Abstract
Objective: To determine the incidence, prevalence, and mortality rates of ALS in the province of Modena, Northern Italy, from 1990 through 1999.
Methods: A retrospective epidemiologic study was conducted, ascertaining cases from all neurologic centers and hospitals of the province, death certificates, and the Italian ALS Association, section of Modena. All clinical records were reviewed, and only patients fulfilling the El Escorial revised diagnostic criteria were included.
Results: During the period considered (1990 to 1999), 143 residents (67 men and 76 women) entered the study. The average annual incidence was 2.16 per 100,000, with a peak in the age class of 75 to 79 years. Mean prevalence rate was 4.02 per 100,000, and mean mortality rate was 1.69 per 100,000. The incidence rate remained constant over time, whereas the prevalence and mortality rates increased owing to a rise in survival time (ALS mean duration was 17.38 months in 1990, 43.18 months in 1999). In the mountainous areas, where agricultural work is more common, the incidence, prevalence, and mortality rates were higher than in urban areas and the disease onset occurred 10 years later. Of the risk factors examined, only agricultural work and rural residence were significant.
Conclusions: The incidence, prevalence, and mortality rates agree with those in recent Italian surveys and with most international studies, but the distribution of cases varied with higher rates in mountainous areas. Further prospective studies are required.
Except for some endemic regions in the Western Pacific, the frequency of ALS is relatively uniform worldwide, with mean annual incidence values ranging from 0.31 (Japan)1 to 2.6 (Varmland County, Sweden)2 and prevalence values from 0.8 (Mexico City)3 to 8.5 (Varmland County, Sweden)2 per 100,000 persons.4-6⇓⇓
All morbidity and mortality studies show a male preponderance, suggesting sex as a possible risk factor. Incidence rates are age-specific with a peak between 70 and 79 years,4,6,7⇓⇓ except in one US study.8
An increase in ALS mortality rates has been described in Europe and in the United States,9,10⇓ but it is uncertain whether this trend is real or simply reflects demographic changes or methodologic improvements in case collection.11
Epidemiologic surveys in Italy indicate mean annual incidence rates from 0.44 in Palermo (Sicily)12 to 2.5 in Piemonte and Valle D’Aosta13 and prevalence from 1.56 in Sardinia14 to 5.4 in Reggio Emilia15 per 100,000 population. We performed a retrospective survey to clarify the clinical and epidemiologic features of ALS in the province of Modena (Emilia–Romagna region, Italy) over a period of 10 years (1989 to 1999). A previous survey carried out in the same province16 during the period 1976 to 1986 showed a mean incidence rate of 0.78 and a prevalence of 2.35 per 100,000 population; the highest rates were recorded in the paving-tile industry district, at risk for occupational lead exposure.
The present epidemiologic study focused on ALS clinical characteristics and possible risk factors in our province, such as work, previous injury, and environmental agents. We determined ALS incidence, prevalence, and mortality rates to verify the geographic distribution and the possible variations of ALS frequency over time and with age.
Materials and methods.
Area of investigation.
The province of Modena is situated in northeastern Italy, between 44°50′ and 44°13′ latitude north and 11°40′ and 10°36′ longitude east, and extends over 2,690 km2. From south to north, the province includes three different geographic and economic regions (mountainous, urban, and flat areas), seven public health districts, and 47 municipalities. During the period of the survey, the resident population increased by 4.28% from 600,118 in 1989 to 625,786 in 1999. This increase was due mainly to immigrant influx from other Italian provinces (mainly southern Italy) and from foreign countries (mainly North Africa), with a migration balance of +2,801 calculated at December 31st, 1999. The 1999 birth rate was +8.8 per 1,000, whereas the mortality rate was 10.7 per 1,000, with a negative balance of −1,157 persons. At the last 1991 census, 48.8% of the inhabitants were employed: 47.7% in industry, 6.2% in agriculture, and 46.1% in other occupations.17
Case collection and ascertainment.
