Abstract
Background
In most patients, a soft tissue sarcoma is sporadic and not related to a specific known cause; however, demographic, environmental, and lifestyle factors may be linked to its development. Alcohol consumption, a major risk factor for oncogenesis, has increased, particularly among females, and it might be a risk factor for soft tissue sarcoma, with potential differences in the association based on the biological differences between males and females. Nevertheless, there is a lack of research data to determine the association between alcohol consumption and soft tissue sarcoma. Because soft tissue sarcoma often has poor oncologic and functional outcomes once it develops, identifying controllable factors for prevention would be beneficial.
Questions/purposes
(1) Is there a dose-response association between overall alcohol consumption and the incidence of soft tissue sarcoma? (2) Are there associations between the amount of alcohol consumption per occasion and drinking frequency with the incidence of soft tissue sarcomas?
Methods
This was a retrospective, population-based comparative study using the National Health Insurance Service database, which offers large-scale data from a relatively ethnically homogeneous Korean population, along with comprehensive health information. The database includes demographic, socioeconomic, health checkups, social behavior surveys, and claims data. We screened 4,234,415 people 20 years or older who underwent health checkups in 2009. Soft tissue sarcoma was defined as ICD-10 codes C47 or C49 and the registration code for cancer (V193), with at least two outpatient claims or more than one inpatient claim per year. Among the screened individuals, we excluded 7% (286,384) because of incomplete data, and we excluded 0.02% (198) with soft tissue sarcoma diagnosed before the index year. To better explore the association, we excluded 0.2% (10,088) of patients who died or developed soft tissue sarcoma in the index year. Finally, we included 3,937,745 participants (2,148,348 males and 1,789,397 females) and followed them until December 31, 2020 (mean follow-up 10 ± 1 years). The mean daily alcohol consumption was calculated using the drinking frequency (number of days per week) and the mean amount consumed on each occasion (the number of glasses [8 grams of ethanol per glass]), based on the concept of a standard drink in Korea. Based on the ethanol consumption, alcohol drinking levels were divided into three categories: individuals who did not drink, those who drank < 30 grams per day of ethanol, and those who drank ≥ 30 grams per day of ethanol. The soft tissue sarcoma incidence was calculated by dividing the number of events by the total person-years of follow-up. To address our primary study question, which was about the association of soft tissue sarcoma incidence and overall alcohol consumption, the analysis model was adjusted for age (years), smoking status (nonsmoker, past smoker, and current smoker), regular exercise (yes versus no), and metabolic syndrome (yes versus no). To address our secondary outcome, which was about associations between the amount of alcohol consumption per occasion and drinking frequency with the incidence of soft tissue sarcomas, alcohol consumption was divided into drinking frequency and amount of alcohol intake per occasion. Among the participants, 969 (males n = 550, females n = 419) were diagnosed with soft tissue sarcoma during the follow-up period, resulting in an incidence of 2.43 (males 2.55, females 2.30) per 100,000 person-years. To identify a monotonic dose-response association, we considered not only the statistical significance for individual exposure groups, but also the overall consistent directional trend in association across all groups.
Results
Compared with the individuals who did not drink (reference), alcohol consumption was not associated with an increased incidence of soft tissue sarcoma in overall participants who drank < 30 grams per day and those who drank ≥ 30 grams per day of ethanol (adjusted HR 1.05 [95% confidence interval (CI) 0.9 to 1.22] and adjusted HR 0.92 [95% CI 0.70 to 1.21], respectively; p = 0.58 among three groups) or in males who drank < 30 grams per day and those who drank ≥ 30 grams per day of ethanol (adjusted HR 0.84 [95% CI 0.70 to 1.01] and adjusted HR 0.75 [95% CI 0.56 to 1.00], respectively; p = 0.17 among three groups). In females, compared with individuals who did not drink (reference), soft tissue sarcoma incidence increased in those who drank < 30 grams per day and those who drank ≥ 30 grams per day of ethanol (adjusted HR 1.51 [95% CI 1.20 to 1.9]; p = 0.01 and adjusted HR 2.48 [95% CI 1.17 to 5.27]; p = 0.06, respectively). Although a drinking frequency of 1 to 2 days per week was associated with increased risk of developing a soft tissue sarcoma (adjusted HR 1.61 [95% CI 1.27 to 2.04]; p = 0.003), the HR did not increase with higher drinking frequency across all four groups (adjusted HR 1.21 [95% CI 0.66 to 2.200; p = 0.88 for 3 to 5 days and adjusted HR 1.46 [95% CI 0.47 to 4.56]; p = 0.60 for 6 to 7 days, respectively). However, for females consuming 3 to 4, 5 to 7, and ≥ 14 glasses per occasion, the adjusted HRs were 1.51 (95% CI 1.07 to 2.13; p = 0.09), 1.73 (95% CI 1.16 to 2.58; p = 0.06), and 3.70 (95% CI 1.37 to 9.98; p = 0.03), respectively, and the HR tended to increase with higher consumption levels per occasion across all six groups (adjusted HR 1.30 [95% CI 0.94 to 1.81]; p = 0.09 for 1 to 2 glasses and adjusted HR 1.73 [95% CI 0.81 to 3.68]; p = 0.38 for 8 to 13 glasses).
Conclusion
This nationwide population-based study demonstrated a tendency toward a dose-response relationship between the level of alcohol consumption and the incidence of soft tissue sarcoma among females. These findings suggest that strategies for individuals vulnerable to alcohol-related complications could be considered. These strategies might include campaigns, education programs, and policy interventions; social guidelines to reduce alcohol consumption may be warranted, and alcohol consumption may be considered as a screening factor for soft tissue sarcoma. To clarify whether there is a causal relationship, further research is required on the mechanisms through which alcohol consumption and drinking patterns may contribute to the development of soft tissue sarcoma.
Level of Evidence
Level III, prognostic study.
