm the diagnosis and abstract information on the tumor, including invasiveness and histotype. For a subset of 215 ovarian cancer cases, we compared the histologic type abstracted from the pathology report with a standardized review of pathology slides completed by a gynecologic pathologist. The concordance was 98 percent for invasiveness and 83 percent for histology. Therefore, we used histologic type from the medical record for all cases.
Statistical analysis
Participants accrued person-time from the return date of the 1984 (NHS) or 1991 (NHSII) questionnaire until the date of ovarian cancer diagnosis, diagnosis of any other cancer (except non-melanoma skin cancer), bilateral oophorectomy, pelvic irradiation, death, or the end of follow-up (2012 in NHS and 2011 in NHSII). At baseline, we excluded women with cancer other than non-melanoma skin cancer (NHS N=7,100, NHSII N=1,621), bilateral oophorectomy (NHS N=13,326, NHSII N=2,898), or menopause due to pelvic irradiation (NHS N=89, NHSII N=25) as well as women who died prior to start of follow-up for this analysis (NHS N=1,045, NHSII N=35).
For our primary analysis, we calculated the cumulative average intakes for energy-adjusted dietary total fat, animal fat, vegetable fat, saturated fat, monounsaturated fat, polyunsaturated fat, trans-unsaturated fat, and cholesterol by averaging all intake values from prior FFQs. For any 2-year period in which a FFQ was not included in the biennial questionnaire, we carried forward cumulative average dietary intake. For example, in the NHS the average values from the 1984 and 1986 FFQs were used for 1986–90 follow-up period, the average values from the 1984, 1986, and 1990 FFQs were used for 1990–94 follow-up period, and so on. Due to changes in dietary fat intake over time, we censored participants who did not complete a FFQ until they completed a subsequent FFQ, at which point they re-entered the analysis. Thus, for example, a participant who completed FFQs in 1984 and 1990, but not in 1986, contributed person-time to the 1984–1986 and 1990–1994 follow-up periods, but not the 1986–1990 follow-up period. In secondary analyses, we also examined baseline intake (NHS=1984, NHSII=1991) and the most recent report of dietary intake (e.g., in NHS: 1986 for the 1986–1990 time period, 1990 for the 1990–1994 time period, and so on). Dietary intake was categorized into quartiles based on the distribution of intake in the NHS and NHSII combined. Further, we examined the trends across quartiles of intake by modeling the median values of each quartile as a continuous variable.
We used Cox regression with time-dependent covariates stratified by age and time period to estimate relative risks (RR) and 95% confidence intervals (CIs). We adjusted for established or probable risk factors for ovarian cancer including, family history of ovarian cancer (no, yes), age at menarche (continuous), tubal ligation (no, yes), hysterectomy (no, yes, unknown), unilateral oophorectomy (no, yes, unknown), body mass index (BMI) (kg/m2, continuous), nulliparity (no, yes), parity (continuous), oral contraceptive use (never, <1, 1–5, 5+years), smoking (pack-years, continuous), menopausal status (premenopausal, postmenopausal), estrogen only postmenopausal hormone therapy (HT) use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), estrogen plus progestin HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), other HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), total caloric intake (continuous), lactose intake (g/day, continuous), caffeine intake (mg/day, continuous), fiber intake (g/day, continuous), and physical activity (MET-hrs/week: 0–3.8, >3.8–11.1, >11.1–26.4, >26.4, unknown). In addition, vegetable fat and animal fat were mutually adjusted for each other and saturated fat, polyunsaturated fat, monounsaturated fat, and trans-unsaturated fat were mutually adjusted for each other. All analyses were conducted separately in each cohort as p-values for cohort by dietary fat cross-product terms were less than 0.05 for many dietary fat types. In secondary analysis, we further stratified by menopausal status at diagnosis (premenopausal, postmenopausal) and body mass index at diagnosis (<25 kg/m2, ≥25 kg/m2), and examined the associations by histology (serous/poorly differentiated, non-serous [mucinous, endometrioid, and clear cell]). In analyses stratified by menopausal status, dietary intake was averaged over both premenopausal and postmenopausal periods. To assess the association of red meat and pork, processed meat, poultry, seafood and eggs with ovarian cancer risk, we examined cumulative average intake as described above in our primary analysis. We considered a two-sided p-value of less than 0.05 to be statistically significant and used SAS version 9.4 (SAS Institute, Cary, NC, USA) for all analyses.
Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Results
Throughout follow-up, we identified 700 cases of ovarian cancer in the NHS and 196 in the NHSII. Participant characteristics at the midpoint of follow-up are presented in Table 1 by total fat intake and cohort. Overall, women in the NHS were older and more likely to be postmenopausal than women in the NHSII. Across both cohorts, on average, women who consumed more dietary fat had higher BMI, higher caffeine intake, and lower levels of both physical activity and fiber intake. They were also more likely to be parous and to have undergone a tubal ligation. Throughout follow-up, fat intake changed over time in both the NHS and NHSII (Figure 1). In both populations, total fat, animal fat, and vegetable fat intake decreased over time until 1994 at which point, total fat intake and vegetable fat intake increased over follow-up whereas animal fat intake continued to decrease over time. However, NHSII women reported greater intakes of dietary fat compared to NHS women throughout follow-up. While no significant difference between NHS and NHSII with saturated fat, monounsaturated fat, polyunsaturated fat, trans fat and cholesterol intakes were observed (Supplemental Figure 1), we observed a decreasing trend in trans fat and cholesterol intake over time, as well as a decrease in polyunsaturated fat intake until 1994 at which time intake began increasing.
Table 1.
Selected characteristics by quartile of cumulative average energy-adjusted total fat intake and by cohort in 1996(NHS)/1997(NHSII)
| NHS | NHSII | |||||||
|---|---|---|---|---|---|---|---|---|
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | |
| (N=18,370) | (N=17,263) | (N=14,189) | (N=7,972) | (N=15,284) | (N=16,433) | (N=19,531) | (N=25,801) | |
| Mean (SD) | ||||||||
| Age (years) | 63.4(7.1) | 62.2(7.1) | 61.5(7.1) | 61.0(6.9) | 42.5(4.7) | 42.4(4.6) | 42.5(4.6) | 42.7(4.6) |
| BMI (kg/m2) | 25.7(4.6) | 26.4(4.9) | 26.9(5.2) | 27.5(5.6) | 24.8(5.1) | 25.5(5.4) | 26.0(5.7) | 26.9(6.4) |
| Age at menarche (years) | 12.5(1.4) | 12.6(1.4) | 12.6(1.4) | 12.5(1.4) | 12.4(1.4) | 12.4(1.4) | 12.4(1.4) | 12.4(1.4) |
| Parity^ | 3.1(1.5) | 3.2(1.5) | 3.2(1.5) | 3.2(1.5) | 2.2(0.9) | 2.3(0.9) | 2.3(0.9) | 2.3(0.9) |
| Pack years smoking | 12.3(18.6) | 13.0(19.1) | 14.1(20.5) | 17.8(22.9) | 6.8(9.8) | 6.8(9.8) | 6.9(9.9) | 7.4(10.0) |
| Calories (per day) | 1,723(445) | 1,766(438) | 1,771(448) | 1,711(472) | 1,765(484) | 1,821(485) | 1,828(495) | 1,795(505) |
| Total fat intake (g/day) | 46.0(5.3) | 55.4(1.9) | 61.7(1.9) | 70.1(4.5) | 45.8(5.4) | 55.6(1.9) | 62(1.9) | 72.4(5.8) |
| Animal fat intake (g/day) | 24.8(5.5) | 30.7(4.9) | 34.4(5.5) | 39.4(7.5) | 24.4(6.2) | 31.1(5.2) | 35.3(5.5) | 42(7.6) |
| Vegetable fat intake (g/day) | 21.2(4.8) | 24.7(4.8) | 27.3(5.4) | 30.7(7.3) | 21.5(5.2) | 24.5(5.1) | 26.7(5.4) | 30.4(7.1) |
| Saturated fat intake (g/day) | 15.5(2.6) | 19.2(1.9) | 21.6(2.1) | 24.7(3.1) | 15.6(2.6) | 19.4(1.9) | 21.9(2.1) | 25.9(3.3) |
| Monounsaturated fat intake (g/day) | 17.2(2.3) | 21(1.3) | 23.5(1.4) | 26.9(2.6) | 17.2(2.5) | 21.2(1.4) | 23.8(1.5) | 28.1(2.9) |
| Polyunsaturated intake (g/day) | 8.9(1.6) | 10.3(1.6) | 11.2(1.8) | 12.5(2.4) | 8.7(1.5) | 9.8(1.5) | 10.6(1.7) | 11.9(2.2) |
| Trans fat intake (g/day) | 2.1(0.6) | 2.6(0.6) | 3.0(0.7) | 3.4(0.8) | 2.1(0.6) | 2.6(0.7) | 3(0.7) | 3.7(1) |
| Cholesterol intake (g/day) | 204.2(53) | 235.4(49.9) | 252.2(55) | 274.3(69.3) | 189.1(55.2) | 222.2(47.9) | 240.4(49.5) | 266.3(57.3) |
| Lactose intake (g/day) | 16.5(10.3) | 13.8(8.5) | 11.8(7.6) | 9.9(7.3) | 21.6(13.3) | 19.4(11.5) | 17.5(10.7) | 14.4(9.8) |
| Caffeine intake (mg/day) | 234.5(180.9) | 272.3(185) | 298.6(197.3) | 333.9(218.7) | 215.4(193.7) | 225.4(193.8) | 239.4(198.6) | 267.7(218.3) |
| Fiber intake (g/day) | 20.6(5.3) | 17.8(3.7) | 16.2(3.3) | 14.6(3.0) | 22.7(6.7) | 19.5(4.5) | 18.0(4.0) | 16.0(3.5) |
| Percent | ||||||||
| Nulliparous | 5.7 | 4.7 | 4.6 | 4.7 | 25.3 | 19.5 | 17.2 | 17.1 |
| Oral contraceptives | ||||||||
| Never | 55.5 | 50.3 | 50.2 | 47.3 | 14.7 | 13.9 | 13.6 | 13.1 |
| <1 year | 12.1 | 12.7 | 12.6 | 13.5 | 15.8 | 15 | 15.1 | 14.7 |
| 1–5 years | 18 | 20.3 | 20.1 | 20.5 | 38.1 | 38.2 | 38.6 | 38.1 |
| 5+ years | 14.4 | 16.7 | 17.1 | 18.7 | 31.4 | 32.9 | 32.7 | 34.1 |
| Family history of ovarian cancer | 3.2 | 3 | 3.1 | 3.3 | 2 | 2 | 2.1 | 2.4 |
| Tubal ligation | 18.8 | 21.1 | 22.5 | 23.8 | 21.3 | 22.7 | 25 | 27.4 |
| Hysterectomy | ||||||||
| Yes | 22.3 | 21.8 | 21.3 | 21.2 | 6.2 | 7 | 6.9 | 7.6 |
| Unknown | 1.3 | 1.1 | 1.2 | 1.2 | 2.6 | 2.2 | 2.4 | 2.6 |
| Unilateral oophorectomy | ||||||||
| Yes | 9.1 | 8.4 | 8.2 | 8.3 | 4 | 4.2 | 4 | 4.4 |
| Unknown | 2 | 1.7 | 1.9 | 1.9 | 3.2 | 2.8 | 3 | 3.2 |
| Physical activity (MET-hours/week) | ||||||||
| <3.8 | 18.3 | 21.8 | 26.8 | 33.6 | 12.8 | 16.9 | 20.6 | 26.7 |
| >3.8–11.1 | 21.8 | 25.1 | 26 | 25.1 | 20.7 | 24.7 | 27 | 27.7 |
| >11.1–26.4 | 27.4 | 27 | 24.6 | 20.7 | 27.9 | 28.6 | 26.3 | 23 |
| >26.4 | 26.4 | 21 | 17.4 | 14.8 | 30.9 | 22.6 | 18.4 | 14.5 |
| Unknown | 6.2 | 5.1 | 5.2 | 5.8 | 7.8 | 7.2 | 7.8 | 8.1 |
| Postmenopausal | 90 | 87.2 | 84.7 | 85 | 5 | 4.1 | 4.6 | 4.9 |
| Ever HT use* | 51 | 51.6 | 51.3 | 49.6 | 59.7 | 59.3 | 57.8 | 54.1 |
Among parous only
Among postmenopausal only
All dietary measures are cumulative average energy-adjusted intake
Figure 1.
