Causal associations of cognition, intelligence, education, health and lifestyle factors with cervical spondylosis: a mendelian randomization study

Background: Education, cognition, and intelligence are phenotypically and genetically related. Education has been shown to have a protective effect on the risk of developing cervical spondylosis. However, it is unclear whether cognition and intelligence have independent causal effects on cervical spondylosis, and whether health and lifestyle factors influence this association.

Methods: We first assessed the independent effects of education, cognition, and intelligence on cervical spondylosis by two-sample Mendelian randomization and multivariable Mendelian randomization analysis, and evaluated 26 potential association mediators using two-step Mendelian randomization, and calculated the median proportion.

Results: The results showed that only education had an independent causal effect on cervical spondylosis, and had a protective effect on the risk of cervical spondylosis (β: 0.3395; se: 0.166; p < 0.05; OR:0.71; [95%CI: 0.481–0.943]. Of the 26 potential associated mediators, a factor was identified: SHBG (mediated proportion: 2.5%). Univariable Mendelian randomization results showed that the risk factors for cervical spondylosis were time spent watching TV (OR:1.96; [95%CI: 1.39–2.76]), smoking (OR:2.56; [95%CI: 1.061–1.486]), body mass index (OR:1.26; [95%CI: 1.124–1.418]), percentage of body fat (OR:1.32; [95%CI: 1.097–1.593]), major depression (OR:1.27; [95%CI: 1.017–1.587]) and sitting height (OR:1.15; [95%CI: 1.025–1.291]). Protective factors include computer using (OR:0.65; [95%CI: 0.418–0.995]), sex hormone binding globulin (OR:0.87; [95%CI: 0.7955–0.951]) and high-density lipoprotein (OR:0.90; [95%CI: 0.826–0.990]).

Conclusion: Our findings demonstrate the causal and independent effects of education on cervical spondylosis and suggest that lifestyle media may be a priority target for the prevention of cervical spondylosis due to low educational attainment.

1 Introduction

The process of cervical spondylosis is a natural aging process, which is mainly manifested by progressive degenerative changes in the components of the cervical spine such as the intervertebral disc, ligament flavum, lamina, facet joint, etc (Bernabeu-Sanz et al., 2020; Kuo and Tadi, 2023). Congenital spinal stenosis, severe spinal trauma, and certain sports are the causes of early onset or accelerated progression of cervical spondylosis (Kelly et al., 2012; Lu et al., 2017). This process is often associated with aging and occurs after the age of 50, with the most commonly affected segments C6-7, followed by C5-6. The most common symptom is neck pain, which has a lifetime morbidity of up to 86.8%, which, along with low back pain, is the main cause of years lived with disability (Hoy et al., 2010; Hurwitz et al., 2018; Kuo and Tadi, 2023). In recent years, studies have pointed out that factors such as smoking, obesity, exercise, lifestyle and cardiovascular disease may be associated with the risk of cervical spondylosis (Theodore, 2020; Sun et al., 2023). With socio-economic development, the health resources and lifestyles people receive can be influenced by cognition, intelligence, and education. In addition, cognition, intelligence, and education have been shown to be phenotypically and genetically associated (Wang et al., 2023), as well as for hypertension and some cardiometabolic mediators. Education has now been shown to protect against the risk of cervical spondylosis (Sun et al., 2023). However, it is unclear whether cognition and intelligence have independent causal effects on cervical spondylosis, and whether health and lifestyle factors influence this association.

As a causal reasoning method, Mendelian randomization (MR) identifies Genome-wide single-nucleotide polymorphisms (SNPs) by using genome-wide association studies (GWAS). SNPs were used as genetic tools to explore the causal relationship between exposure of interest and outcomes. MR avoids the impact of reverse causal, environmental, and behavioral factors that can occur in observational studies (Lawlor et al., 2008). Multivariable Mendelian randomization (MVMR) is an extension of MR that allows different genetic variants of different exposures with correlation to be placed in the same study to determine the independent effects of different exposures on outcomes and to adjust potential pleiotropy (Sanderson et al., 2019).