Introduction
Soft tissue sarcoma may be associated with the effects of demographic, environmental, and lifestyle factors, although most are believed to be sporadic; some are caused with radiotherapy, inherited cancer syndrome, and a damaged lymph system [1, 7, 12]. Alcohol, overweight and obesity, tobacco, late maternal first pregnancy, viral infections (herpes zoster, chickenpox, and mumps), occupational exposures (farmer, building caretakers, and military personnel), and chemical exposure (Agent Orange, chlorophenol, and cutting oil) may contribute to soft tissue sarcoma risk [4, 13, 14, 24].
Worldwide, alcohol consumption stands as a risk factor for oncogenesis and related deaths [11]. Ethanol in alcoholic beverages and acetaldehyde, its major metabolite, are categorized as Group 1 (carcinogenic to humans) by the International Agency for Research on Cancer [9]. Nevertheless, global per capita alcohol consumption has been increasing, and the upward trend poses a potential risk of escalating the overall burden of cancers [11]. In particular, the Korea National Health and Nutrition Examination Survey reported that the prevalence of high-risk drinking, defined as the proportion of people age 19 years or older who drink ≥ 2 times per week and consume a large amount of alcohol each time (a mean of ≥ 7 standard drinks for males and ≥ 5 standard drinks for females per episode) was 3.4% in 2005 and 8.4% in 2018 among females, implying a pronounced rising trend [25].
Therefore, alcohol consumption might be a risk factor for soft tissue sarcoma, and considering the biological differences between males and females, such as the percentage of body water, and the blood alcohol concentrations and distribution volume of alcohol after drinking, there may be differences in the association. However, there is a lack of research data to determine the association between alcohol consumption and soft tissue sarcoma. Because soft tissue sarcoma often has poor oncologic and functional outcomes once it develops, identifying controllable factors for prevention, such as daily alcohol consumption, alcohol intake per occasion, and drinking frequency, would be beneficial.
Using nationwide claims data in Korea, we asked: (1) Is there a dose-response association between overall alcohol consumption and the incidence of soft tissue sarcoma? (2) Are there associations between the amount of alcohol consumption per occasion and drinking frequency with the incidence of soft tissue sarcomas?
Patients and Methods
Study Design and Setting
This was a retrospective population-based comparative study. Using the National Health Insurance Service (NHIS) database, which represents the Korean population, we analyzed 4.2 million adults ages 20 years or older. The NHIS serves as a single health insurance provider to all citizens and provides biannual regular free health checkups. This database includes demographic, socioeconomic, health checkup, and claims data on all insured people. The checkup programs include data for medical history, family history, height, weight, waist circumference, blood pressure, laboratory tests, as well as hearing and visual checkups. Also, in 2009, according to the activation of the Basic Act on Health Examination, social behavior surveys including smoking, alcohol consumption, and exercise were added. Therefore, we included individuals who underwent NHIS health checkups between January 1, 2009, and December 31, 2009, and tracked them until December 31, 2020, to identify the incidence of soft tissue sarcoma. The participation rate for NHIS health checkups in the index year 2009 was 67% of those age 20 years or older who were eligible.
Participants
Soft tissue sarcoma was defined as ICD-10 codes C47 or C49 and the registration code for cancer (V193), with at least two outpatient claims or more than one inpatient claim per year. We initially screened 4,234,415 people ages 20 years or older who underwent health checkups in 2009. Among them, we excluded 7% (286,384) of individuals because of incomplete data, and we excluded 0.02% (198) of patients who were already diagnosed with soft tissue sarcoma before the index year. In addition, to better explore the association, we excluded 0.2% (10,088) of patients who died or developed soft tissue sarcoma in the index year. Finally, we included 3,937,745 participants (2,148,348 males and 1,789,397 females) and followed them until December 31, 2020 (mean follow-up time 10 ± 1 years) (Fig. 1). Participants who died before the development of soft tissue sarcoma were censored. Given that the NHIS database encompasses the entire population and systematically collects both healthcare claims and mortality data, loss to follow-up for reasons other than death is exceedingly rare. From the baseline characteristics of male (Table 1) and female (Table 2) participants, differences in variables, including age, smoking, obesity, and metabolic syndrome, were observed between groups with different levels of alcohol consumption, supporting the need to adjust for these variables.
Fig. 1.
This Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) flow chart demonstrates participants’ enrollment.
Table 1.
