Abstract
Introduction
Observational epidemiological studies have revealed that multiple risk factors may be associated with the risk of obstructive sleep apnea (OSA). However, the causal relationship between them remains largely unknown. We aimed to investigate the causal relationship between potential risk factors and OSA risk.
Methods
A two-sample Mendelian randomization approach was used to evaluate the causal association of 42 risk factors with OSA risk. Summary data on OSA were obtained from a recently published genome-wide association study including 16,761 patients with OSA.
Results
Suggestive associations with increased risk were observed for body mass index (OR = 2.19, 95% CI 1.99–2.42, P < 0.01), childhood body mass index (OR = 1.46, 95% CI 1.15–1.83, P < 0.01), overweight (OR = 1.43, 95% CI 1.17–1.74, P < 0.01), smoking initiation (OR = 1.27, 95% CI 1.09–1.49, P < 0.01), gastroesophageal reflux disease (OR = 1.60, 95% CI 1.40–1.82, P < 0.01), and depression (OR = 1.28, 95% CI 1.06–1.53, P = 0.01). Age at first birth (OR = 0.88, 95% CI 0.82–0.94, P < 0.01) and education (OR = 0.76, 95% CI 0.66–0.87, P < 0.01) were significantly associated with decreased risk of OSA.
Conclusions
We found a causal effect for several potential risk factors on OSA risk, including obesity, smoking, education, age at first birth, gastroesophageal reflux disease, and depression.
Keywords: Risk factor, obstructive sleep apnea, Mendelian randomization study, obesity, smoking
PLAIN LANGUAGE SUMMARY
We found that obesity in childhood and adulthood, smoking, gastroesophageal reflux disease, and depression raised the risk of obstructive sleep apnea (OSA). Conversely, higher education level and older age at first birth lowered the OSA risk. Our findings provide strong evidence that these factors likely have a causal relationship with OSA. Understanding these causes can help guide efforts to prevent and manage this common sleep disorder.
ARTICLE HIGHLIGHTS
We investigated causal relationships between potential risk factors and the risk of obstructive sleep apnea (OSA).
Most of the 42 risk factors investigated in this study have not been estimated using a Mendelian randomization approach.
Significant relationships were found between obesity factors, smoking, gastroesophageal reflux disease, depression, and an increased risk of OSA.
We also found evidence of significant correlations between the age at first birth, education level, and decreased OSA risk.
Further research is needed to elucidate the potential mechanisms that mediate the connection between these risk factors and OSA.
1. Introduction
Obstructive sleep apnea (OSA) is a severe sleep-related breathing disorder, and the prevalence of OSA is increasing exponentially [1]. OSA is characterized by repeated episodes of airflow cessation, which lead to recurrent oxygen desaturation, sleep fragmentation, exhaustion, and poor attention [2]. These symptoms can affect quality of life and increase the risk of traffic and other accidents [3]. Research indicates that approximately 1 billion people worldwide have OSA, which severely affects human health [4].
OSA is affected by a variety of risk factors, such as obesity, sex, alcohol consumption, smoking, serum neuronal PAS domain protein 2, uric acid, chronic obstructive pulmonary disease, and nasal disease [5–9]. Additionally, gastroesophageal reflux disease (GERD) has been suggested as a potential risk factor for OSA [10]. Interestingly, one study reported that a healthy lifestyle was inversely associated with OSA risk [11]. Recent studies have revealed that obesity and education are causally associated with the risk of OSA [12,13]. However, for most potential risk factors, evidence that reliably establishes a causal relationship is lacking.
Most evidence on causality between potential risk factors and OSA risk comes from observational studies, which are susceptible to confounding and reverse causation [14]. Mendelian randomization (MR) is a tool designed to investigate the causal relationship between risk factors and disease using genetic variations as instrumental variables, which represent the target exposure and serve as the basis for group allocation in a natural randomized trial [15]. MR analyses using such instrumental variables resemble randomized clinical trials and are less susceptible to confounding and reverse causation, consequently generating robust causal evidence that can be useful in promoting OSA prevention and clinical management [16].
In this study, we aimed to explore the possible causal association between potential risk factors and the risk of OSA.
2. Methods
2.1. Single-nucleotide polymorphisms (SNPs) for potential risk factors
Figure 1 shows a flowchart of the study design. We conducted a search using the terms (obstructive sleep apnea) AND (risk factor). SNPs associated with 42 potential risks were identified in previously published genome-wide association studies (GWAS) including participants of European descent. We constructed genetic tools by obtaining SNPs showing robust (p < 5 × 10−8) and independent associations (r2 < 0.001). The GWAS summary data harmonization process primarily involved: (1) data preparation and quality control, (2) allele matching across datasets, (3) handling of palindromic SNPs, (4) strand flipping, (5) removal of invalid SNPs and SNPs with allele conflicts, and (6) outcome data extraction. SNPs with an F-statistic <10 were excluded to ensure the strength of the selected instruments. The F-statistic functions as a quality metric in MR: values <10 indicate a risk of weak instrument bias, and values >10 confirm the reliability of instrument strength. We used the following formulas: F = R2 × (N − 2)/(1 − R2), and R2 = 2 × MAF × (1 − MAF) × beta2 [16,17]. Ethics approval was obtained from its institutional review board.