Energy‐adjusted dietary fat intake (g/day) in the NHS and NHSII by year. In both NHS and NHSII, total fat, animal fat and vegetable fat intake decreased until 1994 at which point, total fat and vegetable fat intake increased over follow‐up and animal fat intake continued to decrease. NHS women reported lower intakes of dietary fat compared to NHSII throughout follow‐up.
In the NHS, women in the highest quartile of cumulative average dietary animal fat intake had a 57% higher risk of ovarian cancer compared to those in the lowest quartile (95%CI: 1.20, 2.06, p-trend<0.01) after adjusting for multiple potential confounders, including dietary fiber intake (Table 2). Similarly, cumulative average cholesterol intake was positively associated with ovarian cancer risk; those in the highest quartile of intake had a 35% greater risk of ovarian cancer compared to those in the lowest quartile (95%CI: 1.08, 1.69, p-trend=0.01). In contrast, there were no significant associations between ovarian cancer risk and animal fat intake (RR comparing extreme quartiles=0.77, 95%CI: 0.48, 1.24, p-trend=0.44) or dietary cholesterol intake (RR comparing extreme quartiles=0.96, 95%CI: 0.63, 1.48, p-trend=0.98) in the NHSII. The p-heterogeneity across cohorts was 0.03 and 0.32, respectively. There were no significant associations between ovarian cancer risk and intake of total fat, vegetable fat, saturated fat, monounsaturated fat, polyunsaturated fat, and trans-unsaturated fat in either cohort (p-trends≥0.05). Furthermore, there were no significant associations seen when examining the association of total omega-3, long-chain omega-3, total omega-6 polyunsaturated fatty acids and the ratio of omega-3 to omega-6 polyunsaturated fatty acids with ovarian cancer risk (Supplemental Table 1).
Table 2.
Relative risks (RR) and 95% confidence intervals (CI) for the association between cumulative average energy-adjusted dietary fat and cholesterol intake and ovarian cancer risk (NHS/NHSII)
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | p-trend | p-het* | |
|---|---|---|---|---|---|---|
| Total fat | ||||||
| Range (g/day) | <53 | 53-<59 | 59-<66 | ≥66 | ||
| NHS | ||||||
| Cases/PY | 232/467,790 | 203/421,010 | 163/347,743 | 102/233,682 | ||
| RR (95%CI) | 1 (Ref) | 1.07(0.88,1.30) | 1.11(0.90,1.37) | 1.14(0.88,1.47) | 0.25 | 0.04 |
| RR (95%CI) + fiber | 1 (Ref) | 1.12(0.91,1.36) | 1.19(0.95,1.49) | 1.25(0.95,1.64) | 0.08 | 0.04 |
| NHS2 | ||||||
| Cases/PY | 38/260,155 | 50/303,353 | 53/368,889 | 55/496,316 | ||
| RR (95%CI) | 1 (Ref) | 1.07(0.69,1.63) | 0.92(0.60,1.42) | 0.70(0.45,1.09) | 0.06 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.12(0.73,1.74) | 1.00(0.64,1.57) | 0.79(0.49,1.29) | 0.22 | |
| Animal fat | ||||||
| Range (g/day) | <27 | 27-<32 | 32-<37 | ≥37 | ||
| NHS | ||||||
| Cases/PY | 214/434,469 | 193/402,406 | 160/351,609 | 133/281,742 | ||
| RR (95%CI) | 1 (Ref) | 1.09(0.89,1.33) | 1.15(0.93,1.43) | 1.36(1.07,1.74) | 0.01 | 0.02 |
| RR (95%CI) + fiber | 1 (Ref) | 1.17(0.95,1.44) | 1.27(1.01,1.61) | 1.57(1.20,2.06) | <0.01 | 0.03 |
| NHS2 | ||||||
| Cases/PY | 53/286,589 | 41/319,744 | 52/373,249 | 50/449,131 | ||
| RR (95%CI) | 1 (Ref) | 0.69(0.46,1.05) | 0.80(0.54,1.20) | 0.68(0.45,1.02) | 0.12 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.74(0.48,1.14) | 0.88(0.58,1.36) | 0.77(0.48,1.24) | 0.44 | |
| Vegetable fat | ||||||
| Range (g/day) | <22 | 22-<27 | 27-<31 | ≥31 | ||
| NHS | ||||||
| Cases/PY | 201/443,319 | 202/398,256 | 160/352,559 | 137/276,092 | ||
| RR (95%CI) | 1 (Ref) | 1.10(0.90,1.34) | 0.96(0.77,1.20) | 1.10(0.87,1.39) | 0.65 | 0.58 |
| RR (95%CI) + fiber | 1 (Ref) | 1.12(0.92,1.37) | 1.00(0.80,1.24) | 1.17(0.92,1.48) | 0.36 | 0.53 |
| NHS2 | ||||||
| Cases/PY | 29/289,684 | 47/329,760 | 62/370,232 | 58/439,039 | ||
| RR (95%CI) | 1 (Ref) | 1.29(0.81,2.07) | 1.38(0.87,2.18) | 0.99(0.61,1.60) | 0.63 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.31(0.82,2.1) | 1.41(0.89,2.23) | 1.02(0.63,1.66) | 0.75 | |
| Saturated fat | ||||||
| Range (g/day) | <18 | 18-<20 | 20-<23 | ≥23 | ||
| NHS | ||||||
| Cases/PY | 230/454,898 | 198/408,313 | 164/346,710 | 108/260,306 | ||
| RR (95%CI) | 1 (Ref) | 1.14(0.92,1.42) | 1.26(0.98,1.63) | 1.27(0.92,1.76) | 0.1 | 0.15 |
| RR (95%CI) + fiber | 1 (Ref) | 1.18(0.95,1.47) | 1.32(1.02,1.72) | 1.35(0.97,1.88) | 0.05 | 0.16 |
| NHS2 | ||||||
| Cases/PY | 52/270,111 | 34/314,901 | 53/377,112 | 57/466,590 | ||
| RR (95%CI) | 1 (Ref) | 0.59(0.37,0.94) | 0.86(0.54,1.37) | 0.81(0.46,1.42) | 0.79 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.62(0.39,0.99) | 0.92(0.57,1.48) | 0.88(0.50,1.58) | 0.94 | |
| Monounsaturated fat | ||||||
| Range (g/day) | <19.7 | 19.7-<22.5 | 22.5-<25.4 | ≥25.4 | ||
| NHS | ||||||
| Cases/PY | 239/477,126 | 188/431,927 | 179/350,948 | 94/210,224 | ||
| RR (95%CI) | 1 (Ref) | 0.90(0.72,1.13) | 1.10(0.84,1.44) | 0.97(0.68,1.39) | 0.84 | 0.02 |
| RR (95%CI) + fiber | 1 (Ref) | 0.91(0.73,1.14) | 1.11(0.85,1.46) | 0.98(0.68,1.40) | 0.80 | 0.02 |
| NHS2 | ||||||
| Cases/PY | 35/250,051 | 52/292,783 | 53/372,402 | 56/513,478 | ||
| RR (95%CI) | 1 (Ref) | 1.15(0.72,1.84) | 0.91(0.53,1.55) | 0.64(0.33,1.24) | 0.09 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.17(0.73,1.88) | 0.93(0.55,1.59) | 0.65(0.34,1.27) | 0.10 | |
| Polyunsaturated fat | ||||||
| Range (g/day) | <9.4 | 9.4-<10.7 | 10.7-<12.3 | ≥12.3 | ||
| NHS | ||||||
| Cases/PY | 174/400,965 | 194/370,479 | 166/355,033 | 166/343,748 | ||
| RR (95%CI) | 1 (Ref) | 1.23(0.99,1.52) | 1.13(0.90,1.44) | 1.28(0.98,1.67) | 0.12 | 0.04 |
| RR (95%CI) + fiber | 1 (Ref) | 1.21(0.98,1.51) | 1.11(0.88,1.41) | 1.25(0.96,1.63) | 0.18 | 0.03 |
| NHS2 | ||||||
| Cases/PY | 38/331,015 | 53/355,957 | 66/363,957 | 39/377,784 | ||
| RR (95%CI) | 1 (Ref) | 1.23(0.79,1.90) | 1.41(0.90,2.21) | 0.80(0.47,1.37) | 0.33 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.19(0.77,1.84) | 1.36(0.86,2.13) | 0.76(0.44,1.30) | 0.24 | |
| Trans fat | ||||||
| Range (g/day) | <2.1 | 2.1-<2.6 | 2.6-<3.3 | ≥3.3 | ||
| NHS | ||||||
| Cases/PY | 221/406,162 | 204/384,842 | 161/367,601 | 114/311,621 | ||
| RR (95%CI) | 1 (Ref) | 0.97(0.79,1.20) | 0.85(0.68,1.08) | 0.78(0.59,1.03) | 0.06 | 0.6 |
| RR (95%CI) + fiber | 1 (Ref) | 1.00(0.81,1.23) | 0.89(0.70,1.12) | 0.82(0.61,1.10) | 0.13 | 0.61 |
| NHS2 | ||||||
| Cases/PY | 57/310,581 | 52/333,201 | 39/361,919 | 48/423,012 | ||
| RR (95%CI) | 1 (Ref) | 0.93(0.62,1.40) | 0.69(0.43,1.09) | 0.88(0.53,1.45) | 0.51 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.97(0.65,1.46) | 0.73(0.45,1.17) | 0.95(0.56,1.59) | 0.73 | |
| Cholesterol | ||||||
| Range (g/day) | <196 | 196-<230 | 230-<270 | ≥270 | ||
| NHS | ||||||
| Cases/PY | 192/365,238 | 169/355,329 | 141/359,794 | 198/389,864 | ||
| RR (95%CI) | 1 (Ref) | 0.97(0.79,1.20) | 0.86(0.69,1.08) | 1.29(1.04,1.60) | 0.03 | 0.28 |
