Article Text
PDF
Abstract
Aims Follicle-stimulating hormone (FSH) is associated with higher risks of metabolic syndrome and diabetes in menopausal women. We aimed to investigate whether FSH was associated with the lipid profile in women older than 55 years.
Design The data were obtained from a cross-sectional study.
Participants Our data were from the Survey on Prevalence in East China for Metabolic Diseases and Risk Factors (China, including Shanghai and Zhejiang, Jiangxi and Anhui provinces). A total of 1795 women older than 55 years were selected.
Methods Morning serum sex hormones and lipid profiles were measured. Linear and logistic regression analyses were used to analyse the data.
Results Lower FSH was associated with lower high-density lipoprotein cholesterol (HDL-C) and higher triglycerides (TG), total cholesterol (TC)/HDL-C ratio and low-density lipoprotein cholesterol (LDL-C)/HDL-C ratio (all p for trend <0.05) after adjusting for age and other sex hormones. After further adjustment for body mass index, diabetes and hypertension, the associations of FSH with the lipid profile weakened, but the associations of FSH quartiles with HDL-C and the TC/HDL-C ratio were still significant (both p for trend <0.05). Compared with women in the highest FSH quartile, the odds of low HDL-C (HDL-C<1.04 mmol/L) in women in the lowest FSH quartile were 5.25 (95% CI 1.60 to 17.26) (p for trend <0.05) in the fully adjusted model, and the odds of TC≥6.22 mmol/L, TGs≥2.26 mmol/L and LDL-C≥4.14 mmol/L were not significant. Luteinising hormone did not show a significant association with dyslipidaemia.
Conclusion Lower FSH was associated with a worse lipid profile in women older than 55. Diabetes, adiposity and hypertension mostly explained the association of FSH with TGs and the LDL-C/HDL-C ratio but only partially explained the associations of FSH with HDL-C and the TC/HDL-C ratio.
- Lipid disorders
- General diabetes
- General endocrinology
Data availability statement
Data are available on reasonable request. No data are available.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
First, the study presented in detail the association between follicle stimulating hormone and the lipid profile for the first time in a large sample and focused on several possible explanatory factors.
Second, the study was conducted with participants from the general population as opposed to those from a clinical population.
Because the nature of this study was cross-sectional, a causal relationship could not be determined among the variables. Moreover, final menstruation information was not obtained, so based on previous studies and the natural menopause age in Chinese women, we used an age proxy of 55 years. Meanwhile, caused by the sample size, we need to verify our results with a larger sample size in the future.
Introduction
Atherosclerotic cardiovascular disease (CVD) ranks highest among the leading causes of death and disability-adjusted life-years lost in developed countries, most developing countries and China.1 ,2 It is known that elevated serum concentrations of total cholesterol (TC) or low-density lipoprotein cholesterol (LDL-C) and triglycerides (TG) and decreased concentrations of high-density lipoprotein cholesterol (HDL-C) are major risk factors for CVD.3–5 The overall prevalence of dyslipidaemia reached 53% and 34.0% in the USA6 and China, respectively.2 Additionally, postmenopausal females have a higher prevalence of dyslipidaemia than that in males.2 Thus, abnormal lipid profiles are a major concern globally, especially in postmenopausal women, who are considered a special population in European and American dyslipidaemia management guidelines.7 8
Endogenous sex hormones mainly act on the reproductive system, but their association with lipid metabolism is being revealed. In postmenopausal women, endogenous testosterone (T) levels may be part of a proatherogenic profile.9 10 Higher sex hormone binding globulin levels are associated with favourable lipid profiles.11 12 However, until now, the association between follicle-stimulating hormone (FSH), the sex hormone released from the pituitary gland, and the lipid profile has not been fully studied in population-based research. A recent study reported that in 400 Chinese postmenopausal women, participants with higher serum FSH had higher TC and LDL-C levels.13 Our previous studies have found that in postmenopausal women, lower FSH was independently associated with diabetes and nonalcoholic fatty liver disease, which were closely related to abnormal lipid profiles.14 15 Therefore, the relationship between FSH and the lipid profile is controversial in postmenopausal women and should be further analysed.
Using data from an observational investigation named Survey on Prevalence in East China for Metabolic Diseases and Risk Factors (SPECT-China) in 2014, we aimed to analyse the association between FSH and the lipid profile in Chinese women older than 55 years.
