Abstract
Background:
Federal initiatives have recently addressed the sex bias that exists in biomedical and clinical research. However, improvement to the inclusion of sex as a biological variable remains unknown.
Methods:
We performed a 5-year follow-up study of all clinical and biomedical research articles published in 5 surgical journals from 1/1/17 to 12/31/18. Human, animal, and cell subjects were analyzed for study/subject type, sex of participants, sex matching, and sex-based data reporting, analysis, and discussion.
Results:
Comparing 2017–2018 to 2011–2012, slightly more articles reported the sex of the human studied (87% vs 83%, p=0.001). Inclusion of both sexes remained high (94% vs 95%, p=0.22) but sex-based data reporting (36% vs 38%, p=0.17), analysis (35% vs 33%, p=0.39) and discussion of results (10% vs 23%, p<0.0001) remained unchanged or worsened. Regarding animal research, the number of articles that stated the sex studied remained unchanged (79% vs 78%, p=0.67); if stated, slightly more included both sexes (7% vs 3%, p=0.002). Regarding cell research, fewer articles reported the sex of the cells studied (5% vs 24%, p=0.0001); if stated, more articles included both sexes but the difference did not reach statistical significance (25% vs 7%, p=0.34). Sex matching remained poor with only 50% of human, 4% of animal, and 9% of cell studies matching the inclusion of both sexes by at least 50%.
Conclusions:
Sex bias persists in surgical research. The majority of articles failed to report, analyze, or discuss results based on sex which will negatively affect clinical translatability and outcomes of evidence-based medicine.
INTRODUCTION
Sex bias and reporting disparities play a significant role in the evolution of scientific research. The recognition of this bias dates back to 1985 with the release of the Report of the Public Health Service Task Force on Women’s Health Issues. This report outlined sex differences and highlighted the importance of female inclusion in scientific research.1 It wasn’t until 1993, however, that the National Institutes of Health (NIH) Revitalization Act mandated the inclusion of women and minorities in clinical studies and recognized the importance of sex-based analysis of results.2 Following this, the NIH established guidelines on the inclusion of women in human research. These guidelines were amended in 2001 to clarify the expectation of the analysis and sex-based reporting of data in clinical trials, and again in 2017 to mandate valid analysis by sex in all NIH phase III clinical trials.3 Historically, females have been underrepresented in basic science studies with a bias towards males in animal research.4,5 In 2015, the NIH released a policy that outlined the expectation to include sex as a biological variable in project design, analysis, and discussion for all biomedical research, including research that utilizes vertebrate animals or humans.6–9 Research, however, continues to demonstrate the underrepresentation of females in basic science and clinical studies.10–13
The importance of equal representation is supported by sex differences in biomarkers, pharmacokinetics, pharmacodynamics, and even variances in cancer-specific mortality.14–18 Sex hormones, environmental factors, and basic physiology all account for varying outcomes in disease manifestation, medical and surgical outcomes, therapeutic efficacy, and side effects.18–25 This difference was highlighted in 2001 when the US Food and Drug Administration (FDA) withdrew 10 drugs, 8 of which had disproportionate health risks and adverse events for females.26 Sex-dependent outcomes were again demonstrated by the sleep aid, zolpidem tartrate (Ambien), after adverse outcomes were reported in females. These concerning results led the FDA to change the recommended dose of zolpidem for men versus women, with women receiving half the recommended dose as men.27–30 Yet, despite these events, research shows that sex disparities persist.5,10,11,13
We previously showed that sex bias exists in clinical and biomedical research published in 5 surgical journals from 2011 to 2012.10,11 The majority of cell and animal studies reported a significant male bias with the research studies. While the majority of human studies included both males and females, only one-third of the articles included sex-based results, analysis, or discussion of the data. Since these publications, the NIH has mandated that sex as a biological variable should be included in all biomedical and human clinical research. In this study, our objectives were to determine if sex bias was still prevalent in surgical biomedical and human research, and to evaluate any changes to sex-based data reporting over the past 5 years since the NIH policy was released. We aimed to specifically evaluate the presence of sex-based data reporting, analysis, and discussion of sex-based results as well as sex-matching within each study. We hypothesize that males and females continue to be unequally represented in surgical research and that sex as a biological variable still remains underreported.
METHODS
Data Abstraction.
Data were collected as previously described by Yoon et al.11 All original manuscripts published from January 1, 2017 through December 31, 2018 in the Annals of Surgery, American Journal of Surgery, JAMA Surgery, Journal of Surgical Research, and Surgery were reviewed for inclusion by 4 abstracters to complete a 5-year follow-up study from our previously published work and evaluation.10,11 Review articles, letters to the editor, editorials, case reports, systematic reviews, and anatomical sex-related disease studies (ovarian, cervical, testicular, prostate, etc.) were excluded. Text, tables and figures were reviewed. To standardize the process, inter-abstracter agreement was reached prior to review. Internal random review was performed by the first and senior authors to ensure accuracy and minimize person-to-person discrepancies.