| RR (95%CI) + fiber | 1 (Ref) | 0.99(0.8,1.23) | 0.89(0.71,1.12) | 1.35(1.08,1.69) | 0.01 | 0.32 |
| NHS2 | ||||||
| Cases/PY | 54/356,936 | 46/368,329 | 50/365,133 | 46/338,316 | ||
| RR (95%CI) | 1 (Ref) | 0.81(0.55,1.21) | 0.88(0.60,1.31) | 0.87(0.57,1.31) | 0.59 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.86(0.57,1.29) | 0.95(0.64,1.43) | 0.96(0.63,1.48) | 0.98 |
p-het=p-for-heterogeneity by cohort (NHS, NHSII)
PY=person-years
Models adjusted for age (continuous), calendar year (continuous), total caloric intake (continuous), family history of ovarian cancer (no, yes), age at menarche (continuous), tubal ligation (no, yes), hysterectomy (no, yes, unknown), unilateral oophorectomy (no, yes, unknown), BMI (kg/m2, continuous), nulliparity (no, yes), parity (continuous), oral contraceptive use (never, <1, 1–5, 5+yrs), smoking (pack-years, continuous), menopausal status (premenopausal, postmenopausal), estrogen only HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), estrogen + progestin HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), other HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), lactose intake (g/day, continuous), caffeine intake (mg/day, continuous), fiber intake (g/day, continuous), and physical activity (METs/week: 0–3.8, >3.8–11.1, >11.1–26.4, >26.4, unknown). Vegetable fat and animal fat were mutually adjusted for each other. Saturated fat, polyunsaturated fat, monounsaturated fat, and trans-unsaturated fat were mutually adjusted for each other.
We examined the associations of the main food sources of animal fat and cholesterol intake (red meat, processed meat, poultry, seafood, eggs) with ovarian cancer risk. Overall, processed meat and egg intake appear to be the main food groups contributing to the positive associations for animal fat and cholesterol intake observed in NHS. Comparing the top versus bottom quartile intake, the hazard ratios were 1.29 (95% CI 1.00–1.68, p-trend=0.2) for processed meat and 1.30 (95% CI 0.97–1.74, p-trend=0.04) for egg intake (data not shown). In a previous publication, we observed no clear associations for dairy intake, another contributor to animal fat, and ovarian cancer risk.21
When we examined the associations by menopausal status, there was a significant positive association between animal fat intake and ovarian cancer risk in postmenopausal NHS women (RR comparing extreme quartiles=1.57, 95%CI: 1.17, 2.09, p-trend<0.01) (Table 3). The positive association between cholesterol intake and ovarian cancer risk in the NHS was apparent in postmenopausal women (RR comparing extreme quartiles =1.41, 95%CI: 1.11, 1.79, p-trend=0.01), but not premenopausal women (RR comparing extreme quartiles =0.78, 95%CI: 0.42, 1.45, p-trend=0.58). Similar to the overall analysis, we did not observe any significant associations with dietary intake of fat or cholesterol in the NHSII among either pre- or postmenopausal women, although case numbers were limited (N, premenopausal=146; N, postmenopausal=50).
Table 3.
Relative risks (RR) and 95% confidence intervals (CI) for the association between cumulative average energy-adjusted dietary fat and cholesterol intake and ovarian cancer risk by cohort and menopausal status (NHS/NHSII)
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | p-trend | p-het* | |
|---|---|---|---|---|---|---|
| Total fat | ||||||
| Range (g/day) | <53 | 53-<59 | 59-<66 | ≥66 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 13/60,805 | 27/82,098 | 18/88,407 | 24/80,465 | ||
| RR (95%CI) | 1 (Ref) | 1.53(0.78,2.99) | 0.97(0.47,2.00) | 1.41(0.69,2.87) | 0.6 | 0.24 |
| RR (95%CI) + fiber | 1 (Ref) | 1.68(0.84,3.35) | 1.11(0.52,2.38) | 1.76(0.81,3.80) | 0.30 | 0.28 |
| NHS2 | ||||||
| Cases/PY | 31/223,766 | 34/255,139 | 38/307,647 | 43/415,310 | ||
| RR (95%CI) | 1 (Ref) | 0.95(0.58,1.56) | 0.88(0.54,1.44) | 0.72(0.44,1.18) | 0.17 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.97(0.59,1.60) | 0.91(0.54,1.52) | 0.75(0.44,1.30) | 0.27 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 219/406,986 | 176/338,912 | 145/259,337 | 78/153,217 | ||
| RR (95%CI) | 1 (Ref) | 1.02(0.83,1.25) | 1.15(0.92,1.44) | 1.11(0.84,1.47) | 0.27 | 0.11 |
| RR (95%CI) + fiber | 1 (Ref) | 1.05(0.85,1.30) | 1.22(0.96,1.54) | 1.19(0.88,1.61) | 0.13 | 0.12 |
| NHS2 | ||||||
| Cases/PY | 7/36,389 | 16/48,214 | 15/61,242 | 12/81,006 | ||
| RR (95%CI) | 1 (Ref) | 1.44(0.58,3.57) | 1.00(0.39,2.55) | 0.57(0.21,1.55) | 0.09 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.67(0.66,4.20) | 1.30(0.49,3.48) | 0.81(0.28,2.39) | 0.40 | |
| Animal fat | ||||||
| Range (g/day) | <27 | 27-<32 | 32-<37 | ≥37 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 13/51,780 | 19/75,405 | 22/90,331 | 28/94,258 | ||
| RR (95%CI) | 1 (Ref) | 0.94(0.45,1.94) | 1.04(0.51,2.11) | 1.21(0.60,2.45) | 0.48 | 0.25 |
| RR (95%CI) + fiber | 1 (Ref) | 1.06(0.50,2.22) | 1.25(0.59,2.64) | 1.56(0.72,3.42) | 0.20 | 0.29 |
| NHS2 | ||||||
| Cases/PY | 36/229,030 | 29/261,550 | 40/316,030 | 41/395,253 | ||
| RR (95%CI) | 1 (Ref) | 0.72(0.44,1.19) | 0.86(0.54,1.36) | 0.71(0.44,1.15) | 0.26 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.74(0.44,1.24) | 0.89(0.54,1.46) | 0.75(0.44,1.30) | 0.44 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 201/382,689 | 174/327,001 | 138/261,279 | 105/187,484 | ||
| RR (95%CI) | 1 (Ref) | 1.10(0.89,1.35) | 1.16(0.92,1.46) | 1.38(1.07,1.79) | 0.02 | 0.12 |
| RR (95%CI) + fiber | 1 (Ref) | 1.16(0.94,1.45) | 1.27(0.99,1.62) | 1.57(1.17,2.09) | <0.01 | 0.14 |
| NHS2 | ||||||
| Cases/PY | 17/57,559 | 12/58,194 | 12/57,219 | 9/53,878 | ||
| RR (95%CI) | 1 (Ref) | 0.55(0.26,1.20) | 0.63(0.29,1.37) | 0.51(0.21,1.21) | 0.15 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.68(0.30,1.54) | 0.84(0.36,1.97) | 0.76(0.28,2.05) | 0.68 | |
| Vegetable fat | ||||||
| Range (g/day) | <22 | 22-<27 | 27-<31 | ≥31 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 26/93,485 | 23/79,476 | 13/74,614 | 20/64,199 | ||
| RR (95%CI) | 1 (Ref) | 1.08(0.61,1.94) | 0.66(0.33,1.31) | 1.24(0.66,2.33) | 0.81 | 0.89 |
| RR (95%CI) + fiber | 1 (Ref) | 1.11(0.62,1.99) | 0.7(0.35,1.40) | 1.36(0.72,2.61) | 0.58 | 0.91 |
| NHS2 | ||||||
| Cases/PY | 22/263,597 | 41/286,486 | 42/308,069 | 41/343,711 | ||
| RR (95%CI) | 1 (Ref) | 1.67(0.99,2.82) | 1.51(0.89,2.57) | 1.25(0.72,2.17) | 0.82 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.67(0.99,2.83) | 1.51(0.89,2.58) | 1.25(0.72,2.18) | 0.81 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 175/349,834 | 179/318,780 | 147/277,945 | 117/211,893 | ||
| RR (95%CI) | 1 (Ref) | 1.09(0.88,1.35) | 1.01(0.80,1.27) | 1.09(0.85,1.41) | 0.61 | 0.14 |
| RR (95%CI) + fiber | 1 (Ref) | 1.11(0.89,1.37) | 1.04(0.83,1.32) | 1.15(0.89,1.49) | 0.38 | 0.12 |
| NHS2 | ||||||
| Cases/PY | 7/26,087 | 6/43,274 | 20/62,163 | 17/95,327 | ||
