尋找與肌萎縮性側索硬化症相關的過去病史：全人口病例對照研究

全文

(1)國立臺灣師範大學健康促進與衛生教育學系碩士論文指導教授：李子奇博士. 尋找與肌萎縮性側索硬化症相關的過去病史：全人口病例對照研究 Finding diseases associated with amyotrophic lateral sclerosis: A total population-based case-control study. 研究生：胡宸語撰中華民國一 ○ 七年一月.

(2) National Taiwan Normal University Department of Health Promotion and Health Education Master Thesis. Academic Advisor: Dr. Charles Tzu-Chi Lee. Finding diseases associated with amyotrophic lateral sclerosis: A total population-based case-control study. Graduate Student: Chen-Yu Hu January 2018.

(3) Contents Abstract ..................................................................................................................... 1 中文摘要 ................................................................................................................... 5 Chapter 1 Introduction .............................................................................................. 8 1.1 Research background ......................................................................................................... 8 1.2 Study objective .................................................................................................................. 9. Chapter 2 Literature Review ................................................................................... 10 2.1 Introduction to amyotrophic lateral sclerosis (ALS) ....................................................... 10 2.2 Epidemiology and pathology of ALS population ............................................................ 12 2.2.1 Neoplasm (ICD-9: 140–239) ................................................................................13 2.2.2 Disorders of the thyroid gland (ICD-9:240–246) .................................................14 2.2.3 Diabetes mellitus (ICD-9:250) .............................................................................14 2.2.4 Overweightness, obesity, and other hyperalimentation disorders (ICD-9:278) and disorders of lipid metabolism (ICD-9:272) .............................15 2.2.5 Mental disorders (ICD-9:290–319) ......................................................................15 2.2.6 Diseases of the nervous system and sense organs (ICD-9:320–389) ...................16 2.2.7 Diseases of the circulatory system (ICD-9:390–459)...........................................16 2.2.8 Diseases of the respiratory system (ICD-9:460–519)...........................................17 2.2.9 Other diseases of the digestive system and liver (ICD-9:570–573) .....................18 2.2.10 Arthropathies and related disorders (ICD-9:710–739) .......................................18 2.2.11 Head trauma ........................................................................................................18 (open wound of head (ICD-9:873) and contusion of eye and adnexa (ICD-9:921)) ........................................................................................................18 2.2.12 PA .......................................................................................................................18. Chapter 3 Material and Methods ............................................................................. 24 3.1 Ethics statement ............................................................................................................... 24 3.2 Data sources ..................................................................................................................... 24 3.3 Study Design and Population .......................................................................................... 25 3.4 Statistical analysis ........................................................................................................... 28. Chapter 4 Results .................................................................................................... 31 4.1 Sample characteristics ..................................................................................................... 31 4.2 Prior diseases associated with ALS incidence: Univariate analysis ................................ 33 4.2.1 Hereditary and degenerative diseases of the central nervous system (ICD-9: 330–337) ................................................................................................33 4.2.2 Disorders of the peripheral nervous system (ICD-9: 350–359) ...........................34 4.2.3 Diseases of oral cavity, salivary glands, and jaws (ICD-9: 520–529)..................35 4.2.4 Diseases of the musculoskeletal system and connective tissue (ICD-9:710–739) .................................................................................................35 4.2.5 Other diseases associated with ALS incidence.....................................................37 4.3 Diseases associated with ALS incidence: Multivariate analysis ..................................... 42 4.3.1 Disorders of the peripheral nervous system (ICD-9: 350–359) ...........................42 4.3.2 Diseases of the ear and mastoid process (ICD-9: 380–389).................................42 I.

(4) 4.3.3 Diseases of the circulatory system (ICD-9: 390–459)..........................................43 4.3.4 Diseases of the respiratory system (ICD-9: 460–519)..........................................43 4.3.5 Diseases of the oral cavity, salivary glands, and jaws (ICD-9: 520–529) ............43 4.3.6 Diseases of the esophagus, stomach, and duodenum (ICD-9: 530–538) .............44 4.3.7 Diseases of the skin and subcutaneous tissue (ICD-9: 680–709) .........................44 4.3.8 Diseases of the musculoskeletal system and connective tissue (ICD-9: 710–739) ..............................................................................................................44 4.3.9 Injury (ICD-9: 800-999) .......................................................................................45 4.3.10 Other diseases associated with ALS incidence ...................................................45 4.4 Prior diseases associated with ALS incidence: Path analysis .......................................... 50 4.4.1 Prior diseases associated with ALS incidence, determined using path analysis ................................................................................................................50 4.4.2 Positive associations .............................................................................................51 4.4.3 Negative associations ...........................................................................................54 4.4.4 Combination of positive and negative associations into one model .....................57. Chapter 5 Discussion and Conclusion..................................................................... 61 5.1 Hypometabolism as a potential protective factor for ALS .............................................. 61 5.2 Hypermetabolism as a potential risk factor for ALS ....................................................... 63 5.3 Strengths of this study ..................................................................................................... 66 5.4 Conclusion ....................................................................................................................... 66. Chapter 6 Limitations .............................................................................................. 76 Reference ................................................................................................................. 78 Appendix I ............................................................................................................... 84. II.

(5) Table Contents Table 1 Summary of current epidemiological evidence on the medical history of ALS: case–control studies ......................................................... 20 Table 2 Summary of current epidemiological evidence on the medical history of ALS: cohort studies ..................................................................... 22 Table 3 Characteristics of the study patients ................................................ 32 Table 4 Hereditary and degenerative diseases of the central nervous system, ICD-9: 330–337 ............................................................................... 33 Table 5 Disorders of the peripheral nervous system, ICD-9: 350–359 ........ 34 Table 6 Diseases of oral cavity, salivary glands, and jaws, ICD-9: 520–52935 Table 7 Diseases of the musculoskeletal system and connective tissue, ICD-9: 710–739............................................................................................ 36 Table 8 Association of prior diseases with ALS for different periods before ALS diagnosis: univariate analysis .................................................. 38 Table 9 Number of patients with prior diseases associated with ALS for different screening periods before ALS diagnosis: univariate analysis .......................................................................................................... 40 Table 10 Diseases of the ear and mastoid process, ICD-9: 380–389 ........... 42 Table 11 Association of prior diseases with ALS for different periods before ALS diagnosis: multivariate analysis .............................................. 46 Table 12 Number of patients with prior diseases associated with ALS for different screening periods before ALS diagnosis: multivariate analysis ............................................................................................. 48 Table 13 Fit of model of prior diseases positively associated with ALS ..... 52 Table 14 Fit of model of prior diseases negatively associated with ALS ..... 55 Table 15 Prior diseases associated with ALS: model fit summary............... 58 Table 16 Standardized effects of prior diseases positively associated with ALS .................................................................................................. 60 Table 17 Standardized effects of prior diseases negatively associated with ALS .................................................................................................. 60 Table 18 Summary of current epidemiological evidence on ALS, along with our preliminary results: case–control studies ................................... 68 Table 19 Summary of current epidemiological evidence on ALS along with our preliminary results: cohort studies............................................. 73. III.

(6) Figure Contents Figure 1 Different screen time windows employed to reveal the diseases associated with ALS .................................................................... 25 Figure 2 Flowchart of data collection in this study ...................................... 27 Figure 3 Diagrammatic representation of the initial model for negative association diseases ........................................................................ 30 Figure 4 Diagrammatic representation of the initial model for positive association diseases ........................................................................ 30 Figure 5 Prior diseases positively associated with ALS: the path-analysis-corrected model ....................................................... 53 Figure 6 Prior diseases negatively associated with ALS: the path-analysis-corrected model ....................................................... 56 Figure 7 Prior diseases associated with ALS: the path-analysis-corrected model.............................................................................................. 59. IV.

(7) Abstract. Importance: Although many studies over the last 20 years have discovered diseases that may increase or decrease the risk of amyotrophic lateral sclerosis (ALS), inconsistent and ambiguous results muddle the direction of these associations. Objective: We investigated prior diseases associated with ALS using a total population-based medical claims database. Design: This was a total population-based case–control study. The index date was set as the date of diagnosis of ALS. Setting: This study was conducted using the National Health Insurance Research Database (NHIRD) and Serious Disabling Diseases (SDD) database in Taiwan. Participants: We included 705 new ALS cases aged more than 15 years from January 1, 2007, to December 31, 2013, and 14,100 sex-, age-, residence-, and insurance premium-matched controls. Exposure: Prior diseases were stratified as being diagnosed 1, 3, 5, 7, and 9 years prior to the ALS diagnosis date. Diseases were identified using the first 3 digits of International Classification of Diseases, ninth revision (ICD-9). Main Outcome Measure: ALS was classified according to the ICD-9 (335.20) by SDD database from Department of Health and Welfare, Taiwan. 1.