Variables Abstracted.
The following data were abstracted from each article: sex-related disease, study type, subject, single vs multi-center, domestic vs international, number and sex of subjects, specialty/disease subject, sex-based reporting, analysis, and discussion of sex-based results, and sex matching. Sex-related disease was defined as a disease in either all males or all females based on the organ or disease studied (e.g., ovary, testicle, prostate). Male only or female only manuscripts were defined as studies that included only one sex despite the disease process affecting both males and females. Regarding specialties, “general surgery” was defined as manuscripts that focused on general surgical procedures such as hernia repair or gallbladder interventions. “Surgery in general” was defined as manuscripts that studied the practice of surgery without specific focus on disease process or specialty. These variables were compared to our prior study conducted by Mansukhani et al and Yoon et al to compare changes over time.10,11
Matching of Included Subjects by Sex.
Similar to Mansukhani et al10, the degree of sex matching for subjects was included and calculated for each study.
Statistical Analysis.
Fisher’s exact tests were used to examine differences among publications by journal that did or did not state sex, to compare the numbers of males, females, and unspecified sex participants in the studies presented in each journal, between domestic and international publications, and among articles published by the different specialties. The differences in the distribution of sex-matched participants, sex-based reporting, analysis, and discussion of the data by specialty were also assessed using a Fisher’s exact test, as were the differences between years 2011–2012 and 2017–2018. Overall differences between years across all levels of sex-matched participants, sex-based reporting, analysis, and discussion of the data by specialty were assessed via the chi-squared test. The 95% confidence intervals were computed using the binomial method. Significance was assumed for P<0.05. Analyses were conducted using SAS, version 9.4 (SAS Institute Inc), and R 3.6.3.
RESULTS
Total Cohort.
A total of 2,927 manuscripts were reviewed in 5 surgical journals over 2017 and 2018 (Figure 1). As described previously10, articles that qualified as a systematic review, those that studied a sex-related disease defined anatomically, and those that reported a number of subjects as zero were excluded from the analysis. After excluding 37 articles, 2,890 articles remained of which 2,523 included human data, 323 included animals, and 88 included cells.
Figure 1.
Flow diagram. Flow diagram identifying total articles, excluded articles, and final human, animal, and cell articles analyzed for this study.
Trends in Human Studies
Sex of the Participant.
Of the 2,523 human articles analyzed, more articles from 2017–2018 reported the sex of the human studied compared with 2011–2012 (2187/2523, 87% vs 1078/1303, 83%10, p=0.001) (Figure 2A). A comparable percentage of articles stated the sex of the human in 2017–2018 vs 2011–201210 with similar male-only articles (24/2187, 1% vs 17/1078, 2%, p=0.25), female-only articles (119/2187, 5% vs 41/1078, 4%, p<0.05), and articles that included both sexes (2044/2187, 94% vs 1020/1078, 95%, p=0.22) (Figure 2B). Three of the 5 journals improved the percentage of articles that stated the sex of the human studied (Supplemental Table 1). There were marginal changes between domestic (Supplemental Figure 1) and international (Supplemental Figure 2) articles with regard to stating the sex of the human studied between the different time periods.
Figure 2.
Stated sex of participants in the articles. (A) Comparison of articles that reported the sex of the participant in human, animal, and cell articles in 2011–2012 vs 2017–2018. *p<0.001. (B) Breakdown of male-only (M), female-only (F), and male+female (M/F) studies in these articles. *p<0.05, **p<0.005.
Sex-Based Reporting of Data.
Comparable or fewer articles included sex-based reporting of data in 2017–2018 when compared to 2011–2012. Overall, of the 2,523 human articles analyzed, 905 articles reported data based on sex (36% vs 497/1303, 38%10, p=0.17), 872 performed sex-based analysis (35% vs 432/1303, 33%10, p=0.39), and 257 included a discussion of sex-based results (10% vs 299/1303, 23%10, p<0.0001) (Figure 3). This trend was consistent across all journals in 2017–2018. When compared to 2011–2012, recent publications reported fewer sex-based results, and fewer articles discussed sex-based results, than previous years. This decrease in sex-based reporting, analysis, and discussion was consistent among domestic and international studies (Table 1). Despite the majority of specialties more frequently reporting the sex of the participant in 2017–2018 when compared to 2011–2012 (Table 2), there continued to be a wide discrepancy in sex-based reporting of results, analysis, and discussion across surgical specialties (Table 3). Interesting, significantly fewer articles that studied surgical education included sex-based reporting of data in 2017–2018 when compared to 2011–2012 (Table 3).
Figure 3.
Sex-based data reporting in human, animal, and cell studies. (A) Comparison of articles that reported, analyzed, or discussed results based on sex in human, animal, and cell manuscripts in 2011–2012 vs 2017–2018.
