Prevalence of Parkinson Disease Among the Navajo; a Preliminary Examination

Abstract

BACKGROUND

The prevalence of Parkinson disease (PD) varies by geographic location and ethnicity, but has never been studied among the Navajo.

METHODS

Period prevalence was calculated using the number of people diagnosed with PD in the Shiprock Service Unit Indian Health Service database during 1995–1999, 2000–2004, and 2005–2009 as the numerator, and the number seen for any reason as the denominator. Age-standardized rates were calculated using the 2000 US population.

RESULTS

During 2005–2009, 126 people were seen with PD (crude prevalence = 203.7/100,000 population). The age-adjusted rate was 335.9 (95% C. I. 277.8–394.0) overall, 438.5 (95% C.I. 336.5–540.5) in men and 259.7 (95% C.I. 192.8–326.7; p=0.004) in women. The adjusted rate increased with age: 788.8 (95% C.I. 652.0–925.7) for age 40 and above to 1964.9 (95% C.I. 1613.7–2316.1) for age 60 and above. Adjusted rates were 246.6 (95% C.I. 187.2–306.0) in 1995–1999 and 284.7 (95% C.I. 227.0–342.4) in 2000–2004.

CONCLUSION

Parkinson disease appears common among the Navajo. Estimates increased with age and time, and were higher in men. In-person interviews are needed to confirm these estimates, and to determine incidence, quality of care, and risk factors for PD among the Navajo.

INTRODUCTION

The Navajo Nation, with about 180,000 inhabitants, is the largest reservation for American Indians in the U.S. [1]. Health care is provided by the Indian Health Service (IHS), a branch of the Federal Government, but socio-economic and health disparities are widespread [2]. Quantifying the rates of neurological illness among the Navajo might improve allocation of resources and quality of care.

Parkinson disease (PD) is a treatable, but incurable disorder, whose prevalence varies by location and ethnicity [3–5]. The etiology of PD is not well understood. The goal of this study was to obtain preliminary estimates of the burden of PD among the Navajo using existing data.

MATERIALS AND METHODS

On the Navajo Nation, health care is divided into eight administrative service units. The Shiprock Service Unit is the largest with more than 20 communities and a population of approximately 50,000 people [1]. The IHS maintains a database containing information from every clinical encounter in the Service Unit. Patient records were accessed from April 1, 1995 to December 31, 2009.

We calculated period prevalence per 100,000 population by age group and gender for three time periods, 1995–1999, 2000–2004, and 2005–2009. The population was taken as the total number of people with a record within the system during the periods; the number whose primary or secondary diagnosis was coded as Parkinson disease (ICD-9-CM code 332.0) [6] was used as the numerator. The age-gender distribution of the denominator population closely resembled the 2010 U.S. population census for the Shiprock area (efigure). Because the median age of the Navajo Nation is younger than the general US population (24 vs. 35.3 yrs; 2000 census), we directly age-standardized our estimates to the distribution of ages in the 2000 U.S. population and calculated 95% confidence intervals (CI) [7,8]. We compared the age-adjusted rates between genders using a Z test for two independent proportions with appropriate variance formula for weighted proportions. We estimated crude and age-adjusted rates for the three time periods. The University of New Mexico and the Navajo Nation gave IRB approval.

RESULTS

Among 61,853 people who appeared for any type of visit during 2005–2009, 126 carried the diagnosis of PD (crude period prevalence = 203.7/100,000 population). The frequency was 103.1/100,000 in people aged 50–59 years and 4391.0/100,000 in those aged ≥80 years (Table 1). The crude rate was higher in men (231.1/100,000) than women (178.1/100,000) for all ages.

Table 1.

Crude and age-adjusted prevalence rates of Parkinson disease among Navajo Indians, Shiprock Service Unit, by age group and gender, 2005–2009, United States^*.

Age group	Patients	Population	Crude Prevalence Rate	Age-adjusted prevalence (95% CI)
Men
All ages	69	29854	231.1	438.5 (336.5–540.5)
≥40 years	69	9146	754.4	1036.6 (795.4–1277.9)
≥60 years	65	2894	2246.0	2603.7 (1982.0–3225.4)
Women
All ages	57	31999	178.1	259.7 (192.8, 326.7)
≥40 years	56	10798	518.6	605.4 (448.1–762.8)
≥60 years	52	3722	1397.1	1495.6 (1093.4–1897.8)

^*Prevalence rate per 100,000 of corresponding group; CI indicates confidence interval

The age-adjusted rate during this period was 335.9 (95% CI = 277.8–394.0); 788.8 (95% CI = 652.0–925.7) for ages ≥40 years and 1964.9 (95% CI = 1613.7–2316.1) for ages ≥60 years (Table 2). Age adjusted rates were higher in men (438.5; 336.5–540.5) than women (259.7;192.8–326.7; p-value = 0.004; Table 1).

Table 2.