We reviewed the clinical records of all hospitals and neurologic centers of our province, identified by a search based on code 335.2 (ALS) of the International Classification of Diseases (9th rev.; ICD). To consider the differential diagnoses of ALS, codes 335.0 to 335.9 (anterior horn cell diseases), 336.0 to 336.9 (myelopathies and syringomyelia), 344.0 to 344.9 (paralytic syndromes), and 356.0 to 356.9 (polyneuropathies) were checked. We examined the archives of the Laboratory of Neurophysiology of the University of Modena. On request, the ALS Italian Association, section of Modena, and the Rehabilitation Medical Center of Veruno (Novara) gave us a list of all their members. We examined the death certificates issued for residents who died from ALS (code 335.2 of the ICD) during the period of the study. Finally, we contacted the private neurologists of the province, general practitioners, and the families of ALS patients not listed in the above sources.
Only patients fulfilling the El Escorial revised diagnostic criteria18 were included in the study.
Statistical analysis.
Percentages were compared by the χ2 test, and only differences significant at 0.05 were accepted.
Results.
During the period from December 31, 1989 through December 31, 1999, 168 patients were diagnosed with definite or probable ALS.18 Only 143 (67 men and 76 women) residents in the province for at least 1 year at the onset of the symptoms entered the study. The male/female ratio was 0.88, not different from that of the general population of Modena (0.95) (χ2 = 0.19, df = 1, p = NS).
Clinical features.
The clinical features of the patients are listed in table 1. Three main clinical forms were considered: bulbar, spinal with onset at upper limbs, and spinal with onset at lower limbs. They were equally represented in both sexes (χ2 = 3.05, df = 2, p = NS). All the cases were sporadic ALS, except for four familial ALS cases. The mean age at onset was 63.07 years (range 26 to 90 years), without increase over time. During the 10 years of the survey, we detected a decreasing and swinging trend: 62.29 years in 1990, 68.64 years in 1994, and 50.5 years in 1999. Nevertheless, if we consider 5-year periods, the mean ages at onset become similar: 63.85 years for the first period (1990 to 1994) and 62.35 years for the second (1995 to 1999).
Table 1 Clinical characteristics of patients
Often the disease began with more than one symptom, including limb weakness (76.93%), limb amyotrophy (10.48%), dysarthria (27.97%), dysphagia (23.08%), dysphonia (16.78%), cramps (15.38%), fasciculations (18.18%), weight loss, weakness, falls, and emotional lability (10.49%).
Risk factors.
Possible risk factors are summarized in table 2. Exposure to possible risk factors was disclosed by analyzing patients’ clinical data. This method may have underestimated risk factors because these clinical records are devised for purposes other than epidemiologic research. We took into account possible risk factors for ALS patients and for the reference population, represented by the Modena population at the last census of 1991, including previous injury or infections, thyroid disorders and other diseases preceding the diagnosis, neurosurgery, textile work, exposure to heavy metals, electromagnetic fields, electrical current, and travel in Asian countries. None of these risk factors was relevant because of their low frequency among our ALS cases. As far as injury is concerned, 26 (18.18%) patients reported fractures or severe traumatic complications not more than 30 years before disease onset. Various somatic areas were involved by trauma, unrelated to the site of disease onset (χ2 = 4.89, df = 8, p = NS). Physical activity, agricultural work, and residence in the mountains were the only relevant risk factors. Among our patients with ALS, 54 (37.76%) performed hard physical work and 31 (21.68%) were farmers. This datum is underestimated as in 24.48% of patients, the occupation was unknown. Despite this, at the last census (1991), in the province of Modena, only 6.26% of the working population (employed or unemployed persons during the week before the census) were employed in farming (χ2 = 57.91, df = 1, p < 0.00). Finally, 26 patients (18.18%) resided in the mountainous areas, whereas in our province, the percentage of residents in mountains was 9.7 to 9.9% of the total provincial population (χ2 = 9.58, df = 1, p < 0.02).