(8) Statistics analysis: Chi-squared or t-test was used to examine differences in demographic characteristics between new patients with ALS and controls. Prior diseases were first screened using conditional logistic regression model. The false discovery rate (FDR)-adjusted P value was then reported to avoid inflating false positives. A disease with FDR-adjusted P<0.10 was considered as a significantly associated disease. After the significant prior diseases were found using univariate analysis, multivariate analysis was performed using stepwise selection to evaluate the association between these diseases and the risk of ALS. We also used the path analysis to analyze the pathway between prior diseases and ALS. The effects of prior diseases on ALS were also estimated. Result: The mean (± standard deviation) age of patients with ALS was 56.65 ± 11.38 years. Of 705 patients with ALS, 400 were males and 305 were females (1.3:1). Only 50 ALS cases lived in rural areas. Nearly half of the ALS cases were under social welfare or supported by family members. In this study, 28 prior diseases were associated with ALS, including 17 positive and 11 negative associations. Diseases positively associated with ALS were the following: general medical examination (ICD-9: V70, OR=1.3, before 9 years), internal derangement of knee (ICD-9: 717, OR=1.9, before 5 years), intervertebral disc disorders (ICD-9: 722, OR=1.4, before 5 years), contusion of the face, scalp, and neck (ICD-9: 920, OR=1.5, before 5 years), nerve root and plexus disorders (ICD-9: 353, OR=1.5, before 3 years), symptoms concerning nutrition, metabolism, and development (ICD-9: 783, OR=2.1, before 3 years), need for prophylactic vaccination and inoculation against certain diseases (ICD-9: V04, OR=1.4, before 1 year), other extrapyramidal disease and abnormal movement disorders (ICD-9:333, OR=1.9, before 1 years), mononeuritis of the upper limb and mononeuritis 2.

(9) multiplex (ICD-9: 354, OR=1.3, before 1 year), mononeuritis of the lower limb (ICD-9: 355, OR=1.5, before 1 year), hereditary and idiopathic peripheral neuropathy (ICD-9: 356, OR=1.7, before 1 year), inflammatory and toxic neuropathy (ICD-9: 357, OR=1.7, before 1 year), occlusion of cerebral arteries (ICD-9: 434, OR=1.6, before 1 year), asthma (ICD-9: 493, OR=1.3 before 1 year), other cellulitis and abscess (ICD-9: 682, OR=1.3, before 1 year), spondylosis and allied disorders (ICD-9:721, OR=1.2, before 1 year), and disorders of the muscle, ligament, and fascia (ICD-9:728, OR=1.4, before 1 year. Diseases negatively associated with ALS were as follows: diabetes mellitus (ICD-9:250, OR=0.7 , before 5 years), other disorders of the skin and subcutaneous tissue (ICD-9:709, OR=0.7 , before 5 years), other forms of chronic ischemic heart disease (ICD-9:414, OR=0.8 , before 3 years), diseases of the hard tissues of the teeth (ICD-9:521, OR=0.8 , before 3 years), duodenal ulcer (ICD-9:532, OR=0.7 , before 3 years), carbuncle and furuncle (ICD-9:680, OR=0.7 , before 3 years), contusion of the lower limb (ICD-9:924, OR=0.8 , before 3 years), follow-up examination (ICD-9:V67, OR=0.7 , before 1 year), otitis media (ICD-9:382, OR=0.6 , before 1 year), disorders of function of the stomach (ICD-9:536, OR=0.8 , before 1 year), rheumatoid arthritis and other inflammatory polyarthropathies (ICD-9:714, OR=0.7 , before 1 year). Moreover, path analysis showed that the 11 negative association diseases can be considered as diabetes mellitus and its comorbidities. The 17 positive association diseases can be considered as metabolic syndrome, neuroinflammation, head trauma, sports injuries, infections, and their comorbidities. Conclusion: Our results supported the hypothesis that prior diseases for ALS were hypermetabolic syndrome diseases. Moreover, the hypometabolic syndrome 3.

(10) diseases may have a beneficial effect on ALS incidence.. Keywords: amyotrophic lateral sclerosis, National Health Insurance Research Database, Serious Disabling Disease Database, risk disease, case–control study, cohort study. 4.

(11) 中文摘要. 研究重要性：在過去二十多年，許多研究發現與肌萎縮性側索硬化症（amyotrophic lateral sclerosis, ALS）有相關之疾病危險因子，但是卻少有研究將這些疾病在 ALS 發病之前的罹患時間納入討論。研究目標：探討 ALS 診斷日之前的疾病罹患情況來揭示與 ALS 有關之過去病史。研究設計：本研究為全人口病例對照研究。研究起始點為第一次 ALS 診斷日。研究資料：本研究使用全民健保資料庫（National Health Insurance Research Database, NHIRD）以及重大傷病資料庫（Serious Disabling Disease Database, SDD）。研究對象：從 2007 年 1 月 1 日至 2013 年 12 月 31 日止，共有 705 位 15 歲以上之 ALS 新發病例。將這些 ALS 病例以性別、年齡、居住地、以及投保薪資進行配對，配對比為 1:20，得到 14,100 位對照組。研究暴露：篩選出 ALS 診斷日 1 年前、3 年前、5 年前、7 年前，以及 9 年前之過去病史，藉以評估該疾病發生是否為 ALS 的獨立變項。我們排除在首次診斷出 ALS 之前的 1、1~2、1~4、1~6、1~8 年間的疾病紀錄；疾病定義採用國際疾病分類碼第九版（International Classification of Diseases, 9th revision, ICD-9）的前三位數。 5.

(12) 主要結果測量：利用台灣衛生福利部之重大傷病資料庫（SDD）中的疾病紀錄，ALS 的 ICD-9 病例編碼為 335.20。統計分析：以卡方檢定或 T 檢定檢查 ALS 新發病例與對照組間的人口學特徵之差異。首先使用單變數條件邏輯式廻歸模型篩選出過去病史，再藉由偽發現率（false discovery rate, FDR）調整 p 值，避免膨脹偽陽性。以偽發現率（FDR）調整後的 p 值<0.10 作為相關疾病的初步篩選依據。除了經由單變數分析篩選疾病外，另外也進行多變數分析來評估先前疾病與 ALS 罹患風險之間的關聯。之後使用路徑分析（Path Analysis）來分析 ALS 病例及過去病史的關係網絡並討論之疾病間的影響力。結果：因配對研究設計的緣故，ALS 患者平均年齡 56.65 ± 11.38 歲（平均值±標準差）與對照組相同；在 705 例 ALS 患者中，男性 400 例，女性 305 例（男女比為 1.3：1）；50 例患者住在鄉村，近一半的 ALS 病例為接受社會福利補助或家庭的依附投保成員。本研究多變數分析結果共篩選出 28 項過去病史與 ALS 有顯著相關，其中正相關疾病有 17 項，負相關疾病有 11 項。一般檢查（ICD-9:V70， OR=1.3，九年以前），耳朵疾病（ICD-9:388，OR=1.6，九年以前），膝內障礙（ICD-9:717，OR=1.9，五年以前），椎間盤疾患（ICD-9:722， OR=1.4，五年以前），頭臉頸部狀挫傷（ICD-9:920，OR=1.5，五年以前），神經根及神經叢疾患（ICD-9:353，OR=1.5，三年以前），營養、新陳代謝和發育的症狀（ICD-9:783，OR=2.1，三年以前），病毒性疫苗接種（ICD-9:V04，OR=1.4，一年以前），其他錐體外疾病及不正常動作疾病（ICD-9:333，OR=1.9，一年以前），上肢神經炎（ICD-9:354，OR=1.3，. 6.

(13) 一年以前），下肢單一神經炎（ICD-9:355，OR=1.5，一年以前），末梢神經病變（ICD-9:356，OR=1.7，一年以前），發炎性及中毒性神經病變（ICD-9:357，OR=1.7，一年以前），腦動脈阻塞（ICD-9:434，OR=1.6，一年以前），氣喘（ICD-9:493，OR=1.3，一年以前），蜂窩性組織炎（ICD-9:682，OR=1.3，一年以前），脊椎關節病（ICD-9:721，OR=1.2，一年以前），肌肉、韌帶及筋膜疾患（ICD-9:728，OR=1.4，一年以前）與 ALS 呈顯著正相關。另外，與 ALS 呈顯著負相關的過去疾病有糖尿病（ICD-9:250，OR=0.7，五年以前），皮膚和皮下組織障礙（ICD-9:709，OR=0.7，五年以前），慢性缺血性心臟病（ICD-9:414，OR=0.8，三年以前），牙齒硬組織疾病（ICD-9:521，OR=0.8，三年以前），十二指腸潰瘍（ICD-9:532，OR=0.7，三年以前），癰及癤（ICD-9:680，OR=0.7，三年以前），下肢挫傷（ICD-9:924， OR=0.8，三年以前），追蹤檢查（ICD-9:V67，OR=0.7，一年以前），中耳炎（ICD-9:382，OR=0.6，一年以前），胃功能障礙（ICD-9:536，OR=0.8，一年以前），類風溼性關節炎（ICD-9:714，OR=0.7，一年以前）。路徑分析顯示與 ALS 有負相關的 11 個疾病可歸納為糖尿病及其相關併發症；與 ALS 有正相關的 17 個疾病可歸納為新陳代謝異常、神經發炎、頭部外傷、運動傷害、感染等。結論：通過這些結果，我們可以假設高代謝綜合症疾病是 ALS 的可能成因，其保護因子為代謝綜合症疾病。. 關鍵字：肌萎縮性側索硬化症（ALS）、全民健保資料庫（NHIRD）、重大傷病資料庫（SDD）、疾病危險因子、病例對照研究、縱貫性研究. 7.