Materials and methods
Study population
SPECT-China was a cross-sectional investigation (ChiCTR-ECS-14005052, www.chictr.org.cn). Briefly, a stratified cluster sampling by urban/rural areas and economic status was used to select a representative sample from the general population at 16 sites across urban and rural areas in Shanghai, Zhejiang, and Jiangxi Province. In urban areas, one city with low economic status and one with high economic status were randomly selected. Then three districts were randomly selected from each city, and one community was randomly sampled from each district. In rural areas, six villages with low economic status and six with high economic status were randomly selected.16 Detailed sampling information was described in a previous study.17 The data were from the SPECT-China study from February 2014 to May 2016. In total, 6899 participants were recruited. Ninety-seven percent of Chinese women above 55 are postmenopausal.18 In our sample, there were 1863 women who were older than 55 and were not using hormone replacement therapy. Women with FSH<25.0 IU/L (n=42), missing FSH values (n=6) and taking lipid-lowering drugs (n=20) were excluded. Finally, the current study was based on a total of 1795 women older than 55 years.
Data collection
Clinical, anthropometric and laboratory measurements
All participants were interviewed about their demographic characteristics, medical history and lifestyle risk factors. Body weight and height were measured with a Seca scale (Hamburg, Germany) and a fixed stadiometer. Body mass index (BMI) was calculated using the measured height and fasting body weight. Blood pressure was measured with the use of standard methods described previously.19
After fasting for at least 8 hours, blood samples were collected from 7:00 to 10:00 hours. They were stored at −20℃ when collected and shipped within 2–4 hours of collection by air on dry ice to a central laboratory, which was certified by the College of American Pathologists. Glycated haemoglobin (HbA1c) was assessed by high-performance liquid chromatography (MQ-2000PT, China). Fasting plasma glucose, TC, TG, HDL-C and LDL-C were measured by the hexokinase method, the Cholesterol oxidase-peroxidase-4-aminoantipyrine and phenol (CHOD-PAP) method, the Glycerol phosphate oxidase-peroxidase-4-aminoantipyrine and phenol (GPO-PAP) method, and homogeneous methods, respectively (AU 680 analyser, Beckman Coulter, USA).
Total T, oestradiol (E2), FSH and luteinising hormone (LH) were measured by chemiluminescence (SIEMENS Immulite 2000, Germany). The minimal detectable limits for total T, E2, FSH and LH were 0.7 nmol/L, 73.4 pmol/L and 0.1 IU/L, respectively. Samples with values below the minimal detectable limit were given a value midway between zero and the minimal detectable limit for the analyses.20 Both the interassay and intra-assay coefficients of variation for each hormone were less than 10%.
Definition of variables
Elevated TC (TC≥6.22 mmol/L), elevated TG (TG≥2.26 mmol/L), elevated LDL-C (LDL-C≥4.14 mmol/L) and low HDL-C (HDL-C<1.04 mmol/L) were defined according to the National Cholesterol Education Program-Adult Treatment Panel III.21 Meanwhile, the population who already use dyslipidaemia medication was excluded from this study.
Diabetes was defined as a previous diagnosis by healthcare professionals, fasting plasma glucose 7.0 mmol/L or higher, or HbA1c 6.5% or higher. Hypertension was identified by systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or a self-reported previous diagnosis of hypertension by a physician.22 23
Statistical analysis
All statistical analyses were performed with IBM SPSS Statistics, V.22 (IBM). All analyses were two sided. A p<0.05 indicated a significant difference unless otherwise noted. Continuous variables were expressed as the mean (SD), and categorical variables were described as a percentage (%).
The association of FSH (independent variable) with LDL-C, HDL-C, non-HDL-C, TC, TG, TC/HDL-C ratio and LDL-C/HDL-C ratio (dependent variables) was assessed by linear regression. The base model (model 1) included terms for age, T, E2 and LH. Model 2 included terms for model 1, BMI, hypertension and diabetes. Since TGs were nonnormally distributed, they were log transformed. The results were expressed as unstandardised coefficients (SEs).
FSH and LH were divided into quartiles, with the first quartile representing the lowest one and the fourth quartile the highest. ORs and 95% CIs were calculated using multinomial logistic regression to determine the risk of elevated TC, TG and LDL-C and low HDL-C for each quartile of FSH and LH, using the highest quartile as the reference. Models were the same as those in linear regression.
Patient and public involvement
Patients and the public were not involved in the development of research questions, design of the study, recruitment and conduct of the study or dissemination of the study results.
Results
Characteristics of the study population
The characteristics of participants by FSH quartiles are summarised in online supplemental table 1. The quartile ranges of FSH were ≤49.3, 49.4–63.4, 63.5–81.2 and ≥81.3 Through FSH quartiles, women in the lower quartile had significant increasing trends of HDL-C, non-HDL-C and LH (p for trend<0.05). TG, TC/HDL-C ratio, LDL-C/HDL-C ratio, BMI and prevalence of diabetes and hypertension significantly decreased in the lower FSH quartile group (p for trend<0.05). LDL-C and TC did not show a significant trend (p for trend>0.05).