Baseline characteristics of enrollment of NHIS male cohort in 2009
| Characteristic | Alcohol intake | p valuea | ||
| Individuals who did not drink (n = 696,960) | Individuals who drank < 30 g/day of ethanol (n = 1,158,250) | Individuals who drank ≥ 30 g/day of ethanol (n = 293,138) | ||
| Age groups in years | < 0.001 | |||
| < 40 | 29 (202,088) | 41 (475,351) | 36 (105,012) | |
| 40-64 | 53 (367,004) | 51 (595,472) | 56 (164,342) | |
| ≥ 65 | 18 (127,868) | 8 (87,427) | 8 (23,784) | |
| Low-income statusb | 18 (122,551) | 13 (156,115) | 15 (43,045) | < 0.001 |
| Smoking status | < 0.001 | |||
| Nonsmokers | 47 (325,733) | 24 (282,989) | 15 (44,400) | |
| Past smokers | 22 (156,494) | 26 (303,259) | 25 (72,464) | |
| Current | 31 (214,733) | 49 (572,002) | 60 (176,274) | |
| Regular exercisec | 19 (132,879) | 20 (234,787) | 21 (60,171) | < 0.001 |
| Obesity | 35 (246,937) | 37 (426,595) | 42 (124,200) | < 0.001 |
| Metabolic syndrome | < 0.001 | |||
| Waist circumference | 21 (149,455) | 20 (234,511) | 26 (76,367) | |
| HDL-C | 28 (197,463) | 20 (234,060) | 17 (49,846) | |
| Triglyceride | 38 (265,546) | 41 (473,936) | 50 (146,089) | |
| Blood pressure | 47 (330,760) | 48 (556,072) | 57 (165,695) | |
| Glucose | 35 (245,115) | 35 (400,858) | 42 (123,041) | |
| Chronic kidney disease | 7 (51,321) | 6 (67,004) | 5 (13,640) | < 0.001 |
| Age in years | 49 ± 15 | 44 ± 13 | 45 ± 13 | < 0.001 |
| Height in cm | 169 ± 7 | 170 ± 6 | 171 ± 6 | < 0.001 |
| Weight in kg | 68 ± 11 | 70 ± 10 | 71 ± 11 | < 0.001 |
| BMI in kg/m2 | 24.0 ± 3.1 | 24.1 ± 3.0 | 24.5 ± 3.1 | < 0.001 |
| Waist circumference in cm | 83 ± 8 | 83 ± 8 | 85 ± 8 | < 0.001 |
| Fasting glucose in mg/dL | 99 ± 27 | 98 ± 24 | 102 ± 28 | < 0.001 |
| Systolic blood pressure in mm Hg | 124 ± 14 | 125 ± 14 | 127 ± 15 | < 0.001 |
| Diastolic blood pressure in mm Hg | 77 ± 10 | 78 ± 10 | 80 ± 10 | < 0.001 |
| Total cholesterol in mg/dL | 193 ± 36 | 195 ± 36 | 196 ± 37 | < 0.001 |
| HDL-C in mg/dL | 51 ± 27 | 53 ± 23 | 56 ± 27 | < 0.001 |
| LDL-C in mg/dL | 115 ± 40 | 112 ± 37 | 106 ± 40 | < 0.001 |
| Estimated GFR in mL/minute/1.73m2 | 86 ± 48 | 88 ± 54 | 90 ± 49 | < 0.001 |
| Triglyceride in mg/dLd | 120.8 (120.6-121.0) | 128.9 (128.8-129.1) | 148.1 (147.8-148.5) | < 0.001 |
Data presented as % (n), mean ± SD, or mean (95% CI). GFR = glomerular filtration rate.
p values refer to comparisons from the baseline characteristics of male participants, where differences in variables including age, smoking, obesity, and metabolic syndrome were observed between groups with different levels of alcohol consumption.
Low-income status was defined as the lowest quantile of income-based insurance contributions and being the recipient of the Medical Aid Program.
Regular exercise was defined as vigorous-intensity exercise ≥ 3 times per week or moderate-intensity exercise ≥ 5 times per week.
Triglyceride was expressed as geometric mean and 95% CI to represent the characteristics of skewed data.
Table 2.
Baseline characteristics of enrollment of NHIS female cohort in 2009
| Characteristic | Alcohol intake | p valuea | ||
| Individuals who did not drink (n = 1,338,781) | Individuals who drank < 30 g/day of ethanol (n = 430,865) | Individuals who drank ≥ 30 g/day of ethanol (n = 19,751) | ||
| Age groups in years | < 0.001 | |||
| < 40 | 18 (237,150) | 43 (186,676) | 50 (9786) | |
| 40-64 | 63 (843,408) | 52 (225,620) | 48 (9410) | |
| ≥ 65 | 19 (258,223) | 4 (18,569) | 3 (555) | |
| Low-income statusb | 24 (323,573) | 27 (115,491) | 30 (5888) | < 0.001 |
| Smoking status | < 0.001 | |||
| Nonsmoker | 97 (1,298,237) | 89 (384,594) | 63 (12,354) | |
| Past smoker | 1 (14,934) | 4 (15,962) | 8 (1655) | |
| Current | 2 (25,610) | 7 (30,309) | 29 (5742) | |
| Regular exercisec | 16 (209,159) | 16 (67,704) | 16 (3131) | < 0.001 |
| Obesity | 29 (392,040) | 21 (90,771) | 25 (4855) | < 0.001 |
| Metabolic syndrome | < 0.001 | |||
| Waist circumference | 20 (262,048) | 12 (51,076) | 16 (3154) | |
| HDL-C | 40 (540,891) | 24 (104,601) | 19 (3765) | |
| Triglyceride | 28 (378,281) | 16 (68,825) | 22 (4378) | |
| Blood pressure | 40 (544,855) | 26 (112,367) | 31 (6028) | |
| Glucose | 28 (377,472) | 21 (91,060) | 25 (5031) | |
| Chronic kidney disease | 9 (117,577) | 5 (22,952) | 5 (905) | < 0.001 |
| Age in years | 52 ± 14 | 41 ± 13 | 39 ± 13 | < 0.001 |
| Height in cm | 156 ± 6 | 158 ± 6 | 159 ± 6 | < 0.001 |
| Weight in kg | 57 ± 8 | 57 ± 8 | 58 ± 9 | < 0.001 |
| BMI in kg/m2 | 23.4 ± 3.3 | 22.6 ± 3.3 | 22.9 ± 3.5 | < 0.001 |
| Waist circumference in cm | 77 ± 9 | 74 ± 9 | 75 ± 9 | < 0.001 |
| Fasting glucose in mg/dL | 96 ± 22 | 93 ± 17 | 94 ± 20 | < 0.001 |
| Systolic blood pressure in mm Hg | 121 ± 16 | 117 ± 14 | 119 ± 15 | < 0.001 |
| Diastolic blood pressure in mm Hg | 75 ± 10 | 73 ± 10 | 75 ± 10 | < 0.001 |
| Total cholesterol in mg/dL | 198 ± 38 | 190 ± 36 | 160 ± 36 | < 0.001 |
| HDL-C in mg/dL | 59 ± 32 | 63 ± 27 | 67 ± 27 | < 0.001 |
| LDL-C in mg/dL | 118 ± 39 | 109 ± 37 | 101 ± 38 | < 0.001 |
| Estimated GFR in mL/minute/1.73m2 | 86 ± 34 | 90 ± 43 | 92 ± 39 | < 0.001 |
| Triglyceride in mg/dLd | 99.9 (99.8-100.0) | 84.5 (84.4-84.7) | 96.5 (95.7-97.2) | < 0.001 |
Data presented as % (n), mean ± SD, or mean (95% CI).
p values refer to comparisons from the baseline characteristics of male participants, where differences in variables including age, smoking, obesity, and metabolic syndrome were observed between groups with different levels of alcohol consumption.