(14) Chapter 1 Introduction 1.1 Research background Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease. Respiratory failure typically follows within 2–5 years of the onset of symptoms, necessitating mechanical ventilation.1 ALS is a rare disease, which makes estimation of its incidence and prevalence difficult without significant pathology. The incidence of ALS remained stable during 1999– 2008, with approximately 100 new cases per year. The mortality rate is more than 50% in the 5 years following ALS diagnosis.2,3 Epidemiology studies have concluded that some diseases are associated with the pathogenesis of ALS: they found inverse associations with diabetes4-6 and alcohol use disorder (AUD)7 and positive associations with psychiatric disorders,8 vascular diseases,9-11 cancer,12,13 and autoimmune diseases.14 By contrast, other studies have discovered no association of diabetes15,16 or cancer17 with ALS. Elucidating the association between diseases and medication use prior to the onset of ALS could lend support to the theory that specific subpopulations are at more risk of developing ALS and provide new insight into shared pathogenic mechanisms. Some studies18-22 have suggested that high physical activity (PA) level or physical trauma, especially if first trauma occurs before the age of 55,22 are associated with higher ALS incidence, even after excluding traumas that occur during the 5 years before ALS diagnosis. Head trauma induces glutamate excitotoxicity, is associated with mitochondrial dysfunction and neuroinflammation, and causes endoplasmic reticulum stress—all of which have been implicated in ALS pathogenesis.23 Because the results of studies are inconsistent and the exact cause of ALS remains unclear, we used a nationwide database in an attempt to reveal associations between prior. 8.

(15) medical conditions and ALS. Studies have rarely considered the screening time window used to exclude patients who develop a disease soon before diagnosis of ALS, which has introduced ambiguity into interpretation of the studies’ results: does a prior disease have a protective or detrimental effect on ALS development and does ALS have a protective or detrimental effect on the development of other diseases. In this study, we used the National Health Insurance Research Database (NHIRD) to perform a case–control study; prior diseases were screened based on whether they were diagnosed within 1, 3, 5, 7, and 9 years before ALS diagnosis. Significant diseases occurring long before ALS diagnosis may be less ambiguously established as protective or risk factors of ALS, whereas those occurring closer to ALS diagnosis may have developed due to the protective or detrimental role of ALS. Therefore, we considered the period between the diagnosis of the candidate disease and that of ALS and conducted a total-population-based case–control study to screen individual diseases for their association, as well as direction of association, with ALS risk in Taiwan.. 1.2 Study objective Epidemiology studies have suggested the influence of some diseases, such as cancer, diabetes, and psychiatric disorders, on the pathogenesis of ALS. Elucidating the association between diseases and medication use prior to ALS diagnosis could lend support to the theory that specific subpopulations are at risk of developing ALS and provide new insights into shared pathogenic mechanisms. Thus, we considered the period between the diagnosis of the candidate disease and that of ALS to determine their association, as well as direction of association, with ALS.. 9.

(16) Chapter 2 Literature Review 2.1 Introduction to amyotrophic lateral sclerosis (ALS) ALS is a relatively rare but fatal motor neuron disease that has a substantial economic and social burden; it is among the most rapidly progressive forms of neurodegenerative disease. ALS involves a heterogeneous mixture of upper and lower motor neuron loss with involvement of extra motor pathways that have a clinicopathologic overlap with frontotemporal dementia.24 Patients experience signs and symptoms of progressive muscle atrophy and weakness, increased fatigue, and difficulty swallowing, which typically lead to respiratory failure and death. Median survival is 3–5 years from symptom onset; only 20% of those diagnosed survive more than 5 years. The disorder is sporadic in 95% of patients.25 ALS is a fatal, paralytic, neurodegenerative disorder that is driven by complex and insufficiently understood pathophysiologic mechanisms. Clinically, ALS is rarely difficult for a neurologist to recognize, yet considerable interpatient variability exists for what is considered to be a singular disorder. The variability manifests neuroanatomically, wherein the weakness may begin in nearly any voluntary muscle and subsequently migrate throughout the nervous system. Several factors have been associated with survivability in ALS. Advanced age and bulbar onset have been linked with poor survivability in large epidemiologic studies.26,27 Specific clinical phenotype variants also display different rates of progression, with pure upper motor neuron (primary lateral sclerosis) and pure lower motor neuron (primary muscular atrophy) disorders tending to progress at slower rates. Considerable variation also exists in the 10% of 10.

(17) patients whose disease is caused by genetic mutations. Some of the fastest ALS progressions are observed inpatients with the Ala4Val mutation of the superoxide dismutase (SOD1) gene, whereas patients with the Phe64Leu SOD1 mutation present with very slowly progressing motor neuropathy. Even more phenotypic variability is seen with hexanucleotide expansion in the chromosome 9 open reading frame 72 (C9orf72) gene.28-32 ALS is mainly a clinical diagnosis, and lack of early diagnosis leads to a mean time from symptom onset to diagnosis of approximately 1 year.33 The World Federation of Neurology created the El Escorial criteria based on the presence and distribution of upper and lower motor neuron signs.34 Imaging, usually magnetic resonance imaging, is performed on the brain, spine, or both for assessing structural abnormalities, such as spinal stenosis or brainstem tumors, that can produce some of the symptoms and signs associated with ALS. A useful diagnostic approach was detailed by Shook and Pioro (2009).35 Nerve conduction studies are performed to exclude multifocal motor neuropathy and identify other neuropathies or sensory involvement, such as Kennedy disease. Needle EMG is used to detect the presence and distribution of fasciculation potentials, positive waves, and fibrillation potentials caused by denervation and enlarged remodeled motor unit potentials consistent with reinnervation. Needle EMG may also detect features of myopathies, such as inclusion body myositis, that can masquerade as a lower motor neuron presentation of ALS.. 11.

(18) 2.2 Epidemiology and pathology of ALS population Epidemiological studies on ALS have been largely conducted on populations of European descent. Incidence and prevalence of ALS are similar in Europe and North America.36 A large epidemiological study in 2010 examining 6 population-based registries of 3 European countries (Ireland, the United Kingdom, and Italy) for the period 1998–1999, covering as many as 24 million individuals, estimated the ALS incidence in Europe as 2.16 per 100,000 individuals.37 Within the population aged 18 years or older, the annual incidence was 3 for men and 2.4 for women (ratio of men to women, 1.4:1). A study from 2014 reported the prevalence in the United States to be 3.9 per 100,000 individuals.17 ALS incidence increases with age, especially after 40 years, reaching a peak at 70–74 years for men and 65–69 years for women. Diagnoses in patients over 80 years represent only 6.6% of the incident population, and those over 85 years, 1.9%.38 Because of either fast progression of the disease or because of other comorbid diseases, death occurs a short interval after onset of ALS, often before a diagnosis is made. The diagnosis age (mean ± standard deviation (SD)) of ALS in Taiwan was 56.6 ± 13.2 years during 1999–2008, with no significant sex difference (men, 57.1 ± 13.3 vs women, 55.9 ± 13.1).39 Although the exact pathogenesis of ALS is unclear, numerous researchers have attempted to determine the risk and protective factors of ALS. A summary of the epidemiological evidence regarding the medical history of ALS is presented in Table 1 and Table 2. We searched many online databases, including Medline, PubMed, the British Medical Journal, and Elsevier, for articles published during the period December 2011 to May 2017 on ALS and associated diseases. The articles were screened based on title, abstract, and keywords,. 12.

(19) including amyotrophic lateral sclerosis, ALS, and prior diseases. We also searched the reference lists of these articles for additional studies. Finally, we extracted 14 case–control studies and 8 cohort studies. A total of 12 influential factors were identified: diseases of the respiratory system; diseases of the nervous system and sense organs; head trauma; mental disorders; diabetes mellitus; overweightness, obesity, and other hyperalimentation disorders; diseases of the circulatory system; neoplasm; liver diseases; disorders of the thyroid gland; arthropathies and related disorders; and PA. We browsed the articles and tried to identify their research methods. However, none of the studies screened prior diseases on the basis of the time of their occurrence before date of ALS diagnosis. This results in ambiguity, making it impossible to state whether a disease was a protective or risk factor for ALS or ALS was an influential factor for that disease. In this study, we used the NHIRD to perform a case–control study, and cases of prior diseases were screened out based on whether they were diagnosed within 1, 3, 5, 7, and 9 years before ALS diagnosis. If a disease developed long before ALS, this would less ambiguously establish its role as a protective or risk factor of ALS. Diseases that developed closer to ALS diagnosis may indicate the role of ALS in their occurrence. Next, the 12 factors associated with ALS as deemed from previous studies are discussed. The International Classification of Disease, ninth revision (ICD-9) codes for each factor are provided.. 2.2.1 Neoplasm (ICD-9: 140–239) In 2013, Freedman et al.40 reported that ALS mortality was not significantly associated with incidence of total cancers (standardized. 13.