Supplemental material
Association of FSH with lipid profile (continuous variables)
Table 1 summarises the results of the linear regression models studying the association of FSH with LDL-C, HDL-C, non-HDL-C, TC, TG, TC/HDL-C ratio and LDL-C/HDL-C ratio. In the base model (table 1, model 1), higher FSH quartiles were associated with higher HDL-C and lower log-TG, TC/HDL-C ratio and LDL-C/HDL-C ratio (all p for trend<0.05). LDL-C, non-HDL-C and TC did not show a significant association with FSH quartiles. After further adjustment for metabolic factors (BMI, diabetes and hypertension), the associations of FSH with log TG and LDL-C/HDL-C ratio weakened such further that no significant relationship was observed (table 1, model 2). The association of FSH quartiles with HDL-C and the TC/HDL-C ratio was still significant (both p for trend<0.05) (table 1, model 2). Moreover, LH quartiles did not show a significant association with these lipid indices in either model (table 2).
Association of circulating FSH with lipid profile (continuous variables)
Association of circulating LH with lipid profile (continuous variables)
Prevalence of dyslipidaemia
The prevalence of elevated LDL-C through FSH quartiles was 10.9%, 10.0%, 7.8% and 8.3%, respectively, and that of elevated TC was 23.1%, 20.9%, 17.4% and 21.3%. The prevalence of elevated LDL-C and TC did not have a significant trend across FSH quartiles (figure 1).
Prevalence of abnormal lipid profile according to FSH quartiles. FSH, follicle-stimulating hormone; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglycerides.
Association of FSH with dyslipidaemia (categorical variables)
Table 3 shows the association of FSH with elevated TC (TC≥6.22 mmol/L), elevated TG (TG≥2.26 mmol/L), elevated LDL-C (LDL-C≥4.14 mmol/L) and low HDL-C (HDL-C<1.04 mmol/L) by logistic regression analyses. In model 1 adjusted for age and other sex hormones, compared with women in the highest quartile of FSH, the ORs of elevated LDL-C, low HDL-C, elevated TGs and elevated TC in women in the lowest FSH quartile were 1.00 (95% CI 0.53 to 1.88), 6.41 (95% CI 1.97 to 20.92), 1.37 (95% CI 0.83 to 2.26) and 0.92 (95% CI 0.59 to 1.44), respectively. Only low HDL-C showed a significant association with FSH quartiles (p for trend<0.05). Further adjusting for other metabolic factors (BMI, diabetes and hypertension) weakened the association between FSH and low HDL-C, but it still had significance (OR 5.25, 95% CI 1.60 to 17.26) (table 3, model 2). LH quartiles did not show a significant association with dyslipidaemia in either model (table 4). A previous study found a significant association between FSH and diabetes,15 and we further performed diabetes and nondiabetes subgroup analyses. The results showed that the associations between FSH and HDL-C were significant in participants without diabetes (p for trend<0.05) and marginal in participants with diabetes (p for trend 0.063).
ORs (95% CI) for association of circulating FSH with dichotomised lipid profile
ORs (95% CI) for association of circulating LH with dichotomised lipid profile
Discussion
In women older than 55 with a high probability of postmenopause, we discovered that participants with lower FSH had lower HDL-C and higher TG, TC/HDL-C ratio and LDL-C/HDL-C ratio. Metabolic factors including diabetes, BMI and hypertension mostly explained the association of FSH with TGs and the LDL-C/HDL-C ratio but only partially explained the associations of FSH with HDL-C and the TC/HDL-C ratio. To the best of our knowledge, this is the first study to detect the association between FSH and lipid profile in detail in a large population sample.
Dyslipidaemia is a well-known independent risk factor for CVD and cerebral ischaemic stroke.24 With rapid socioeconomic development, urbanisation, and lifestyle and diet changes, the prevalence of dyslipidaemia has also been rapidly increasing.25 Dyslipidaemia is closely related to other metabolic comorbidities, such as diabetes, obesity and fatty liver. For example, the features of diabetic dyslipidaemia are increased TG concentration, reduced HDL-C concentration, and increased LDL-C concentration.26 Non-alcoholic fatty liver disease (NAFLD) is also characterised by atherogenic dyslipidaemia and HDL dysfunction.27 According to recent studies, FSH was found to be significantly associated with metabolic diseases.13–15 28 Our previous studies showed that compared with postmenopausal women in the highest quartile of FSH, the odds of diabetes and NAFLD in postmenopausal women in the lowest quartile of FSH was increased by 96% and 84%–194%, respectively.14 15 In this study, we further demonstrated that lower FSH was significantly associated with lower HDL-C and higher TG, TC/HDL-C ratio and LDL-C/HDL-C ratio after adjusting for age, total T, E2 and LH. These results agree with recent studies about the association of FSH with metabolic syndrome.28 29 The above information lends credibility to the findings on FSH and lipid profiles.