Table 1:
Changes to Sex-Based Reporting by Study Location or Journal.
| Study Type or Journal | Total Articles | Sex-Based Data Reporting N (%) | P value | Statistical Analysis of Sex-Based Results N (%) | P value | Discussion of Sex-Based Results N (%) | P value | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | ||||
| Domestic | 771 | 1751 | 283 (37) | 616 (35) | NS | 248 (32) | 594 (34) | P=0.002 | 179 (23) | 181 (10) | P<0.0001 |
| International | 532 | 772 | 214 (40) | 289 (37) | P<0.0001 | 184 (25) | 278 (36) | P=0.006 | 120 (23) | 79 (10) | P<0.0001 |
| American Journal of Surgery | 392 | 622 | 110 (28) | 140 (22) | NS | 103 (26) | 131 (20) | NS | 71 (18) | 50 (8) | P<0.0001 |
| Annals of Surgery | 273 | 473 | 109 (40) | 154 (32) | P<0.05 | 100 (37) | 148 (31) | NS | 60 (22) | 43 (9) | P<0.0001 |
| Journal of Surgical Research | 116 | 653 | 47 (41) | 228 (35) | NS | 33 (28) | 211 (33) | NS | 25 (22) | 59 (9) | P<0.0001 |
| JAMA Surgery | 256 | 210 | 114 (45) | 93 (45) | NS | 99 (39) | 85 (42) | NS | 73 (29) | 16 (8) | P=0.0003 |
| Surgery | 266 | 565 | 117 (44) | 269 (47) | NS | 97 (37) | 273 (48) | P=0.002 | 70 (26) | 89 (16) | P=0.0004 |
Table 2:
Difference in Stating Sex of Participant by Specialty.
| Specialty | Total Articles | Sex Stated N (%) | Sex Not Stated N (%) | P value | |||
|---|---|---|---|---|---|---|---|
| 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | ||
| Bariatric | 33 | 54 | 29 (88) | 52 (96) | 4 (12) | 3 (4) | NS |
| Breast | 56 | 121 | 37 (66)) | 106 (88) | 19 (34) | 15 (12) | P=0.0002 |
| Burn | 21 | 19 (90) | 15 (12) | ||||
| Cardiac | 18 | 30 | 12 (67) | 27 (90) | 6 (33) | 3 (10) | NS |
| Colorectal | 137 | 213 | 122 (89) | 197 (92) | 15 (11) | 16 (8) | NS |
| Endocrine | 126 | 174 | 110 (87) | 163 (94) | 16 (13) | 11 (6) | NS |
| General Surgery | 205 | 460 | 178 (87) | 392 (85) | 27 (13) | 68 (15) | NS |
| Neurosurgery | 7 | 6 (86) | 1 (14) | ||||
| Otolaryngology/HNS | 4 | 4 (100) | 0 (0) | ||||
| Orthopedics | 9 | 7 (78) | 2 (22) | ||||
| Pediatric Surgery | 34 | 132 | 27 (79) | 113 (86) | 7 (21) | 19 (14) | NS |
| Plastic Surgery | 22 | 20 (91) | 2 (9) | ||||
| Surg Education/Training | 98 | 208 | 52 (53) | 100 (48) | 46 (47) | 108 (52) | NS |
| Surgery in General | 72 | 49 | 56 (78) | 41 (84) | 16 (22) | 8 (16) | NS |
| Thoracic | 39 | 51 | 37 (95) | 48 (94) | 2 (5) | 3 (6) | NS |
| Trauma/critical care | 118 | 292 | 93 (79) | 259 (89) | 25 (21) | 33 (11) | NS |
| Transplant | 49 | 88 | 43 (88) | 82 (93) | 6 (12) | 6 (7) | P<0.05 |
| Urology | 3 | 3(100) | 0 (0) | ||||
| Vascular | 42 | 65 | 36 (86) | 60 (92) | 6 (14) | 5 (8) | NS |
Individual data was not available for specialties that contain a grey box over 2011–12 data. Data was arranged alphabetically.