(20) mortality ratio (SMR) = 1.00), a significantly elevated risk of ALS death among survivors of melanoma (SMR = 1.49), tongue cancer (SMR = 2.57), or significantly reduced ALS death risk among prostate cancer survivors (SMR = 0.86); in 2014,17 they reported the hazard ratio (HR) of ALS after all cancers were combined in the 10-year follow-up period (HR = 0.99) and after a 1-year follow-up period (HR = 1.98). Fang et al.29 and Seelen et al.30 reported that a previous cancer was not associated with subsequent risk of ALS. Gibson et al.13 observed significantly elevated HRs for salivary gland cancer (HR = 5.27) and testicular cancer (HR = 3.82), and the only site-specific cancer to result in a significantly decreased risk of ALS was lung cancer (HR = 0.23).. 2.2.2 Disorders of the thyroid gland (ICD-9:240–246) Turner et al.14 found that ALS was associated with a prior diagnosis—at least 1 year before ALS diagnosis—of myxedema (risk ratio (RR) = 1.2).. 2.2.3 Diabetes mellitus (ICD-9:250) Turner et al.14 also discovered significantly more ALS cases than expected among patients with a prior diagnosis of young-onset diabetes (younger than 30 years; RR = 3.94). Sun et al.6 estimated the incidence density of ALS and calculated the relative HRs of ALS in relation to diabetes. Patients with diabetes had a significantly elevated HR of 1.35; a higher covariate-adjusted HR was noted in men (HR = 1.48), especially men aged less than 65 years (HR = 1.67). Mariosa et al.5 conducted a population-based case–control study and explored the association between diabetes and ALS risk (OR = 0.79); the. 14.

(21) association was stronger for patients with type II diabetes (OR = 0.66) and for younger individuals (<50 years). Preexisting type I diabetes was associated with a higher ALS risk (OR = 5.38). Kioumourtzoglou et al.4 found the odds ratio for ALS to be 0.61 (95% CI: 0.46–0.80) for diabetes.. 2.2.4 Overweightness, obesity, and other hyperalimentation disorders (ICD-9:278) and disorders of lipid metabolism (ICD-9:272) Sutedja et al.9 analyzed blood samples of 303 patients with ALS and compared the resultant data with prospectively collected data from 2100 population-based controls; they found that patients with ALS had a lower BMI (OR = 0.9), a lower low-/high-density lipoprotein (LDL/HDL) ratio (women, OR = 0.4; men, OR = 0.5), and lower homocysteine levels (women, OR = 0.9; men, OR = 0.9). Seelen et al.41 reported that patients with ALS have a lower prevalence of hypercholesterolemia than normal adults (OR = 0.76). Kioumourtzoglou et al.4 found the odds ratio for ALS to be 0.81 (95% CI: 0.57–1.16) for obesity. Hollinger et al.42 determined that antecedent obesity (BMI > 30) was correlated with earlier ALS onset (OR = 1.994) and that hyperlipidemia (OR = 0.416) and hypertension (OR = 0.499) were significantly associated with late ALS onset.. 2.2.5 Mental disorders (ICD-9:290–319) Turner et al.8 used a hospital record linkage database and searched for patients hospitalized with a diagnosis of mental disorder; they found that a diagnosis of depression (RR = 1.50) was significantly associated with a diagnosis of ALS ≥5 years later. 15.

(22) Ji et al.7 examined the Swedish Hospital Discharge and Outpatient Registers to investigate ALS incidence among individuals with AUD; they found that ALS incidence among individuals with AUD was lower than that among individuals without AUD (standardized incidence ratio (SIR) = 0.54).. 2.2.6 Diseases of the nervous system and sense organs (ICD-9:320–389) In 2012, Turner et al.43 reported that Parkinson disease (RR = 0.95) and multiple sclerosis (RR = 0.95) were not associated with a higher risk of ALS; however, in 2013, they discovered that multiple sclerosis (RR = 4.25) and myasthenia gravis (RR = 5.02) were significantly correlated with ALS.14 Phukan et al.44 found that 13.8% of patients with ALS fulfilled the Neary criteria for frontotemporal dementia, 34.1% without evidence of dementia fulfilled the recently published consensus criteria for cognitive impairment, 14% had evidence of cognitive impairment without executive dysfunction, and almost half the cohort (46.9%) had no cognitive abnormality. Hollinger et al.42 determined that non-ALS neurological diseases (OR = 0.097) are significantly associated with delayed ALS onset.. 2.2.7 Diseases of the circulatory system (ICD-9:390–459) Sutedja et al.9 found that lower homocysteine levels were correlated with lower risk of ALS (women, OR = 0.9, 95% CI: 0.9–1.0; men, OR = 0.9, 95% CI: 0.8–0.9). Turner et al.43 discovered that the rate ratio for ALS in the non-CHD cohort compared with the CHD cohort was 1.14. Turner et al.11 reported an increased rate ratio for ALS in patients with. 16.

(23) prior arteriovenous malformation (AVM) (RR = 2.69), all stroke (RR = 1.38), and transient ischemic attack (TIA) (RR = 1.47). Kioumourtzoglou et al.45 found that the adjusted OR for ALS among patients with any cardiovascular disease (CVD) admission at least 3 years prior to the date of ALS diagnosis was 1.15 (95% CI: 1.04–1.27); for atherosclerosis, 1.36 (95% CI: 1.02–1.80); and for ischemic heart disease, 1.14 (95% CI: 0.99–1.31). No association with hypertension (OR = 1.10, 95% CI: 0.90–1.31) or cerebrovascular diseases (OR = 0.95, 95% CI: 0.77–1.17) was identified. In 2017, Visser et al.46 reported that parents of patients with ALS were less likely to die from CVD compared with parents of controls (OR = 0.78; 95% CI: 0.64–0.94) and less likely to die because of severe trauma (OR = 2.30; 95% CI: 1.17–4.46); other diseases considered with those of the kidney, liver, or lung (OR = 2.09; 95% CI: 1.27–3.40).. 2.2.8 Diseases of the respiratory system (ICD-9:460–519) Asthma (ICD-9:493) Turner et al.14 reported an RR of 1.60 for ALS in patients with asthma aged 5–54 years, whereas Mitchell et al.47 detected a lower association of ALS with adult asthma (OR = 0.39). Chronic obstructive pulmonary disease (COPD) (ICD-9:490–496) Mitchell et al.47 detected a weak association of ALS with COPD (OR = 0.23). Kioumourtzoglou et al.45 performed analyses by stratifying patients with ALS by COPD status: the OR in patients with ALS but without COPD was 1.15 (95% CI: 1.04–1.28) and that in patients with both ALS and COPD was 1.18 (95% CI: 0.80–1.73), indicating a lack of a. 17.

(24) significant difference.. 2.2.9 Other diseases of the digestive system and liver (ICD-9:570–573) Hollinger et al.42 found that liver diseases (OR = 0.389) were associated with increased survival duration in patients with ALS.. 2.2.10 Arthropathies and related disorders (ICD-9:710–739) Turner et al.14 reported that ALS was associated with (at least 1 year) prior diagnosis of polymyositis (RR = 6.05).. 2.2.11 Head trauma (open wound of head (ICD-9:873) and contusion of eye and adnexa (ICD-9:921)) Seals et al.22 found a borderline association between any trauma and ALS (OR = 1.09); first trauma before the age of 55 years was associated with ALS (OR = 1.22), whereas first trauma occurring at older ages was not (OR = 0.97). Pupillo et al.48 discovered that a history of disabling traumatic events was predominant among patients with ALS (OR = 1.54) and the odds ratio was significant even after adjustment. Seelen et al.41 identified head trauma as being associated with ALS (OR = 1.95).. 2.2.12 PA Harwood et al.20 discovered that participation in an extra 10 kJ/kg/d of PA (OR = 1.47) was consistently associated with an increased risk of ALS, and an extra 10 min/d of vigorous PA was also associated with higher odds of ALS (OR = 1.03). 18.

(25) Fang et al.19 observed that the fastest skiers (HR = 2.08) and skiers who had completed 4 races during the study period both had a higher risk of ALS (HR = 1.88) than those who participated in only one race; those who finished within 180% of the winning time had a lower risk (HR = 0.46). Longinetti et al.49 found that men with physical fitness in the highest tertile have a high risk of ALS (HR = 1.42–1.75).. 19.

(26) Table 1 Summary of current epidemiological evidence on the medical history of ALS: case–control studies Research Sutedja et al. (2011)9 Phukan et al.(2012)44 Fang et al.(2013)50 Seelen et al.(2014)41 Kioumourtzoglou et al. (2015)4. Setting Study design Netherlands A case-control studies of questionnaire and blood samples. Irish A prospective population-based study Sweden A register-based case-control study Netherlands A population-based case-control study Danish A population-based nested case-control study. Mariosa et al.(2015)5. Sweden A population-based case-control study. Mitchell et al. (2015)47 Pupillo et al.(2016)48. US A case-control study Italy, Ireland, France, United Kingdom, Serbia A population-based case-control study. Harwood et al.(2016)20. UK A case-control study. Duration. Sample. July 2004-July 2009. Blood samples were 303 ALS and 2100 controls. 160 ALS and 110 controls. 1987-2009. 5481 ALS and 27405 controls 722 ALS and 2268 controls. 1982-2009. 3650 ALS and 36500 controls. Jan 1991 - Dec 2010 diabetes duration (0-2, 2-4, 4-6, 6-8, 8-10, 10+ years). 5108 ALS cases and 25540 controls. 1288 ALS and 7561 controls Feb 2008 - Apr 20012 censored 5 years before the date of symptoms onset (for controls the same date of their matched cases). 575 ALS cases and 1150 controls. 175 ALS and 317 controls. 20.