| RR (95%CI) | 1 (Ref) | 0.41(0.13,1.23) | 0.84(0.34,2.07) | 0.38(0.15,0.98) | 0.09 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.44(0.14,1.35) | 0.94(0.38,2.34) | 0.45(0.17,1.17) | 0.17 | |
| Saturated fat | ||||||
| Range (g/day) | <18 | 18-<20 | 20-<23 | ≥23 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 9/49,879 | 21/75,745 | 27/92,705 | 25/93,445 | ||
| RR (95%CI) | 1 (Ref) | 1.71(0.73,3.98) | 2.21(0.89,5.49) | 2.14(0.73,6.31) | 0.22 | 0.38 |
| RR (95%CI) + fiber | 1 (Ref) | 1.84(0.78,4.34) | 2.45(0.97,6.18) | 2.43(0.82,7.23) | 0.16 | 0.41 |
| NHS2 | ||||||
| Cases/PY | 40/220,588 | 17/259,343 | 43/317,581 | 46/404,351 | ||
| RR (95%CI) | 1 (Ref) | 0.38(0.21,0.70) | 0.90(0.53,1.53) | 0.81(0.42,1.55) | 0.91 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.39(0.21,0.71) | 0.91(0.53,1.57) | 0.83(0.42,1.61) | 0.85 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 221/405,019 | 177/332,568 | 137/254,004 | 83/166,861 | ||
| RR (95%CI) | 1 (Ref) | 1.10(0.88,1.38) | 1.21(0.92,1.58) | 1.24(0.88,1.76) | 0.17 | 0.17 |
| RR (95%CI) + fiber | 1 (Ref) | 1.13(0.90,1.42) | 1.25(0.95,1.64) | 1.30(0.91,1.85) | 0.11 | 0.19 |
| NHS2 | ||||||
| Cases/PY | 12/49,524 | 17/55,558 | 10/59,531 | 11/62,239 | ||
| RR (95%CI) | 1 (Ref) | 1.13(0.49,2.63) | 0.67(0.24,1.85) | 0.70(0.22,2.30) | 0.39 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.37(0.58,3.21) | 0.82(0.30,2.29) | 0.97(0.29,3.18) | 0.71 | |
| Monounsaturated fat | ||||||
| Range (g/day) | <19.7 | 19.7-<22.5 | 22.5-<25.4 | ≥25.4 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 19/73,443 | 18/88,611 | 27/88,122 | 18/61,597 | ||
| RR (95%CI) | 1 (Ref) | 0.73(0.35,1.53) | 1.23(0.54,2.81) | 1.09(0.36,3.28) | 0.67 | 0.14 |
| RR (95%CI) + fiber | 1 (Ref) | 0.76(0.36,1.59) | 1.29(0.57,2.95) | 1.14(0.38,3.38) | 0.62 | 0.15 |
| NHS2 | ||||||
| Cases/PY | 26/216,921 | 39/247,670 | 39/311,742 | 42/425,530 | ||
| RR (95%CI) | 1 (Ref) | 1.29(0.75,2.23) | 1.05(0.56,1.95) | 0.77(0.35,1.68) | 0.32 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.29(0.75,2.23) | 1.05(0.56,1.96) | 0.78(0.36,1.69) | 0.32 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 220/403,683 | 170/343,316 | 152/262,826 | 76/148,627 | ||
| RR (95%CI) | 1 (Ref) | 0.92(0.73,1.16) | 1.10(0.82,1.46) | 0.97(0.66,1.42) | 0.88 | 0.06 |
| RR (95%CI) + fiber | 1 (Ref) | 0.92(0.73,1.17) | 1.10(0.83,1.47) | 0.97(0.66,1.43) | 0.86 | 0.07 |
| NHS2 | ||||||
| Cases/PY | 9/33,130 | 13/45,113 | 14/60,660 | 14/87,948 | ||
| RR (95%CI) | 1 (Ref) | 0.91(0.36,2.30) | 0.63(0.22,1.78) | 0.42(0.12,1.52) | 0.14 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.12(0.43,2.91) | 0.81(0.28,2.36) | 0.58(0.16,2.10) | 0.29 | |
| Polyunsaturated fat | ||||||
| Range (g/day) | <9.4 | 9.4-<10.7 | 10.7-<12.3 | ≥12.3 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 17/69,607 | 16/68,916 | 19/75,794 | 30/97,457 | ||
| RR (95%CI) | 1 (Ref) | 0.92(0.45,1.91) | 1.14(0.56,2.33) | 1.42(0.68,2.98) | 0.25 | 0.23 |
| RR (95%CI) + fiber | 1 (Ref) | 0.89(0.43,1.84) | 1.08(0.53,2.21) | 1.31(0.62,2.76) | 0.36 | 0.23 |
| NHS2 | ||||||
| Cases/PY | 31/296,975 | 42/304,560 | 43/298,161 | 30/302,167 | ||
| RR (95%CI) | 1 (Ref) | 1.30(0.80,2.11) | 1.34(0.80,2.23) | 0.91(0.50,1.66) | 0.63 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.29(0.79,2.11) | 1.33(0.79,2.23) | 0.90(0.49,1.66) | 0.61 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 157/331,358 | 178/301,563 | 147/279,239 | 136/246,291 | ||
| RR (95%CI) | 1 (Ref) | 1.25(0.99,1.56) | 1.12(0.87,1.44) | 1.25(0.94,1.66) | 0.23 | 0.04 |
| RR (95%CI) + fiber | 1 (Ref) | 1.24(0.98,1.55) | 1.11(0.86,1.43) | 1.23(0.92,1.63) | 0.29 | 0.04 |
| NHS2 | ||||||
| Cases/PY | 7/34,041 | 11/51,398 | 23/65,796 | 9/75,617 | ||
| RR (95%CI) | 1 (Ref) | 1.06(0.40,2.86) | 1.56(0.60,4.03) | 0.55(0.17,1.74) | 0.23 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.98(0.37,2.64) | 1.31(0.50,3.42) | 0.39(0.12,1.27) | 0.07 | |
| Trans fat | ||||||
| Range (g/day) | <2.1 | 2.1-<2.6 | 2.6-<3.3 | ≥3.3 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 12/46,221 | 18/65,131 | 25/87,844 | 27/112,578 | ||
| RR (95%CI) | 1 (Ref) | 1.05(0.49,2.25) | 1.04(0.48,2.22) | 0.79(0.34,1.82) | 0.44 | 0.49 |
| RR (95%CI) + fiber | 1 (Ref) | 1.12(0.52,2.42) | 1.14(0.52,2.47) | 0.89(0.38,2.09) | 0.62 | 0.51 |
| NHS2 | ||||||
| Cases/PY | 43/241,641 | 33/270,738 | 33/305,277 | 37/384,207 | ||
| RR (95%CI) | 1 (Ref) | 0.74(0.46,1.19) | 0.64(0.38,1.07) | 0.64(0.36,1.13) | 0.14 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.74(0.45,1.20) | 0.64(0.38,1.08) | 0.63(0.35,1.14) | 0.15 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 209/359,941 | 186/319,710 | 136/279,756 | 87/199,043 | ||
| RR (95%CI) | 1 (Ref) | 0.97(0.78,1.20) | 0.84(0.66,1.08) | 0.80(0.58,1.08) | 0.09 | 0.36 |
| RR (95%CI) + fiber | 1 (Ref) | 0.99(0.80,1.23) | 0.86(0.67,1.11) | 0.83(0.60,1.13) | 0.17 | 0.34 |
| NHS2 | ||||||
| Cases/PY | 14/68,940 | 19/62,463 | 6/56,643 | 11/38,805 | ||
| RR (95%CI) | 1 (Ref) | 1.69(0.78,3.68) | 0.69(0.23,2.07) | 2.07(0.73,5.86) | 0.35 | |
| RR (95%CI) + fiber | 1 (Ref) | 2.15(0.96,4.79) | 1(0.32,3.09) | 3.24(1.08,9.73) | 0.09 | |
| Cholesterol | ||||||
| Range (g/day) | <196 | 196-<230 | 230-<270 | ≥270 | ||
| Premenopausal | ||||||
| NHS | ||||||
| Cases/PY | 18/47,014 | 18/61,449 | 11/78,904 | 35/124,407 | ||
| RR (95%CI) | 1 (Ref) | 0.73(0.37,1.43) | 0.37(0.17,0.80) | 0.73(0.40,1.33) | 0.43 | 0.64 |
| RR (95%CI) + fiber | 1 (Ref) | 0.75(0.38,1.50) | 0.39(0.18,0.85) | 0.78(0.42,1.45) | 0.58 | 0.59 |
| NHS2 | ||||||
| Cases/PY | 40/297,569 | 34/307,521 | 37/306,972 | 35/289,801 | ||
| RR (95%CI) | 1 (Ref) | 0.83(0.52,1.32) | 0.85(0.54,1.34) | 0.82(0.51,1.32) | 0.46 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.85(0.53,1.36) | 0.88(0.55,1.40) | 0.86(0.53,1.41) | 0.61 | |
| Postmenopausal | ||||||
| NHS | ||||||
| Cases/PY | 174/318,224 | 151/293,880 | 130/280,891 | 163/265,457 | ||
| RR (95%CI) | 1 (Ref) | 0.98(0.78,1.22) | 0.93(0.74,1.18) | 1.36(1.08,1.71) | 0.02 | 0.72 |
| RR (95%CI) + fiber | 1 (Ref) | 1.00(0.80,1.25) | 0.96(0.76,1.22) | 1.41(1.11,1.79) | 0.01 | 0.76 |
| NHS2 | ||||||
| Cases/PY | 14/59,368 | 12/60,808 | 13/58,161 | 11/48,515 | ||
| RR (95%CI) | 1 (Ref) | 0.84(0.38,1.83) | 0.98(0.45,2.15) | 1.04(0.46,2.39) | 0.85 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.00(0.45,2.25) | 1.24(0.55,2.82) | 1.43(0.59,3.45) | 0.37 |
p-het=p-for-heterogeneity by cohort (NHS, NHSII)
PY=person-years
