Incidence and mortality of ALS: a 42-year population-based nationwide study

Abstract

Background and aim

Recent studies have suggested that the incidence rate (IR) and the rate of death (MR) of amyotrophic lateral sclerosis (ALS) are increasing. Still, it remains unclear whether this is due to improved case ascertainment or represents a true increase. We examined the development in the incidence and mortality of ALS in Denmark for 42 years.

Methods

We retrieved individual-level data of all patients aged above 18 years with first-time ALS diagnosed at any Danish department of neurology. The IR and MR were calculated based on data from 1980 to 2021, stratified by gender and age.

Results

We identified 5,943 patients with ALS and identified a total of 5,069 deaths in the nationwide population. Overall, the IR was 3.4 per 100,000 persons per year (95% CI 3.4–3.5). ALS incidence rose gradually during the study period, and the IR was 2.8 times higher (95% CI 2.4–3.2) when comparing the latest period (2018–2021) with the first (1980–1983). Parallel to the IR, the MR increased over time and was associated with male gender and rose with age at diagnosis, peaking in the 70–79-year age group.

Conclusion

In Denmark, the IR and MR of ALS increased threefold from 1980 to 2021, with steadily increasing risk related to male gender and in particular to higher age. Considering our aging societies, the number of elderly patients with ALS can be expected to increase considerably.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00415-024-12743-0.

Introduction

Amyotrophic lateral sclerosis (ALS), the most common motor neuron disease in adulthood [1], affects both upper and lower motor neurons, leading to death due to respiratory failure [2]. Hereditary ALS forms, for which a genetic mutation often can be demonstrated, account only for 5–10% of cases [3]. The underlying etiology of sporadic ALS remains poorly understood and identification of risk factors for ALS has proven difficult [1]. Several environmental [4], occupational [5], lifestyle-related [6], traumatic [7], and hormonal [8] causes have been proposed without any conclusive evidence.

Across the world, the reported incidence rates (IR) of ALS have varied considerably [9]. In a recent meta-analysis including 44 studies covering 45 geographical areas, pooled IRs of ALS ranged from 0.5 per 100,000 person-years (PY) in Asia to 2.4 per 100,000 PY in Western Europe [10]. Using information from population-based motor neuron disease registries, the incidence of ALS was 3.8 per 100,000 person-years in Stockholm (2014) and Scotland (2015–2017) [11, 12]. Based on an extensive effort of information collection through different sources, a slightly lower incidence of ALS was reported in Nordland, Norway, during 2000–2015 (2.1 per 100 000 person-years) [13] and in Friuli-Venezia Giulia, Italy during 2002–2014 (2.8 per 100 000 person-years) [14]. Several studies have suggested an increase in incidence and mortality rates over time [9]. This increase could be explained both by methodological differences including study design, data sources, and case definition as well as due to a true change in incidence in different populations, geographical areas, and periods. The rarity of ALS implies that reliable epidemiologic estimates require long observation periods. Hence, to provide a more accurate picture of the development of ALS over time, we used validated nationwide data to obtain robust estimates of ALS incidence and mortality in 42 years, stratified by year of diagnosis, gender, and age at diagnosis.

Methods

Setting and data sources

Denmark has 6.0 million citizens (December 2023) [15] and since 1968, all Danish citizens have been registered in the Central Office of Civil Registration and have been given a unique 10-digit civil registry number (CPR number) [16]. This number contains information on birth date and gender and enables unique identification and matching of registry data at the individual level [16, 17]. The registration includes historical records on all individuals who have disappeared, immigrated, or emigrated [16]. The Danish National Health Service provides all Danish residents with free access (tax-funded) to medical care registered in one nationwide system, The Danish National Patient Registry (DNPR). The DNPR was established in 1977 and contains data on all hospital admissions, admission and discharge dates, hospital department, primary discharge diagnosis code (the primary reason for hospitalization), and supplementary diagnosis codes [18]. Hospital data are recorded prospectively for reimbursement purposes independently of specific research questions [18]. Medical diagnoses have been registered in the DNPR using the International Classification of Disease, version 8 (ICD-8) from 1977 through 1993 and the International Classification of Disease, version 10 (ICD-10) from 1994 onwards [19]. Data on the Danish general population size was obtained from Statistics Denmark [15].

The ALS disease-modifying treatment, riluzole, was approved in Denmark in June, 1996 [20]. Using the Danish National Prescription Registry [21], we identified all individual-level riluzole prescriptions among ALS patients, redeemed at community pharmacies in Denmark. Prescription data in the Danish National Prescription Registry were complete for all inhabitants in Denmark from 1994 [21].