Low-income status was defined as the lowest quantile of income-based insurance contributions and being the recipient of the Medical Aid Program.
Regular exercise was defined as vigorous-intensity exercise ≥ 3 times per week or moderate-intensity exercise ≥ 5 times per week.
Triglyceride was expressed as geometric mean and 95% CI to represent the characteristics of skewed data.
Descriptive Data
Among the 3,937,745 eligible participants, with a mean age of 47 ± 14 years, and 46% (1,799,163 of 3,937,745) residing in metropolitan cities, 969 (males n = 550, females n = 419) were diagnosed with soft tissue sarcoma during the follow-up period of 10 ± 1 years, resulting in an incidence of 2.43 (males 2.55, females 2.30) per 100,000 person-years.
Variables
We calculated the mean daily alcohol consumption using drinking frequency (number of days per week) and the mean amount consumed on each occasion (the number of standard drinks in Korea, including 8 grams of ethanol per glass) [10]. We acquired information about frequency and amount of alcohol intake from the standardized self-reported questionnaire at enrollment: (1) How many days a week do you drink alcohol on average? (2) How much do you usually drink in a day in terms of glasses when you drink alcohol, regardless of the type of drink? Then, to standardize the alcohol consumption, the amount was converted into grams of ethanol per day based on the concept of a standard drink. We applied the concept of a standard drink to ensure consistency across different types of alcohol. Because beer, wine, and spirits contain varying amounts of alcohol, this conversion allows us to compare equal quantities of alcohol regardless of the beverage. Although the WHO defines a standard drink as containing 10 grams of pure alcohol, the definition varies across countries and institutions because of the lack of global consensus. In the United States, one standard drink (or one alcoholic drink equivalent) contains approximately 14 grams of pure alcohol, while in the Republic of Korea, it is defined as 7 to 8 grams. Based on the ethanol consumption, alcohol drinking levels were divided into three categories: individuals who did not drink, those who drank < 30 grams per day, and those who drank ≥ 30 grams per day of ethanol [23].
All of the following laboratory tests and medical conditions were recorded only at the time of entry into the database. Age; sex; low-income status; smoking; regular exercise; height; weight; waist circumference; BMI; medical conditions including obesity, metabolic syndrome, and chronic kidney disease; and laboratory results including fasting glucose, total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and estimated glomerular filtration rate were the variables of this study. Age was stratified to young adults (younger than 40 years), middle-age adults (between 40 and 64 years of age), and older adults (65 years or older). Data on smoking status as a nonsmoker, past smoker, and current smoker were collected. Regular exercise was defined as at least 20 minutes of vigorous-intensity exercise ≥ 3 times a week or at least 30 minutes of moderate-intensity exercise ≥ 5 times a week. Low-income status was defined as the lowest quantile of income-based insurance contributions and being a recipient of the Medical Aid Program. Medical conditions were based on NHIS health checkups and claims data. Hypertension was defined as a systolic blood pressure of ≥ 140 mm Hg or a diastolic blood pressure of ≥ 90 mm Hg, or at least one claim of ICD-10 codes of I10-13 and I15 with a prescription of hypertension medication in the index year. Diabetes mellitus was defined as fasting glucose of ≥ 126 mg/dL or at least one claim for a diabetes medication prescription with ICD-10 codes E11-E14 in the index year. Dyslipidemia was defined as a serum total cholesterol level of ≥ 240 mg/dL or at least one claim for a prescription of dyslipidemia medications under the ICD-10 code E78. Metabolic syndrome was defined as having ≥ 3 of the five following categories: waist circumference (male ≥ 90 cm, female ≥ 85 cm), HDL-C (male ≤ 40 mg/dL, female ≤ 50 mg/dL, or participant taking dyslipidemia medications), triglyceride (≥ 150 mg/dL or participant taking dyslipidemia medications), blood pressure (systolic blood pressure ≥ 130 mm Hg or diastolic blood pressure ≥ 85 mm Hg, or participant taking hypertension medications), and glucose level (fasting glucose ≥ 100 mg/dL or participant taking diabetes medications).
Primary and Secondary Study Outcomes
Our primary study goal was to investigate whether there is a dose-response association between overall alcohol consumption and the incidence of soft tissue sarcoma. To achieve this, we adjusted for the covariates of age, smoking, regular exercise, and metabolic syndromes, which are known to have the most impact, and conducted statistical analyses.
Our secondary study goals were to examine the associations between the amount of alcohol consumed per occasion and the frequency of drinking with the incidence of soft tissue sarcomas. These were also explored through statistical analyses under covariate adjustment.
Ethical Approval
This study was approved by the institutional review board of St. Vincent’s Hospital, the Catholic University of Korea (VC23ZISI0194), and exempted from further review. Informed consent was waived because of the data publicity and deidentification before the analysis. All analysis was conducted in accordance with the Declaration of Helsinki.
Statistical Analysis
The participants’ characteristics were presented using descriptive statistics. We analyzed comparisons of numeric data with an independent t-test, and we analyzed categorical data with a chi-square test. The soft tissue sarcoma incidence was calculated by dividing the number of events by the total person-years of follow-up. We performed multivariable Cox proportional hazards regression analyses to estimate the HRs and 95% confidence intervals (CIs). While a formal test was conducted to validate the proportional hazards assumption for the Cox regression model, the standard error was estimated to be very small because of the large sample size. To evaluate soft tissue sarcoma risk by mean alcohol consumption, we divided the participants into male and female groups, and we constructed three models for each sex: Model 1 was unadjusted, Model 2 was adjusted for age (years), and Model 3 was adjusted for age (years), smoking status (nonsmoker, past smoker, and current smoker), regular exercise (yes versus no), and metabolic syndrome (yes versus no). Because the follow-up for all participants began in the same year, we determined to adjust for age as a covariate rather than to use age as a time scale. In addition, we conducted a Kaplan-Meier estimation by the level of alcohol consumption to present the cumulative incidence, facilitating an intuitive interpretation of the results. Given the study design and sample size, a combination of the Kaplan-Meier curve and Cox proportional hazards regression analysis was employed rather than more complex nonlinear models, such as fractional polynomials or restricted cubic splines. Further analysis of alcohol consumption was also conducted, dividing alcohol consumption into drinking frequency and amount of alcohol intake per occasion. All statistical analyses were performed using SAS, version 9.4 (SAS Institute), and results with two-sided p values of < 0.05 were considered statistically significant. An overall monotonic pattern, where the response consistently increases or decreases with increasing exposure, is generally regarded as strong evidence for causality. Therefore, to identify a dose-response association, we considered not only the statistical significance of HRs for each group but also the overall consistent directional trend of HR changes across all groups.