In this study, we explored the independent causal relationship between cognition, intelligence and education and cervical spondylosis by using MR. Besides, the role of different lifestyle and health factors in the pathogenesis of cervical spondylosis is of great concern to us. Understanding this topic can improve the understanding of the etiological risks of cervical spondylosis, thereby providing prevention and intervention strategies for cervical spondylosis.

2 Methods

2.1 Study design

In this study, we first evaluated the causal relationship between cognition, intelligence, education and cervical spondylosis through univariable Mendelian randomization (UVMR) and MVMR. UVMR results showed that cognition, intelligence, education and cervical spondylosis were all causally related. MVMR results showed that only education had an independent causal relationship with cervical spondylosis and inter-regulated with intelligence and cognition. We then screened for two factors associated with cervical spondylosis and calculated the mediating effect using TWO-STEP-MR (Figure 1).

Figure 1

Figure 1. Flow chart of this article. (A) Study design: Firstly, the three exposure factors (education, cognition and intelligence) and cervical spondylosis were analyzed by UVMR and MVMR. Determine that education consistently has a causal relationship with cervical spondylosis, whether or not adjusted for other exposures. Secondly, determine the intermediary factors. (B) Mediator selection (Bernabeu-Sanz et al., 2020): There is a causal relationship between exposure and the mediator, and the effect of exposure on the mediator should be unidirectional (Kuo and Tadi, 2023). There is always a causal relationship between the mediator and the outcome, whether adjusted for the outcome or not (Kelly et al., 2012). There is always a causal relationship between exposure and outcome. SHBG: Sex hormone binding globulin; BMI: Body mass index; BFP: Body fat percentage; BMD: Bone mineral density; JIMP: Job involves heavy manual or physical work; JIWS: Job involves mainly walking or standing; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol.

2.2 Data sources

In this study, the source of data for exposure, outcomes, and mediators was genome-wide association studies (GWASs). A total of 30 GWAS data were used for UVMR and MVMR testing (Table 1; Supplementary Table S5). All studies were limited to people of European ancestry to reduce bias due to population differences.

Table 1

Table 1. The selected genome-wide association studies (GWAS) databases were shown in table. A total of 30 GWAS data were used and all studies were limited to people of European ancestry to reduce bias due to population differences. SHBG: Sex hormone binding globulin; BMI: Body mass index; BFP: Body fat percentage; BMD: Bone mineral density; JIMP: Job involves heavy manual or physical work; JIWS: Job involves mainly walking or standing; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol.

2.2.1 Exposure and outcome

The Intelligence Genetic instruments were selected from a GWAS analysis of 269,867 European individuals. Genetic instruments for education and cognition were selected from GWAS analyses involving 766,345/25,841 European individuals. For outcomes, we searched the Finngen Biobank for data on cervical spondylosis, and a GWAS analysis covering 171,956 European individuals was selected. First, to ensure that the genetic tools were strong, we extracted SNPs with relevant p values (p < 5 × 10⁻⁸) from the above GWAS datasets. Later, in order to avoid inaccurate results caused by the calculation of related SNP, linkage disequilibrium clumping (R² ≤ 0.001, window size = 10,000 kb) was performed for each MR Exposure. When R² > 0.001, SNPs with larger associated p-values will be deleted to ensure the accuracy of the results. When there are no specific SNPs in the exposure factor dataset, we use proxy SNPs by labeling LD (minimum LD Rsp value = 0.8; MAF threshold = 0.3).