Study population

Identification of patients with ALS

Adult patients (≥ 18 years) diagnosed for the first time with ALS were identified in the DNPR between January 1, 1980, and December 31, 2021. We restricted patient ascertainment to individuals diagnosed at an inpatient or outpatient clinic at departments of neurology in Denmark offering ALS care [22]. Primary discharge diagnoses from this case definition have a documented high positive predictive value (PPV) [23, 24]. Details including codes used to identify ALS cases were previously described [24]. In our prior validation study, we accepted the following phenotypes as correctly coded ALS cases: classical ALS, ALS with frontotemporal dementia, progressive muscle atrophy, and progressive bulbar palsy. Among the 25 patients whose ALS diagnosis was rejected, less than five cases (< 2%) had primary lateral sclerosis.

Statistical analyses

The average annual ALS IR and mortality rate (MR) were estimated based on the corresponding population (≥ 18 years) of Denmark obtained from Statistics Denmark [15] and 95% CIs were calculated assuming a Poisson distribution. IRs and MRs were also calculated for consecutive 2-year periods. Differences in IRs and MRs by year of ALS diagnosis (nine 4/5-year periods), gender, and age were evaluated by calculating incidence rate ratios (IRR) and mortality rate ratios (MRR), respectively.

All patients diagnosed with ALS were followed from the date of the first hospital contact for ALS until death, migration, or until the end of follow-up (42 years). We used (1-the survival function) to compute cumulative mortality.

We reran all IR and IRR analyses in a sensitivity analysis, including a subset of ALS patients who in addition to the previously described ALS case definition were treated with riluzole. This population was restricted to patients diagnosed with ALS from 1997 to 2021 because riluzole was approved in June 1996. All analyses were performed using Stata Software (StataCorp. 2023. Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC).

Results

Descriptive data

The final study population included 5,943 adult ALS patients from 1980 to 2021 (Table 1). Approximately 80% were aged between 50 and 79 years, 54.6% were male, and the mean age at diagnosis was 65 years (range, 18–96 years).

Table 1.

ALS incidence rates and incidence rate ratios

Category	ALS patients	IR/100000 PY	IRR
	N (%)	(95% CI)	(95% CI)
Total	5943	3.36 (3.28–3.45)	–
Males	3248 (54.7)	3.75 (3.63–3.89)	1.25 (1.19–1.32)
Females	2695 (45.3)	2.99 (2.88–3.11)	Reference
Year of diagnosis
1980–1983	270 (4.5)	1.61 (1.42–1.82)	Reference
1984–1988	321 (5.4)	1.62 (1.45–1.81)	1.01 (0.85–1.19)
1989–1993	395 (6.6)	1.93 (1.74–2.13)	1.20 (1.02–1.40)
1994–1998	665 (11.2)	3.21 (2.97–3.46)	1.99 (1.73–2.30)
1999–2003	708 (11.9)	3.38 (3.14–3.64)	2.10 (1.82–2.42)
2004–2008	913 (15.4)	4.33 (4.05–4.62)	2.69 (2.34–3.09)
2009–2013	992 (16.7)	4.56 (4.28–4.86)	2.83 (2.47–3.25)
2014–2017	852 (14.3)	4.72 (4.41–5.05)	2.93 (2.55–3.37)
2018–2021	827 (13.9)	4.45 (4.15–4.76)	2.76 (2.40–3.18)
Age at diagnosis
18–39	180 (3.0)	0.26 (0.23–0.31)	0.05 (0.04–0.06)
40–49	401 (6.7)	1.28 (1.16–1.41)	0.24 (0.21–0.27)
50–59	1075 (18.1)	3.83 (3.61–4.07)	0.71 (0.64–0.79)
60–69	1927 (32.4)	8.22 (7.85–8.59)	1.52 (1.38–1.67)
70–79	1812 (30.5)	10.85 (10.35–11.36)	2.00 (1.82–2.21)
≥ 80	548 (9.2)	5.42 (4.97–5.89)	Reference

ALS amyotrophic lateral sclerosis, CI confidence interval, IR incidence rate, IRR incidence rate ratio, PY person years