(27) Setting Study design Atlanta A case-control study. Duration. Seals et al.(2016)22. Denmark A case-control study. Kioumourtzoglou et al.(2016)45. Denmark A case-control study. Visser et al.(2017)46. Netherlands A case-control study Australia A case-control study. Jan 1982 - Dec 2009 up to 1 year after diagnosis of ALS, up to 1, 1-5, and more than 5 years prior to diagnosis 1982-2009 Detected the odds ratio for CVD and cause-specific CVD, for three, five and 10 years prior to the index date May 2010- April 2014. Research (continued) Hollinger et al.(2016)42. Timmins et al.(2016)51. Sample 839 ALS with antecedent conditions (hypertension, hyperlipidemia, diabetes, obesity, asthma, arthritis, chronic obstructive pulmonary disease (COPD), thyroid, kidney, liver, and other non-ALS neurological diseases) and 600 ALS without antecedent conditions 3650 ALS cases and 365000 controls. 3182 ALS cases and 318200 controls. 487 ALS cases and 1092 controls 58 ALS cases compared Australian population norms.. 21.

(28) Table 2 Summary of current epidemiological evidence on the medical history of ALS: cohort studies Research Turner et al. (2012)43. Fondell et al.(2013)52 Freedman et al.(2013)40. Turner et al.(2013)14. Setting Study design England A record-linkage cohort study. Duration. Sample. 3 hospital admission datasets, from two record linkage systems, were used. 1st for Oxford Record Linkage Study (ORLS), comprising of Oxford National Health Service (NHS) region from 1963-1998 (ORLS1) and from 1999-2008 (ORLS2). 2nd for English National Linked Hospital Episode Statistics spanning 1999-2008. US A follow-up cohort study US A population-based cohort study. 1992-2008. England A record-linkage cohort study. 1999 - 2011. ORLS1_ALS_No CHD: 205 ORLS1_ALS_CHD:53 ORLS2_ALS_ No CHD:77 ORLS2_ALS_ CHD:40 England_ ALS_No CHD:1351 England_ ALS_CHD:1313 ORLS1_PD_No CHD:1407 ORLS1_PD_CHD:622 ORLS2_PD_No CHD:455 ORLS2_PD_CHD:520 England_PD_No CHD:8040 England_PD_CHD:12433 ORLS1_MS_No CHD:367 ORLS1_MS_HD:33 ORLS2_MS_No CHD:127 ORLS2_MS_CHD:41 England_MS_No CHD:2035 England_MS_CHD:848 51529(men, aged 46-81 years); 42 men were diagnosed with ALS 2.7 million cancer survivors who survived at least 1 year following cancer diagnosis; 1216 ALS deaths were reported among 1 year survivors of cancer over 16.6 million person-years of follow-up All England record-linkage database. 1973-2007. 22.

(29) Research (continued) Freedman et al.(2014)17. Sun et al.(2014)6 Turner et al. (2015)11. Ji et al.(2016)7 Gibson et al.(2016)13 Fang et al.(2016)19 Turner et al.(2016)8. Setting Study design US A population-based study. Duration. Sample. 1992–2005. Taiwan A population-based cohort study England A record-linkage cohort study. 2000-2008. 303 ALS cases were ascertained in cancer patients (2,154,062 person-years) and 246 ALS cases (2,467,634 person-years) in the reference population 615492 diabetic; 614835 controls. Sweden A national cohort study Utah A population-based cohort study Sweden a case-control cohort study England A case-control cohorts study. 1973-2010. 1999 - 2011. 1966-2013. All arteriovenous malformation (AVM) (n=12220; 6345 AVM, cerebral vessels; 872AVM, pre-cerebral vessels;6640AVM, peripheral) All stroke (n=1004289; 179717 Haemorrhagic; 526575 Ischaemic; 389919 Unspecified; 59227 subarachnoid haemorrhage(SAH); 265676 Transient ischaemic attack(TIA)) 7965 ALS (136 with AUD versus 7829 without AUD) 1081 ALS (114 with cancer). 1989-2010. 212246 skier (ALS=39) and 508176 controls. 1999-2011 Diagnosis of ALS (<1 Year Later, 1-4 Years Later, and ≥5Years Later). schizophrenia (narrowly defined), 151,026; schizophrenia (broadly defined), 224,978; bipolar (narrowly defined), 85,799; bipolar (broadly defined), 89,469; depression, 843,820; anxiety, 356,602; reference cohort, 8,679,473. 23.

(30) Chapter 3 Material and Methods 3.1 Ethics statement The Institutional Review Board (IRB) of National Taiwan Normal University approved this study (Protocol Number: 201603HM004). Written consent from the study patients was not required because data were collected from the NHIRD, which comprises deidentified secondary data released for research purposes. The IRB formally waived the need for consent.. 3.2 Data sources We performed this case–control study using information from the NHIRD.53 Outpatient care, hospital inpatient care, ambulatory care, dental service, and prior medical condition details were provided by the National Health Research Institute. In 1995, the National Health Insurance (NHI) program in Taiwan, which is a single-payer insurance system operated by the government, was established to ensure the health of the entire nation and prevent social problems caused by poverty and disease. By December 2010, 23.074 million people were enrolled nationwide, a coverage rate of 99.6%. Registration of all cases of severely disabling diseases (SDDs), such as chronic renal failure, myasthenia gravis, cancer, and ALS, are required by the Bureau of NHI before SDD certification can be granted. In 2008, 37,099 medical doctors and 553 neurology specialists were registered in Taiwan. In addition, 790,621 people had SDD certificates in 2008, which constituted 3.4% of the total population. The characteristics of the NHI program include the following: payroll-related premiums shared by government, employers, and employees; requirement of copayment for ambulatory care, inpatient care,. 24.

(31) and medicines; and a fee-for-service under the global budget. Furthermore, enrollment in the NHI program is mandatory. The NHIRD provides patient medical records dating back to 1996. We screened out prior diseases based on whether they were diagnosed within 1, 3, 5, 7, and 9 years from before the date of ALS diagnosis. Hence, only ALS cases of diagnoses made during 2007–2013 were included in this study to ensure each patient’s 9-year history of diseases before ALS diagnosis could be traced (Figure 1). The NHIRD contains the following information: demographic characteristics (age, sex, and income); diagnosis code according to the ICD-9; procedure; claimed prescriptions; and prior diseases. Personal identity is encrypted for privacy protection, but all information in the NHIRD could be cross-linked to each data set, with unique and double-identified secondary data released to the public for research purposes.. Figure 1 Different screen time windows employed to reveal the diseases associated with ALS. 3.3 Study Design and Population ALS cases were identified according to the ICD-9 code 335.20. The study period was from 2007 to 2013, and the NHIRDs published since 25.

(32) 1997 were used for analysis. The medical claims data from 1997 to 2006 were used to verify that patients with ALS in this study were new cases of ALS. The ALS diagnosis was based on the decision of the in-charge clinical neurologist, and the medical records of patients were sent to the NHI Bureau. Another group of neurologists of the NHI Bureau verified the medical records of ALS diagnosis to confirm a diagnosis. Only patients with ALS who had SDD certification were included. Patients with SDD certificates are eligible for exemption from insurance premiums and copayments. SDD certificate approval requires a strict evaluation by the Ministry of Health and Welfare, Executive Yuan, in Taiwan. In this study, all ALS cases were verified by linking encrypted identification numbers with SDD certificates. Overall, 705 new cases of ALS diagnosed from January 1, 2003, to December 31, 2013, were included and were compared with 14,100 sex-, age-, residence-, and insurance-premium-matched controls. Controls with birthdays as close as possible to their case equivalents were collected. Residence was categorized as rural or urban. Insurance premium was a proxy indicator of economic status and was classified into 1 of 4 categories: fixed premium and dependent, less than New Taiwan Dollar (NTD) 20,000 monthly, NTD 20,000–39,999 monthly, and NTD 40,000 or more monthly (1 USD = 32.1 NTD in 2008). The fixed premium group comprised people receiving social welfare support and included veterans and people on low income. The dependent insurance premium group comprised people with family members who did not have a job or income. We focused on the relationship between ALS and prior diseases. We included all patients with ALS aged more than 15 years, and data of first date of ALS diagnosis were assigned as the index data. If patients had been diagnosed with other motor neuron diseases (ICD-9: 335.21, 335.22, 335.23, 335.24, 335.29), those patients were excluded. 26.

(33) The study patients were enrolled by applying the following inclusion/exclusion criteria (Figure 2): Step 1: Patients with ALS (ICD-9: 335.20) diagnosed during 1997–2013 and an SDD certificate were considered. Step 2: The index date was the date of first inpatient or outpatient ALS diagnosis. Step 3: From these patients, we extracted cases with date of first inpatient or outpatient ALS diagnosis between January 2007 and December 2013. Newly diagnosed patients with ALS were defined as those who had no ALS diagnosis from 1997 to 2006. Step 4: If age at ALS diagnosis was less than 15 years, the case was excluded. Step 5: Controls from the general population were matched exactly in sex, age, residence, and insurance premium.. Figure 2 Flowchart of data collection in this study. 27.