Results
Association Between Overall Alcohol Consumption and Soft Tissue Sarcoma Incidence
Compared with the individuals who did not drink (reference), soft tissue sarcoma incidence was not increased among overall participants who drank < 30 grams per day and those who drank ≥ 30 grams per day of ethanol after the adjustment of covariates of age, smoking, regular exercise, and metabolic syndrome (adjusted HR 1.05 [95% CI 0.9 to 1.22] and adjusted HR 0.92 [95% CI 0.70 to 1.21], respectively; p = 0.58 among three groups).
In the male group, the incidence of soft tissue sarcoma was not increased with the mean alcohol consumptions of < 30 grams per day and ≥ 30 grams per day of ethanol (adjusted HR 0.84 [95% CI 0.70 to 1.01] and adjusted HR 0.75 [95% CI 0.56 to 1.00], respectively; p = 0.17 among three groups). However, in the female group, compared with those who did not drink (reference), the incidence of soft tissue sarcoma increased with the mean alcohol consumption of < 30 grams per day and ≥ 30 grams per day of ethanol after adjustment (adjusted HR 1.51 [95% CI 1.20 to 1.9] and adjusted HR 2.48 [95% CI 1.17 to 5.27], respectively; p = 0.01 among three groups) (Table 3). A Kaplan-Meier curve for the incidence of soft tissue sarcoma according to mean alcohol consumption in each sex based on Model 1 has been provided (Supplemental Fig. 1; http://links.lww.com/CORR/B436). Individuals who did not drink had the lowest 10-year survival probability (0.99969 [95% CI 0.99965 to 0.99974]), followed by those who drank < 30 g/day (0.99973 [95% CI 0.9997 to 0.99977]), whereas those who drank ≥ 30 g/day demonstrated the highest survival probability (0.99976 [95% CI 0.9997 to 0.99982]) (p < 0.001). In Model 1, covariates, including age, may have acted as confounders, potentially masking or weakening the effect of alcohol consumption, implying that adjusting for them resulted in a more precise association estimation. Specifically, because age is a variable related to both alcohol consumption and soft tissue sarcoma incidence, adjusting for age enables a clearer evaluation of alcohol’s independent effect. The Kaplan-Meier curve for soft tissue sarcoma incidence according to alcohol consumption in all, male, and female participants (p = 0.58, p = 0.17, and p = 0.01, respectively) from Model 3, adjusted for age, smoking status, regular exercise, and metabolic syndrome, showed increased soft tissue sarcoma incidence according to overall alcohol consumption in females (Fig. 2).
Table 3.
Incidence and association of soft tissue sarcoma with alcohol consumption (grams of pure ethanol per day)
| Alcohol intake | Incidence rate (95% CI)a | Adjusted HR (95% CI) | p value |
| Total | 0.58 | ||
| Individuals who did not drink (n = 2,035,741) | 2.62 (2.4-2.85) | 1 (Reference) | |
| Individuals who drank < 30 g/day (n = 1,589,115) | 2.25 (2.02-2.48) | 1.05 (0.9-1.22) | |
| Individuals who drank ≥ 30 g/day (n = 312,889) | 2.16 (1.65-2.7) | 0.92 (0.70-1.21) | |
| Male | 0.17 | ||
| Individuals who did not drink (n = 696,960) | 3.29 (2.87-3.72) | 1 (Reference) | |
| Individuals who drank < 30 g/day (n = 1,158,250) | 2.23 (1.97-2.5) | 0.84 (0.70-1.01) | |
| Individuals who drank ≥ 30 g/day (n = 293,138) | 2.07 (1.56-2.61) | 0.75 (0.56-1.00) | |
| Female | 0.01 | ||
| Individuals who did not drink (n = 1,338,781) | 2.29 (2.04-2.54) | 1 (Reference) | |
| Individuals who drank < 30 g/day (n = 430,865) | 2.28 (1.85-2.73) | 1.51 (1.20-1.9) | |
| Individuals who drank ≥ 30 g/day (n = 19,751) | 3.45 (0.99-6.42) | 2.48 (1.17-5.27) |
The analysis model was adjusted for age, smoking status, regular exercise, and metabolic syndrome.
Per 100,000 person-years.
Fig. 2.
Kaplan-Meier curve for soft tissue sarcoma incidence according to alcohol consumption in (A) all, (B) male, and (C) female participants, adjusted for age, smoking status, regular exercise, and metabolic syndrome (p = 0.58, p = 0.17, and p = 0.01, respectively). A color image accompanies the online version of this article.