Incidence rates

The ALS IR was calculated for the period 1980–2021: a total of 5,943 unique adult individuals received an ALS diagnosis, and the Danish adult population accumulated during this period to a total of 176,612,807. The overall incidence rate was 3.4 per 100,000 PY (95% CI 3.4–3.5). ALS incidence gradually rose during the study period (Fig. 1a), and the IRR had clearly increased (IRR; 2.8; 95% CI 2.4–3.2) when comparing the latest period (2018–2021) with the first (1980–1983) (Table 1). This increased IR over time applied to both males and females (Fig. 1a). Over time, strongly increased IRs were observed for ALS patients aged 70 years and above (Fig. 2a–c). Comparing the latest period with the first, the IRR was 2.9 (95% CI 2.2–3.8) for the age group 70–79 years and the IRR was 5.4 (95% CI 2.9–11.2) for the age group ≥ 80 years. No statistically significant differences in IRRs were observed in younger age groups when comparing the latest period with the first. The incidence rose with age at diagnosis, particularly after the age of 45 in both males and females (Fig. 3a), and peaked in the age group 70–79 years (IR; 10.9; 95% CI 10.4–11.4) with a clearly increased IRR (IRR; 2.0; 95% CI 1.8–2.2) comparing the age group of 70–79-year-olds with patients aged ≥ 80 years.

Fig. 1.

a ALS incidence rates per 100,000 from 1980 to 2021 in total and for males and females. ALS amyotrophic lateral sclerosis, IR incidence rate, PY person years. b ALS mortality rates per 100,000 from 1980 to 2021 in total and for males and females. ALS amyotrophic lateral sclerosis, MR mortality rate, PY person years

Fig. 2.

a ALS incidence rate stratified by age and year of diagnosis. b Male ALS incidence rate stratified by age and year of diagnosis. c Female ALS incidence rate stratified by age and year of diagnosis. d ALS mortality rate stratified by age and year of diagnosis. ALS amyotrophic lateral sclerosis, IR incidence rate, MR mortality rate, PY person years

Fig. 3.

a ALS incidence rate per 100,000 during the period from 1980 to 2021 stratified by gender and age. b ALS mortality rate per 100,000 during the period from 1980 to 2021 stratified by gender and age. ALS amyotrophic lateral sclerosis, IR incidence rate, MR mortality rate, PY person years

Mortality

A total of 5,069 adult patients with ALS died during the period from 1980 to 2021. The overall mortality rate was 2.9 per 100,000 PY (95% CI 2.8–3.0). Parallel to the IR, the MR gradually rose with time (Fig. 1b), with the highest MRR in the latest period compared with the first (MR of 2018–2021 versus MR of 1980–1983: MRR, 4.9; 95% CI 3.8–5.3). However, due to the study period starting in 1980, the MR during 1980–1983 included fewer registered ALS deaths compared to the following. Still, the MRR was 3.2 (95% CI 2.8–3.7) comparing the MR of 2018–2021 with the MR of 1984–1988. The MRs gradually increased during the entire study period for both genders (Fig. 1b) and across all age groups, yet the highest MRR was observed for patients aged ≥ 80 years (MRR, 7.0; 95% CI 3.7–14.9) (Fig. 2d). The MRR for males was higher than for females (MRR, 1.2; 95% CI 1.1–1.2). Like the IR, the MR rose with age at diagnosis in both genders (Fig. 3b) and peaked in the age group 70–79 years (MR; 9.8; 95% CI 9.4–10.3). The MRR was 2.0 (95% CI 1.8–2.2) comparing the age group of 70–79-year-olds with patients aged ≥ 80 years.

During the entire study period, ALS patients were followed for a median of 1.4 (interquartile range (IQR): 0.6–3.1) years and the cumulative mortality was 37.4% (95% CI 36.1–38.6%) and 71.7%, 95% (CI 70.6–72.9%) after 1 year and 3 years of follow-up, respectively (Fig. 4a). After the first two years of follow-up, the cumulative mortality incline decreased rapidly (Fig. 4a). Cumulative mortality after 1 and 3 years was highest during the first two decades (Fig. 4b) and was overall higher for females. Finally, cumulative mortality after 1 and 3 years increased strongly with age at diagnosis (Fig. 4c). Thus, the 1-year cumulative mortality was only 5.1% in the group aged 18–39 years but 62.6% in patients aged ≥ 80 years.