Models adjusted for age (continuous), calendar year (continuous), total caloric intake (continuous), family history of ovarian cancer (no, yes), age at menarche (continuous), tubal ligation (no, yes), hysterectomy (no, yes, unknown), unilateral oophorectomy (no, yes, unknown), BMI (kg/m2, continuous), nulliparity (no, yes), parity (continuous), oral contraceptive use (never, <1, 1–5, 5+yrs), smoking (pack-years, continuous), menopausal status (premenopausal, postmenopausal), estrogen only HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), estrogen + progestin HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), other HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), lactose intake (g/day, continuous), caffeine intake (mg/day, continuous), fiber intake (g/day, continuous), and physical activity (METs/week: 0–3.8, >3.8–11.1, >11.1–26.4, >26.4, unknown). Vegetable fat and animal fat were mutually adjusted for each other. Saturated fat, polyunsaturated fat, monounsaturated fat, and trans-unsaturated fat were mutually adjusted for each other.
In stratified analysis by BMI, we observed a positive association between cholesterol intake and ovarian cancer risk in women with a BMI ≥25 kg/m2 in NHS (RR comparing extreme quartiles = 1.73, 95% CI: 1.23,2.43, p-trend<0.01), but not for women with a BMI <25 kg/m2 (comparable RR = 1.15, 95% CI: 0.84, 1.57, p-trend=0.67; p-interaction=0.11) (Supplemental Table 2). While, several types of fat intake in NHS were more strongly associated with ovarian cancer risk in women with a BMI ≥25 kg/m2 than <25 kg/m2, none were significant. In NHSII, saturated fat intake was more strongly associated with a BMI ≥25 kg/m2 (p-interaction=0.01).
We next examined the associations by tumor histology (serous vs non-serous). The positive association between animal fat intake and risk in the NHS was similar for both serous tumors (RR comparing extreme quartiles=1.56, 95%CI: 1.11, 2.20, p-trend=0.01; n=451 cases) and non-serous tumors (RR=1.88, 95%CI: 1.07, 3.31, p-trend=0.03; n=144 cases) in full multivariable models (Table 4). In addition, the positive association between cholesterol intake and risk in the NHS was similar for both serous and non-serous tumors (RR comparing extreme quartiles for serous tumors=1.38, 95%CI: 1.04, 1.83, p-trend=0.05 and comparable RR for non-serous tumors=1.48, 95%CI: 0.92, 2.39, p-trend=0.08). In the NHS, we noted a significant positive association between saturated fat intake and risk of non-serous tumors (RR comparing extreme quartiles=2.07, 95%CI: 1.04, 4.12, p-trend=0.03), but not serous tumors (RR comparing extreme quartiles=1.23, 95%CI: 0.80, 1.88, p-trend=0.30). There were no significant associations by histotype in the NHSII.
Table 4.
Relative risks (RR) and 95% confidence intervals (CI) for the association between cumulative average energy-adjusted dietary fat and cholesterol intake and ovarian cancer risk by cohort and histology
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | p-trend | p-het* | |
|---|---|---|---|---|---|---|
| Total fat | ||||||
| Range (g/day) | <53 | 53-<59 | 59-<66 | ≥66 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 145/467,790 | 152/421,010 | 97/347,743 | 57/233,682 | ||
| RR (95%CI) | 1 (Ref) | 1.30(1.02,1.64) | 1.09(0.83,1.43) | 1.05(0.75,1.47) | 0.7 | 0.23 |
| RR (95%CI) + fiber | 1 (Ref) | 1.33(1.04,1.7) | 1.14(0.85,1.52) | 1.11(0.78,1.59) | 0.52 | 0.24 |
| NHS2 | ||||||
| Cases/PY | 15/260,155 | 28/303,353 | 25/368,889 | 23/496,316 | ||
| RR (95%CI) | 1 (Ref) | 1.60(0.85,3.03)) | 1.25(0.65,2.41) | 0.95(0.48,1.90) | 0.53 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.73(0.90,3.33) | 1.41(0.70,2.84) | 1.14(0.53,2.45) | 0.92 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 46/467,790 | 30/421,010 | 37/347,743 | 31/233,682 | ||
| RR (95%CI) | 1 (Ref) | 0.75(0.47,1.19) | 1.10(0.70,1.73) | 1.34(0.82,2.20) | 0.2 | 0.25 |
| RR (95%CI) + fiber | 1 (Ref) | 0.82(0.51,1.32) | 1.26(0.78,2.04) | 1.60(0.93,2.76) | 0.06 | 0.26 |
| NHS2 | ||||||
| Cases/PY | 16/260,155 | 14/303,353 | 20/368,889 | 27/496,316 | ||
| RR (95%CI) | 1 (Ref) | 0.70(0.34,1.45) | 0.83(0.42,1.63) | 0.73(0.38,1.43) | 0.5 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.67(0.32,1.41) | 0.78(0.38,1.57) | 0.67(0.32,1.38) | 0.38 | |
| Animal fat | ||||||
| Range (g/day) | <27 | 27-<32 | 32-<37 | ≥37 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 133/434,469 | 132/402,406 | 105/351,609 | 81/281,742 | ||
| RR (95%CI) | 1 (Ref) | 1.21(0.94,1.54) | 1.24(0.95,1.62) | 1.38(1.02,1.88) | 0.03 | 0.23 |
| RR (95%CI) + fiber | 1 (Ref) | 1.28(0.99,1.65) | 1.35(1.01,1.81) | 1.56(1.11,2.20) | 0.01 | 0.27 |
| NHS2 | ||||||
| Cases/PY | 23/286,589 | 18/319,744 | 27/373,249 | 23/449,131 | ||
| RR (95%CI) | 1 (Ref) | 0.80(0.43,1.49) | 1.13(0.64,2.01) | 0.93(0.50,1.72) | 0.97 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.90(0.47,1.73) | 1.35(0.72,2.55) | 1.19(0.58,2.44) | 0.46 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 38/434,469 | 38/402,406 | 32/351,609 | 36/281,742 | ||
| RR (95%CI) | 1 (Ref) | 1.09(0.69,1.72) | 1.08(0.66,1.75) | 1.49(0.90,2.47) | 0.15 | 0.19 |
| RR (95%CI) + fiber | 1 (Ref) | 1.22(0.76,1.96) | 1.27(0.75,2.14) | 1.88(1.07,3.31) | 0.03 | 0.21 |
| NHS2 | ||||||
| Cases/PY | 20/286,589 | 15/319,744 | 21/373,249 | 21/449,131 | ||
| RR (95%CI) | 1 (Ref) | 0.62(0.31,1.23) | 0.83(0.44,1.58) | 0.67(0.35,1.30) | 0.39 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.58(0.28,1.17) | 0.75(0.38,1.48) | 0.58(0.28,1.21) | 0.26 | |
| Vegetable fat | ||||||
| Range (g/day) | <22 | 22-<27 | 27-<31 | ≥31 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 126/443,319 | 139/398,256 | 104/352,559 | 82/276,092 | ||
| RR (95%CI) | 1 (Ref) | 1.23(0.96,1.58) | 1.02(0.78,1.34) | 1.09(0.80,1.47) | 0.83 | 0.55 |
| RR (95%CI) + fiber | 1 (Ref) | 1.25(0.98,1.60) | 1.05(0.80,1.38) | 1.13(0.83,1.54) | 0.63 | 0.53 |
| NHS2 | ||||||
| Cases/PY | 14/289,684 | 21/329,760 | 32/370,232 | 24/439,039 | ||
| RR (95%CI) | 1 (Ref) | 1.29(0.65,2.58) | 1.70(0.88,3.29) | 0.97(0.47,1.99) | 0.78 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.32(0.66,2.64) | 1.75(0.90,3.39) | 1.02(0.49,2.11) | 0.90 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 49/443,319 | 37/398,256 | 26/352,559 | 32/276,092 | ||
| RR (95%CI) | 1 (Ref) | 0.86(0.55,1.33) | 0.66(0.40,1.07) | 1.07(0.66,1.73) | 0.91 | 0.94 |
| RR (95%CI) + fiber | 1 (Ref) | 0.88(0.57,1.37) | 0.70(0.43,1.15) | 1.18(0.72,1.93) | 0.78 | 0.98 |
| NHS2 | ||||||
| Cases/PY | 13/289,684 | 16/329,760 | 21/370,231.50 | 27/439,039 | ||
| RR (95%CI) | 1 (Ref) | 0.98(0.46,2.08) | 1.07(0.52,2.22) | 1.01(0.49,2.09) | 0.95 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.98(0.46,2.07) | 1.07(0.51,2.21) | 1.00(0.48,2.08) | 0.97 | |
| Saturated fat | ||||||
| Range (g/day) | <18 | 18-<20 | 20-<23 | ≥23 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 142/454,898 | 141/408,313 | 104/346,710 | 64/260,306 | ||
| RR (95%CI) | 1 (Ref) | 1.28(0.98,1.66) | 1.26(0.91,1.73) | 1.16(0.76,1.75) | 0.43 | 0.66 |
| RR (95%CI) + fiber | 1 (Ref) | 1.32(1.01,1.73) | 1.32(0.95,1.83) | 1.23(0.80,1.88) | 0.30 | 0.68 |
| NHS2 | ||||||
| Cases/PY | 22/270,111 | 16/314,901 | 25/377,112 | 28/466,590 | ||