Table 3:
Variations to Sex-Based Data Reporting Across Surgical Specialties in 2017–2018.
| Specialty | Total Articles | Sex-Based Data Reporting N (%) | P value | Statistical Analysis of Sex-Based Results N (%) | P value | Discussion of Sex-Based Results N (%) | P value | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | 2011–2012 | 2017–2018 | ||||
| Breast | 56 | 121 | 3 (5) | 6 (5) | NS | 3 (5.4) | 4 (3) | NS | 1 (2) | 5 (4) | NS |
| Surg Education/Training | 98 | 208 | 22 (23) | 26 (13) | 0.03 | 19 (19) | 27 (13) | NS | 18 (18) | 21 (10) | NS |
| Orthopedics | 9 | 2 (22) | 2 (22) | 1 (11) | |||||||
| Plastic Surgery | 22 | 5 (23) | 4 (18) | 1 (5) | |||||||
| General Surgery | 205 | 460 | 68 (33) | 140 (30) | NS | 61 (30) | 136 (30) | NS | 42 (21) | 43 (9) | 0.0001 |
| Urology | 3 | 1 (33.3) | 0 (0) | 0 (0.0) | |||||||
| Trauma/critical care | 118 | 292 | 40 (34) | 105 (36) | NS | 35 (30) | 94 (32) | NS | 23 (20) | 26 (9) | 0.004 |
| Transplant | 49 | 88 | 20 (41) | 32 (36) | NS | 13 (27) | 32 (36) | NS | 10 (20) | 8 (9) | NS |
| Cardiac | 18 | 30 | 3 (17.7) | 11 (36.7) | NS | 3 (17.7) | 11 (36.7) | NS | 2 (11.8) | 4 (13.3) | NS |
| Burn | 21 | 8 (38) | 8 (38) | 1 (5) | |||||||
| Vascular | 42 | 65 | 15 (36) | 26 (40) | NS | 12 (29) | 27 (42) | NS | 9 (21) | 13 (20) | NS |
| Pediatric Surgery | 34 | 132 | 13 (32) | 53 (40) | NS | 10 (29) | 51 (39) | NS | 8 (24) | 16 (12) | NS |
| Colorectal | 137 | 213 | 69 (50) | 89 (42) | NS | 66 (48) | 91 (43) | NS | 34 (25) | 29 (14) | 0.01 |
| Neurosurgery | 7 | 3 (43) | 3 (43) | 0 (0.0) | |||||||
| Surgery in General | 72 | 49 | 24 (33) | 21 (43) | NS | 23 (32) | 22 (45) | <0.0001 | 16 (22) | 9 (18) | 0.03 |
| Endocrine | 126 | 174 | 60 (48) | 75 (43) | NS | 53 (42) | 79 (42) | NS | 49 (39) | 23 (13) | <0.0001 |
| Surgical Oncology | 250 | 471 | 124 (50) | 226 (48) | NS | 102 (41) | 213 (45) | NS | 64 (26) | 45 (10) | <0.0001 |
| Bariatric | 33 | 54 | 5 (15) | 26 (48) | 0.002 | 5 (15) | 24 (44) | 0.005 | 5 (15) | 7 (13) | NS |
| Otolaryngology/HNS | 4 | 3 (50) | 2 (50) | 0 (0) | |||||||
| Thoracic | 39 | 51 | 18 (46) | 27 (53) | NS | 18 (46) | 2 (53) | NS | 11 (28) | 5 (10) | 0.03 |
Individual data was not available for specialties that contain a grey box over 2011–12 data. Data was arranged in ascending order by percentage of sex-based reporting in 2017–2018.
Variations in Sex Distribution.
There were 213,500,066 humans studied in the 5 surgical journals in 2017–2018, almost 100 million more than studied in 2011–2012.10 Out of these participants, significantly less were male (78,880,581, 37% vs 45,638,833/115,377,213, 40% in 2011–201210, p<0.0001), significantly less were female,(105,254,080, 49% vs 58,805,665/115,377,213, 51%10, p<0.0001), and significantly more participants were unspecified (29,365,405,14% vs 10,459,739/115,377,213, 9%10, p<0.0001) as compared to 2011–2012 (Figure 4A).
Figure 4:
Absolute number of male or female human or animal participants. (A) Absolute number of male, female, or unspecified human subjects in 2011–2012 vs 2017–2018. *p<0.001. Numbers are represented as millions. (B) Absolute number of male, female, or unspecified animal subjects in 2011–2012 vs 2017–2018. *p<0.05, ***p<0.001.
There was a wide distribution in the number of males (23%-55%), females (42%-61%) and unspecified subjects (2%-35%) among surgical journals. Four of the journals included more male participants, while 3 included more unspecified subjects (p<0.0001) (Supplemental Table 2).
Significant variations in sex distribution between specialties persisted in 2017–2018 which was consistent with findings from 2011–2012. Pediatric surgery, thoracic surgery, and trauma/critical care continued to publish articles with >50% male participates in 2017–2018 (p<0.0001) with the addition of vascular, urology, and burns. Endocrine, bariatrics, and breast surgery continued to use >50% female participants (p<0.0001). In contrast, cardiac surgery in 2017–2018 had >50% female patients, which varied from 2011–2012 when cardiac surgery studied >50% male subjects (p<0.0001).10 There were no specialties that did not specify the sex in >50% of study subjects, which was an improvement from 2011–2012 when transplant, general surgery, and surgical education had >50% unspecified subjects.10
Variations in Sex Matching.