Figure 1.
A schematic summary of the study design.
2.2. OSA population
A recently published GWAS including participants of European descent from the FinnGen study was used to determine the outcome of OSA risk [12]. We analyzed a total of 16,761 patients with OSA and 201,194 controls. Patients with OSA were classified according to the International Classification of Diseases, Ninth Revision (ICD-9: 3472 A) and Tenth Revision (ICD-10: G47.3). Ethics approval was obtained from the relevant ethics committee.
2.3. Statistical analysis
We used R version 4.0.3 with the TwoSampleMR package to perform MR analysis [17]. The inverse-variance weighting (IVW), weighted median, and MR-Egger methods were used to assess causality between potential risk factors and OSA [16,18]. The estimates of causal effect and equivalents to beta coefficients were calculated, which were then converted to odds ratios (ORs). Horizontal pleiotropy was assessed using MR-Egger regression, a core approach in MR for detecting and adjusting horizontal pleiotropy. When the MR-Egger intercept differs from zero (P < 0.05), this indicates horizontal pleiotropy or a violation of the MR assumption [19]. Sensitivity analysis was performed by removing SNPs each time to explore whether the estimates from inverse-variance weighting analysis were biased owing to a single SNP. We used leave-one-out analysis to systematically identify outlier SNPs and remove them from the final analysis. Power calculations for the MR analysis were performed using a web application (https://shiny.cnsgenomics.com/mRnd/) [20].
3. Results
In this study, we investigated 42 potential risk factors for OSA, including nine obesity-related factors, five behavioral factors, 13 factors related to serum or plasma, and 15 disease-related factors. The number of SNPs, the proportion of variance explained (R2) by the SNPs, and the F statistics for the instrument are shown in Table 1. Two-sample MR analyses showed that eight potential risk factors had a causal effect on OSA risk; 34 risk factors were not associated with OSA risk (Table 2). Additionally, power calculations were carried out using a sample size of 217,955 (16,761 OSA cases and 201,194 controls) in the outcome datasets and a prevalence of type 1 error of 0.05, which provided sufficient statistical power (>80%) to detect a causal effect for eight potential risk factors on the risk of OSA (Table 3).
Table 1.
42 risk factors for obstructive sleep apnea included in Mendelian randomization analysis.
| Reported risk factor | Sample size | R 2 | F statistic | References (PMID or Website) |
|---|---|---|---|---|
| Body fat percentage | 100716 | 0.005 | 523.94321 | 26833246 |
| Body mass index | 681275 | 0.051 | 36847.641 | 30124842 |
| Childhood body mass index | 39620 | 0.230 | 11833.948 | 33045005 |
| Childhood obesity | 13848 | 0.078 | 1171.3536 | 22484627 |
| Overweight | 158855 | 0.034 | 5642.1848 | 23563607 |
| Height | 253288 | 0.135 | 39361.025 | 25282103 |
| Waist circumference | 104079 | 0.013 | 1370.8217 | 25673412 |
| Hip circumference | 99562 | 0.020 | 2042.2043 | 25673412 |
| Waist-to-hip ratio | 99076 | 0.020 | 2032.2353 | 25673412 |
| Alcoholic drinks per week | 335394 | 0.060 | 21408 | 30643251 |
| Cigarettes smoked per day | 249752 | 0.080 | 21717.391 | 30643251 |
| Smoking initiation | 607291 | 0.080 | 52807.739 | 30643251 |
| Age at first birth | 542901 | 0.058 | 33426.902 | 34211149 |