2.2.2 Mediators

Through the search of the existing literature, it is shown that smoking, obesity, exercise, lifestyle and cardiovascular disease may be associated with the incidence of cervical spondylosis (Theodore, 2020; Sun et al., 2023). Therefore, we chose smoking (smoking initiation, smoking heaviness); obesity (body mass index (BMI), body fat percentage (BFP)); leisure activity (moderate to vigorous physical activity levels); lifestyle (job involves heavy manual or physical work (JIMP), job involves mainly walking or standing (JIMS), time spent using computer, time spent watching television (TV), playing computer games, alcohol drinking, coffee intake, household income, sitting height); cardiovascular factors (total cholesterol, triglycerides, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C)); In addition, we selected factors related to the body [diabetes (type 1/2 diabetes); sex hormones (sex hormone binding globulin (SHBG), testosterone, estrogen); Bone-related (osteoporosis, bone mineral density (BMD), serum 25-Hydroxy vitamin D); and major depression. We analyzed exposure (education), outcomes (cervical spondylosis) and mediators according to the following principles (Bernabeu-Sanz et al., 2020): There is a causal relationship between exposure and the mediator, and the effect of exposure on the mediator should be unidirectional (Kuo and Tadi, 2023). There is always a causal relationship between the mediator and the outcome, whether adjusted for the outcome or not (Kelly et al., 2012). There is always a causal relationship between exposure and outcome (Lu et al., 2017). The causal direction between exposure and mediation, between mediation and outcome, should be consistent. The mediating factors was SHBG.

2.3 Statistic analysis

All analyses MR satisfy three key assumptions (Bernabeu-Sanz et al., 2020): the genetic variants must be strongly associated with exposure and must be associated with at least one exposure in the MVMR (Kuo and Tadi, 2023); Genetic variants was not associated with any confounding factors associated with expose and outcome (Kelly et al., 2012); Genetic variants does not affect the results, unless the effect on cervical spondylosis is after each exposure (Hemani et al., 2018a). Inverse variance weighted (IVW): IVW is the most important MR analysis and forces the linear regression intercept to be set to 0. In IVW, it is assumed that the individual genetic variants are independent of each other. When there is a plurality of potentials, results can be skewed (Hemani et al., 2018b). MR-Egger: This method does not enforce a linear regression intercept of 0, so the intercept can be used to assess whether genetic variants has directional pleiotropism for the results (Bowden et al., 2017). Furthermore, the MR-Egger method is less accurate than the weighted median because it may be affected by peripheral genetic variation. Weighted median: When >50% of the information is useful for the analysis of valid instrumental variables, this method provides estimates that are compatible with the final effect. MR-PRESSO: It is currently widely used to test horizontal pleiotropy (Verbanck et al., 2018). We also applied the Q′ heterogeneity statistic to assess heterogeneity between instruments and condition F statistic to test instrument validity (Bowden et al., 2019). When F > 10, we can consider the instrument validity to be high.

UVMR: We used UVMR to estimate the effects of cognition, intelligence, and education on cervical spondylosis, respectively. For UVMR, we determined IVW as the main basis for MR Analysis. The weighted median and MR-Egger based regression method are combined to ensure the accuracy of the conclusion. In addition, horizontal pleiotropy was tested using the MR-PRESSO method. MVMR: It is an extension of MR by combining multiple potentially relevant exposure-related genetic variants to calculate the independent effect of each exposure on the outcome (Sanderson, 2021). For MVMR, we used IVW results as the main basis for MVMR, and used Q’ heterogeneity statistics to assess inter-tool heterogeneity.

Mediation Analyses: To further determine the overall effect rate of education on cervical spondylosis, we performed mediation analysis using the two-step-MR approach. First, we estimate the causal effect of exposure factors on the mediator (β1) using UVMR. Then, the MVMR method was used to estimate the causal effect of mediator on cervical spondylosis (β2). The indirect effect is obtained by multiplying (β1Xβ2) by the total effect. In this study, the association between all exposures and outcome phenotypes was statistically significant at p < 0.05. All of our analyses were performed using the “TwoSampleMR”, “MRPRESSO”, “MRPracticals” R packages.

3 Results

3.1 Casual associations of cognition, intelligence, education with cervical spondylosis

We carried out UVMR analysis on the causal relationship between cognition, intelligence, education and cervical spondylosis respectively: longer education time (β: 0.604; se: 0.094; p < 0.05; OR:0.54; [95%CI: 0.454–0.657]), higher intelligence (β: 0.43; se: 0.10; p < 0.05; OR:0.65; [95%CI: 0.533–0.796]), better cognition (β: 0.423; se: 0.090; p < 0.05; OR:0.66; [95%CI: 0.549–0.783]) had protective effects on the incidence of cervical spondylosis.