(34) 3.4 Statistical analysis Chi-squared tests or t-tests were used to examine differences in demographic characteristics between patients newly diagnosed with ALS and controls. First, prior diseases were screened using a univariate conditional logistic regression model. To evaluate whether a disease was an independent factor for ALS, we excluded the past 1, 1–2, 1–4, 1–6, and 1– 8 years of disease before the first ALS diagnosis. The false discovery rate54 (FDR) -adjusted p value was then calculated, which avoided inflating the number of false positives. An FDR-adjusted p < 0.10 was considered significant. Second, we performed multivariate analysis by stepwise selection to evaluate the association between prior diseases and the risk of ALS. Analyses were performed using SAS version 9.3 (SAS Institute Inc., Cary, NC, USA). We also employed path analysis using AMOS 21 (IBM Corp., 2012) to analyze the pathway between ALS and other diseases (錯誤! 找不到參照來源。 and Figure 3 ). The effects of prior diseases and ALS were also estimated. Significant associations were separated into positive and negative associations. Diagrams of the initial models of positive and negative association diseases are presented in 錯誤! 找不到參照來源。 and Figure 3 , respectively. Studies have rarely used path analysis to examine the disease and comorbidity associated with ALS. The major challenge in investigation of multiple prior diseases/conditions relates to the complexity and collinearities of the relationships among these variables, which cannot be unraveled using standard statistical analyses. The path analysis model is a powerful statistical tool that combines factor analysis and mathematical modeling to test hypotheses consisting of interacting variables and 28.

(35) pathways with reference to substantive theory. This method is widely employed in the psychological, social, educational, and management fields, where indicator variables such as attitude, performance, and personality traits, which cannot be readily measured, must be derived from questionnaires. The interdependent nature of these variables is very similar to the complexity of a biological system, characterized by interactions among endogenous and exogenous processes. Hence, path analysis can be a useful research tool to complement the more conventional experimental or cohort studies when an attempt I smade to understand biological phenomena and human diseases. We employed this model to evaluate the relationship between ALS and prior diseases. Using the path analysis model correction procedure, modification indices (MIs) were used to assess the local model fit, after which the initial model was accordingly adjusted one disease at a time for improving the goodness-of-fit. We included diseases significantly associated with ALS through multivariate analysis for the path analysis. After these correction steps, we also checked the global model fit indices. Bagozzi and Yi55 considered that if the sample size is sufficiently large, X2/df can be employed to exclude the effect of the complexity and the chi-squared value can be replaced to assess the model. The acceptable value of X2/df was 1–5. After the previous procedures were performed for positive and negative association diseases, we combined the corrected models. Again using the path analysis model correction, we modified the combined model to obtain the final model.. 29.

(36) Figure 3 Diagrammatic representation of the initial model for negative association diseases. Figure 4 Diagrammatic representation of the initial model for positive association diseases. 30.

(37) Chapter 4 Results 4.1 Sample characteristics The clinicodemographic characteristics of the included patients are summarized in Table 3. Because of the exact matching design employed, the age, sex, residence, and insurance premium distributions of the ALS and non-ALS groups were equivalent. The mean age of the patients with ALS was 56.65 ± 11.38 years, identical to that of the controls. Among the 705 patients newly diagnosed with ALS, 400 were man and 305 women (1.3:1). According to a study of urbanization stratification in Taiwan,56 townships are classified into 7 types from highly urbanized city to remote town. A total of 90.64% of the patients with ALS lived in highly urbanized cities. Furthermore, nearly half of the cases were patients receiving social support or who were a dependent member in a family.. 31.

(38) Table 3 Characteristics of the study patients ALS,. Non-ALS,. n=705 (%). n=14100 (%). Female. 305 (43.26). 6100 (43.26). Male. 400 (56.74). 8000 (56.74). 15-24. 8 (1.13). 160 (1.13). 25-44. 93 (13.19). 1841 (13.06). 45-64. 436 (61.84). 8735 (61.95). 65+. 168 (23.83). 3364 (23.86). Highly Urbanized City. 639 (90.64). 12780 (90.64). Median Urbanized City. 33 (4.68). 660 (4.68). New Town. 21 (2.98). 420 (2.98). General Town. 10 (1.42). 200 (1.42). Agricultural Town. 2 (0.28). 40 (0.28). 320 (45.39). 6400 (45.39). Less than NTDB 20,000. 189 (26.81). 3780 (26.81). NTD20,000 ~39,999. 150 (21.28). 3000 (21.28). NTD40,000 or more. 46 (6.52). 920 (6.52). Characteristic. P value. Sex >0.99. ALS diagnosis age (year) >0.99. ResidenceA >0.99. Insurance premium Fixed premium and dependent. >0.99. A. No patients lived in aging or remote towns, the other two classes of township.. B. 1 USD = NTD 32.1 in 2008.. 32.

(39) 4.2 Prior diseases associated with ALS incidence: Univariate analysis There were 40 prior diseases (FDR < 0.1) that were significantly associated with ALS, with 36 positive and 4 negative associations (Table 8 and Table 9).. 4.2.1 Hereditary and degenerative diseases of the central nervous system (ICD-9: 330–337) We screened out 4 hereditary and degenerative diseases of the central nervous system (ICD-9: 332, 333, 334, and 336) (Table 4). Parkinson’s disease diagnosis earlier than 1 year before ALS diagnosis was associated with ALS (OR = 4). The estimated adjusted OR for ALS for any CVD admission earlier than at least 3 years prior to ALS diagnosis was 1.15 (95% CI: 1.04-1.27). The associations were similar for both sexes (Pinteraction = 0.87). However, the estimated ORs were attenuated when we considered CVDs that developed earlier than at least 5 or 10 years prior to the index date; Significant associations of CVDs with ALS were only identified when 1-year screening was employed. In other words, we did not find significant evidence that these diseases are risk or protective factors for ALS.. Table 4 Hereditary and degenerative diseases of the central nervous system, ICD-9: 330–337 ICD-9 code 330. Content. 331. Other cerebral degenerations. Cerebral degenerations usually manifest in childhood. 33.

(40) 332. Parkinson's disease. 333. Other extrapyramidal disease and abnormal movement disorders. 334. Spinocerebellar disease. 335. Anterior horn cell disease. 336. Other diseases of spinal cord. 337. Disorders of the autonomic nervous system. 4.2.2 Disorders of the peripheral nervous system (ICD-9: 350–359) Of the 10 main disorders of the peripheral nervous system, we screened out eight (Table 5). Five of the 10 were significantly associated with ALS, even when they occurred as long as 5 years before the index date. Muscular dystrophies and other myopathies (ICD-9:359) (Table 8) were highly associated with ALS, with an estimated OR of more than 20 in each screening period. The results imply that disorders of the peripheral nervous system are risk or protective factors for ALS.. Table 5 Disorders of the peripheral nervous system, ICD-9: 350–359 ICD-9 code 350. Content. 351. Facial nerve disorders. 352. Disorders of other cranial nerves. 353. Nerve root and plexus disorders. 354. Mononeuritis of upper limb and mononeuritis multiplex. 355. Mononeuritis of lower limb. 356. Hereditary and idiopathic peripheral neuropathy. 357. Inflammatory and toxic neuropathy. 358. Myoneural disorders. 359. Muscular dystrophies and other myopathies. Trigeminal nerve disorders. 34.

(41) 4.2.3 Diseases of oral cavity, salivary glands, and jaws (ICD-9: 520–529) The teeth diseases (ICD-9: 521, 522, 523, and 525) (Table 6) were negatively associated with ALS. The pulp, periapical, gingival, and periodontal diseases were also negatively associated with ALS when 7-year screening was employed.. Table 6 Diseases of oral cavity, salivary glands, and jaws, ICD-9: 520–529 ICD-9 code 520. Content. 521. Diseases of hard tissues of teeth. 522. Diseases of pulp and periapical tissues. 523. Gingival and periodontal diseases. 524. Dentofacial anomalies, including malocclusion. 525. Disorders of tooth development and eruption. Other diseases and conditions of the teeth and supporting structures. 526. Diseases of the jaws. 527. Diseases of the salivary glands. 528. Diseases of the oral soft tissues, excluding lesions specific for gingiva and tongue Diseases and other conditions of the tongue. 529. 4.2.4 Diseases of the musculoskeletal system and connective tissue (ICD-9:710–739) We screened out 9 musculoskeletal and connective tissue diseases (ICD-9: 715, 716, 721, 722, 723, 724, 728, 729, and 738) (Table 7). Of those that remained, most were associated with ALS when they occurred earlier than 5 years before the index date. These results imply that musculoskeletal and connective tissue diseases could be risk or protective 35.

(42) factors for ALS. Table 7 Diseases of the musculoskeletal system and connective tissue, ICD-9: 710–739 ICD-9 Content code 710 Diffuse diseases of connective tissue 711. Arthropathy associated with infections. 712. Crystal arthropathies. 713. Arthropathy associated with other disorders classified elsewhere. 714. Rheumatoid arthritis and other inflammatory polyarthropathies. 715. Osteoarthrosis and allied disorders. 716. Other and unspecified arthropathies. 717. Internal derangement of knee. 718. Other derangement of joint. 719. Other and unspecified disorders of joint. 720. Ankylosing spondylitis and other inflammatory spondylopathies. 721. Spondylosis and allied disorders. 722. Intervertebral disc disorders. 723. Other disorders of cervical region. 724. Other and unspecified disorders of back. 725. Polymyalgia rheumatica. 726. Peripheral enthesopathies and allied syndromes. 727. Other disorders of synovium, tendon, and bursa. 728. Disorders of muscle, ligament and fascia. 729. Other disorders of soft tissues. 730. Osteomyelitis, periostitis, and other infections involving bone. 731 732. Osteitis deformans and osteopathies associated with other disorders classified elsewhere Osteochondropathies. 733. Other disorders of bone and cartilage. 36.

