This study utilized a robust methodology to determine the prevalence of essential tremor in an urban black African community in sub-Saharan Africa, where data are sparse regarding the epidemiology of movement disorders including ET. The sample size was modest, and was dictated by financial and logistic constraints, as well as a commitment to allow complete case ascertainment using a manageable population size. The study instrument for screening for ET had several advantages, in that the sensitivity (100%), specificity (95.2%) and number needed to diagnose ET (1.05) significantly improved the likelihood of complete ascertainment of cases of ET in the participants. Furthermore, face-to-face neurological diagnosis remains the clinical gold standard for the diagnosis of ET, and was employed in this study to ensure diagnostic certainty.
The crude and age-standardized prevalence data obtained here fall within the range previously reported from other populations (4.0 – 39.2/1000), but are higher than the values from earlier studies based in a rural community in the same ethno-geographical zone, but published three decades ago by Osuntokun et al.
]. The earlier study reported a crude prevalence rate (0.1/1000) ten times lower than the present study. Although there is a paucity of comparative data from other African countries, the studies by Haimanot et al.
] (1990) and Dotchin et al.
] (2008) reported lower prevalence rates of 5/100,000 and 82/100,000 respectively from communities in Ethiopia and Tanzania (both in east Africa) respectively. In contrast to these African studies, our prevalence rates are lower than that reported from a recent study in Sile, Turkey where the crude prevalence of ET was documented as 30/1000 in persons above 18 years of age [23
]. It is also lower than the prevalence rate documented from a community based study in Manhattan New York [24
However, direct comparison of these data presupposes that the population structures of these countries are similar, and the lack of age-adjustment in the earlier West African study [25
] and that of Haimanot [16
] limit the conclusions that can be derived regarding differences in prevalence. Though the study by Dotchin et al.
] employed age – adjustment this was done to United Kingdom population of 2001 which may not reflect WHO new world population.
World-wide, the prevalence of ET varies from not being present at all [25
] to being common [26
]. Differences in prevalence estimates have been attributed to differences in study design, study population and methodology and diagnostic criteria. Many of the initial studies on ET prevalence relied on screening instruments only though in the last two decades most of the population studies have incorporated clinical examination.
Other reasons for varied worldwide prevalence could be differences in exposure to risk and aetiological factors. Aging is the most consistent documented risk factor associated with increasing prevalence of ET [27
] a finding also corroborated by this study. Genetics have also been implicated in ET as demonstrated by clustering of the condition within families [28
]. About a third of the participants with ET in this study had a positive family history of tremor in a first degree relative. It is conceivable that undiagnosed ET and recall bias would have lowered the reporting of a positive family history. Environmental factors (such as harmane) have been linked to ET, although the etiologic association is confounded by the possibility of a genetic risk modifier [29
A gender predilection has previously been reported in some ET studies, while others such as ours have not found any male to female differences in the prevalence of ET [4
]. The basis of a gender predisposition is unproven and there are no strong biological indications that could explain such findings. However, environmental peculiarities relating to occupational and recreational exposures differ by gender, and are possible contributors to gender differences.
The clinical phenotype of ET in this study is consistent with descriptions from other population-based studies. The predominant tremor location in our cohort is similar to previously documented reports with upper limb tremor being the most common feature. However, compared to other population based studies, none of our cohort had a head tremor. Dotchin et al.
] in Tanzania found head tremor in 54% of their cohort, while Louis et al.
] reported head tremor in 27% of their patients. There are some possible explanations for this discrepancy. Most importantly, the initial screening questions did not specifically ask for the presence of head tremor, but rather hand tremor, and so it is entirely feasible that this inadvertently screened out persons with isolated head tremor as the ET phenotype. However the second screening questionnaire specifically screened for head tremor in addition to other tremor locations. This was in addition to using a face to face examination to ascertain the nature of the tremor. In this case ascertainment stage, none of the 36 cases had combined head and hand tremors, all had hand tremors only. ET can manifest as isolated hand tremors, isolated head tremors, or a combination of hand tremors with head tremors. The number of ET cases encountered was small (36), and mainly of a mild nature (in about three quarters of the cases) and this could also partly account for the absence of the concomitant hand and head tremors. In general, in our clinical practice, we more commonly encounter isolated hand tremors, combined hand and head tremors, and less commonly isolated head tremors, and so our finding may simply be a reflection of the relative frequencies of these phenotypes. The higher proportion of mild cases in this study may indicate the greater likelihood of finding milder, less disabling ET in the community (in contrast to a clinic-based study).
We could not assess drug responsiveness in this study because of the low treatment rate (1/36 only on treatment) and a concomitant large treatment gap in our cohort with ET. As many as 75 - 99% of patients with ET detected through population-based studies are reported to be previously undiagnosed and untreated [4
]. Our findings corroborate the previous documentation of a large treatment gap in ET in the developing world alluded to in a previous hospital-based study in a tertiary hospital in Nigeria which highlighted the low hospital frequency of ET over a 25-year period (only 10 ET cases out of 2.1million patients) and proposed a probable high treatment gap considering the presumption that ET is a relatively common disorder [31
]. Several factors including differing severity and impact of ET on activities of daily living, non-recognition or poor awareness of ET as a neurological or medical disorder amenable to treatment, and limited access to care may contribute to this gap.
We found that a large proportion of ET cases reported that their tremors began on one side but however did not find any remarkable gross asymmetry in the tremor severity in any of the cases on clinical examination, We suggest that, as often encountered in the clinic setting, tremors may be more noticeable to patients in the dominant hand despite being present in both hands, because of the associated interference of tremors with carrying out actions, and thus attribute this finding to misperception of the nature of onset. Furthermore, due to the long standing nature of ET tremors, historical accounts of site of onset may not be accurate. Also, although ET is predominantly regarded as a symmetrical disease, there is clear evidence that it also presents asymmetrically, and indeed that mild asymmetry is a fundamental property of ET [32
]. Putzke et al.
, in a long-term follow-up study, reported that about 55% of ET cases in their series presented with asymmetrical disease, and that the occurrence of asymmetric ET may predict disease progression [33
Our study has certain limitations relating to the size of the population screened and invariably, the population structure. The population screened (3000) is lower than that from previous studies, but we anticipate that the rigorous methodology employed ensured that our data are credible and truly representative of the current prevalence of ET in our community. We would have ideally screened a larger population, but were limited by the attendant cost implications. The proportions of persons in the age groups is a reflection of the Nigerian population structure which, according to current official statistics (2011 estimates), is comprised as follows (combined rural and urban figures): <15 years (40.9%), 15 – 64 years (55.9%), 65 years and above (3.1%), and median age (19.2 years) [34
]. Our study was conducted in an urban centre and the smaller percentage of urban-dwelling elderly (>65 years – 2.2%) is expected. We however provide age adjustment to the WHO New World population and complete data for adjustment to any population structure to enable comparison of our findings with those of other researchers. Finally, we note the possibility of persons with very mild tremors having screened negative in the initial stage if the tremors were not obvious to the participants themselves.