MVMR results showed that there was a statistically significant causal relationship between education and cervical spondylosis, although it was accompanied by intelligence (β: 0.464; se: 0.161; p < 0.05; OR:0.63; [95%CI: 0.431–0.827]), cognition (β: 0.326; se: 0.158; p < 0.05; OR:0.72; [95%CI: 0.498–0.945]) or both (β: 0.3395; se: 0.166; p < 0.05; OR:0.71; [95%CI: 0.481–0.943]). After adjustment for education and cognition, the causal relationship between intelligence and cervical spondylosis was no longer statistically significant. The causal relationship between cognition and cervical spondylosis was statistically significant when only adjusted for education (β: 0.362; se: 0.145; p < 0.05; OR:0.70; [95%CI: 0.498 to 0.894]), which was not statistically significant when adjusted for intelligence or both (Figure 2).

Figure 2

Figure 2. MR estimates of the causal associations of exposures, candidate mediators with cervical spondylosis. (A) Causal relationship between the exposures (education, cognition and intelligence) and cervical spondylosis, whether or not adjusted for other exposures. (B) UVMR analysis was performed on 26 candidate mediators and cervical spondylosis, and nine mediators were identified as having causal relationship with cervical spondylosis. OR indicates odds ratio and plots (bars) represent OR (95% CI). SHBG: Sex hormone binding globulin; BMI: Body mass index; BFP: Body fat percentage. HDL-C: high-density lipoprotein cholesterol.

3.2 Influence of candidate mediating factors on cervical spondylosis

Through MR analysis, we selected 26 candidate mediators and performed causal analysis of education with mediators, mediators with cervical spondylosis. Then, nine media were identified. Smoking was associated with cervical spondylosis (Smoking initiation had an adverse effect on cervical spondylosis (β: 0.228; se: 0.086; p < 0.05; OR:1.25; [95%CI: 1.061–1.486]), there was no causal relationship between smoking heaviness and cervical spondylosis (β: 0.072; se: 0.068; p > 0.05; OR:1.07; [95%CI: 0.940–1.228]). Time spent using computer was a protective factor for cervical spondylosis (β: 0.438; se: 0.221; p < 0.05; OR:0.65; [95%CI: 0.418–0.995]) and time spent watching TV was a risk factor for cervical spondylosis (β:0.672; se: 0.174; p < 0.05; OR:1.96; [95%CI: 1.39–2.76]). In addition, BMI (β:0.233; se: 0.059; p < 0.05; OR:1.26; [95%CI: 1.124–1.418]), BFP (β:0.279; se: 0.095; p < 0.05; OR:1.32; [95%CI: 1.097–1.593]), major depression (β:0.239; se: 0.113; p < 0.05; OR:1.27; [95%CI: 1.017–1.587]) and sitting height (β:0.140; Se: 0.059; p < 0.05; OR:1.15; [95%CI: 1.025–1.291]) were risk factors for cervical spondylosis. And SHBG (β: 0.140; se: 0.046; p < 0.05; OR:0.87; [95%CI: 0.795–0.951]) and HDL-C (β: 0.101; se: 0.046; p < 0.05; OR:0.90; [95%CI: 0.826–0.990]) were protective factors of cervical spondylosis (Figure 2). Otherwise, no causal relationship was found between exercise (leisure activities, physical activity); lifestyle (JIMP, JIMS, alcohol drinking, coffee intake, household income); cardiovascular factors (total cholesterol, triglycerides, LDL-C); type 1/2 diabetes; testosterone, estrogen; osteoporosis, bone mineral density, serum 25-Hydroxyvitamin vitamin D and cervical spondylosis (Supplementary Tables S1–S3).