More articles matched the sex of the participant by at least 50% (example: 2 males to 1 female or visa versa) in 2017–2018 when compared to 2011–2012 (50%, n=1266 vs 45%, n=598/1303, p<0.0001).10 Only 1% of articles (n=19) matched the sex of the participant by 100% (example: 1 male to 1 female), down from 2% (n=25/1303) in the years prior (p=0.02) (Figure 5A).10 There was also variation in sex matching across surgical specialties in which colorectal surgery, pediatric surgery, and surgical oncology continued to have the greatest percentage of articles that matched sex by at least 50%.
Figure 5:
Sex matching of participants in the articles. (A) Percentage of sex matching in human articles in 2011–2012 vs 2017–2018. *p<0.05, **p<0.001. (B) Direct comparison of sex matching among human, animal, and cell studies in 2017–2018.
Trends in Animal and Cell Studies
Sex of the Participant.
Of the 2,927 articles reviewed in 2017–2018, 367 (12.5%) used animals and/or cells, with 323 reporting the use of animals (88% vs 531/618, 86% in 2011–201211) and 88 reporting the use of cells (24% vs 118/618,19%).11 Of the 323 articles that contained animals in 2017–2018, a comparable number of articles reported data based on sex vs those reported in 2011–2012 (256/323, 79% vs 414/531, 78%11, p=0.67) (Figure 2A). Fewer were male-only studies (181/256, 71% vs 331/531, 80%11, p=0.03), more were female-only studies (58/256, 23% vs 70/531, 17%11, p=0.001), and more included both male and female animals (17/256, 7% vs 11/531, 3% in 2011–201211, p=0.002) (Figure 2B). Of the 88 articles that contained cells in 2017–2018, significantly fewer articles stated the sex of the cells studied compared to 2011–2012 (4/88, 5% vs 28/118, 24%11, p=0.0001) (Figure 2A). Out of the 4 articles that stated the sex of the cells in 2017–2018, there was not a significant difference between male-only studies (2/4, 50% vs 20/28, 71%11, p=0.57) or female-only studies (1/4, 25% vs 6/28, 21%11, p=0.99), but a significant difference existed in studies that included both males and females (1/4, 25% vs 2/28, 7%11, p=0.0001).
Regarding journal analysis, we excluded JAMA Surgery due to the limited number of animal and/or cell articles (n=1/206, 0.4%), which was also done in 2011/12. Similar to 2011–201211, the remaining 4 journals utilized a majority of male-only animal studies (Supplemental Table 3) with varying distribution of sex-based reporting. A comparable number of domestic animal studies reported the sex in 2017–2018 (87/115, 76% vs 145/185, 78%11, p<0.0001) consistent across male-only (70, 64% vs 103/145, 71%11, p=0.28), female-only (31, 29% vs 35/145, 24%11 p=0.47), and male/female articles (8, 7% vs 7/135, 5%11p=0.43) (Supplemental Figure 1A–B). There were similar rates of sex-based reporting amongst international studies (135, 82% vs 272/346, 79%11, p=0.43) with variability regarding sex inclusion between years (Supplemental Figure 2A–B) and continued male-only animal majority (p=0.01).
Sex-Based Reporting of Data.
Out of the 260 animal and/or cell articles that stated the sex in 2017–2018, significantly more articles studied both males and females (21, 8% vs 13/419, 3%11, p=0.0001). However, sex-based reporting, analysis, and discussion of results remained minimal when categorized by animal only (2%, 1%, 1%) (n=5, 3, 3) and cell only (6%, 6%, 1%) (n= 5, 5, 1) articles (Figure 3).
Variations in Sex Distribution.
Overall, there were less animals studied in 2017–2018 compared to 2011–2012 (n=16,004 vs 23,385). Participant inclusion remained mostly male (n=11,041, 69% vs 15,884 68%11, p=0.50), with no change to the percent of female animals (2,124, 13% vs 2,975, 13%, p=0.11), and minimal change to unspecified participants (2,839, 18% vs 4,526, 19%, p<0.0001) (Figure 4B).
The absolute number of male and female animal subjects varied between journals (excluding JAMA Surgery with animal/cell manuscript n=1) (Supplemental Table 3). Overall, all 4 surgical journals utilized >80% male subjects.
In the 2017–2018 animal cohort, surgical oncology, transplant, vascular surgery, general surgery, and trauma/critical care published the majority of animal articles, all of which utilized >50% male animals.
Variations in Sex Matching.
Regarding animal studies, only 14 of 323 (4%) articles matched animal sex by >50% and only 7 articles (2%) matched the animals by 100%.