Prevalence rates by time period

Time Period	Patients	Population	Crude prevalence*	Age-adjusted prevalence (95% CI)*
1995–1999
All ages	66	54584	120.9	246.6 (187.2–306.0)
≥ 40 years	66	13944	473.3	582.4 (442.1–722.7)
≥60 years	62	4511	1374.4	1443.9 (1085.4–1802.5)
2000–2004
All ages	94	59229	158.7	284.7 (227.0–342.4)
≥ 40 years	90	17450	515.8	654.0 (518.9–789.0)
≥60 years	85	5546	1532.6	1637.6 (1290.4–2316.1)
2005–2009
All ages	126	61853	203.7	335.9 (277.8–394.0)
≥ 40 years	125	19944	626.8	788.8 (652.0–925.7)
≥60 years	117	6616	1768.4	1964.9 (1613.7–2316.1)

^*prevalence per 100,000 population

Navajo Indian, Shiprock Service Unit; CI indicates confidence interval

Differences by gender persisted when we analyzed the three time periods 1995–1999, 2000–2004, and 2005–2009 separately (Figure). Age-adjusted prevalence estimates increased over time; 246.6 in the first period to 335.9 in the third period (Table 2).

Figure 1.

Crude prevalence by time period

Age-adjusted prevalence of Parkinson Disease by gender group and time period

DISCUSSION

Parkinson disease appears to be common among the Navajo. The adjusted rate of 335.9/100,000 over the most recent five-year period is at least as high as many other studies [3,4,9–11]. These findings are consistent with a recent examination of Parkinson disease among American Indians using nationwide data from the Indian Health Service. In that study, rates were highest among those living in the Southwest, where the Navajo along with other American Indian tribes reside [12]. Our estimates increased with age, especially after age 50. The rates were higher among men than women, and the age-adjusted prevalence appeared to increase with time.

Genetic or environmental risks could contribute to PD among the Navajo. Exposure to metals and pesticides has been implicated in development of Parkinson disease [5]. Many people in this community have been exposed to metals through work in uranium mines and drinking contaminated water. Sheep herding used to be common on the reservation and sheep were bathed in pesticide solutions containing organophosphates. This process, known as sheep-dipping, has been linked to high human exposures and neurological injury in other populations [13]. The high male-to-female prevalence ratio we report could reflect these environmental exposures but also may indicate bias in diagnosis, particularly underdiagnosis in elderly Navajo women. The increasing prevalence rates over time may demonstrate improved case ascertainment or a true increase in prevalence.

Parkinson disease is acknowledged to have important genetic influences that vary according to ethnic background [14]; Asian populations, from whom today’s American Indians are thought to have originated, may carry different susceptibility factors than Caucasian populations [15]. Heredity likely plays a meaningful role in other diseases with high prevalence in American Indians [16], related to lower genetic diversity from centuries of living in relative isolation [17]; analysis of genetic risks in homogeneous populations could produce a meaningful contribution to the understanding and eventually treatment of PD.

How do the rates we report compare to other populations? A study that calculated 4-year period prevalence in a diverse population reported overall age-adjusted prevalence rates of 130/100,000 among Hispanics, 116/100,000 among Caucasians, and 57/100,000 among African Americans [3]; period prevalence was higher in men than women among Caucasians and Hispanics, but not African Americans. Our estimate of PD in the Navajo is higher than the rates in that study, but it was conducted nearly 20 years ago. Other reports have also shown varying PD prevalence by race and gender [18], and PD has been documented in Native people across the world [4]. Twelve high-quality studies from the U.S. and Europe yielded point prevalence rates that ranged from 950/100,000 to 4380/100,000 in largely Caucasian populations aged 65 years and above [9]. Our estimates of PD for men and women in this age group are in the middle of that range, but we could not calculate point prevalence in our study so the data are not strictly comparable.

Our analyses included all patients receiving free and universal health care in an administrative unit of the IHS. The IHS database is a better measure of population health than most clinic-based records system. The database captures the vast majority of people who are eligible to use the health system because the poverty and geographic isolation of the reservation means that those living there seek their medical care at reservation facilities. The age-gender distribution of those included in the database closely resembles the estimates of the age-gender distribution of the Shiprock population based on census [1].

This preliminary examination is the first study of PD in Navajo Indians. Despite the novelty of this report, it has important limitations. The data are from a single region and may not reflect the rates among the Navajo population residing elsewhere. They are based on records from a clinical database; mild cases and cases in the elderly may have been missed, underestimating the true rates [5]. Some of the cases likely represent related disorders, not PD itself; diagnostic and coding errors may have inflated our estimates. In a chart review of 175 charts by a movement disorder specialist (PG), we found accurate documentation of PD according to research criteria [19] in 138 charts (79%). We focused our analyses on the most recent time period when documentation appeared most accurate. However, the complete medical record was not available in many cases. These gaps in documentation prevented us from being able to estimate the rate of misdiagnosis. Finally, we could not determine whether a person had died by a specific date so we could not calculate point prevalence. This meant that it was difficult to directly compare our data to other studies that present point prevalence estimates of PD.

The prevalence among the Navajo in this preliminary analysis warrants further investigation. A full epidemiologic study that includes standardized patient interviews, diagnostic verification of PD and its variants, better documentation of deaths, and quantification of the intensity and time course of the exposures is needed. Environmental or genetic risk assessment in this unique population could lead to a better understanding of PD and its causes. Examination of quality of care could determine whether social or financial factors influence the diagnosis or care of Navajo people with PD.