| Years of schooling | 766345 | 0.023 | 17719.82 | 30038396 |
| C-Reactive protein level | 204402 | 0.069 | 15148.872 | 30388399 |
| Serum 25-Hydroxyvitamin D | 417580 | 0.044 | 19219.071 | 32242144 |
| Total testosterone (man) | 425097 | 0.170 | 87067.651 | 32042192 |
| HDL cholesterol | 94595 | 0.081 | 8337.3591 | 24097068 |
| LDL cholesterol | 94595 | 0.139 | 15271.111 | 24097068 |
| Total cholesterol | 94595 | 0.088 | 9127.3947 | 24097068 |
| Triglycerides | 94595 | 0.048 | 4769.395 | 24097068 |
| Adiponectin | 39883 | 0.048 | 2010.8067 | 22479202 |
| Leptin | 27947 | 0.010 | 277.92911 | 32917775 |
| Fasting glucose | 200622 | 0.007 | 1414.2397 | 34059833 |
| Fasting insulin | 151013 | 0.003 | 424.01805 | 34059833 |
| HbA1c | 46368 | 0.048 | 2337.7815 | 20858683 |
| Serum urate | 69374 | 0.007 | 489.02719 | 23263486 |
| Allergic rhinitis | 212387 | 0.026 | 5598.6325 | https://www.finngen.fi/en |
| Asthma (childhood) | 300671 | 0.256 | 103456 | 30929738 |
| Asthma (adult) | 300671 | 0.667 | 602240.91 | 30929738 |
| Chronic sinusitis | 167849 | 0.027 | 4646.8158 | https://www.finngen.fi/en |
| Nasal polyp | 167849 | 0.181 | 37206.85 | https://www.finngen.fi/en |
| Chronic tonsils and adenoids | 167849 | 0.033 | 5737.2708 | https://www.finngen.fi/en |
| Gastroesophageal reflux disease | 473524 | 0.036 | 17446.401 | 34187846 |
| Barrett’s esophagus | 43071 | 0.068 | 3123.5151 | 34187846 |
| Type 2 diabetes | 655666 | 0.230 | 195847.69 | 30054458 |
| Breast cancer | 139274 | 0.165 | 27520.814 | 29059683 |
| Prostate cancer | 140254 | 0.402 | 94301.309 | 29892016 |
| Rheumatoid arthritis | 8383 | 0.005 | 38.63091 | 24390342 |
| Bipolar disorder | 51710 | 0.049 | 2675.6769 | 31043756 |
| Major depression | 500199 | 0.018 | 8943.7643 | 30718901 |
| Polycystic ovary syndrome | 113238 | 0.077 | 9418.7718 | 30566500 |
HDL: high density lipoprotein; LDL: low density lipoprotein; HbA1c: hemoglobin A1c.
Table 2.
Inverse-variance weighted (IVW), weighted median and MR-egger analysis estimates for the association of 42 risk factors with OSA risk.
| Risk factor | N SNPS |
Inverse variance weighted |
MR-egger |
Weighted median |
|||
|---|---|---|---|---|---|---|---|
| OR (95% CI) |
P | OR (95% CI) |
P | OR (95% CI) |
P | ||
| Body fat percentage | 10 | 1.72 (0.97–3.05) |
0.06 | 29.77 (4.55–194.92) |
0.01 | 1.38 (0.92–2.06) |
0.12 |
| Body mass index | 467 | 2.19 (1.99–2.42) |
<0.01 | 1.95 (1.51–2.51) |
<0.01 | 2.06 (1.8–2.36) |
<0.01 |
| Childhood body mass index | 16 | 1.46 (1.15–1.83) |
<0.01 | 1.1 (0.43–2.86) |
0.84 | 1.29 (1.07–1.56) |
0.01 |
| Childhood obesity | 5 | 1.18 (0.97–1.44) |
0.09 | 1.85 (0.36–9.5) |
0.51 | 1.08 (0.98–1.18) |
0.12 |
| Overweight | 14 | 1.43 (1.17–1.74) |
<0.01 | 2.72 (1.54–4.82) |
<0.01 | 1.34 (1.12–1.61) |
<0.01 |
| Height | 349 | 0.94 (0.89–1) |
0.06 | 0.98 (0.84–1.15) |
0.81 | 0.92 (0.85–1) |
0.04 |
| Waist circumference | 26 | 1.00 (0.84–1.2) |
0.96 | 0.89 (0.32–2.46) |
0.82 | 0.98 (0.78–1.23) |
0.85 |
| Hip circumference | 31 | 0.92 (0.79–1.08) |
0.33 | 0.7 (0.32–1.52) |
0.37 | 0.93 (0.77–1.13) |
0.47 |
| Waist-to-hip ratio | 6 | 0.74 (0.43–1.27) |
0.27 | 119.93 (2.08–6914.21) |
0.08 | 0.76 (0.44–1.33) |
0.34 |
| Alcoholic drinks per week | 33 | 1.32 (0.86–2.02) |
0.21 | 2.14 (0.81–5.65) |
0.14 | 1.34 (0.81–2.21) |
0.26 |
| Cigarettes smoked per day | 22 | 0.98 (0.88–1.08) |
0.67 | 0.93 (0.78–1.11) |
0.42 | 0.96 (0.85–1.08) |