DISCUSSION
We present evidence of continued sex bias in surgical research with marginal changes over the past 5-year period. Overall, the percentage of human and animal articles that stated the sex of the participant remained similar, while fewer cell-based articles stated the sex of the cell. The majority of human articles that stated the sex of the participant still included both sexes, while animal and cell articles included mostly males. Despite NIH mandates, there was not a significant improvement in the number of articles that reported, analyzed, or discussed results based on sex. Only marginal changes to sex matching of participants was observed in clinical and biomedical studies and overall remained poor. Despite the overall general equality with respect to the number of human subjects included, there continues to be significant bias in data reporting. Sex continues to be an important biological variable that remains poorly represented in the reporting, analysis, and discussion of data pertaining to surgical research today.
A major concern lies in the fact that sex bias is still very prevalent in basic science research. This early bias undoubtedly sets the stage for introduction of continued bias in human studies and clinical trials which only perpetuates the problem of female underrepresentation and limits widespread clinical translatability, especially considering instances where males and females have been proven to respond to medication differently.31 In 2015, the NIH Office of Research on Women’s Health stated their goals in the strategic plan of 2020 to “increase sex differences research in basic science studies” and “incorporate these findings of sex/gender differences in design and application of therapeutic drugs.”32 This plan also individualized objectives to include identification of sex differences in gene expression and specific exploration of sex differences in cell lines and animal models.32 In an effort to standardize sex-based reporting, the Sex and Gender Equity in Research (SAGER) guidelines33 were published in 2016 but our research, and that of others, continues to highlight the disproportionally male-only animal models used in studies today. The majority of cell-based studies in 2017–2018 (95%) and 2011–2012 (76%) did not state the sex of the cell studied. This bias continued among animal studies which, again in 2017–2018, utilized more male animals overall and included more male-only manuscripts.11 The vast majority of animal and cell studies did not report data based on sex (98% and 94%, respectively) and even less frequently analyzed or discussed sex-based results. Hughes investigated the prevalence of male-only drug studies and noted marginal improvement in male-only rat manuscripts over a 10-year period (82% to 85%).13 Ramirez et al found no change in sex-based reporting in cardiovascular journals and significantly more male-only manuscripts over a similar time period.34 Unfortunately, there have been minimal improvements regarding sex bias in basic science with continued male dominance and underreporting of sex, with the majority of cell studies failing to state the sex of the cell studied. This limits the clinical translatability of early studies since the vast majority of surgical human studies are performed on both males and females.
We are already aware that sex bias exists in human clinical research and clinical trials.10,35,36 In an effort to combat this bias, the NIH amended their inclusion of women policy in 2017 to mandate sex-based analysis in clinical trials.3 Our research shows that despite the continued inclusion of both males and females in human surgical studies, the utilization of sex as a biological variable in human research is not improving. Over the past 5 years we observed various small modifications but no significant improvement in reporting or analysis by sex. One of our most concerning findings was the limited number of manuscripts that reported, analyzed, and discussed sex-based results. Additionally, we found subjects like surgical education to include significantly fewer articles that stated the sex of the human included and reported data based on sex. This could hold significant implications when training future surgeons considering the possible differences in male and female learning styles and potentially limit the benefit of these methods of education. These findings are not specific to the field of surgery. Prakash et al specifically investigated bias in clinical trials and found sex bias exists equally among both industry-funded and NIH-funded trials.35 Geller et al echoed these findings noting that only 26% of randomized control trials included sex as a variable or statistically analyzed results based on sex, with no significant improvement in reporting between 2004 and 2009.36 There was little improvement in sex inclusion, utilization of sex as an independent variable, or reporting sex as a primary outcome in emergency department studies published 2014–2017 when compared to 2011.24 Concerningly, we have even found evidence of bias in manuscripts that study disease with female prevalence such as thyroid or cardiovascular disease, with the majority of these studies only utilizing male animals.11 The established implications of this bias remain numerous and little progress has been made to combat this disparity in research reporting and analysis.
There is continued hope for the future. In 2019 the NIH, along with the Office of Research on Women’s Health, released their updated strategic plan for female inclusion in research. This plan includes specific goals in basic, preclinical, and clinical research environments spanning a period of 5 years (2019–2023)37 with integration and advancement of sex-based research specifically relating to women’s health and treatment tailored to female-specific pathophysiology and therapeutic response.38 We support these goals and hope for continued mandates and strict regulations regarding sex-based data reporting and analysis. We suggest that all journals require authors to state the sex of the cell, animal, or human studied or to provide justification for a single-sex study, and urge the NIH to improve oversight to current requirements of sex disclosure in grant submissions. We urge research teams conducting cell, animal, and human research to understand the importance of sex as a biological variable and to place immediate attention on sex matching and inclusion in research studies.
Our study is not without limitations. Similar to our review in 2011–2012, this data is not inclusive of all surgical literature or specialties published during this time period. Additionally, we were unable to differentiate between sex and gender and assumed the reports were again based on phenotypic sex alone. We did not analyze results based on funding source and therefore were not able to draw conclusions based solely on the NIH mandates and articles reviewed for our report. We did not abstract whether a study was retrospective versus prospective, as our intent was to abstract the same data that was collected in the 2011–2012 study. Lastly, we were not able to perform a deeper analysis into determining if the percent of sex matching of subjects matched the sex prevalence of a specific disease. While this type of analysis would be very interesting, it is beyond the scope of this current study.