Soft Tissue Sarcoma Incidence by Drinking Frequency and Amount of Alcohol Consumption Per Occasion
The frequency and amount of alcohol consumption were not associated with increased sarcoma development in males. In females, although drinking frequency of 1 to 2 days per week was associated with increased risk of developing a soft tissue sarcoma (adjusted HR 1.61 [95% CI 1.27 to 2.04]; p = 0.003), the HR did not increase with higher drinking frequency across all four groups. However, the amount of alcohol consumption (3 to 4, 5 to 7, and ≥ 14 glasses per occasion) was associated with increased soft tissue sarcoma risk (adjusted HR 1.51 [95% CI 1.07 to 2.13]; p = 0.09, adjusted HR 1.73 [95% CI 1.16 to 2.58]; p = 0.06, and adjusted HR 3.70 [95% CI 1.37 to 9.98]; p = 0.03, respectively), and the HR tended to increase with higher amounts of alcohol consumption per occasion in all six groups (Table 4). The incidence and association of soft tissue sarcoma with overall alcohol consumption, drinking frequency, and amount of alcohol per occasion, based on Models 1 and 2, as well as Model 3, have been provided (Supplemental Tables 1 and 2; http://links.lww.com/CORR/B437). In the subgroup analysis by covariates, the absence of chronic kidney disease was associated with increased soft tissue sarcoma incidence by alcohol consumption among females (Supplemental Tables 3 and 4; http://links.lww.com/CORR/B437).
Table 4.
Incidence and association of soft tissue sarcoma with the drinking frequency and amount of alcohol per occasion
| Sex | Incidence rate (95% CI)a | Adjusted HR (95% CI) | p value |
| Frequency (day/week) | |||
| Male | 0.16 | ||
| None (n = 690,795) | 3.32 (2.89-3.76) | 1 (Reference) | |
| 1-2 (n = 996,804) | 2.12 (1.84-2.4) | 0.84 (0.69-1.02) | |
| 3-5 (n = 381,397) | 2.11 (1.67-2.58) | 0.72 (0.56-0.93) | |
| 6-7 (n = 79,352) | 3.54 (2.23-4.99) | 0.88 (0.60-1.32) | |
| Female | 0.03 | ||
| None (n = 1,336,053) | 2.28 (2.03-2.54) | 1 (Reference) | |
| 1-2 (n = 386,894) | 2.39 (1.91-2.89) | 1.61 (1.27-2.04) | |
| 3-5 (n = 57,035) | 1.88 (0.86-3.08) | 1.21 (0.66-2.20) | |
| 6-7 (n = 9415) | 3.15 (0-7.34) | 1.46 (0.47-4.56) | |
| Amount (number of drinks) | |||
| Male | 0.02 | ||
| None (n = 633,722) | 3.39 (2.94-3.85) | 1 (Reference) | |
| 1-2 (n = 159,186) | 1.95 (1.32-2.64) | 0.63 (0.43-0.91) | |
| 3-4 (n = 285,705) | 2.92 (2.33-3.55) | 0.98 (0.76-1.26) | |
| 5-7 (n = 537,086) | 2.17 (1.78-2.58) | 0.81 (0.64-1.02) | |
| 8-13 (n = 367,958) | 2.26 (1.78-2.77) | 1.00 (0.77-1.31) | |
| ≥ 14 (n = 164,691) | 1.19 (0.72-1.73) | 0.53 (0.33-0.84) | |
| Female | 0.09 | ||
| None (n = 1,288,199) | 2.32 (2.06-2.58) | 1 (Reference) | |
| 1-2 (n = 115,855) | 2.19 (1.55-2.88) | 1.30 (0.94-1.81) | |
| 3-4 (n = 115,855) | 2.26 (1.59-3) | 1.51 (1.07-2.13) | |
| 5-7 (n = 115,855) | 2.27 (1.43-3.19) | 1.73 (1.16-2.58) | |
| 8-13 (n = 34,603) | 1.97 (0.56-3.65) | 1.73 (0.81-3.68) | |
| ≥ 14 (n = 9278) | 4.19 (1.05-8.37) | 3.70 (1.37-9.98) |
The analysis model was adjusted for age, smoking status, regular exercise, and metabolic syndrome.
Per 100,000 person-years.
Discussion
Soft tissue sarcoma occurs sporadically and is not attributed to a specific known cause in most patients. Nevertheless, demographic, environmental, and lifestyle factors may contribute to its development. Alcohol consumption, a well-established factor in oncogenesis, has increased, particularly among females, and it might play a role in occurrence of soft tissue sarcoma [11]. Yet, research data on the association between alcohol consumption and soft tissue sarcoma remain limited. Given its poor outcome once diagnosed, identifying modifiable factors for prevention would be valuable. We found that the incidence of soft tissue sarcoma in females who drank alcoholic beverages was higher compared with those who did not drink. Further analysis showed that the amount of alcohol consumption per occasion, not the frequency, tended to be associated with an increased incidence of soft tissue sarcoma in females. Proactive public health measures, including awareness campaigns, educational programs, policy initiatives, and social guidelines to reduce alcohol consumption, may be justified, while alcohol consumption could also be considered a potential factor in screening for soft tissue sarcoma.
Limitations
This study had some limitations. Primarily, the use of data from a claims database may have influenced the accuracy of the results. Although NHIS data have gained prominence in public health research, the primary intent of this database was not research but expense reimbursement and medical regulation [11]. Because of these inherent traits of secondary data, operational definitions based on disease registration codes as well as the number of claims could lead to underestimation, posing a potential bias. As discordance between the first and final histologic diagnoses of soft tissue sarcoma is common [20], misregistration may initially occur. Nevertheless, the number of new patients with a soft tissue sarcoma derived from the NHIS database on the basis of our operational definition was roughly consistent with those from the reports of cancer statistics in Korea [16]. Second, lifestyle information, specifically alcohol consumption patterns, relied on self-reported questionnaires, possibly leading respondents to understate their actual drinking levels. Additionally, because alcohol intake was assessed at a singular instance, it might not accurately represent the ongoing and consistent alcohol exposure over time. Moreover, it was not possible to distinguish between individuals who did not drink alcohol in the index year but drank before it and those who never drank alcohol. To minimize the impact of alcohol exposure before and after the index year, we adjusted the model using relevant covariates such as age. This was a methodological decision in line with previous studies that employed the same criteria [18, 26]. Third, our population-based study was conducted exclusively on Korean participants and may not have represented the influence of alcohol on soft tissue sarcoma risk across different racial and geographic populations. To verify our findings, larger-scale research involving diverse ethnicities is essential. Fourth, a challenge arose in pinpointing participants who moved to other countries during the follow-up period, making it impractical to exclude them from the study. Fifth, this research could not investigate certain potential influencing factors such as a family history of cancer predisposition syndromes, a medical history of precursor lesions of soft tissue sarcoma, previous cancers and related treatments, and dietary habits. While most soft tissue sarcomas tend to sporadically arise, unlike carcinomas, and thus their impact is likely minimal, it would have been beneficial to examine these factors as well [7]. Last, we were unable to determine the histologic subtype, despite the potential variability in the effects of alcohol on the risk of incidence based on each subtype. Soft tissue sarcomas are subclassified into more than 70 subtypes, most of which are extremely rare [7], posing difficulties in conducting large-scale studies.