0.46 |
| Smoking initiation | 84 | 1.27 (1.09–1.49) |
<0.01 | 2.48 (1.13–5.43) |
0.03 | 1.26 (1.05–1.51) |
0.01 |
| Age at first birth | 54 | 0.88 (0.82–0.94) |
<0.01 | 0.87 (0.66–1.16) |
0.36 | 0.9 (0.82–0.99) |
0.03 |
| Years of schooling | 299 | 0.76 (0.66–0.87) |
<0.01 | 0.56 (0.33–0.97) |
0.04 | 0.75 (0.63–0.91) |
<0.01 |
| C-reactive protein level | 53 | 1.04 (0.95–1.15) |
0.39 | 0.99 (0.85–1.14) |
0.86 | 1.02 (0.93–1.11) |
0.73 |
| Serum 25-Hydroxyvitamin D | 101 | 1.06 (0.93–1.2) |
0.38 | 1.06 (0.87–1.3) |
0.56 | 1.06 (0.87–1.27) |
0.58 |
| Total testosterone (man) | 151 | 0.87 (0.72–1.05) |
0.15 | 0.81 (0.57–1.13) |
0.22 | 1.06 (0.8–1.4) |
0.69 |
| HDL cholesterol | 86 | 0.99 (0.92–1.06) |
0.69 | 1.01 (0.9–1.13) |
0.87 | 0.95 (0.86–1.05) |
0.33 |
| LDL cholesterol | 69 | 0.99 (0.94–1.04) |
0.62 | 0.98 (0.91–1.05) |
0.54 | 0.98 (0.91–1.05) |
0.59 |
| Total cholesterol | 78 | 1.00 (0.94–1.07) |
0.95 | 0.96 (0.87–1.07) |
0.51 | 0.98 (0.89–1.08) |
0.68 |
| Triglycerides | 55 | 0.98 (0.9–1.08) |
0.72 | 0.92 (0.8–1.07) |
0.29 | 0.93 (0.83–1.05) |
0.26 |
| Adiponectin | 14 | 1.04 (0.95–1.15) |
0.37 | 1.12 (0.98–1.27) |
0.12 | 1.09 (0.97–1.22) |
0.16 |
| Leptin | 5 | 0.79 (0.62–1.01) |
0.06 | 0.29 (0.06–1.48) |
0.23 | 0.82 (0.64–1.04) |
0.11 |
| Fasting glucose | 63 | 0.85 (0.71–1.03) |
0.1 | 1.21 (0.87–1.67) |
0.26 | 0.95 (0.73–1.25) |
0.74 |
| Fasting insulin | 38 | 0.88 (0.57–1.36) |
0.57 | 0.84 (0.21–3.43) |
0.81 | 1 (0.63–1.58) |
0.99 |
| HbA1c | 11 | 0.97 (0.74–1.28) |
0.84 | 0.80 (0.39–1.63) |
0.56 | 1.09 (0.75–1.58) |
0.66 |
| Serum urate | 2 | 0.96 (0.91–1.01) |
0.13 | NA | NA | NA | NA |
| Allergic rhinitis | 4 | 1.06 (0.95–1.18) |
0.33 | 0.77 (0.42–1.4) |
0.48 | 1.06 (0.94–1.2) |
0.31 |
| Asthma (childhood) | 94 | 1.00 (0.99–1) |
0.26 | 1.00 (0.97–1.03) |
0.86 | 1.00 (0.99–1) |
0.59 |
| Asthma (adult) | 40 | 1 (0.99–1) |
0.61 | 1.04 (0.97–1.12) |
0.26 | 1 (1–1.01) |
0.63 |
| Chronic sinusitis | 5 | 1.06 (0.96–1.18) |
0.27 | 0.86 (0.64–1.16) |
0.4 | 1.11 (0.97–1.26) |
0.13 |
| Nasal polyp | 8 | 1.05 (0.99–1.11) |
0.08 | 1.05 (0.89–1.24) |
0.59 | 1.05 (1–1.11) |
0.07 |
| Chronic tonsils and adenoids | 13 | 0.99 (0.84–1.17) |
0.92 | 1.23 (0.69–2.17) |
0.50 | 0.99 (0.85–1.16) |
0.94 |
| Gastroesophageal reflux disease | 74 | 1.60 (1.40–1.82) |
<0.01 | 2.6 (1.2–5.61) |
0.02 | 1.43 (1.24–1.65) |
<0.01 |
| Barrett’s esophagus | 16 | 1.07 (0.99–1.14) |
0.07 | 1 (0.59–1.67) |
0.99 | 1.05 (0.96–1.15) |
0.31 |
| Type 2 diabetes | 114 | 1.03 (0.98–1.08) |
0.24 | 0.98 (0.87–1.1) |
0.72 | 0.97 (0.91–1.04) |
0.43 |
| Breast cancer | 129 | 1.01 (0.97–1.06) |
0.64 | 0.99 (0.91–1.09) |
0.91 | 1 (0.94–1.06) |
0.96 |
| Prostate cancer | 109 | 0.97 (0.94–1) |
0.09 | 0.98 (0.92–1.04) |
0.46 | 0.97 (0.92–1.01) |
0.16 |
| Rheumatoid arthritis | 34 | 0.99 (0.96–1.03) |
0.60 | 0.98 (0.89–1.09) |
0.78 | 1.01 (0.96–1.07) |
0.66 |
| Bipolar disorder | 13 | 1.03 (0.94–1.13) |
0.51 | 0.84 (0.48–1.45) |
0.54 | 1.1 (0.98–1.24) |
0.11 |
| Major depression | 47 | 1.28 (1.06–1.53) |
0.01 | 0.88 (0.3–2.62) |
0.82 | 1.27 (1.03–1.57) |
0.02 |
| Polycystic ovary syndrome | 13 | 1.05 (0.97–1.14) |
0.22 | 1.29 (0.87–1.91) |
0.22 | 1.00 (0.92–1.10) |
0.90 |
HDL: high density lipoprotein; LDL: low density lipoprotein; OR: odds ratio; HbA1c: hemoglobin A1c; SNPs: single nucleotide polymorphism; MR: Mendelian randomization. Bold values indicate statistically significant results (p<0.05).