Table 2 Exposure to possible risk factors
Incidence.
Incident cases were considered all the new ALS cases diagnosed every year. In the decade of the survey, 132 new cases (63 men and 69 women) were identified in our province, with a mean incidence rate of 2.16 per 100,000 population (2.05 men and 2.27 women). The male/female rate ratio was 0.91, probably due to a predominance of women (male/female ratio of 0.75 in 1999) in the population aged over 65 years in our province. The incidence rates, age and gender adjusted to the 1995 Italian population (direct method), were 1.99 per 100,000 for men and 1.85 per 100,000 for women, without differences (χ2 = 0.16, df = 1, p < NS). Moreover, ALS incidence in men and women did not differ in any of the age classes, nor did the difference between men and women. Mean incidence rate in the period 1990 to 1994 was 2.25, whereas in 1995 to 1999, it was 2.08, without increase over time (figure 1). No cases were found before 24 years, and the incidence increased with age, with a peak of 11.23 per 100,000 between 75 and 79 years (figure 2). In the oldest age group, the disease occurred more rarely. In mountainous areas, the mean incidence rate was 3.67, whereas it was 2.14 in the urban and 1.27 in the flat areas (figure 3). This different geographic distribution is found even considering only the population older than 44 years. This procedure is important because the old age index (percent ratio between the population aged over 65 years and the population under 14 years) is 200 to 300% in the mountains and lower (<100%) in the urban and flat areas. In the mountainous area, the peak ALS frequency was at 60 to 64 years, whereas in the urban and flat areas, it was 15 years later (75 to 79 years old) (figure 4).
Figure 1. ALS incidence (diamonds), prevalence (squares), and mortality (triangles) rates per 100,000 per year in Modena in the period 1989 to 1999.
Figure 2. ALS incidence (diamonds), prevalence (squares), and mortality (triangles) rates per 100,000 per age class in Modena in the period 1989 to 1999.
Figure 3. ALS incidence per year and geographic areas in Modena in the period 1989 to 1999. Open columns = flat area; gray columns = urban area; black columns = mountainous area.
Figure 4. ALS incidence and mortality rates per 100,000 per age class and geographic area in Modena in the period 1989 to 1999. Diamonds = incidence rates in flat area; squares = incidence rates in urban area; filled triangles = incidence rates in mountainous area; open triangles = mortality rates in flat area; circles = mortality rates in urban area; squares = mortality rates in mountainous area.
Prevalence.
Prevalent cases were considered the patients with ALS at December 31st of every year, from 1989 through 1999. Twenty-six of 143 patients with ALS died in the year of diagnosis, so that prevalent cases remained 117 (55 men and 62 women). The mean prevalence rate was 4.02 per 100,000 population (4.10 men and 3.94 women; sex ratio of 1.02), with an increase from 2.17 at December 31, 1989 (13 cases) to 4.63 per 100,000 population at December 31, 1999 (29 cases) (see figure 1). The highest prevalence rate (15.66) was observed from 75 to 79 years (see figure 2). In the mountainous area, the mean prevalence rate was 6.21, with a peak at 65 to 69 years, whereas in urban and flat areas, it was 3.86 and 3.35, with a peak between 75 and 79 years.
Mortality.
Mortality was calculated on the basis of the deaths caused by ALS for each year. On December 31, 1999, 29 of 143 ALS patients were still alive, and 2 had emigrated. ALS was the cause of death in 103 of 112 cases (43 men and 60 women), whereas 9 patients died from other causes (neoplasm, AIDS, suicide, heart failure, stroke, and pancreatitis). The mean mortality rate was 1.69 per 100,000 inhabitants (1.45 men and 1.92 women), with a peak of 2.59 in 1998. Moreover, the mean mortality rate showed a trend toward a longitudinal increase over time, varying from 1.33 in 1990 to 1991 to 2.17 in 1998 to 1999 (63.95%) (see figure 1).