CONCLUSIONS
Sex bias still exists in biomedical and clinical surgical research. There have been inconsistent and marginal changes to sex inclusion over the last 5 years with continued unequal representation of males and females, underreporting of sex as a biological variable, and male dominance in animal and cell studies. This disparity is a continued concern to all aspects of medicine and requires significant attention as we move into the future.
Supplementary Material
ACKNOWLEDGEMENTS
We thank Neel Mansukhani and Dustin Yoon for assistance with the follow-up study, data collection templates, and general support. We thank Deborah Hepp for assistance with proofing the manuscript.
FUNDING/SUPPORT
This work was supported in part by the National Institutes of Health grant numbers 2T32GM008450–26, T32GM086330–08, and T32HL007149–42.
Footnotes
COI/DISCLOSURES
The authors have no related conflicts of interest to declare.
REFERENCES
- 1.Women’s health. Report of the Public Health Service Task Force on Women’s Health Issues. Public Health Rep. 1985;100(1):73–106. [PMC free article] [PubMed] [Google Scholar]
- 2.Freedman LS, Simon R, Foulkes MA, et al. Inclusion of women and minorities in clinical trials and the NIH Revitalization Act of 1993--the perspective of NIH clinical trialists. Control Clin Trials. 1995;16(5):277–285; discussion 286–279, 293–309. [DOI] [PubMed] [Google Scholar]
- 3.Amendment: NIH Policy and Guidelines on the Inclusion of Women and Minorities as Subjects in Clinical Research. In: Health NIo, ed. Notice Number: NOT-OD-18–0142017. [Google Scholar]
- 4.Beery AK, Zucker I. Sex bias in neuroscience and biomedical research. Neurosci Biobehav Rev. 2011;35(3):565–572. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lee SK. Sex as an important biological variable in biomedical research. BMB Rep. 2018;51(4):167–173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Klein SL, Schiebinger L, Stefanick ML, et al. Opinion: Sex inclusion in basic research drives discovery. Proc Natl Acad Sci U S A. 2015;112(17):5257–5258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Clayton JA, Collins FS. Policy: NIH to balance sex in cell and animal studies. Nature. 2014;509(7500):282–283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Enhancing Reproducibility through Rigor and Transparency. National Institutes of Health;2015. [Google Scholar]
- 9.Consideration of Sex as a Biological Variable in NIH-funded Research. National Institutes of Health;2015. [Google Scholar]
- 10.Mansukhani NA, Yoon DY, Teter KA, et al. Determining If Sex Bias Exists in Human Surgical Clinical Research. JAMA Surg. 2016;151(11):1022–1030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Yoon DY, Mansukhani NA, Stubbs VC, Helenowski IB, Woodruff TK, Kibbe MR. Sex bias exists in basic science and translational surgical research. Surgery. 2014;156(3):508–516. [DOI] [PubMed] [Google Scholar]
- 12.Tisosky A, Logan C, Brook EM, Xu J, Matzkin E. Reporting the Influence of Sex in Research: Trends at AAOS Annual Meetings. J Am Acad Orthop Surg. 2019;27(3):e112–e117. [DOI] [PubMed] [Google Scholar]
- 13.Hughes RN. Sex still matters: has the prevalence of male-only studies of drug effects on rodent behaviour changed during the past decade? Behav Pharmacol. 2019;30(1):95–99. [DOI] [PubMed] [Google Scholar]
- 14.Franconi F, Campesi I. Pharmacogenomics, pharmacokinetics and pharmacodynamics: interaction with biological differences between men and women. Br J Pharmacol. 2014;171(3):580–594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Franconi F, Campesi I. Sex Impact on Biomarkers, Pharmacokinetics and Pharmacodynamics. Curr Med Chem. 2017;24(24):2561–2575. [DOI] [PubMed] [Google Scholar]
- 16.Gandhi M, Aweeka F, Greenblatt RM, Blaschke TF. Sex differences in pharmacokinetics and pharmacodynamics. Annu Rev Pharmacol Toxicol. 2004;44:499–523. [DOI] [PubMed] [Google Scholar]
- 17.Freire AC, Basit AW, Choudhary R, Piong CW, Merchant HA. Does sex matter? The influence of gender on gastrointestinal physiology and drug delivery. Int J Pharm. 2011;415(1–2):15–28. [DOI] [PubMed] [Google Scholar]
- 18.Mori K, Mostafaei H, Enikeev DV, et al. Differential Effect of Sex on Outcomes after Radical Surgery for Upper Tract and Bladder Urothelial Carcinoma: A Systematic Review and Meta-Analysis. J Urol. 2020:101097JU0000000000000788. [DOI] [PubMed] [Google Scholar]