Association Between Overall Alcohol Consumption and Soft Tissue Sarcoma Incidence
We found that, in females, the incidence of soft tissue sarcoma increased with higher levels of mean alcohol consumption, while it was not associated with an increased incidence in the male group. Taking the current study’s findings, public health campaigns could raise awareness of the potential association between high overall alcohol consumption and soft tissue sarcoma development in individuals vulnerable to alcohol-related complications, promote moderate drinking, and encourage health authorities to update alcohol consumption guidelines to emphasize restrictions. Additionally, medical professionals could be trained to assess drinking habits and provide counseling on possible soft tissue sarcoma risk through enhanced regular screenings and consultations. Policy and regulatory improvements could also be made by adding warning labels to alcoholic beverages and expanding community health education and support programs to reduce alcohol consumption. Our findings differ from the conclusions of previous studies, which have stated that the consumption of alcoholic beverages did not appear to affect the risk of soft tissue sarcoma [8, 22]. While ethanol was not considered a carcinogen but rather a co-carcinogen in the past, it is now regarded as a direct carcinogen [21]. Although not fully understood, the relevant mechanism for alcohol-mediated cancer initiation may include genotoxic actions of acetaldehyde, cytochrome P450 2E1–mediated generation of reactive oxygen species, aberrant metabolism of folate and retinoids, increased estrogen, and genetic polymorphisms [4, 21]. Alcohol-induced epigenetic mechanisms, including DNA methylation, may also contribute to the DNA hypermethylation status of transcription regulators, thus leading to tumorigenesis [4, 15, 19, 21]. Meanwhile, heavy drinking tendency may contribute to the incidence of soft tissue sarcomas, particularly among females. Females eliminate a similar amount of alcohol per unit body weight as males [6, 17] but more alcohol per unit of lean body mass per hour [2, 6]. However, female lower percentage of body water leads to higher blood alcohol concentrations, compared with males with similar body weight, even after consuming the same quantity of alcohol [5, 6]. Moreover, the smaller distribution volume of alcohol in females is linked to prolonged periods of elevated ethanol blood levels [3]; thus, females may have higher vulnerability to the effects of alcohol compared with males [6]. Females face a greater risk of alcohol-related cirrhosis and other liver diseases. Additionally, alcohol-related cognitive decline and brain shrinkage occur earlier in females. Excessive drinking heightens the risk of heart muscle damage in females at lower consumption levels and over a shorter duration. Nonetheless, little is known about the potential mechanisms behind the association between alcohol intake and soft tissue sarcoma [4]. To clarify the causal relationship, further research should be conducted on the mechanisms through which overall alcohol consumption may contribute to the development of soft tissue sarcoma.
Soft Tissue Sarcoma Incidence by Drinking Frequency and Amount of Alcohol Consumption Per Occasion
We found that the HR increased with greater amounts of alcohol consumed weekly, and in particular, those who engaged in binge drinking (≥ 14 drinks per occasion) experienced a higher risk of developing a soft tissue sarcoma. In light of these results, campaigns could prompt individuals vulnerable to alcohol-related complications to reconsider their drinking patterns, encourage moderation in alcohol consumption, and advocate for updated guidelines, too. Training for medical professionals, counseling provision, and policy and regulatory adjustments could also be implemented, as mentioned above. Consistent with our study, hospital-based case-control studies [8, 22] showed that the frequency of alcohol consumption is not associated with soft tissue sarcoma incidence, while, to the best of our knowledge, there has been no research specifically regarding the association between alcohol consumption per occasion and soft tissue sarcoma incidence. To identify a causal link, additional studies are required to explore the processes through which alcohol intake per occasion may influence the occurrence of soft tissue sarcoma as well.
Conclusion
This nationwide population-based study demonstrated a tendency toward a dose-response relationship between the level of alcohol consumption and the risk of soft tissue sarcoma among females. These findings suggest that strategies for individuals vulnerable to alcohol-related complications could be considered. These strategies might include campaigns, education programs, and policy interventions; social guidelines to reduce alcohol consumption may be warranted, and alcohol consumption may be considered as a screening factor for soft tissue sarcoma. To clarify whether there is a causal relationship, the mechanisms underlying the association between alcohol intake and drinking patterns and soft tissue sarcoma development should also be explored.
Acknowledgments
We thank MD Candidates Jewoo Lee and Minpyo Lee for their technical assistance in editing.
Footnotes
A Note from the Editor-in-Chief: This article was presented at the 2024 European Musculo-Skeletal Oncology Society Meeting.
The institution of one or more of the authors (MWJ) has received, during the study period, funding from St. Vincent’s Hospital, The Catholic University of Korea, Research Institute of Medical Science Foundation (SVHR-BD-2023-09).
The first two authors contributed equally to this manuscript.
Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members.
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
Ethical approval for this study was obtained from the Catholic University of Korea St. Vincent’s Hospital Institutional Review Board (Project number VC23ZISI0194).
This work was performed at St. Vincent’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea.
Contributor Information
Kyungdo Han, Email: hkd917@naver.com.
Yoon Joo Cho, Email: choyj8908@gmail.com.
Nicholas Matthew Bernthal, Email: NBernthal@mednet.ucla.edu.
