The incidence, prevalence and treatment of narcolepsy in New Zealand

Narcolepsy is a chronic sleep disorder characterised by the irrepressible need to sleep during daytime hours (excessive daytime sleepiness [EDS]).¹ It is clinically divided into two forms: type 1 and type 2. Type 1, affecting up to 60% of patients, is further characterised by cataplexy—presenting as sudden, brief episodes of muscle weakness or loss of muscle tone occurring during wakefulness.² All patients may otherwise experience hallucinations on falling asleep/waking, disrupted nighttime sleep and/or episodes of sleep paralysis.² Global prevalence estimates suggest that narcolepsy affects 0.025–0.05% of the world’s population.³ The condition typically onsets between 10 and 20 years of age, with an average diagnostic delay of 8–12 years.³

The gold-standard diagnostic test for narcolepsy remains the multiple sleep latency test (MSLT).⁴ This measures how fast patients fall asleep under controlled conditions as well as the speed at which they enter REM sleep—a phase of sleep normally observed at the end of a 60-to-90-minute sleep cycle.⁴ Primarily, a diagnosis of narcolepsy requires EDS to be present every day for at least 3 months and an MSLT result of ≤8 minutes with ≥2 sleep-onset entries into the REM sleep phase (SOREMPs).² Because other sleep disorders can produce EDS, patients also require baseline polysomnography to rule out other conditions. This is an overnight sleep test that measures brain activity, blood oxygen levels, heart rate, breathing and movements of the eyes and legs.⁴ Idiopathic hypersomnolence (IH), obstructive sleep apnoea (OSA) and chronic sleep restriction are the key differentials to exclude before diagnosing narcolepsy.⁴

A combination of symptom-managing medications and lifestyle changes can help improve patient wellbeing.⁵ However, even with treatment, many feel disabled by their condition.⁶ EDS and the other symptoms of narcolepsy can have significant impacts on the psycho-social-behavioural outcomes of patients. Patients often face barriers in achieving their social, educational, career and financial potential, especially if delays in accessing diagnosis and effective treatment occur.⁷ Furthermore, due to the presence of involuntary sleep episodes, patients are at high risk for car accidents and other injuries.⁸ Considering the young age of onset, it is clear that prompt recognition, diagnosis and treatment are important for improving patient wellbeing in the long term.

In this regard, the last 5 years have shown significant advancements in the medications available for symptom control.⁵ As research in this area accelerates, it is important to understand where each country stands in diagnosing and treating this condition. While several international studies have explored the prevalence, incidence and treatment of narcolepsy, minimal research has been conducted in New Zealand. Therefore, an up-to-date perspective is highly necessary. To achieve this, national data regarding testing and diagnosis between 2021 and 2023 were collated at the WellSleep centre in Wellington from all New Zealand centres offering formal testing for narcolepsy. This data was then compared to the prescription funding requests—special authorities (SAs)—processed by Pharmac (the national regulatory authority for medical funding) in the same period. Demographic trends in these SA applications were further described according to region, age, gender and ethnicity. This was done to help identify at-risk populations and to ensure that healthcare resources are being allocated appropriately based on specific community needs. This is particularly important for Māori, who often face inequities in healthcare access and outcomes, making it essential to understand how narcolepsy may uniquely affect them.

Method

Testing and diagnostic outcomes data for 2021–2023

All centres that conduct testing for narcolepsy (the NZ Respiratory & Sleep Institute [NZRSI], Auckland; WellSleep, Wellington; and the hospitals of Auckland City, Waikato, Christchurch and Dunedin) were contacted and asked to answer the following:

1.      How many MSLTs did you perform in 2021, 2022 and 2023?

2.      Based on the national MSLT reporting criteria, how many tests were positive for narcolepsy versus idiopathic hypersomnolence (IH) versus OSA/other diagnoses/non-diagnostic outcomes?

These data were used to estimate the incidence and prevalence of narcolepsy. These values were calculated as follows:

1.      Incidence = (number of cases ÷ [total New Zealand population × study period in years]) × 100,000.

2.      Prevalence = ([incidence ÷ 100,000] × [New Zealand average lifespan − average age of onset for narcolepsy in New Zealand]) × 100,000.