3.3 Influence of education on candidate mediating factors

Among them, education is a protective factor for smoking heaviness (β: 0.373; se: 0.032; p < 0.05; OR:0.69; [95%CI: 0.646–0.734]) and BFP (β: 0.290; se: 0.019; p < 0.05; OR:0.75; [95%CI: 0.720–0.777]). Besides, education is a risk factor for time spent using computer (β: 0.337; se: 0.016; p < 0.05; OR:1.40; [95%CI: 1.359–1.444]), time spent watching TV (β: 0.401; se: 0.013; p < 0.05; OR:0.67; [95%CI: 0.653–0.687]), BMI (β: 0.344; se: 0.026; p < 0.05; OR:0.71; [95%CI: 0.673–0.746]), HDL(β: 0.144; se: 0.034; p < 0.05; OR:1.15; [95%CI: 1.080–1.233]), major depression(β: 0.241; se: 0.036; p < 0.05; OR:0.79; [95%CI: 0.732–0.843]), SHBG(β: 0.109; se: 0.048; p < 0.05; OR:1.12; [95%CI: 1.016–1.225]) and sitting height (β: 0.150; se: 0.026; p < 0.05; OR:1.16; [95%CI: 1.103–1.222]) (Supplementary Table S4).

3.4 Influence of mediating factors on cervical spondylosis with adjustment for education

It was estimated that the proportion of educational mediated effects on cervical spondylosis through SHBG as (2.5%; [95%CI:0.2%-4.9%]) (Figure 3).

Figure 3

Figure 3. SHBG-related analysis results. (A) UVMR assessing the causal association between education and SHBG; (B) MVMR assessing the causal association between SHBG and CS with adjustment for education; (C) Proportion of SHBG in the causal relationship between education and CS; (D) Forest plot (causal association between SHBG and CS, adjusted for education). SHBG: Sex hormone binding globulin; CS cervical spondylosis.

4 Discussion

In this study, we assessed the causal association of cognitive, intellectual, educational, different lifestyles, health factors with cervical spondylosis by MR. Cognition, intelligence, and education have proven to be closely linked and inseparable (Savage et al., 2018; Wang et al., 2023). Studies have shown that higher educational attainment is a protective factor for cervical spondylosis (Sun et al., 2023). Factors such as smoking, obesity, exercise, lifestyle and cardiovascular disease may be related to the risk of cervical spondylosis (Theodore, 2020; Sun et al., 2023). At present, the mediating role of cardiovascular factors in education and cervical spondylosis has been demonstrated (Sun et al., 2023). However, the effects of different lifestyle and health factors on cervical spondylosis have not been elaborated by studies. This article will focus on the influence of different lifestyle and health factors on cervical spondylosis. In addition, this study expands on previous research by increasing the evidence on the total causal effect of cognitive function on cervical spondylosis, and for the first time identifies higher education as a protective factor for cervical spondylosis independent of intelligence and cognition, providing new insights into the possible pathogenesis of cervical spondylosis.

First, we conducted MR analysis of the association of cognition, intelligence, education with cervical spondylosis. The results pointed out that all three were causally related to the onset of cervical spondylosis. However, when the three are adjusted or any of the two, only education have an independent causal association with cervical spondylosis. Compared with the heritability of cognition and intelligence, educational attainment is more susceptible to other factors, which has lasting effects on people’s socioeconomic resources and status, as well as their lifestyle (Lawrence, 2017). In addition, educational attainment can be used as a proxy indicator of knowledge acquisition and health opportunities in later life. This suggests that preventive measures for cervical spondylosis can be taken by improving educational policies.