| RR (95%CI) | 1 (Ref) | 0.82(0.41,1.65) | 1.31(0.65,2.65) | 1.59(0.70,3.65) | 0.16 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.87(0.43,1.77) | 1.43(0.70,2.96) | 1.80(0.77,4.24) | 0.10 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 46/454,898 | 34/408,313 | 32/346,710 | 32/260,306 | ||
| RR (95%CI) | 1 (Ref) | 1.05(0.64,1.73) | 1.30(0.74,2.31) | 1.99(1.01,3.93) | 0.04 | 0.49 |
| RR (95%CI) + fiber | 1 (Ref) | 1.08(0.65,1.78) | 1.35(0.75,2.41) | 2.07(1.04,4.12) | 0.03 | 0.49 |
| NHS2 | ||||||
| Cases/PY | 19/270,111 | 14/314,901 | 20/377,112 | 24/466,590 | ||
| RR (95%CI) | 1 (Ref) | 0.61(0.29,1.27) | 0.78(0.36,1.69) | 0.67(0.27,1.66) | 0.54 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.58(0.28,1.24) | 0.74(0.34,1.62) | 0.62(0.24,1.58) | 0.46 | |
| Monounsaturated fat | ||||||
| Range (g/day) | <19.7 | 19.7-<22.5 | 22.5-<25.4 | ≥25.4 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 152/477,126 | 135/431,927 | 112/350,948 | 52/210,224 | ||
| RR (95%CI) | 1 (Ref) | 1.01(0.77,1.33) | 1.06(0.75,1.49) | 0.83(0.52,1.32) | 0.56 | 0.06 |
| RR (95%CI) | 1 (Ref) | 1.02(0.77,1.34) | 1.07(0.76,1.51) | 0.83(0.52,1.33) | 0.58 | 0.06 |
| NHS2 | ||||||
| Cases/PY | 18/250,051 | 27/292,783 | 22/372,402 | 24/513,478 | ||
| RR (95%CI) | 1 (Ref) | 1.15(0.59,2.23) | 0.72(0.33,1.58) | 0.57(0.22,1.49) | 0.14 | |
| RR (95%CI) | 1 (Ref) | 1.18(0.61,2.30) | 0.75(0.34,1.63) | 0.59(0.22,1.53) | 0.15 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 50/477,126 | 27/431,927 | 42/350,948 | 25/210,224 | ||
| RR (95%CI) | 1 (Ref) | 0.63(0.37,1.08) | 1.27(0.71,2.26) | 1.19(0.56,2.52) | 0.38 | 0.46 |
| RR (95%CI) + fiber | 1 (Ref) | 0.64(0.38,1.09) | 1.29(0.72,2.29) | 1.20(0.56,2.56) | 0.35 | 0.46 |
| NHS2 | ||||||
| Cases/PY | 13/250,051 | 15/292,783 | 22/372,402 | 27/513,478 | ||
| RR (95%CI) | 1 (Ref) | 0.83(0.37,1.85) | 0.96(0.41,2.23) | 0.67(0.24,1.90) | 0.47 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.82(0.37,1.84) | 0.95(0.41,2.22) | 0.67(0.24,1.89) | 0.46 | |
| Polyunsaturated fat | ||||||
| Range (g/day) | <9.4 | 9.4-<10.7 | 10.7-<12.3 | ≥12.3 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 119/400,965 | 126/370,479 | 106/355,033 | 100/343,748 | ||
| RR (95%CI) | 1 (Ref) | 1.16(0.89,1.51) | 1.03(0.77,1.39) | 1.09(0.78,1.52) | 0.78 | 0.1 |
| RR (95%CI) + fiber | 1 (Ref) | 1.15(0.88,1.49) | 1.02(0.76,1.37) | 1.06(0.76,1.49) | 0.90 | 0.1 |
| NHS2 | ||||||
| Cases/PY | 20/331,015 | 23/355,957 | 32/363,957 | 16/377,784 | ||
| RR (95%CI) | 1 (Ref) | 1.09(0.58,2.04) | 1.49(0.79,2.82) | 0.75(0.34,1.64) | 0.55 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.06(0.56,1.99) | 1.44(0.76,2.72) | 0.70(0.32,1.56) | 0.46 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 34/400,965 | 35/370,479 | 35/355,033 | 40/343,748 | ||
| RR (95%CI) | 1 (Ref) | 1.22(0.75,2.00) | 1.34(0.80,2.25) | 1.67(0.95,2.94) | 0.07 | 0.23 |
| RR (95%CI) + fiber | 1 (Ref) | 1.20(0.73,1.98) | 1.31(0.78,2.21) | 1.62(0.91,2.87) | 0.10 | 0.23 |
| NHS2 | ||||||
| Cases/PY | 14/331,015 | 20/355,957 | 25/363,957 | 18/377,784 | ||
| RR (95%CI) | 1 (Ref) | 1.13(0.55,2.32) | 1.31(0.63,2.71) | 0.81(0.35,1.86) | 0.53 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.14(0.56,2.35) | 1.33(0.64,2.77) | 0.83(0.35,1.93) | 0.56 | |
| Trans fat | ||||||
| Range (g/day) | <2.1 | 2.1-<2.7 | 2.7-<3.3 | ≥3.3 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 142/406,162 | 129/384,842 | 109/367,601 | 71/311,621 | ||
| RR (95%CI) | 1 (Ref) | 0.99(0.77,1.28) | 0.95(0.71,1.26) | 0.82(0.57,1.17) | 0.26 | 0.39 |
| RR (95%CI) + fiber | 1 (Ref) | 1.01(0.78,1.31) | 0.97(0.73,1.31) | 0.85(0.59,1.23) | 0.38 | 0.4 |
| NHS2 | ||||||
| Cases/PY | 29/310,581 | 26/333,201 | 14/361,919 | 22/423,012 | ||
| RR (95%CI) | 1 (Ref) | 1.03(0.58,1.81) | 0.58(0.28,1.18) | 1.00(0.48,2.07) | 0.8 | |
| RR (95%CI) + fiber | 1 (Ref) | 1.08(0.61,1.92) | 0.62(0.30,1.29) | 1.10(0.52,2.33) | 0.99 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 36/406,162 | 48/384,842 | 35/367,601 | 25/311,621 | ||
| RR (95%CI) | 1 (Ref) | 1.26(0.80,2.00) | 0.91(0.54,1.54) | 0.70(0.38,1.32) | 0.15 | 0.57 |
| RR (95%CI) + fiber | 1 (Ref) | 1.32(0.83,2.10 | 0.97(0.57,1.66) | 0.77(0.40,1.46) | 0.25 | 0.56 |
| NHS2 | ||||||
| Cases/PY | 21/310,581 | 16/333,201 | 18/361,919 | 22/423,012 | ||
| RR (95%CI) | 1 (Ref) | 0.72(0.36,1.44) | 0.70(0.34,1.45) | 0.81(0.36,1.81) | 0.73 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.70(0.35,1.41) | 0.67(0.32,1.41) | 0.77(0.34,1.76) | 0.66 | |
| Cholesterol | ||||||
| Range (g/day) | <196 | 196-<230 | 230-<270 | ≥270 | ||
| Serous | ||||||
| NHS | ||||||
| Cases/PY | 121/365,238 | 116/355,329 | 90/359,794 | 124/389,864 | ||
| RR (95%CI) | 1 (Ref) | 1.08(0.83,1.40) | 0.91(0.69,1.21) | 1.34(1.02,1.76) | 0.07 | 0.18 |
| RR (95%CI) + fiber | 1 (Ref) | 1.09(0.84,1.42) | 0.93(0.70,1.24) | 1.38(1.04,1.83) | 0.05 | 0.19 |
| NHS2 | ||||||
| Cases/PY | 26/356,936 | 21/368,329 | 27/365,133 | 17/338,316 | ||
| RR (95%CI) | 1 (Ref) | 0.82(0.46,1.46) | 1.11(0.64,1.92) | 0.77(0.41,1.45) | 0.62 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.87(0.48,1.57) | 1.20(0.68,2.14) | 0.86(0.44,1.66) | 0.88 | |
| Non-serous | ||||||
| NHS | ||||||
| Cases/PY | 37/365,238 | 29/355,329 | 26/359,794 | 52/389,864 | ||
| RR (95%CI) | 1 (Ref) | 0.80(0.49,1.31) | 0.70(0.42,1.16) | 1.35(0.86,2.14) | 0.15 | 0.91 |
| RR (95%CI) + fiber | 1 (Ref) | 0.84(0.51,1.39) | 0.74(0.44,1.25) | 1.48(0.92,2.39) | 0.08 | 0.95 |
| NHS2 | ||||||
| Cases/PY | 19/356,936 | 18/368,329 | 17/365,133 | 23/338,316 | ||
| RR (95%CI) | 1 (Ref) | 0.87(0.45,1.67) | 0.79(0.40,1.55) | 1.13(0.60,2.13) | 0.67 | |
| RR (95%CI) + fiber | 1 (Ref) | 0.86(0.44,1.66) | 0.78(0.39,1.54) | 1.11(0.57,2.14) | 0.72 |
p-het=p-for-heterogeneity by cohort (NHS, NHSII)
PY=person-years
Age-adjusted models adjusted for age (continuous), calendar year (continuous), total caloric intake (continuous), family history of ovarian cancer (no, yes), age at menarche (continuous), tubal ligation (no, yes), hysterectomy (no, yes, unknown), unilateral oophorectomy (no, yes, unknown), BMI (kg/m2, continuous), nulliparity (no, yes), parity (continuous), oral contraceptive use (never, <1, 1–5, 5+yrs), smoking (pack-years, continuous), menopausal status (premenopausal, postmenopausal), estrogen only HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), estrogen + progestin HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), other HT use (never, past <5 years, past 5+ years, current <5 years, current 5+ years), lactose intake (g/day, continuous), caffeine intake (mg/day, continuous), fiber intake (g/day, continuous), and physical activity (METs/week: 0–3.8, >3.8–11.1, >11.1–26.4, >26.4, unknown). Vegetable fat and animal fat were mutually adjusted for each other. Saturated fat, polyunsaturated fat, monounsaturated fat, and trans-unsaturated fat were mutually adjusted for each other.