FSH is mainly secreted by the pituitary gland, and it regulates the development and function of reproductive organs after being released into the blood; therefore, it is not surprising that FSH receptors (FSHRs) are mainly located throughout the reproductive system.30 However, FSHRs have also been found in nonreproductive organs and tissues such as mouse osteoclasts,31 human blood vessels,32 chicken adipose tissue33 and human liver,13 exerting nonreproductive functions. Interestingly, a recent study showed that FSH interacted with FSHR in hepatocytes and reduced LDL receptor levels, which subsequently attenuated the endocytosis of LDL-C in a mouse model. This might have led to an elevated circulating LDL-C level in postmenopausal women with a mean age of 48.6 years.13 In chicken and mouse models, studies have also found that FSH promotes lipid biosynthesis in adipose tissue and visceral fat accumulation by upregulating FSHR mRNA expression and the Gai/Ca2+/cAMP-response-element-binding protein pathway.33 34 These results seem to contradict those of previous epidemiological studies. A negative correlation between FSH concentrations and BMI values was observed.22 35–37 Additionally, every 1 SD decrement of FSH is associated with a 3.83-fold increased risk of metabolic syndrome, in which adiposity is the core element.28 Moreover, in observational cohort studies, in obese women, the increase in FSH was significantly attenuated, especially after the final menstrual period, and weight loss increased FSH among overweight and obese postmenopausal women.38 In regard to opposing study results, there is one explanation: FSH may induce fat accumulation and lipid biosynthesis in the early stage of postmenopause, but the development of metabolic diseases, especially obesity, would conversely suppress the release of FSH and LH. This hypothesis needs further investigation for confirmation.
It is not surprising that common metabolic factors, including diabetes, BMI and hypertension, largely mediated the association between FSH and the lipid profile. Other endogenous sex hormones may also have a role in this association. For example, total testosterone may induce an increase in TC, LDL-C or TG and a decrease in HDL-C,10 39 although some studies found no such effect.11 40 Oestrogen hormone replacement therapy was shown to ameliorate an unfavourable lipid profile.41 42 However, it is worth mentioning that in our study, even after adjustment for total T, E2 and LH, FSH was still significantly associated with some elements of the lipid profile. Additionally, considering the independent association of FSH and HDL-C, some internal relationship might exist, which needs further exploration.
Our study43 had some strengths. First, the study presented in detail the association between FSH and the lipid profile for the first time in a large sample and focused on several possible explanatory factors. Second, the study was conducted with participants from the general population as opposed to those from a clinical population. However, our study also has some limitations. Because the nature of this study is cross-sectional, a causal relationship could not be determined. Moreover, final menstruation information was not obtained, and we used an age proxy of 55 years to define postmenopause based on previous studies and natural menopause age in Chinese women. Meanwhile, caused by the sample size, we need to verify our results with a larger sample size in the future.
Conclusions
Lower FSH was associated with lower HDL-C and higher TG, TC/HDL-C ratio and LDL-C/HDL-C ratio in women older than 55. Diabetes, BMI and hypertension mostly explained the association of FSH with TGs and the LDL-C/HDL-C ratio but only partially explained the associations of FSH with HDL-C and the TC/HDL-C ratio.
Data availability statement
Data are available on reasonable request. No data are available.
Ethics statements
Patient consent for publication
Ethics approval
The study protocol was approved by the ethics committee of Shanghai Ninth People’s Hospital, Shanghai JiaoTong University School of Medicine (No.SH9H-2013-86). All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Declaration of Helsinki of 1975, as revised in 2008. Informed consent was obtained from all patients included in the study.
Acknowledgments
The authors thank Xiaojin Wang and Bingshun Wang from the Department of Biostatistics, Shanghai Jiaotong University School of Medicine, Shanghai, China for data processing and Weiping Tu, Bin Li and Ling Hu for helping organise this investigation.The authors also thank all team members and participants from Shanghai, Zhejiang and Jiangxi Province in the SPECT-China study.The manuscript has been preprinted at the following link: https://www.researchsquare.com/article/rs-860860/v1.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
WZ and YC contributed equally.
Contributors YL, NW and HZ conceived and designed the research; WZ and YC analysed the data and drafted the manuscript. WZ, YC, XZ, YC, BH, QL, FX, HZ, NW and YL contributed to the interpretation of the data and the editing of the manuscript. YL, NW and HZ critically revised the manuscript for important intellectual content. All authors have approved the final version to be published. As the guarantor of this work, YL had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Funding This study was supported by the National Natural Science Foundation of China (82120108008) and the Science and Technology Commission of Shanghai Municipality (22015810500, 20015800400).
Disclaimer The funders played no role in the design or conduct of the study, collection, management, analysis, or interpretation of data or in the preparation, review, or approval of the article.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.