- 19.Exploring the biological contributions to human health: does sex matter? J Womens Health Gend Based Med. 2001;10(5):433–439. [DOI] [PubMed] [Google Scholar]
- 20.Mnatzaganian G, Hiller JE, Braitberg G, et al. Sex disparities in the assessment and outcomes of chest pain presentations in emergency departments. Heart. 2020;106(2):111–118. [DOI] [PubMed] [Google Scholar]
- 21.Huckaby LV, Seese LM, Sultan I, et al. The Impact of Sex on Outcomes after Revascularization for Multivessel Coronary Disease. Ann Thorac Surg. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zhou W, Li Y, Liu X, et al. Sex-specific relationship between adult height and the risk of stroke: A dose-response meta-analysis of prospective studies. J Clin Hypertens (Greenwich). 2019;21(2):262–270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Westerman S, Wenger N. Gender Differences in Atrial Fibrillation: A Review of Epidemiology, Management, and Outcomes. Curr Cardiol Rev. 2019;15(2):136–144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Safdar B, Ona Ayala KE, Ali SS, et al. Inclusion of Sex and Gender in Emergency Medicine Research-A 2018 Update. Acad Emerg Med. 2019;26(3):293–302. [DOI] [PubMed] [Google Scholar]
- 25.de Lacy N, McCauley E, Kutz JN, Calhoun VD. Sex-related differences in intrinsic brain dynamism and their neurocognitive correlates. Neuroimage. 2019;202:116116. [DOI] [PubMed] [Google Scholar]
- 26.GAO-01–286R Drugs Withdrawn From Market. In:2001.
- 27.Farkas RH, Unger EF, Temple R. Zolpidem and driving impairment--identifying persons at risk. N Engl J Med. 2013;369(8):689–691. [DOI] [PubMed] [Google Scholar]
- 28.Irwin DE, Palmer LA. When Gender Matters: Impact of Fda Safety Warning on Dispensed Zolpidem Dose. Value Health. 2014;17(7):A415. [DOI] [PubMed] [Google Scholar]
- 29.Kuehn BM. FDA warning: Driving may be impaired the morning following sleeping pill use. JAMA. 2013;309(7):645–646. [DOI] [PubMed] [Google Scholar]
- 30.Roth T, Steinberg F, Singh NN, Moline M. Gender influences on efficacy and safety of sublingual zolpidem tartrate for middle-of-the-night awakening in insomnia. Hum Psychopharmacol. 2014;29(1):25–30. [DOI] [PubMed] [Google Scholar]
- 31.Anderson GD. Sex and racial differences in pharmacological response: where is the evidence? Pharmacogenetics, pharmacokinetics, and pharmacodynamics. J Womens Health (Larchmt). 2005;14(1):19–29. [DOI] [PubMed] [Google Scholar]
- 32.Office of Research on Women’s Health NIH, US Department of Health and Human Services. Moving Into The Future With New Dimensions And Strategies: A Vision For Women’s Health Research. 2015. [Google Scholar]
- 33.De Castro P, Heidari S, Babor TF. Sex And Gender Equity in Research (SAGER): reporting guidelines as a framework of innovation for an equitable approach to gender medicine. Commentary. Ann Ist Super Sanita. 2016;52(2):154–157. [DOI] [PubMed] [Google Scholar]
- 34.Ramirez FD, Motazedian P, Jung RG, et al. Sex Bias Is Increasingly Prevalent in Preclinical Cardiovascular Research: Implications for Translational Medicine and Health Equity for Women: A Systematic Assessment of Leading Cardiovascular Journals Over a 10-Year Period. Circulation. 2017;135(6):625–626. [DOI] [PubMed] [Google Scholar]
- 35.Prakash VS, Mansukhani NA, Helenowski IB, Woodruff TK, Kibbe MR. Sex Bias in Interventional Clinical Trials. J Womens Health (Larchmt). 2018;27(11):1342–1348. [DOI] [PubMed] [Google Scholar]
- 36.Geller SE, Koch AR, Roesch P, Filut A, Hallgren E, Carnes M. The More Things Change, the More They Stay the Same: A Study to Evaluate Compliance With Inclusion and Assessment of Women and Minorities in Randomized Controlled Trials. Acad Med. 2018;93(4):630–635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Advancing Science for the Health of Women: The 2019–2023 Trans-NIH Strategic Plan for Women’s Health Research. Bethesda, MD: The National Institutes of Health Office of Research on Women’s Health;2019. [Google Scholar]
- 38.Arnegard ME, Whitten LA, Hunter C, Clayton JA. Sex as a Biological Variable: A 5-Year Progress Report and Call to Action. J Womens Health (Larchmt). 2020. [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.