Fig. 4.

a Cumulative mortality in total and for men and women. b Cumulative mortality stratified by decennials. c Cumulative mortality stratified by age at diagnosis. ALS amyotrophic lateral sclerosis

Sensitivity analyses

In the subset of riluzole-treated ALS patients diagnosed between 1997 and 2021 (N = 2,765), 55.0% were male and the mean age at the time of diagnosis was 66 years (range, 19–94 years). The overall ALS IR was 2.5 (95% CI 2.5–2.6). Although the study period included only the most recent 25 years, IR clearly rose with time (Fig. S1), with an increased IRR when comparing the latest period with the first (IR of 2018–2021 versus IR of 1997–2000: IRR; 1.7; 95% CI 1.5–2.0). Increasing IRs over time were observed for both genders (Fig. S1) but not across all age groups. Increased IR was only observed in patients ≥ 60 years (Figure S2) when comparing the latest period with the first. The IRR for males was higher than for females being 1.27 (95% CI 1.18–1.37). The IR increased with age and peaked in the age group 70–79 years (Fig. S3).

Discussion

Our study demonstrates a steadily increasing ALS IR during the past 42 years in Denmark. This 2.8 times increase over time applies to both genders and in particular to ALS patients aged ≥ 70 years. Inclinations within the same strata were observed for the ALS MR and for the subgroup of ALS patients diagnosed from 1997 to 2021 and treated with riluzole. The IRs calculated based on this subpopulation did not represent the entire ALS population. However, the diagnostic accuracy of the riluzole-treated sub-population was very high (PPV = 98.7; 95% CI 95.5–99.6) [24] and the indication for prescribing riluzole remained unchanged since approval in 1996 [20]. Thus, the increasing IR in the period 1997–2021 for the riluzole-treated sub-group supports our main finding demonstrating a steady increase in the IR in the entire ALS population.

The overall ALS incidence in Denmark (3.4 per 100,000 PY) was within the range reported from other European countries (2.1–3.8 per 100,000 PY) [9, 14, 25, 26]. A recent (2023) literature review from USA [26] reported conflicting evidence on whether the incidence of ALS was increasing over time, as different diagnostic criteria were used and data being collected over a limited time (up to 4 years) [27–30]. However, we have demonstrated a substantial increase in the ALS incidence also based on the ICD-10 coding which has been used only since 1994 and forward. Recently, more studies have suggested an increase in ALS incidence and mortality rates over time [9]. In our study, the ascending ALS incidence and mortality were in particular evident for ALS patients aged ≥ 70 years. Since aging is the primary risk factor for most neurodegenerative diseases including Alzheimer’s, Parkinson’s disease, and ALS [31, 32], an increasing number of elderly people in the general population over the study period may contribute to the total increased IR and MR. From 1980 to 2021, the Danish population increased across all age groups except for the youngest age group (18–39 years: −0.5% increase), and the greatest increase was observed for individuals aged 70 years and above (70–79 years: 57% increase and ≥ 80 years: 77% increase) [15]. Hence, due to a continuously aging population, the incidence of ALS can be expected to increase considerably in the future [33–36]. In a Danish population-based study the socioeconomic consequences of ALS for patients were evaluated between 1998 and 2009 [37]. Both direct (health sector costs including medication, aids, appliances) and indirect (productivity loss) costs related to ALS were included and the mean annual costs per patient were 23,094 Euro (direct costs encompassed 17,418 Euro). Considering the highly increased number of ALS patients over the past 42 years (1980–1983: N = 270; 2018–2021: N = 827), the costs for ALS care have more than tripled in Denmark. Our estimates and estimates from other Western countries are essential for the planning of future health services for elderly people with ALS, as governments will face rapidly increasing requirements for treatment, rehabilitation, palliation, and support services [26, 38].

The higher ALS incidence among men has been a consistent finding across studies [3, 9]. This difference between genders may be due to occupational exposures that occur more commonly among men [39], a neuroprotective effect of estrogen, or other unidentified sex differences [40]. Similarly to our findings, ALS incidence was previously characterized as increasing with age until the age of 75 years followed by a decline [39]. This decline in incidence may be due to challenges ascertaining cases among the elderly. Thus, more likely older patients may be referred to non-neurological services, and ALS may go undiagnosed due to generalized weakness and wasting being regarded as a normal part of aging, or may be ignored due to multiple medical problems [3]. In our study, high age was proven to be a poor prognosis predictor. Many previous studies have confirmed that older age was associated with a poorer prognosis [41–43], and several explanations have been proposed. One reason may be a lower reserve capacity of motor neurons with advancing age, shown in histopathological studies [44]. Furthermore, rapid progression of regional dysfunction was found in many physiological studies reporting a decline of swallowing and respiratory function as well as increased limb and trunk muscle weakness resulting from aging [45–47]. Moreover, elderly patients with ALS often have comorbidities preventing them from being offered invasive respiratory treatment [48]. The survival of ALS varies considerably, about 10% of ALS patients have a slowly progressive form with a survival of 10 years or longer [25]. Yet, the vast majority of ALS patients have a very limited life span following diagnosis [25]. In our cohort, the cumulative mortality after 3 years was 71.7%, being highest during the first two decades, in females, and strongly increased with age at diagnosis in line with recent studies, reporting longer survival associated with male gender [12] and younger age at onset of diagnosis [49, 50]. In the most recent decades, the survival of ALS patients in our cohort may be extended as a consequence of more commonly used invasive respiratory support [25]. Mean survival time from symptom onset to death has been shown to be longer than five years if ALS patients were treated with tracheostomy [13].