The mean age at death was 68.29 years (67.67 men and 68.73 women), and the mean duration of illness was 28.74 months, without differences between sexes. The disease duration progressively increased over time, from 17.38 months in patients who died in 1990 to 43.18 months in patients who died in 1999. Mortality rates increased with patients’ age, and average mortality rate was maximum at 75 to 79 years, with a peak of 9.42 (10.05 men and 9.00 women) (see figure 2). During the period 1990 to 1999, the mean mortality rate increased more in men than in women (50.83 and 33.57%).
The geographic distribution of mortality rates is uneven in the three regions of the province, with the highest rates (3.50/100,000) in the mountainous area with respect to the flat (1.15) and urban (1.55) areas, even considering only people older than 44 years. In the mountainous area, the highest mortality rate was at 60 to 64 years, whereas in other areas, the peak was in the age class 75 to 79 years (see figure 4).
Discussion.
A male preponderance in ALS, with sex ratios ranging from 1.2 to 2.6, is reported in most literature data, suggesting a possible unknown risk factor in several countries.6 Like other investigators,19-21⇓⇓ we found a lower male/female ratio (0.88) due to the structure of our provincial population, characterized by a prevalence of elderly women. In fact, age- and sex-adjusted incidence rates show values of 1.99 per 100,000 for men and 1.85 per 100,000 for women with a male/female ratio of 1.07 (reference population: Italian population in 1995). The clinical picture of ALS in our cases agrees with that in the literature reports, but we found a higher incidence of bulbar ALS, possibly because of the female predominance in our cases.22,23⇓ In 35.66% of our patients, the onset symptoms of ALS were bulbar, particularly in women (women, 60.78%; men, 39.22%) and in the older age group, according to previous data.24,25⇓ The most common clinical subtype of ALS was spinal ALS (60.84%)26-28⇓⇓; at variance with other reports,23,24⇓ onset in the lower limbs (36.36%) was more frequent than in the upper limbs (24.48%).
The average time lapse from first symptoms to diagnosis was 9.02 months, explained by the insidious onset of the disease, the lack of pain, and the need to exclude all ALS-mimicking diseases.29 In previous surveys, time to diagnosis varied from 11 to 19.97 months,26,28,30-33⇓⇓⇓⇓⇓ suggesting different levels of health education in the population and medical services efficiency. Age at onset influences the time to diagnosis, as indicated by the shortest time (6.3 months) recorded in patients older than 75 years, probably owing to the more rapid progression of disease. For ALS with onset at older than 60 years, time to diagnosis was 8.7 months, perhaps due to the impact of muscle weakness in still-working patients, whereas it was longer (10.2 months) in patients aged 60 to 75 years, in whom the symptoms could be attributed to old age.
Age at onset ranged from 25 to 90 years (average 63.07 years), without significant differences between the sexes, and was similar in the three clinical forms, according to previous data.34,35⇓ At variance with other studies36,37⇓ reporting an increase in age at onset from 1971, in the decade of our survey, the mean age at onset showed a swinging trend, reaching the highest value in 1994 (68.64 years) and the lowest in 1999 (50.5 years). This apparent decrease could be ascribed to the low number of cases with ALS onset in 1999 (many patients with the first ALS symptoms in 1999 would be diagnosed in the following years). On grouping the data into two 5-year periods, the mean age at onset was 63.85 years in 1990 to 1994 and 62.35 years in 1995 to 1999, without wide variations over time.
Accurate identification of risk factors in a rare disease such as ALS is very difficult, both for retrospective and for prospective epidemiologic surveys. Case-control prospective studies are often influenced by defects in case collection, as only selected cases, not representative of the general population, are included.38-41⇓⇓⇓ Conversely, retrospective studies are based only on clinical records, which are not targeted to epidemiologic purposes, leading to incomplete information.