Table 3.
Power for conventional Mendelian randomization analysis.
| Risk factor | R 2 | OR | Sample size | Proportion of cases | No. of required for 80% | Power |
|---|---|---|---|---|---|---|
| Body mass index | 0.051 | 2.19 | 217955 | 0.077 | 1671 | 1 |
| Childhood body mass index | 0.230 | 1.46 | 217955 | 0.077 | 2402 | 1 |
| Overweight | 0.034 | 1.43 | 217955 | 0.077 | 18540 | 1 |
| Smoking initiation | 0.080 | 1.27 | 217955 | 0.077 | 19657 | 1 |
| Age at first birth | 0.058 | 0.88 | 217955 | 0.077 | 129819 | 0.95 |
| Years of schooling | 0.023 | 0.76 | 217955 | 0.077 | 80068 | 1 |
| Gastroesophageal reflux disease | 0.036 | 1.60 | 217955 | 0.077 | 9119 | 1 |
| Major depression | 0.018 | 1.28 | 217955 | 0.077 | 81326 | 1 |
OR: odds ratio.
Significant associations with an increased risk of OSA were observed for obesity factors, including body mass index (BMI; OR = 2.19, 95% CI 1.99–2.42, P < 0.01), childhood BMI (OR = 1.46, 95% CI 1.15–1.83, P < 0.01), and overweight (OR = 1.43, 95% CI 1.17–1.74, P < 0.01). Egger regression did not show evidence of directional pleiotropy for BMI (intercept = 0.002; P = 0.32) and childhood BMI (intercept = 0.019; P = 0.57) (Additional file: Table S1). Scatter plots of the effects of SNPs on potential risk factors versus their effects on OSA are shown in Additional file: Figures S1–S14. Leave-one-out sensitivity analysis showed that the association between obesity factors and OSA was not substantially driven by any individual SNP (Additional file: Tables S1–S3 and Figures S1–S14). In the context of MR analysis, the obesity factors (BMI, childhood BMI and overweight) do not constitute multicollinearity in conventional regression, as each exposure utilizes distinct genetic instruments. Furthermore, sensitivity analyses confirmed the robustness of our causal estimates.
Genetically predicted GERD (OR = 1.60, 95% CI 1.40–1.82, P < 0.01) and depression (OR = 1.28, 95% CI 1.06–1.53, P = 0.01) were significantly associated with an increased risk of OSA. When we included SNPs associated with lifestyle, smoking initiation (OR = 1.27, 95% CI 1.09–1.49, P < 0.01) was associated with an increased risk of OSA. On the contrary, age at first birth (AFB; OR = 0.88, 95% CI 0.82–0.94, P < 0.01) and years of schooling (OR = 0.76, 95% CI 0.66–0.87, P < 0.01) were associated with a decreased risk of OSA. AFB refers to a woman’s age at the time of her first live birth, and years of schooling in MR studies was defined herein as the total duration of successfully completed formal full-time education. MR-Egger regression did not show evidence of directional pleiotropy in GERD, depression, smoking initiation, AFB, and years of schooling (Additional file: Table S1).
No association was predicted between body fat percentage, childhood obesity, height, waist circumference, hip circumference, waist–hip ratio, alcoholic drinks per week, cigarettes smoked per day, C-reactive protein level, serum 25-hydroxyvitamin D, total testosterone (men), high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, total cholesterol, triglycerides, adiponectin, leptin, fasting glucose, fasting insulin, glycated hemoglobin, serum urate, allergic rhinitis, asthma, chronic sinusitis, nasal polyps, chronic tonsil and adenoid disease, Barrett’s esophagus, type 2 diabetes mellitus (T2DM), breast cancer, prostate cancer, rheumatoid arthritis, bipolar disorder, polycystic ovary syndrome, and OSA risk (Table 2).
4. Discussion
Most of the 42 risk factors investigated in this study have not been estimated using an MR approach. The present study findings revealed significant relationships between obesity factors, smoking, GERD, depression, and an increased risk of OSA. We also found evidence of significant correlations between AFB, education, and a decreased risk of OSA.