The availability of both long-term incidence and mortality data are the major strengths of our study. The main limitation is the retrospective nature and prospective ALS registries, such as PARALS and the Irish ALS Register are preferable [51, 52]. Still, the tax-financed healthcare system in Denmark largely eliminates referral bias and other selection problems. Moreover, information bias including investigator or recall bias is unlikely, as data was collected for purposes unrelated to our study. Despite the strengths of using national registries, the data are of limited value, if the diagnostic accuracy is low. However, in Denmark the general validity of the DNPR is high [18], and we have found a very high PPV of the ALS discharge code in the DNPR (> 90%) [23, 24]. Still, data on undiagnosed patients with ALS was not assessed and consequently, the negative predictive value, sensitivity, and specificity of the ALS diagnosis were not estimated [53]. In our prior validation study [24], the highest PPV of the ALS diagnosis was observed in patients aged ≥ 70 years (PPV = 93.8; 95% CI 86.2–97.3) and the lowest in patients aged < 60 years (PPV = 83.4; 95% CI 73.3–90.7). Thus, possibly more patients were correctly registered with ALS in the later time periods. Moreover, it is possible that improved ALS awareness over time increased the referral rate from general practitioners and/or general neurology departments to ALS centers, producing increased patient capture in later time periods [24]. On the other hand; due to severe progressive symptoms and clinical signs in ALS, probably only few patients remained undiagnosed, and overall misdiagnoses were infrequent [23, 24]. Yet the minor group of long-term survivors may represent some patients misdiagnosed with ALS. Exact diagnoses of patients misdiagnosed with ALS were presented previously [24]. A related limitation was the switch in 1994 from ICD-8, which had unique ALS codes, to ICD-10, where other motor-neuron diseases were included. To accommodate this change, our case definition included only the following DG122 sub–codes consistent with ALS: DG122C: atrophia musculorum Duchenne–Aran, DG122F: paralysis spinalis progressiva, and DG122G: amyotrophic lateral sclerosis. Nevertheless, the IR and MR increase in 1994 (Fig. 1a, b) may at least partly be due to this change in diagnostic code. Still, limiting our analysis to the period after 1994, a clearly increased IRR (IRR, 1.7; 95% CI 1.5–1.9) remains, comparing the latest period with the period 1994–1998. Moreover, limitations of our study included restricted access to diagnostic characteristics (neurological findings including cognitive tests, cerebrospinal fluid analysis, imaging, genetic data, and neurophysiological examinations). Therefore, we were unable to distinguish -and stratify for ALS subtypes. Finally, the Danish population is homogeneous (> 90% Caucasian), and our findings may thus not apply to other ethnic groups.

Conclusion

The Danish IR and MR of ALS steadily increased over 42 years, with the greatest increase observed for ALS patients aged ≥ 70 years. Important risk increases were related to male gender and particularly to higher age. These findings were supported by similar observations for the subgroup of only riluzole-treated ALS patients. Considering increasingly aging societies, the total number of elderly patients with ALS is expected to increase considerably. Hence, our findings underscore the need for equitable healthcare resource planning for elderly patients to reduce associated morbidity and mortality. The scarcity of established modifiable risks for ALS development demonstrates that new knowledge is required to develop effective prevention and treatment strategies.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This work was supported by Health Research Fund of Central Denmark Region, A3520, Lotte Levison, Institut for Klinisk Medicin, Aarhus Universitet, Aarhus Universitetshospital.

Data availability

Data are pseudonymized centrally at the Danish Health Data Authority. Individual-level data will not be made publicly available in accordance with Danish law.

Declarations

Conflicts of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Ethical approval

The study was approved by The Central Jutland Regional Committee on Health Research Ethics (record number 1–45–70–63–21). Approval from the Ethical Committee is not required to conduct registry-based studies in Denmark.⁵⁴