Trauma has been implicated in the pathogenesis of motor neuron disease,40,42,43⇓⇓ with a topographic relation between the site of injury and clinical onset. In our survey, traumatic events preceded ALS onset (up to 30 years before) in 18.18% of cases, thereby not supporting a relation between the site of trauma and the site of disease onset.44 If trauma and motor neuron disease are related, trauma may be caused by the disease, as the weakness increases the risk of falls and injury. In a disease with unknown pathogenesis, like ALS, patients pay particular attention to even slight trauma, overestimating its importance.
Only 5.59% of our patients with ALS underwent neurosurgery, suggesting a lack of evidence for its role in the development of the disease.40,43⇓
Physically strenuous jobs,39,45⇓ like textile work46 and exposure to heavy metals,47 electromagnetic fields,48 electrical current,38,42,49,50⇓⇓⇓ and chemicals,40 are cited as possible risk factors for ALS. At least 37.76% of our patients performed heavy physical work. According to previous studies,39,40,51-53⇓⇓⇓⇓ our major finding is the high percentage (21.68%) of patients employed in agricultural work. At the last census of 1991, only 6.26% of the working population in the province of Modena was represented by farmers.19 The high frequency of ALS among agricultural workers implicates an environmental agent such as heavy work or toxic substances like pesticides. Agricultural work is done mainly by the rural population, which is more affected by ALS than urban residents.53 Accordingly, patients residing in mountainous areas comprised 18.18% of total cases, whereas only 9.7 to 9.9% of the population of Modena resided in these areas.54
A greater association between ALS and internal medicine diseases is reported in the literature than in our data: 17% for thyroid diseases55 and 5% for gammopathies.56 Nevertheless, a case-control study on ALS and thyroid disorders failed to demonstrate any association.57
A previous survey performed in the province of Modena during the years 1976 to 198616 showed a mean annual incidence rate of 0.78 and a mean prevalence rate of 2.35 per 100,000 population. The current survey disclosed higher incidence (2.16/100,000) and prevalence data (4.02/100,000), probably reflecting more careful case ascertainment, as a real increase in disease frequency seems unlikely.
Our results agree with those from previous studies in Italy (incidence: 0.4412 to 2.513; prevalence: 1.5614 to 5.415), Europe (Norway, Sweden, Finland, Ireland), and North America (Canada, United States),4 suggesting a relatively uniform ALS distribution worldwide.
In the last 20 years, ALS incidence seemed to increase in the province of Reggio Emilia, bordering on the province of Modena,15 in agreement with other data.4,36,58⇓⇓ As ALS is an age-related disease, the increase may be partly ascribed to aging of the population or, alternatively, improved diagnoses, increased attention to the elderly, or accuracy in case collection and treatment of data.1,5,32,37,59,60⇓⇓⇓⇓⇓ In our survey, the incidence of ALS did not change considerably over time, whereas the prevalence more than doubled, probably because of the increase in patients’ survival. Incidence values were very low before 45 years, progressively increased after 45 years, reached a peak at 75 to 79 years, and declined thereafter.1,7,32,60⇓⇓⇓ The only discordant data were the oldest age classes8,61⇓ (>80 years), in which statistical estimates are based on small case series. Furthermore, the incidence decrease in older age classes might be due to the faster course of disease in the elderly and to misattribution of ALS complaints to old age.
In our survey, ALS incidence and prevalence rates per 100,000 population were higher in mountainous areas with respect to flat and urban ones.32,39,53⇓⇓ We do not ascribe this different geographic distribution to the prevalence of older people in mountainous areas, because this result persists even considering only people older than 45 years. Moreover, it seems unlikely that the distribution of ALS is due to different diagnostic accuracy, as the mean incidence rate was minimal in the districts where the only two neurologic centers of the province are situated. In advanced ALS, symptoms always require medical assistance and compel patients to look for care, facilitating diagnosis. The higher rates in the mountainous areas may be explained by the high frequency of agricultural work and greater exposure to possible environmental risk factors. We cannot exclude a possible role of genetic aggregation, even if only two of four familial ALS in our survey resided in the mountains.