Another finding of this study was to identify and quantify the mediating role of different lifestyle and health factors in the association between education and cervical spondylosis. Through MR analysis, UVMR results showed a causal relationship between smoking, time spent using computer, time spent on TV, sitting height, BFP, major depression, BMI, HDL, SHBG and cervical spondylosis in 26 candidate mediators. The effect of smoking, BMI on cervical spondylosis is the same as that of SUN Y et al. (Sun et al., 2023). These results support the influence of lifestyle on the incidence of cervical spondylosis. Interestingly, the effect of watching TV on cervical spondylosis was different from that of computer using. The reason why watching TV is a risk factor for cervical spondylosis may be affected by the following factors. Firstly, the distance between the TV and the people is far, many people are basically in a state of poor posture when watching TV on the sofa or chair, and the bad posture of watching TV may last for a long time. Poor posture may be a factor that increases the risk of cervical spondylosis. Secondly, some people fall asleep while watching TV, whether lying on the sofa or sleeping in a chair, people are also in a bad posture. All these may increase the risk of cervical spondylosis. Different from television, the current stage of the computer monitor can be freely adjusted the height and angle of the monitor, people can adjust the position of the monitor according to the real-time situation when using the computer. This results in the user not being in a state of poor neck posture for a long time. In addition, with the increase of computer display screens, people need to move their necks frequently when using computers to ensure that they can get all the content on the monitor. This activity may be unconscious, but it also reduces the incidence of cervical spondylosis to a certain extent. The specific reasons for this phenomenon need to be confirmed by further research. Furthermore, factors such as lifestyle (leisure activities, long standing time and walking, amount of exercise, work intensity, drinking alcohol, coffee intake, household income), health factors (type 1/2 diabetes, osteoporosis, bone mineral density, serum vitamin D) were not found to have a clear causal relationship with cervical spondylosis. Promoting lifestyle changes through health education and developing good lifestyle habits may be used as a preventive measure for the risk of cervical spondylosis.

In further studies, when adjusted for education, only SHBG was still statistically significant in relation to cervical spondylosis. Education prevents the risk of cervical spondylosis by reducing SHBG. They mediated the protective effect of education by 2.5%. The discovery of SHBG is of great interest. SHBG is a protein produced by the liver that binds to and transports steroids such as estradiol and testosterone in plasma (Xing et al., 2022). Plasma concentrations of SHBG were positively correlated with estrogen and thyroid hormones, and negatively correlated with androgens (Selby, 1990). This may suggest to some extent that the effects of SHBG on education and cervical spondylosis are sex-related. However, we found that there was no causal relationship between androgens and cervical spondylosis, and the effect of estrogen on cervical spondylosis cannot be carried out because a suitable GWAS dataset has not been found, and further research is needed to explore the relationship. Excitingly, plasma concentrations of SHBG were shown to be associated with insulin concentrations and obesity (Xing et al., 2022). Our study found that there is no clear causal relationship between cervical spondylosis and type 1 and 2 diabetes. Studies have shown that SHBG is negatively correlated with BMI (Nokoff et al., 2019). After our MR analysis of SHBG, BMI and cervical spondylosis, we found that there was no clear causal relationship between SHBG and BMI, so BMI could not be used as an intermediary factor of SHBG for cervical spondylosis. It is a good research direction to further explore the relationship between SHBG and cervical spondylosis in follow-up research.

To ensure the accuracy of the results, we used the MVMR sensitivity analysis to support the IVW estimates and tried to ensure that each analysis accommodated different assumptions about genetic pleiotropy. In addition, we screen mediators through rigorous criteria to reduce reverse causality between intermediaries and education, thus ensuring accurate and reasonable results. However, our study still has some limitations. Although we focused on cervical spondylosis, whether this result is universal for different types and causes of cervical spondylosis still needs further discussion. Secondly, the study population had certain limitations, only for individuals of European ancestry, and the generalization of the results needs further research. In addition, although all the selected people are European, there are still regional differences between countries, suggesting that there may still be population heterogeneity in this study. Thirdly, the heterogeneity of SNPs can skew and affect the results. In future studies, we need more diverse populations, larger sample sizes and more diverse research methods to further validate this result.

5 Conclusion

In summary, this study elaborates the causal protective effect of education independent of intelligence and cognitive ability on cervical spondylosis, and outlines the causal mediating factors (SHBG) and factors that are causally associated with the onset of cervical spondylosis. This study further complements the causal evidence for the etiology of cervical spondylosis, and provides new ideas for prevention and intervention to control cervical spondylosis and its related disease burden.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Author contributions

ZR: Writing–original draft, Writing–review and editing. XC: Writing–original draft, Writing–review and editing. JX: Writing–review and editing. TN: Writing–review and editing. HL: Funding acquisition, Supervision, Writing–review and editing.