In sensitivity analyses, there were no significant associations of dietary intake of fat or cholesterol at baseline with ovarian cancer risk, with the exception of polyunsaturated fat. Polyunsaturated fat intake at baseline was positively associated with risk in the NHS only (RR comparing extreme quartiles in the NHS =1.48, 95%CI: 1.10, 1.98, p-trend=0.01 and comparable RR in the NHSII = 0.89, 95%CI: 0.54, 1.48, p-trend=0.56) (Supplemental Table 3). When we examined recent dietary intake, women in the highest quartile of recent cholesterol intake had a significantly higher risk compared to those in the lowest quartile in the NHS only (RR comparing extreme quartiles in the NHS=1.35, 95%CI: 1.07, 1.70, p-trend=0.02 and comparable RR in the NHSII = 0.89, 95%CI: 0.57, 1.38, p-trend=0.54) (Supplemental Table 4).
Discussion
We examined the associations between long-term dietary intake of total fat, animal fat, vegetable fat, saturated fat, monounsaturated fat, polyunsaturated fat, trans-unsaturated fat, and cholesterol in two large prospective cohorts, the NHS and NHSII. We did not observe an association with total fat intake consistent with most15, 16, but not all18, prior cohort studies. Overall, we observed a positive association between animal fat and cholesterol intake in the NHS; however, the risk estimates for these factors were inverse, although not statistically significant, in the NHSII. Intake of other types of dietary fat was not related to risk in either cohort.
The inconsistent observations by cohort in our study may be attributed to the disparate age, adiposity, and menopausal status distributions of the two cohorts such that NHS participants are approximately 20 years older on average than NHSII participants at the midpoint of follow-up (1996/1997) and predominantly postmenopausal, while NHSII participants generally had higher BMI. However, the results were still different by cohort even when we stratified by menopausal status, although this analysis was limited by the number of premenopausal cases in the NHS and postmenopausal cases in the NHSII, and BMI. Another possible explanation for the inconsistent results by cohort may be varying associations by histotype, which has been observed for other ovarian cancer risk factors, since a higher proportion of ovarian tumors in the NHSII were non-serous than in the NHS.22 However the associations between animal fat and cholesterol intake in each cohort did not appreciably differ by histotype, with NHS suggesting positive associations for serous and non-serous disease and NHSII showing inverse or null relationships. Similarly, dietary fat and ovarian cancer relative risks did not vary by histotype in several previous cohort studies.15, 16, 18 The observed differences by cohort may be due to chance or reflect differences between birth cohorts that we were not able to address in our analyses, particularly given the very different intakes observed across the two populations.
Our observation of an increased risk with higher intakes of animal fat in the NHS is consistent with results from the US NIH-AARP Diet and Health Study, a prospective cohort of women over 50 years of age, but not with results from other cohorts such as EPIC and the Netherlands Cohort Study.16–18 The pooled analysis of 12 cohort studies suggested that very high intake of animal fat at baseline was associated with a non-statistically significant higher ovarian cancer risk.15 Our observation of an increased risk of ovarian cancer with higher intakes of cholesterol is consistent with suggestive evidence from EPIC, but was not supported by other prior prospective studies.15, 16 Though biologic data support a role for cholesterol in ovarian carcinogenesis by altering tumor growth and influencing progesterone synthesis,2, 23 dietary intake of cholesterol is not strongly correlated with cholesterol levels in circulation24, which may better reflect the relevant exposure. A small prospective epidemiologic study examining the association between circulating cholesterol and ovarian cancer risk observed a positive association with over a three-fold increased risk of ovarian cancer comparing the top tertile vs. bottom of cholesterol levels; whereas another small prospective study reported no significant association.25, 26 Additional research on circulating cholesterol levels and ovarian cancer risk is warranted, particularly given the lipid dysregulation found in ovarian tumors.27
A potential reason for the inconsistent results in observational studies of diet and ovarian cancer is that most studies rely on one measurement of diet at study entry, which may not represent long-term intake, particularly later in follow-up. As we observed in the NHS and NHSII, participants’ average fat intake overall as well as their intake of animal and vegetable fat and specific types of fatty acids changed dramatically over time. This observation suggests that as study follow-up time increases, baseline diet may be less representative of participants’ diets. In fact, when we examined the associations with ovarian cancer risk for baseline intake, we observed that baseline intake of polyunsaturated fat was significantly positively associated with risk in the NHS, but the associations with animal fat and cholesterol were null. Interestingly, an association with polyunsaturated fat and ovarian cancer risk has been observed in several15, 16, 18, but not all,28 prospective cohort analyses that used baseline diet data, although the reason for this is not clear.
Our study has some limitations. Since dietary intake and data on ovarian cancer risk factors were self-reported, there is likely some measurement error. However, as all data were collected prospectively, any measurement error should be non-differential. Further, in our primary analyses, we calculated cumulative average intake from multiple FFQs, which should reduce potential random measurement error. The NHS and NHSII are comprised of predominantly white women, which may limit generalizability. The primary strengths of our study include prospective repeated assessment of diet and potential confounders, large cohort sizes, as well as detailed covariate and tumor data.
In the NHS, we observed significant positive associations between long-term dietary intake of animal fat and cholesterol and ovarian cancer risk; however we did not observe significant associations in the NHSII, where participants were younger at enrollment and born later in time. Further research is needed to understand the differences across cohorts. Pooled analyses of multiple large cohort studies with updated dietary intake data would allow for associations between dietary fat and ovarian cancer risk to be assessed by potential modifying factors such as menopausal status as well as by tumor type.
Supplementary Material
Novelty and Impact :
Dietary fat may alter ovarian cancer risk by increasing circulating estrogen levels. However, results from studies of dietary fat and ovarian cancer risk have been inconsistent. In this analysis of data from two large, prospective cohorts, higher animal fat and cholesterol intake were associated with increased ovarian cancer risk only among women in the older cohort. Further pooled studies are needed to assess factors that might modify this relationship, such as menopausal status and histotype.
Acknowledgments
We would like to thank the participants and staff of the Nurses’ Health Study and Nurses’ Health Study II for their valuable contributions as well as the following state cancer registries for their help: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, and WY. The authors assume full responsibility for analyses and interpretation of these data.
Financial support: This work was supported by an Ovarian Cancer Research Fund Ann Schreiber Mentored Investigator Award (to MSR) as well as by the National Institutes of Health (UM1 CA186107, P01 CA87969, UM1 CA176726).
Abbreviations:
- BMI
Body mass index
- CI
Confidence intervals
- FFQ
Food frequency questionnaire
- HT
Hormone therapy
- NHS
Nurses’ Health Study
- RR
Relative risk
Footnotes
The authors have no disclosures or conflicts of interest.