Among our patients, ALS incidence reached a peak at 60 to 64 years in the mountains, whereas it was reached 15 years later (75 to 79 years) in other areas of the province. Young persons may also be exposed to risk factors connected to farming in mountain areas, whereas in the flat and urban areas, only older people had been employed in agriculture that was abandoned long ago. The advanced onset of the disease in the mountains might also suggest an anticipation effect due to genetic aggregation: Familial ALS cases are reported to have earlier onset and longer survival.25,62⇓
Despite our method of ascertaining patient deaths, about 25% of ALS patients were lost because the reported cause of death was a terminal complication (e.g., bronchopneumonia).63
Our mean mortality rate per 100,000 population was 1.69, one of the highest values reported in Italy14 and comparable with those of Scotland, Canada,64 the United States,65 England and Wales,66 and France.58 These mortality results resemble the incidence results regarding the distribution throughout the province as well as the relation with age and sex. In particular, we found a peak mortality between 60 and 64 years in mountainous areas and between 75 and 79 years in flat and urban areas. Even the male/female mortality rate (0.88) is close to the male/female incidence rate (0.91).62
From 1990 to 1999, we found an increase in annual ALS mortality in agreement with recent reports67 and at variance with past surveys reporting stable mortality rates worldwide .22 The increased ALS mortality in Italy and in other countries may be explained by the increase in incidence, spurious effects of methodologic variables, aging of the population, and decreasing mortality from competing diseases. The increase in ALS mortality was not related to an increase in incidence, as incidence rates remained stable over time. The increase in ALS mortality and prevalence, in spite of the steadiness of incidence, may be paradoxically explained by increased mean survival of patients: 1.4 years for patients who died in 1990 and 3.6 years for those who died in 1999. The longer course of disease may determine the delay of patients’ death in the last years of the survey. The increase in survival could be due to improved care and to lower mortality from other diseases, better controlled to date,11 like cerebrovascular diseases.68
Considering the steadiness of incidence, changes in methodologic procedures could explain the increased mortality. Improvements in diagnosis due to better health services, more resources for ALS diagnostics in the elderly, and changes in ICD coding and in the reliability of the death certificates may be responsible for an increased number of detected deaths for ALS.
Our results indicate that the increase in ALS mortality is related to age. During the decade of the survey, the mortality rates remained unchanged in patients younger than 50 years, whereas the rates increased from 3.67 (1990 to 1991) to 6.85 (1998 to 1999) in patients aged 54 years or older.
This increase has been related to an increase in the population at risk for the disease.9,68⇓ In fact, more people reach older ages because of the improvement in the underlying health of the general population in the last decades. As a consequence, the susceptible population is less liable to early death from other diseases, with a higher mean life expectancy and therefore a greater mortality from ALS, particularly at an advanced age. A similar increase in mortality has been described worldwide67 for other neurodegenerative diseases like PD. The deterministic and competitive model of mortality of the Gompertzian analysis could explain the increase in mortality from neurodegenerative diseases and therefore also from ALS.11,65,67-69⇓⇓⇓⇓
In conclusion, ALS mortality, incidence, and prevalence in the province of Modena are similar to those reported in Northern Europe and North America.
Acknowledgments
Supported by a fellowship to J.M. by the Italian Association of ALS, section of Modena.
Acknowledgment
The authors thank Anne Collins for editing the manuscript.
- Received October 29, 2001.
- Accepted October 1, 2002.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Topics Discussed
Alert Me
Recommended articles
Incidence, Prevalence, and Geographical Clustering of Motor Neuron Disease in the Netherlands
Temporal and geographic variation in United States motor neuron disease mortality, 1969–1998
Increasing mortality from amyotrophic lateral sclerosis in Norway?
Increasing stroke incidence and prevalence of risk factors in a low-income Chinese population