Funding

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This research was founded by Natural Science Foundation of China (grant no. 82201360) and by Guangdong Basic and Applied Basic Research Foundation (Nr. 2021A1515110188).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fgene.2024.1297213/full#supplementary-material

References

Bernabeu-Sanz, A., Molla-Torro, J. V., Lopez-Celada, S., Moreno Lopez, P., and Fernandez-Jover, E. (2020). MRI evidence of brain atrophy, white matter damage, and functional adaptive changes in patients with cervical spondylosis and prolonged spinal cord compression. Eur. Radiol. 30 (1), 357–369. doi:10.1007/s00330-019-06352-z

PubMed Abstract | CrossRef Full Text | Google Scholar

Bowden, J., Del Greco, M. F., Minelli, C., Davey Smith, G., Sheehan, N., and Thompson, J. (2017). A framework for the investigation of pleiotropy in two-sample summary data Mendelian randomization. Stat. Med. 36 (11), 1783–1802. doi:10.1002/sim.7221

PubMed Abstract | CrossRef Full Text | Google Scholar

Bowden, J., Del Greco, M. F., Minelli, C., Zhao, Q., Lawlor, D. A., Sheehan, N. A., et al. (2019). Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int. J. Epidemiol. 48 (3), 728–742. doi:10.1093/ije/dyy258

PubMed Abstract | CrossRef Full Text | Google Scholar

Hemani, G., Bowden, J., and Davey Smith, G. (2018a). Evaluating the potential role of pleiotropy in Mendelian randomization studies. Hum. Mol. Genet. 27 (R2), R195–R208. doi:10.1093/hmg/ddy163

PubMed Abstract | CrossRef Full Text | Google Scholar

Hemani, G., Zheng, J., Elsworth, B., Wade, K. H., Haberland, V., Baird, D., et al. (2018b). The MR-Base platform supports systematic causal inference across the human phenome. Elife 7, e34408. doi:10.7554/eLife.34408

PubMed Abstract | CrossRef Full Text | Google Scholar

Hoy, D. G., Protani, M., De, R., and Buchbinder, R. (2010). The epidemiology of neck pain. Best. Pract. Res. Clin. Rheumatol. 24 (6), 783–792. doi:10.1016/j.berh.2011.01.019

PubMed Abstract | CrossRef Full Text | Google Scholar

Hurwitz, E. L., Randhawa, K., Yu, H., Cote, P., and Haldeman, S. (2018). The Global Spine Care Initiative: a summary of the global burden of low back and neck pain studies. Eur. Spine J. 27 (Suppl. 6), 796–801. doi:10.1007/s00586-017-5432-9

PubMed Abstract | CrossRef Full Text | Google Scholar

Kelly, J. C., Groarke, P. J., Butler, J. S., Poynton, A. R., and O'Byrne, J. M. (2012). The natural history and clinical syndromes of degenerative cervical spondylosis. Adv. Orthop. 2012, 393642. doi:10.1155/2012/393642

PubMed Abstract | CrossRef Full Text | Google Scholar

Kuo, D. T., and Tadi, P. (2023). Cervical spondylosis. StatPearls. Treasure Isl. (FL). eng.

Google Scholar

Lawlor, D. A., Harbord, R. M., Sterne, J. A., Timpson, N., and Davey Smith, G. (2008). Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat. Med. 27 (8), 1133–1163. doi:10.1002/sim.3034

PubMed Abstract | CrossRef Full Text | Google Scholar

Lawrence, E. M. (2017). Why do college graduates behave more healthfully than those who are less educated? J. Health Soc. Behav. 58 (3), 291–306. doi:10.1177/0022146517715671