(43) 734. Flat foot. 735. Acquired deformities of toe. 736. Other acquired deformities of limbs. 737. Curvature of spine. 738. Other acquired deformity. 739. Nonallopathic lesions, not elsewhere classified. 4.2.5 Other diseases associated with ALS incidence Nutritional deficiencies (ICD-9: 269); anxiety, dissociative and somatoform disorders (ICD-9: 300); chronic bronchitis (ICD-9: 491); pneumoconiosis (ICD-9: 505); functional digestive disorders (ICD-9: 564); other congenital musculoskeletal anomalies (ICD-9: 756); symptoms involving the respiratory system and other chest symptoms (ICD-9: 786); and contusion of the eye and adnexa (ICD-9: 921) were associated with ALS when they developed earlier than 5 years before the index date, implicating them as potential risk or protective factors for ALS. Moreover, Kaposi's sarcoma (ICD-9: 176), stroke (ICD-9: 434), asthma (ICD-9: 493), diseases and other conditions of the tongue (ICD-9: 529), symptoms involving the digestive system (ICD-9: 787), and an open wound of the head (ICD-9: 873) were associated with ALS only for the 1-year screening. Thus, these diseases were not implicated in ALS incidence.. 37.

(44) Table 8 Association of prior diseases with ALS for different periods before ALS diagnosis: univariate analysis Odds-ratio. of. disease. before x year of ALS. ICD-9 Disease name. first diagnosis1 x=1 x=3 x=5 x=7 x=9. 176. Kaposi's sarcoma. 40.1. 269. Other nutritional deficiencies. 10. 300. Anxiety, dissociative disorders Parkinson's disease. 332. and. somatoform. 12. 15. 20. 1.5 1.4 1.5 1.4 1.6 4 2.6. 334. Other extrapyramidal disease and abnormal movement disorders Spinocerebellar disease. 336. Other diseases of spinal cord. 10.4. 350. Trigeminal nerve disorders. 4.4 6.2 9.2. 353. Nerve root and plexus disorders. 5.7 3.2 2.8 2.4. 354. Mononeuritis of upper limb and mononeuritis multiplex Mononeuritis of lower limb. 2.2. 4.3 2.5 3. 357. Hereditary and idiopathic peripheral neuropathy Inflammatory and toxic neuropathy. 358. Myoneural disorders. 4.2. 359. Muscular dystrophies and other myopathies. 36.7 28. 434. Occlusion of cerebral arteries (stroke). 1.7. 491. Chronic bronchitis. 1.6 1.6 1.6. 493. Asthma. 1.5. 505. Pneumoconiosis, unspecified. 6.7. 521. Diseases of hard tissues of teeth. 0.7 0.8 0.8. 522. Diseases of pulp and periapical tissues. 0.7 0.7 0.7 0.7. 333. 355 356. 20. 38. 24. 2.6 3.6 6.2. 4.3 3.3 3.4 5.4 5.3. 40.1 30. 20. 1.8. 7.5 7.5 12.

(45) 523. Gingival and periodontal diseases. 0.7 0.8 0.7 0.8. 525. 0.6 0.7. 716. Other diseases and conditions of the teeth and supporting structures Diseases and other conditions of the tongue Functional digestive disorders, not elsewhere classified Osteoarthrosis and allied disorders Other and unspecified arthropathies. 721. Spondylosis and allied disorders. 2.5 1.7 1.6 1.6 1.8. 722. Intervertebral disc disorders. 2.7 2.1 2.2 1.9. 723. Other disorders of cervical region. 1.8. 724. Other and unspecified disorders of back. 1.4 1.3. 728. Disorders of muscle, ligament and fascia. 2.2 1.8 2. 729. Other disorders of soft tissues. 1.8 1.5 1.4 1.4. 738. Other acquired deformity. 2.1. 756. Other congenital musculoskeletal anomalies. 3.4 3.2 3.9. 780. General symptoms. 1.4. 786. 1.4 1.3 1.3. 873. Symptoms involving respiratory system and other chest symptoms Symptoms involving digestive system Other open wound of head. 921. Contusion of eye and adnexa. 4.4. 529 564 715. 787. 1. 3.4 1.3 1.3 1.3 1.4. 1.5. 1.5. 1.4 1.6 1.6. 2.2 2. 1.3 1.5. 1.4 1.7 5.2. Only significant results (FDR < 0.10) are shown in this table. FDR: false. discovery rate.. 39.

(46) Table 9 Number of patients with prior diseases associated with ALS for different screening periods before ALS diagnosis: univariate analysis Number of patients before x year of ALS first diagnosis1 x=1. ICD-9. x=5. ALS (%) Non-ALS (%). ALS (%). x=9 Non-ALS (%). ALS (%). Non-ALS (%). 176. 2 (0.3). 1 (0.0). 6 (0.9). 3 (0.0). 2 (0.3). 2 (0.0). 269. 3 (0.4). 6 (0.0). 3 (0.4). 4 (0.0). 2 (0.3). 2 (0.0). 300. 173 (24.5). 2553 (18.1). 3 (0.4). 8 (0.1). 3 (0.4). 5 (0.0). 332. 15 (2.1). 80 (0.6). 11 (1.6). 75 (0.5). 7 (1.0). 23 (0.2). 333. 14 (2). 111 (0.8). 13 (1.8). 77 (0.6). 6 (0.9). 23 (0.2). 334. 6 (0.9). 336. 7 (1). 14 (0.1). 47 (6.7). 490 (3.5). 20 (2.8) 205 (1.5). 350. 7 (1). 32 (0.2). 65 (9.2). 876 (6.2). 37 (5.3). 353. 67 (9.5). 261 (1.9) 134 (19.0) 1973 (14.0) 73 (10.4) 964 (6.8). 354. 34 (4.8). 326 (2.3). 355. 10 (1.4). 356. 23 (3.3). 110 (0.8). 23 (3.3). 170 (1.2). 357. 26 (3.7). 125 (0.9). 6 (0.9). 13 (0.1). 358. 10 (1.4). 48 (0.3). 8 (1.1). 42 (0.3). 359. 11 (1.6). 6 (0.0). 67 (9.5). 651 (4.6). 434. 36 (5.1). 491. 85 (12.1). 493. 66 (9.4). 932 (6.6) 290 (41.1) 5090 (36.1). 505. 3 (0.4). 9 (0.1) 175 (24.8) 2886 (20.5). 521. 433 (61.4). 9748 (69.1). 522. 229 (32.5). 5661 (40.2) 197 (27.9) 4870 (34.5). 5 (0.0) 118 (16.7) 1634 (11.6). 79 (11.2). 1167 (8.3). 56 (7.9) 665 (4.7). 426 (3). 45 (6.4) 583 (4.1). 77 (0.6) 189 (26.8) 3122 (22.1) 92 (13.1) 1280 (9.1). 435 (3.1) 277 (39.3) 4471 (31.7) 1110 (7.9). 78 (11.1). 402 (57) 8812 (62.5). 40. 1134 (8).

(47) 523. 489 (69.4). 525. 65 (9.2). 1966 (13.9). 529. 9 (1.3). 53 (0.4). 564. 223 (31.6). 3776 (26.8). 715. 206 (29.2). 3310 (23.5). 716. 105 (14.9). 1478 (10.5). 721. 216 (30.6). 2291 (16.3). 722. 119 (16.9). 1003 (7.1). 723. 85 (12.1). 1020 (7.2). 724. 333 (47.2). 5501 (39). 728. 81 (11.5). 780 (5.5). 729. 390 (55.3). 5788 (41.1). 738. 17 (2.4). 166 (1.2). 756. 17 (2.4). 103 (0.7). 780. 385 (54.6). 6653 (47.2). 786. 279 (39.6). 4445 (31.5). 787. 100 (14.2). 1526 (10.8). 873. 40 (5.7). 485 (3.4). 921. 6 (0.9). 28 (0.2). 1. 10773 (76.4) 441 (62.6) 9744 (69.1). Only significant results (FDR < 0.10) are shown in this table. FDR: false. discovery rate.. 41.

(48) 4.3 Diseases associated with ALS incidence: Multivariate analysis There were 28 prior diseases significantly associated with ALS, with 17 positive associations and 11 negative associations (Table 11 and Table 12).. 4.3.1 Disorders of the peripheral nervous system (ICD-9: 350–359). In multivariate analysis, of the 10 main disorders of the peripheral nervous system, 5 were significantly associated with ALS when they occurred earlier than 1 year before the index date, bur for no other screening window. Nerve root and plexus disorders (ICD-9: 353) were associated with ALS when 3-year screening was used, with the estimated OR < 2 for each screening period. The results of the multivariate and univariate analyses were different, implying that disorders of the peripheral nervous system were not dominant factors affecting ALS incidence.. 4.3.2 Diseases of the ear and mastoid process (ICD-9: 380–389). We screened out one of the ear diseases (ICD-9: 382) (Table 10). Suppurative and unspecified otitis media diagnosis earlier than 1-year before ALS diagnosis was negatively associated with ALS (OR = 0.6). Univariate analysis did not show any ear diseases to be associated with ALS, whereas multivariate analysis revealed that suppurative and unspecified otitis media may be a protective factor for ALS.. Table 10 Diseases of the ear and mastoid process, ICD-9: 380–389 42.