References
- 1.Cancer Facts & Figures 2017. 2017, Atlanta: American Cancer Society. [Google Scholar]
- 2.Hill MJ, Goddard P, and Williams RE, Gut bacteria and aetiology of cancer of the breast. Lancet, 1971. 2(7722): p. 472–3. [DOI] [PubMed] [Google Scholar]
- 3.Walker BE, Tumors in female offspring of control and diethylstilbestrol-exposed mice fed high-fat diets. J Natl Cancer Inst, 1990. 82(1): p. 50–4. [DOI] [PubMed] [Google Scholar]
- 4.Prentice RL, Thomson CA, Caan B, Hubbell FA, Anderson GL, Beresford SA, Pettinger M, Lane DS, Lessin L, Yasmeen S, Singh B, Khandekar J, Shikany JM, Satterfield S, Chlebowski RT, Low-fat dietary pattern and cancer incidence in the Women’s Health Initiative Dietary Modification Randomized Controlled Trial. J Natl Cancer Inst, 2007. 99(20): p. 1534–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lubin F, Chetrit A, Modan B, Freedman LS, Dietary intake changes and their association with ovarian cancer risk. J Nutr, 2006. 136(9): p. 2362–7. [DOI] [PubMed] [Google Scholar]
- 6.Zhang M, Yang ZY, Binns CW, Lee AH, Diet and ovarian cancer risk: a case-control study in China. Br J Cancer, 2002. 86(5): p. 712–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Risch HA, Jain M, Marrett LD, Howe GR, Dietary fat intake and risk of epithelial ovarian cancer. J Natl Cancer Inst, 1994. 86(18): p. 1409–15. [DOI] [PubMed] [Google Scholar]
- 8.Merritt MA, Cramer DW, Missmer SA, Vitonis AF, Titus LJ, Terry KL, Dietary fat intake and risk of epithelial ovarian cancer by tumour histology. British Journal of Cancer, 2014. 110(5): p. 1392–1401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pan SY, Ugnat AM., Mao Y, Wen SW, Johnson KC, & Canadian Cancer Registeries Epidemiology Research Group, A case-control study of diet and the risk of ovarian cancer. Cancer Epidemiol Biomarkers Prev, 2004. 13(9): p. 1521–7. [PubMed] [Google Scholar]
- 10.McCann SE, Freudenheim JL, Marshall JR, Graham S, Risk of human ovarian cancer is related to dietary intake of selected nutrients, phytochemicals and food groups. J Nutr, 2003. 133(6): p. 1937–42. [DOI] [PubMed] [Google Scholar]
- 11.Bidoli E, La Vecchia C, Montella M, Dal Maso L, Conti E, Negri E, Scarabelli C, Carbone A, Decarli A, Franceschi S, Nutrient intake and ovarian cancer: an Italian case-control study. Cancer Causes Control, 2002. 13(3): p. 255–61. [DOI] [PubMed] [Google Scholar]
- 12.Huncharek M and Kupelnick B, Dietary fat intake and risk of epithelial ovarian cancer: a meta-analysis of 6,689 subjects from 8 observational studies. Nutr Cancer, 2001. 40(2): p. 87–91. [DOI] [PubMed] [Google Scholar]
- 13.Bertone ER, Rosner BA, Hunter DJ, Stampfer MJ, Speizer FE, Colditz GA, Willett WC, Hankinson SE, Dietary fat intake and ovarian cancer in a cohort of US women. American Journal of Epidemiology, 2002. 156(1): p. 22–31. [DOI] [PubMed] [Google Scholar]
- 14.Kushi LH, Mink PJ, Folsom AR, Anderson KE, Zheng W, Lazovich D, Sellers TA, Prospective study of diet and ovarian cancer. Am J Epidemiol, 1999. 149(1): p. 21–31. [DOI] [PubMed] [Google Scholar]
- 15.Genkinger JM, Hunter DJ, Spiegelman D, Anderson KE, Beeson WL, Buring JE, Colditz GA, Fraser GE, Freudenheim JL, Goldbohm RA, Hankinson SE, Koenig KL, Larsson SC, Leitzmann M, McCullough ML, Miller AB, Rodriguez C, Rohan TE, Ross JA, Schatzkin A, Schouten LJ, Smit E, Willett WC, Wolk A, Zeleniuch-Jacquotte A, Zhang SM, Smith-Warner SA, A pooled analysis of 12 cohort studies of dietary fat, cholesterol and egg intake and ovarian cancer. Cancer Causes Control, 2006. 17(3): p. 273–85. [DOI] [PubMed] [Google Scholar]
- 16.Merritt MA, Riboli E, Weiderpass E, Tsilidis KK, Overvad K, Tjᴓnneland A, Hansen L, Dossus L, Fagherazzi G, Baglietto L, Fortner RT, Ose J, Steffen A, Boeing H, Trichopoulou A, Trichopoulos D, Lagiou P, Masala G, Sieri S, Mattiello A, Tumino R, Sacerdote C, Bueno-de-Mesquita HB, Onland-Moret NC, Peeters PH, Hjjartåker A, Gram IT, Quirós JR, Obón-Santacana M, Molina-Montes E, Huerta Castaño JM, Ardanaz E, Chamosa S, Sonestedt E.m Idahl A, Lundin E, Khaw KT, Wareham N, Travis RC, Rinaldi S, Romieu I, Chajes V, Gunter MJ, Dietary fat intake and risk of epithelial ovarian cancer in the European Prospective Investigation into Cancer and Nutrition. Cancer Epidemiol, 2014. 38(5): p. 528–37. [DOI] [PubMed] [Google Scholar]
- 17.Gilsing AM, Weijenberg MP, Goldbohm RA, van den Brandt PA, Schouten LJ, Consumption of dietary fat and meat and risk of ovarian cancer in the Netherlands Cohort Study. Am J Clin Nutr, 2011. 93(1): p. 118–26. [DOI] [PubMed] [Google Scholar]
- 18.Blank MM, Wentzensen N, Murphy MA, Hollenbeck A, Park Y, Dietary fat intake and risk of ovarian cancer in the NIH-AARP Diet and Health Study. Br J Cancer, 2012. 106(3): p. 596–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rich-Edwards JW, Corsano KA, and Stampfer MJ, Test of the National Death Index and Equifax Nationwide Death Search. Am J Epidemiol, 1994. 140(11): p. 1016–9. [DOI] [PubMed] [Google Scholar]
- 20.Yuan MY, He JR, Chen NN, Lu JH, Shen SY, Xiao WQ, Hu F, Xiao HY, Wu YY, Xia XY, Liu Y, Qiu L, Wu YF, Hu CY, Xia HM, Qui X, Validity and Reproducibility of a Dietary Questionnaire for Consumption Frequencies of Foods during Pregnancy in the Born in Guangzhou Cohort Study (BIGCS). Nutrients, 2016. 8(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Merritt MA, Poole EM, Hankinson SE, Willett WC, Tworoger SS, Dairy food and nutrient intake in different lie periods in relation to risk of ovarian cancer. Cancer Causes Control, 2014. 25(7): p. 795–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wentzensen N, Poole EM, Trabert B, White E, Arslan AA, Patel AV, Setiawan VW, Visvanathan K, Weiderpass E, Adami HO, Black A, Bernstein L, Brinton LA, Buring J, Butler LM, Chamosa S, Clendenen TV, Dossus L, Fortner R, Gapstur SM, Gaudet MM, Gram IT, Hartge P, Hoffman-Bolton J, Idahl A, Jones M, Kaaks R, Kirsh V, Koh WP, Lacey JV, Lee IM, Lundin E, Merritt MA, Onland-Moret NC, Peters U, Poynter JN, Rinaldi S, Robein K, Rohan T, Sandler DP, Schairer C, Schouten LJ, Sjöholm LK, Sieri S, Anthony Swerdlow A, Tjonneland A, Travis R, Trichopoulou A, van den Brandt PA, Wilkens L, Wolk A, Yang HP, Zeleniuch-Jacquotte A, Tworoger SS, Ovarian Cancer Risk Factors by Histologic Subtype: An Analysis From the Ovarian Cancer Cohort Consortium. J Clin Oncol, 2016. 34(24): p. 2888–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Su F, Kozak KR, Imaizumi S, Gao F, Amneus MW, Grijalva V, Ng C, Wagner A, Gough G, Farias-Eisner G, Anantharamaiah GM, Van Lenten BJ, Navab M, Fogelman AM, Reddy ST, Farias-Eisner R, Apolipoprotein A-I (apoA-I) and apoA-I mimetic peptides inhibit tumor development in a mouse model of ovarian cancer. Proc Natl Acad Sci U S A. 107(46): p. 19997–20002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Willett W, Oxford Scholarship Online, and Oxford University Press, Nutritional epidemiology, in Monographs in epidemiology and biostatistics v 40 2013, Oxford University Press,: Oxford: p. 1 online resource. [Google Scholar]
- 25.Hiatt RA and Fireman BH, Serum cholesterol and the incidence of cancer in a large cohort. J Chronic Dis, 1986. 39(11): p. 861–70. [DOI] [PubMed] [Google Scholar]
- 26.Helzlsouer KJ, Alberg AJ, Norkus EP, Morris JS, Hoffman SC, Comstock GW, Prospective study of serum micronutrients and ovarian cancer. J Natl Cancer Inst, 1996. 88(1): p. 32–7. [DOI] [PubMed] [Google Scholar]
- 27.Baenke F, Peck B, Miess H, Schulze A, Hooked on fat: the role of lipid synthesis in cancer metabolism and tumour development. Dis Model Mech, 2013. 6(6): p. 1353–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Merritt MA, Tzuolaki I, Van Den Brandt PA, Schouten LJ, Tsilidis KK, Weiderpass E, Patel CJ, Tjønneland A, Hansen L, Overvad K, His M, Dartois L, Boutron-Ruault MC, Fortner RT, Kaaks R, Aleksandrova K, Boeing H, Trichopoulou A, Lagiou P, Bamia C, Palli D, Krogh V, Tumino R, Ricceri F, Mattiello A, Bueno-de-Mesquita HB, Onland-Moret NC, Peeters PH, Guri Skeie G, Jareid M, Quirós JR, Obón-Santacana M, Sánchez MJ, Chamosa S, Huerta JM, Barricarte A, Dias JA, Sonestedt E, Idahl A, Lundin E, Wareham NJ, Khaw KT, Travis RC, Ferrari P, Riboli E, Gunter MJ, Nutrient-wide association study of 57 foods/nutrients and epithelial ovarian cancer in the European Prospective Investigation into Cancer and Nutrition study and the Netherlands Cohort Study. Am J Clin Nutr, 2016. 103(1): p. 161–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.