Ye Bin Park, Email: par6799@naver.com.
References
- 1.American Cancer Society. Risk factors for soft tissue sarcomas. Available at: https://www.cancer.org/cancer/types/soft-tissue-sarcoma/causes-risks-prevention/risk-factors.html. Accessed September 6, 2024.
- 2.Ammon E, Schäfer C, Hofmann U, Klotz U. Disposition and first-pass metabolism of ethanol in humans: is it gastric or hepatic and does it depend on gender? Clin Pharmacol Ther. 1996;59:503-513. [DOI] [PubMed] [Google Scholar]
- 3.Baraona E, Abittan CS, Dohmen K, et al. Gender differences in pharmacokinetics of alcohol. Alcohol Clin Exp Res. 2001;25:502-507. [PubMed] [Google Scholar]
- 4.Burningham Z, Hashibe M, Spector L, Schiffman JD. The epidemiology of sarcoma. Clin Sarcoma Res. 2012;2:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ely M, Hardy R, Longford NT, Wadsworth ME. Gender differences in the relationship between alcohol consumption and drink problems are largely accounted for by body water. Alcohol Alcohol. 1999;34:894-902. [DOI] [PubMed] [Google Scholar]
- 6.Erol A, Karpyak VM. Sex and gender-related differences in alcohol use and its consequences: contemporary knowledge and future research considerations. Drug Alcohol Depend. 2015;156:1-13. [DOI] [PubMed] [Google Scholar]
- 7.Fletcher C, Baldini E, Blay J, et al. Soft tissue tumours: introduction. In: WHO Classification of Tumours Editorial Board, ed. WHO Classification of Tumours: Soft Tissue and Bone Tumours. 5th ed. International Agency for Research on Cancer; 2020:1-12. [Google Scholar]
- 8.Franceschi S, Serraino D. Risk factors for adult soft tissue sarcoma in northern Italy. Ann Oncol. 1992;3(Suppl 2):S85-S88. [DOI] [PubMed] [Google Scholar]
- 9.International Agency for Research on Cancer. Agents classified by the IARC monographs, volumes 1–135. Available at: https://monographs.iarc.who.int/list-of-classifications. Accessed September 6, 2024. [Google Scholar]
- 10.International Alliance for Responsible Drinking. Drinking guidelines: general population. Available at: https://iard.org/science-resources/detail/Drinking-Guidelines-General-Population. Accessed September 6, 2024.
- 11.Jin EH, Han K, Shin CM, et al. Sex and tumor-site differences in the association of alcohol intake with the risk of early-onset colorectal cancer. J Clin Oncol. 2023;41:3816-3825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Joo MW, Kang YK, Ogura K, et al. Post-radiation sarcoma: a study by the Eastern Asian Musculoskeletal Oncology Group. PLoS One. 2018;13:e0204927. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Joo MW, Lee YS, Chung YG, Lee HK. Sarcomas in teachers using three-dimensional printers: a report of three patients and literature review. Clin Orthop Surg. 2022;14:310-317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kang H, Enzinger FM, Breslin P, Feil M, Lee Y, Shepard B. Soft tissue sarcoma and military service in Vietnam: a case-control study. J Natl Cancer Inst. 1987;79:693-699. [PubMed] [Google Scholar]
- 15.Kim H, Joo MW, Yoon J, et al. Can DNA methylation profiling classify histologic subtypes and grades in soft tissue sarcoma? Clin Orthop Relat Res. 2024;482:1028-1037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Korea Central Cancer Registry, National Cancer Center. Annual report of cancer statistics in Korea 2009-2020. Available at: https://ncc.re.kr/cancerStatsList.ncc. Accessed September 6, 2024.
- 17.Kwo PY, Ramchandani VA, O’Connor S, et al. Gender differences in alcohol metabolism: relationship to liver volume and effect of adjusting for body mass. Gastroenterology. 1998;115:1552-1557. [DOI] [PubMed] [Google Scholar]
- 18.Lee YJ, Han KD, Kim JH. Association among current smoking, alcohol consumption, regular exercise, and lower extremity amputation in patients with diabetic foot: nationwide population-based study. Endocrinol Metab (Seoul). 2022;37:770-780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Liu X, Chen J, Li J, et al. Integrated analysis reveals common DNA methylation patterns of alcohol-associated cancers: a pan-cancer analysis. Front Genet. 2023;14:1032683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ray-Coquard I, Montesco MC, Coindre JM, et al. Sarcoma: concordance between initial diagnosis and centralized expert review in a population-based study within three European regions. Ann Oncol. 2012;23:2442-2449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Seitz HK, Stickel F. Molecular mechanisms of alcohol-mediated carcinogenesis. Nat Rev Cancer. 2007;7:599-612. [DOI] [PubMed] [Google Scholar]
- 22.Serraino D, Franceschi S, Talamini R, Frustaci S, La Vecchia C. Non-occupational risk factors for adult soft-tissue sarcoma in northern Italy. Cancer Causes Control. 1991;2:157-164. [DOI] [PubMed] [Google Scholar]
- 23.Sun-Hong M, Ki-Young K, Hong-Jun C, Jeong-Ah L, Young-Zu K. Association between high risk alcohol consumption and hypertension in Korean people: the Fifth Korea National Health and Nutrition Examination Survey (KNHANES V, 2010-2012). KJFP. 2015;5:781-788. [Google Scholar]
- 24.Tavani A, Soler M, La Vecchia C, Negri E, Gallus S, Franceschi S. Body weight and risk of soft-tissue sarcoma. Br J Cancer. 1999;81:890-892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Trends in the prevalence of high-risk drinking, during 2012-2021. Public Health Weekly Report. 2023;16:376-377. [Google Scholar]
- 26.Yeo Y, Shin DW, Han K, et al. Smoking, alcohol consumption, and the risk of thyroid cancer: a population-based Korean cohort study of 10 million people. Thyroid. 2022;32:440-448. [DOI] [PubMed] [Google Scholar]