(49) ICD-9 code 380. Content. 381. Nonsuppurative otitis media and Eustachian tube disorders. 382. Suppurative and unspecified otitis media. 383. Mastoiditis and related conditions. 384. Other disorders of tympanic membrane. 385. Other disorders of middle ear and mastoid. 386. Vertiginous syndromes and other disorders of vestibular system. 387. Otosclerosis. 388. Other disorders of ear. 389. Hearing loss. Disorders of external ear. 4.3.3 Diseases of the circulatory system (ICD-9: 390–459) We screened out 2 diseases of the circulatory system from our procedure. Chronic ischemic heart disease (ICD-9:414) was negatively associated with ALS when occurring earlier than 3 years before ALS diagnosis. Occlusion of cerebral arteries (ICD-9:434) in both analyses was associated with ALS when occurring earlier than 1 year before (ORunivariate = 1.7; OR multivariate. = 1.6).. 4.3.4 Diseases of the respiratory system (ICD-9: 460–519) Multivariate analysis revealed 1 disease of the respiratory system associated with ALS. In both univariate and multivariate analyses, asthma (ICD-9:493) was significantly associated with ALS when it occurred earlier than 1 year before the index date (OR univariate = 1.5; OR multivariate = 1.3). 4.3.5 Diseases of the oral cavity, salivary glands, and jaws (ICD-9: 520–. 43.

(50) 529) Diseases of the teeth were still negatively associated with ALS. Diseases of the hard tissues of the teeth (ICD-9: 521) were negatively associated with ALS when occurring earlier than 3 years before ALS diagnosis in both univariate and multivariate analyses (OR univariate = 0.8; OR multivariate. = 0.8).. 4.3.6 Diseases of the esophagus, stomach, and duodenum (ICD-9: 530– 538) We found that duodenal ulcer (ICD-9:532) and disorders of function of the stomach (ICD-9:536) were negatively associated with ALS when they occurred earlier than 1 year before the index date.. 4.3.7 Diseases of the skin and subcutaneous tissue (ICD-9: 680–709) Two skin diseases—carbuncles and furuncles (ICD-9: 680) and other disorders of skin and subcutaneous tissue (ICD-9: 709)—were negatively associated with ALS when occurring earlier than 3 years before ALS diagnosis, but cellulitis and abscesses (ICD-9:682) were positively associated with ALS when occurring earlier than 1 year before the index date.. 4.3.8 Diseases of the musculoskeletal system and connective tissue (ICD-9: 710–739) Only 4 of the musculoskeletal system and connective tissue diseases (ICD-9: 714, 717, 722, and 728) were associated with ALS, with all ORs being less than 2. Rheumatoid arthritis and other inflammatory polyarthropathies (ICD-9: 714) were negatively associated with ALS only 44.

(51) when 1-year screening was used.. 4.3.9 Injury (ICD-9: 800-999) Contusion of the face, scalp, and neck except for eye(s) (ICD-9: 920) was positively associated with ALS (OR = 1.5 when occurring earlier than 5 years before). However, contusion of the lower limb, other, and unspecified sites was negatively associated with ALS when occurring earlier than 3 years before (OR = 0.8).. 4.3.10 Other diseases associated with ALS incidence Symptoms concerning nutrition, metabolism, and development (ICD-9: 783) (OR = 2.1) were associated with ALS. Moreover, diabetes mellitus (ICD-9:250) was negatively associated with ALS when developing earlier than 5 years before the index date (OR = 0.7).. 45.

(52) Table 11 Association of prior diseases with ALS for different periods before ALS diagnosis: multivariate analysis Odds-ratio. of. disease. before x year of ALS. ICD-9 Disease name. first diagnosis x=1 x=3 x=5 x=7 x=9. V67. Need for prophylactic vaccination inoculation against certain diseases Follow-up examination. V70. General medical examination. 1.2 1.3 1.2 1.2 1.3. 250. Diabetes mellitus. 0.7 0.7 0.7. 333. Extrapyramidal disease and abnormal movement 1.9 disorders Nerve root and plexus disorders 2.0 1.5. V04. 353. and1.4 0.7. 355. Mononeuritis of upper limb and mononeuritis 1.3 multiplex Mononeuritis of lower limb 1.5. 356. Hereditary and idiopathic peripheral neuropathy 1.7. 357. Inflammatory and toxic neuropathy. 1.7. 382. Suppurative and unspecified otitis media. 0.6. 414. Other forms of chronic ischemic heart disease. 0.8 0.8. 434. Occlusion of cerebral arteries. 1.6. 493. Asthma. 1.3. 521. Diseases of hard tissues of teeth. 0.7 0.8. 532. Duodenal ulcer. 0.8 0.7. 536. Disorders of function of stomach. 0.8. 680. Carbuncle and furuncle. 0.7 0.7. 682. Other cellulitis and abscess. 1.3. 709. Other disorders of skin and subcutaneous tissue. 0.7 0.7 0.7. 714. Rheumatoid arthritis and other inflammatory0.7 polyarthropathies. 354. 46. 2.0.

(53) 717. Internal derangement of knee. 1.5 1.8 1.9. 721. Spondylosis and allied disorders. 1.2. 722. Intervertebral disc disorders. 1.3. 728. Disorders of muscle, ligament, and fascia. 1.4. 783. Symptoms concerning nutrition, metabolism, and 2.2 2.1 development Contusion of face, scalp, and neck except eye(s) 1.7 1.5 1.5. 920 924. Contusion of lower limb and of other and 0.8 0.8 unspecified sites. 47. 1.4.

(54) Table 12 Number of patients with prior diseases associated with ALS for different screening periods before ALS diagnosis: multivariate analysis Number of patients before x year of ALS first diagnosis2 ICD-9. x=1 ALS (%). x=5 Non-ALS (%). ALS (%). x=9 Non-ALS (%). ALS (%). Non-ALS (%). V04. 83 (11.8). 1562 (11.1). 48 (6.8). 872 (6.2). 9 (1.3). 154 (1.1). V67. 5 (0.7). 237(1.7). 2(0.3). 146(1). 1(0.1). 63(0.5). V70. 264 (37.5). 5153(36.6). 198(28.1). 3631(25.8). 81(11.5). 1221(8.7). 250. 97 (13.8). 2383 (16.9). 67 (9.5). 1668 (11.8). 35 (5). 804 (5.7). 333. 14 (2.0). 111 (0.8). 13 (1.8). 77 (0.6). 6 (0.9). 23 (0.2). 353. 67 (9.5). 261 (1.9). 134 (19.0). 1973 (14.0). 73 (10.4). 964 (6.8). 354. 34 (4.8). 326 (2.3). 79 (11.2). 1167 (8.3). 45 (6.4). 583 (4.1). 355. 10 (1.4). 77 (0.6). 189 (26.8). 3122 (22.1). 92 (13.1). 1280 (9.1). 356. 23 (3.3). 110 (0.8). 23 (3.3). 170 (1.2). 7 (1). 23 (0.2). 357. 26 (3.7). 125 (0.9). 6 (0.9). 13 (0.1). 6 (0.9). 23 (0.2). 382. 16 (2.3). 373 (2.7). 13 (1.8). 267 (1.9). 7 (1). 118 (0.8). 414. 66 (9.4). 1500 (10.6). 41 (5.8). 998 (7.1). 19 (2.7). 441 (3.1). 434. 36 (5.1). 435 (3.1). 277 (39.3). 4471 (31.7). 6 (0.9). 99 (0.7). 493. 66 (9.4). 932 (6.6). 290 (41.1). 5090 (36.1). 26 (3.7). 314 (2.2). 521. 433 (61.4). 9748 (69.1). 402 (57). 532. 27 (3.8). 678 (4.8). 19 (2.7). 504 (3.6). 536. 219 (31.1). 4024 (28.5). 174 (24.7). 3234 (22.9). 680. 15 (2.1). 344 (2.4). 11 (1.6). 222 (1.6). 3 (0.4). 80 (0.6). 682. 52 (7.4). 895 (6.4). 31 (4.4). 563 (4). 12 (1.7). 233 (1.7). 709. 32 (4.5). 854 (6.1). 22 (3.1). 644 (4.6). 11 (1.6). 280 (2). 714. 10 (1.4). 225 (1.6). 10 (1.4). 167 (1.2). 5 (0.7). 79 (0.6). 48. 8812 (62.5) 300 (42.6) 6499 (46.1) 12 (1.7). 245 (1.7). 86 (12.2) 1605 (11.4).

(55) 717. 8 (1.1). 144 (1). 8 (1.1). 89 (0.6). 4 (0.6). 38 (0.3). 721. 216 (30.6). 2291 (16.3). 118 (16.7). 1634 (11.6). 56 (7.9). 665 (4.7). 722. 119 (16.9). 1003 (7.1). 67 (9.5). 651 (4.6). 23 (3.3). 282 (2). 728. 81 (11.5). 780 (5.5). 47 (6.7). 490 (3.5). 20 (2.8). 205 (1.5). 783. 11 (1.6). 101 (0.7). 8 (1.1). 53 (0.4). 3 (0.4). 16 (0.1). 920. 18 (2.6). 192 (1.4). 11 (1.6). 133 (0.9). 6 (0.9). 53 (0.4). 924. 82 (11.6). 1740 (12.3). 51 (7.2). 1266 (9). 25 (3.6). 586 (4.2). 49.