Our findings of an increased risk of OSA related to obesity were consistent with those of previous observational and MR studies [12,21,22]. This is supported by the high genetic associations between OSA and BMI. Our results were consistent with the finding that weight loss is an important factor in reducing the OSA severity [23]. OSA, T2DM, polycystic ovary syndrome (PCOS), and cancer share a common risk factor, namely, obesity. Previous research has revealed that T2DM is associated with high-risk OSA [24,25]. Additionally, women with PCOS are at increased risk of developing OSA [26]. An epidemiological study also reported that breast cancer increases the risk of OSA [27]. However, we did not find any evidence of an association between T2DM, glycated hemoglobin, fasting insulin, fasting blood glucose, PCOS, breast cancer, and OSA risk. The findings of this study differ from those of previous observational evidence, primarily stemming from methodological differences; prior studies have largely been observational whereas we investigated causal relationships in this MR analysis. Future prospective multicenter studies are warranted to validate the present findings.
Our study results suggested that smoking was associated with increased OSA risk. Smoking can cause nasal airflow resistance and chronic inflammation of the upper respiratory tract [28]. Long-term smoking can increase the prevalence of moderate or severe OSA [29]. A higher educational attainment was also found to be associated with a lower risk of higher BMI and smoking [30–32]. Higher education levels may lead to having an elevated socioeconomic status and enhanced awareness about OSA, concurrently improving health care accessibility and mitigating risk factors such as smoking and obesity [33]. This cascade effect could facilitate OSA screening and prevention. We found that education was significantly correlated with a decreased risk of OSA. The causal relationship between education and OSA risk has been explored in previous MR studies [13]. A larger sample size could provide greater power to determine such correlations.
Our findings revealed a significant association between AFB and OSA risk. The reason may be that a younger age when first giving birth is associated with increased BMI, which can increase the OSA risk [34]. Pregnancy and childbirth are characterized by adipose tissue expansion, insulin resistance, and inflammation, which may persist postpartum, and younger mothers typically experience a longer cumulative exposure duration [35]. This study also suggested that GERD increased the risk of OSA, which is in line with a previous observational study [36]. Obesity and depression are known to increase the risk of GERD [37,38]. Furthermore, depression was found to increase the risk of OSA. This result leads to a clearer understanding of the relationship between depression and OSA, which may contribute to better prevention and treatment of OSA.
The main strength of this analysis was that we used the MR method to evaluate the causal associations between 42 potential risk factors and OSA risk, which can reduce the confounding and reverse causal bias inherent in observational studies. The present MR study had sufficient power to detect moderate correlations and was unlikely to be affected by weak instrument bias. This study also has some limitations. Only individuals with European ancestry were included in this study. Given the multi-factorial interplay of genetic, environmental, and sociocultural determinants in the pathogenesis of OSA, our results may not translate well to other ancestry groups [33]. However, this can reduce the bias caused by population stratification. Future large-scale, multicenter, multi-ethnic prospective cohort studies are required to validate our findings.
5. Conclusion
In summary, this study revealed a causal effect of several potential risk factors on OSA risk, including obesity, smoking, education level, AFB, GERD, and depression. Further research is needed to elucidate the potential mechanisms that mediate the connection between these risk factors and OSA.
Acknowledgment
We thank Analisa Avila, MPH, ELS, of Liwen Bianji (Edanz) (www.liwenbianji.cn) for editing the language of a draft of this manuscript.
Funding Statement
This work was supported by funds from the Science and Technology Program of Guangzhou (201903010024) and the Chinese National Natural Science Foundation (81902771).
Ethics approval and consent to participate
The summary-level genetic data utilized in this study were obtained from open-access GWAS summary statistic from FinnGen study (www.finngen.fi/en). The original studies received ethical approval from their respective institutional review boards, and participants provided informed consent for data sharing and secondary research.
Author contributions
JC, WXL, PL, YL, HL, and YJ conceived the idea. WXL, PL, YL, JC, and ZYW collected and assembled the data. JC, WXL, YL, and PL conducted the experiments. JC, WXL, PL, YL, HL, ZYW, and YJ analyzed the data and wrote the manuscript. All authors read and agreed with the final manuscript.
Disclosure statement
The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
Data availability statement
The datasets analyzed during the current study are available in the IEU OPEN GWAS (https://gwas.mrcieu.ac.uk/) or PubMed (https://pubmed.ncbi.nlm.nih.gov/) repository, and Additional file 2 contains the corresponding GWAS ID and PMID.
References
Papers of special note have been highlighted as either of interest (*) or of considerable interest (**) to readers.