PubMed Abstract | CrossRef Full Text | Google Scholar

Lu, X., Tian, Y., Wang, S. J., Zhai, J. L., Zhuang, Q. Y., Cai, S. Y., et al. (2017). Relationship between the small cervical vertebral body and the morbidity of cervical spondylosis. Med. Baltim. 96 (31), e7557. The authors report no conflicts of interest. doi:10.1097/MD.0000000000007557

PubMed Abstract | CrossRef Full Text | Google Scholar

Nokoff, N., Thurston, J., Hilkin, A., Pyle, L., Zeitler, P. S., Nadeau, K. J., et al. (2019). Sex differences in effects of obesity on reproductive hormones and glucose metabolism in early puberty. J. Clin. Endocrinol. Metab. 104 (10), 4390–4397. doi:10.1210/jc.2018-02747

PubMed Abstract | CrossRef Full Text | Google Scholar

Sanderson, E. (2021). Multivariable mendelian randomization and mediation. Cold Spring Harb. Perspect. Med. 11 (2), a038984. doi:10.1101/cshperspect.a038984

PubMed Abstract | CrossRef Full Text | Google Scholar

Sanderson, E., Davey Smith, G., Windmeijer, F., and Bowden, J. (2019). An examination of multivariable Mendelian randomization in the single-sample and two-sample summary data settings. Int. J. Epidemiol. 48 (3), 713–727. doi:10.1093/ije/dyy262

PubMed Abstract | CrossRef Full Text | Google Scholar

Savage, J. E., Jansen, P. R., Stringer, S., Watanabe, K., Bryois, J., de Leeuw, C. A., et al. (2018). Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence. Nat. Genet. 50 (7), 912–919. Competing Financial Interests Statement PF Sullivan reports the following potentially competing financial interests: Lundbeck (advisory committee), Pfizer (Scientific Advisory Board member), and Roche (grant recipient, speaker reimbursement). G Breen reports consultancy and speaker fees from Eli Lilly and Illumina and grant funding from Eli Lilly. J Hjerling-Leffler reports interests from Cartana (Scientific Advisor) and Roche (grant recipient). TD Cannon is a consultant to Boehringer Ingelheim Pharmaceuticals and Lundbeck A/S. All other authors declare no financial interests or potential conflicts of interest. doi:10.1038/s41588-018-0152-6

PubMed Abstract | CrossRef Full Text | Google Scholar

Selby, C. (1990). Sex hormone binding globulin: origin, function and clinical significance. Ann. Clin. Biochem. 27 (Pt 6), 532–541. doi:10.1177/000456329002700603

PubMed Abstract | CrossRef Full Text | Google Scholar

Sun, Y., Jin, M., Yu, T., and Zhang, J. (2023). Cardiovascular risk factors mediating the protective effect of education on cervical spondylosis risk. Sci. Rep. 13 (1), 936. doi:10.1038/s41598-023-28153-7

PubMed Abstract | CrossRef Full Text | Google Scholar

Theodore, N. (2020). Degenerative cervical spondylosis. N. Engl. J. Med. 383 (2), 159–168. doi:10.1056/NEJMra2003558

PubMed Abstract | CrossRef Full Text | Google Scholar

Verbanck, M., Chen, C. Y., Neale, B., and Do, R. (2018). Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat. Genet. 50 (5), 693–698. doi:10.1038/s41588-018-0099-7

PubMed Abstract | CrossRef Full Text | Google Scholar

Wang, Y., Ye, C., Kong, L., Zheng, J., Xu, M., Xu, Y., et al. (2023). Independent associations of education, intelligence, and cognition with hypertension and the mediating effects of cardiometabolic risk factors: a mendelian randomization study. Hypertension 80 (1), 192–203. doi:10.1161/HYPERTENSIONAHA.122.20286

PubMed Abstract | CrossRef Full Text | Google Scholar

Xing, C., Zhang, J., Zhao, H., and He, B. (2022). Effect of sex hormone-binding globulin on polycystic ovary syndrome: mechanisms, manifestations, genetics, and treatment. Int. J. Womens Health 14, 91–105. doi:10.2147/IJWH.S344542

PubMed Abstract | CrossRef Full Text | Google Scholar