- 1.Senaratna CV, Perret JL, Lodge CJ, et al. Prevalence of obstructive sleep apnea in the general population: a systematic review. Sleep Med Rev. 2017;34:70–81. doi: 10.1016/j.smrv.2016.07.002 [DOI] [PubMed] [Google Scholar]
- 2.Veasey SC, Rosen IM.. Obstructive sleep apnea in adults. N Engl J Med. 2019;380(15):1442–1449. doi: 10.1056/NEJMcp1816152 [DOI] [PubMed] [Google Scholar]
- 3.Borsoi L, Armeni P, Donin G, et al. The invisible costs of obstructive sleep apnea (OSA): systematic review and cost-of-illness analysis. PLoS One. 2022;17(5):e0268677. doi: 10.1371/journal.pone.0268677 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Benjafield AV, Ayas NT, Eastwood PR, et al. Estimation of the global prevalence and burden of obstructive sleep apnea: a literature-based analysis. Lancet Respir Med. 2019;7(8):687–698. doi: 10.1016/S2213-2600(19)30198-5 ** This reference offers global data on OSA prevalence and its disease burden, which is central to our study’s importance. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Jordan AS, McSharry DG, Malhotra A.. Adult obstructive sleep apnoea. Lancet. 2014;383(9918):736–747. doi: 10.1016/S0140-6736(13)60734-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Karuga FF, Jaromirska J, Sochal M, et al. Association between glucose metabolism, the circadian cycle and hypoxia: evaluation of the NPAS2 and Rev-Erb-α protein serum levels in obstructive sleep apnea patients – a pilot study. Dent Med Probl. 2024;61(3):465–469. doi: 10.17219/dmp/185718 [DOI] [PubMed] [Google Scholar]
- 7.Kanclerska J, Wieckiewicz M, Nowacki D, et al. Sleep architecture and vitamin D in hypertensives with obstructive sleep apnea: a polysomnographic study. Dent Med Probl. 2024;61(1):43–52. doi: 10.17219/dmp/172243 [DOI] [PubMed] [Google Scholar]
- 8.Graizel-Armoni D, Greenbaum T, Brosh T.. Sex difference in the hyoid bone position in adults with obstructive sleep apnea: systematic review and meta-analysis. Dent Med Probl. 2025. [cited 2025 Apr 28]; [4 p.]. doi: 10.17219/dmp/192096 [DOI] [PubMed] [Google Scholar]
- 9.Hansen KK, Hilberg O, Løkke A.. High prevalence of undiagnosed and treatment-requiring obstructive sleep apnea and nocturnal desaturation in outpatients with severe chronic obstructive pulmonary disease. Respir Med. 2024;233:107771. doi: 10.1016/j.rmed.2024.107771 [DOI] [PubMed] [Google Scholar]
- 10.Wu Z-H, Yang X-P, Niu X, et al. The relationship between obstructive sleep apnea hypopnea syndrome and gastroesophageal reflux disease: a meta-analysis. Sleep Breath. 2019;23(2):389–397. doi: 10.1007/s11325-018-1691-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Duan X, Huang J, Zheng M, et al. Association of healthy lifestyle with risk of obstructive sleep apnea: a cross-sectional study. BMC Pulm Med. 2022;22(1):33. doi: 10.1186/s12890-021-01818-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Strausz S, Ruotsalainen S, Ollila HM, et al. Genetic analysis of obstructive sleep apnoea discovers a strong association with cardiometabolic health. Eur Respir J. 2021;57(5):2003091. doi: 10.1183/13993003.03091-2020 [DOI] [PubMed] [Google Scholar]
- 13.Song L, Li H, Wang J, et al. Educational attainment could be a protective factor against obstructive sleep apnea: a study based on Mendelian randomization. J Thorac Dis. 2022;14(1):210–215. doi: 10.21037/jtd-21-945 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.de Araujo Dantas AB, Gonçalves FM, Martins AA, et al. Worldwide prevalence and associated risk factors of obstructive sleep apnea: a meta-analysis and meta-regression. Sleep Breath. 2023;27(6):2083–2109. doi: 10.1007/s11325-023-02810-7 * This reference explores the risk factors for OSA through a meta-analysis, thus providing a basis for our research. [DOI] [PubMed] [Google Scholar]
- 15.Pagoni P, Dimou NL, Murphy N, et al. Using Mendelian randomisation to assess causality in observational studies. Evid Based Ment Health. 2019;22(2):67–71. doi: 10.1136/ebmental-2019-300085 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Burgess S, Butterworth A, Thompson SG.. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–665. doi: 10.1002/gepi.21758 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7:e34408. doi: 10.7554/eLife.34408 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Levin MG, Judy R, Gill D, et al. Genetics of height and risk of atrial fibrillation: a Mendelian randomization study. PLoS Med. 2020;17(10):e1003288. doi: 10.1371/journal.pmed.1003288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bowden J, Davey Smith G, Burgess S.. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–525. doi: 10.1093/ije/dyv080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Brion M-JA, Shakhbazov K, Visscher PM.. Calculating statistical power in Mendelian randomization studies. Int J Epidemiol. 2013;42(5):1497–1501. doi: 10.1093/ije/dyt179 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ioan I, Mulier G, Taytard J, et al. Evaluation of obesity and asthma as risk factors for moderate to severe obstructive sleep apnea in children. J Clin Sleep Med. 2022;18(6):1639–1648. doi: 10.5664/jcsm.9948 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Erridge S, Moussa O, McIntyre C, et al. Obstructive sleep apnea in obese patients: a UK population analysis. Obes Surg. 2021;31(5):1986–1993. doi: 10.1007/s11695-020-05196-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kuna ST, Reboussin DM, Strotmeyer ES, et al. Effects of weight loss on obstructive sleep apnea severity. ten-year results of the sleep AHEAD study. Am J Respir Crit Care Med. 2021;203(2):221–229. doi: 10.1164/rccm.201912-2511OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wondie A, Taderegew MM, Girma B, et al. Obstructive sleep apnea risk and its associated factors among type 2 diabetes mellitus patients at Wolkite University Specialized Hospital, Wolkite, Southern Ethiopia, 2021. A comparative cross-sectional study. Diabetol Metab Syndr. 2022;14(1):157. doi: 10.1186/s13098-022-00931-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Edmonds PJ, Gunasekaran K, Edmonds LC.. Neck grasp predicts obstructive sleep apnea in type 2 diabetes mellitus. Sleep Disord. 2019;2019:3184382–3184386. doi: 10.1155/2019/3184382 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kumarendran B, Sumilo D, O’Reilly MW, et al. Increased risk of obstructive sleep apnoea in women with polycystic ovary syndrome: a population-based cohort study. Eur J Endocrinol. 2019;180(4):265–272. doi: 10.1530/EJE-18-0693 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Madut A, Fuchsova V, Man H, et al. Increased prevalence of obstructive sleep apnea in women diagnosed with endometrial or breast cancer. PLoS One. 2021;16(4):e0249099. doi: 10.1371/journal.pone.0249099 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ioannidou D, Kalamaras G, Kotoulas S-C, et al. Smoking and obstructive sleep apnea: is there an association between these cardiometabolic risk factors? Gender analysis. Medicina (Kaunas). 2021;57(11):1137. doi: 10.3390/medicina57111137 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kim KS, Kim JH, Park SY, et al. Smoking induces oropharyngeal narrowing and increases the severity of obstructive sleep apnea syndrome. J Clin Sleep Med. 2012;8(4):367–374. doi: 10.5664/jcsm.2024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Li Y, Cai T, Wang H, et al. Achieved educational attainment, inherited genetic endowment for education, and obesity. Biodemography Soc Biol. 2021;66:132–144. *This reference proposes an interesting perspective: an association exists between education and OSA. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Wang L, Ren J, Chen J, et al. Lifestyle choices mediate the association between educational attainment and BMI in older adults in China: a cross-sectional study. Front Public Heal. 2022;10:1000953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.McCaffery JM, Papandonatos GD, Lyons MJ, et al. Educational attainment, smoking initiation and lifetime nicotine dependence among male Vietnam-era twins. Psychol Med. 2008;38(9):1287–1297. doi: 10.1017/S0033291707001882 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Iannella G, Pace A, Bellizzi MG, et al. The global burden of obstructive sleep apnea. Diagnostics (Basel). 2025;15(9):1088. doi: 10.3390/diagnostics15091088 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Patchen L, Leoutsakos J-M, Astone NM.. Early parturition: is young maternal age at first birth associated with obesity? J Pediatr Adolesc Gynecol. 2017;30(5):553–559. doi: 10.1016/j.jpag.2016.12.001 [DOI] [PubMed] [Google Scholar]
- 35.Yang HH, Chen GC, Zhou MG, et al. Association of age at first birth and risk of non-alcoholic fatty liver disease in women: evidence from the NHANES. Respir Med. 2023;17(2):303–312. *This reference provides unique insights into the potential relationship between age at first birth and fat deposition, and also offers a theoretical basis for the innovative findings of this study. [DOI] [PubMed] [Google Scholar]
- 36.Gilani S, Quan SF, Pynnonen MA, et al. Obstructive sleep apnea and gastroesophageal reflux: a multivariate population-level analysis. Otolaryngol Head Neck Surg. 2016;154(2):390–395. doi: 10.1177/0194599815621557 [DOI] [PubMed] [Google Scholar]
- 37.Ong J-S, An J, Han X, et al. Multitrait genetic association analysis identifies 50 new risk loci for gastro-oesophageal reflux, seven new loci for Barrett’s oesophagus and provides insights into clinical heterogeneity in reflux diagnosis. Gut. 2022;71(6):1053–1061. doi: 10.1136/gutjnl-2020-323906 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Chen G, Xie J, Ye J, et al. Depression promotes gastroesophageal reflux disease: new evidence based on Mendelian randomization. Turk J Gastroenterol Off J Turkish Soc Gastroenterol. 2023;34(5):457–462. [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.
Data Availability Statement
The datasets analyzed during the current study are available in the IEU OPEN GWAS (https://gwas.mrcieu.ac.uk/) or PubMed (https://pubmed.ncbi.nlm.nih.gov/) repository, and Additional file 2 contains the corresponding GWAS ID and PMID.