All of the sleep disorders mentioned in this study were diagnosed based on clinical criteria and objective sleep measures according to the International Classification of Sleep Disorders-Third Edition (ICSD-3).² All labs provided finalised diagnoses alongside their MSLT results.

Narcolepsy-related prescription data for 2021–2023

Pharmac was contacted by email to obtain data regarding medications used in the treatment of narcolepsy. A formal request under the Official Information Act 1982 was sent to enquiry@pharmac.govt.nz in accordance with their protocols. The email asked the following:

For these medications:

• Modafinil (MF)
• Methylphenidate hydrochloride (MPHC)
• Dexamphetamine sulfate
• Clomipramine hydrochloride

Could you please provide:

1.      The number of SAs for each medication, applied for and approved in relation to narcolepsy from 2021 to 2023. Both:

• Overall (new, repeats, and renewals).
• New (excluding repeats and renewals).

2.      Demographic data (and associated sublevels provided by Pharmac) for the information in (1), broken down by:

• Region: further divided into Northern, Midland, Central and Southern.
• Age: ranging from 10 to 84-years-of-age, plus an “other” category, recorded in 5-year increments (e.g., 10–14, 15–19, etc.).
• Gender: female, male and unknown (encapsulating gender-diverse and non-reported individuals).
• Ethnicity: divided into NZ European, Māori, Pacific peoples, Asian, Indian, Middle Eastern or Latin American (MELAA), other and unspecified.

For each sublevel, the overall and new SA applications were broken down by medication type and provided in a spreadsheet by Pharmac.

Of note, Pharmac requires that all stimulant medications used in the treatment of narcolepsy are funded via the SA application process. Prior to December 2024, these SAs had to be renewed every 2 years. This restriction has since been lifted.⁹ All data used in this study were collected before this change was implemented.

Statistical analysis

As this is a descriptive report, no statistical analyses were performed.

Results

View Table 1–6.

Between 2021 and 2023, New Zealand conducted a total of 346 MSLTs (Table 1). Four of these tests, conducted by Waikato Hospital, had unknown diagnostic outcomes and were therefore excluded, leaving 342 useable outcomes. Among the remaining tests, 57 (16.7%, averaging 19 per year) were positive for narcolepsy, 106 (31.0%, averaging 35.3 per year) for IH and 179 (52.3%, averaging 59.7 per year) for OSA and other diagnoses.

Based on the average number of new narcolepsy cases diagnosed annually, the incidence of narcolepsy between 2021 and 2023 was 0.36 per 100,000 people per year. This was calculated as: Incidence = (57 cases ÷ [5,223,100 total New Zealand population × 3 years]) × 100,000. The prevalence of narcolepsy was therefore 21.9 cases per 100,000 people, calculated as: Prevalence = ([0.36 ÷ 100,000] × [New Zealand average lifespan of 82 years − average age of onset for narcolepsy in New Zealand of 20.7 years¹⁰]) × 100,000.

Overall, Auckland NZRSI (39.2%) and WellSleep Wellington (30.7%) conducted the highest proportions of MSLTs, followed by Auckland City Hospital (13.2%), Christchurch Hospital (11.7%) and Dunedin Hospital (5.3%). In terms of narcolepsy case detection, Auckland NZRSI (18/57; 31.6%), Auckland City Hospital (16/57; 28.1%) and WellSleep Wellington (14/57; 24.6%) diagnosed the most cases. Christchurch (3/57; 5.3%) and Dunedin (6/57; 10.5%) made smaller contributions.

Between 2021 and 2023, a total of 223 new and 762 overall SA applications were submitted to Pharmac, all of which were approved (Table 2). The annual average number of new SA applications was 74.3 per year, which was 3.9 times higher than the average annual case number for narcolepsy (19 cases per year). In comparison, the annual average of overall SA applications (254 per year) was 13.4 times higher than the average annual case number for narcolepsy. MPHC was prescribed slightly more than MF, representing 53.8% of new and 57.9% of overall applications.

When testing centres were grouped by their Health New Zealand – Te Whatu Ora regions, the total number of new applications was lower than overall new SA applications provided in Table 2 (193 vs 223), while the total number of overall applications was higher (839 vs 762) (Table 3). The greatest contribution to new SA applications came from Central (34.2%, covered by WellSleep Wellington), followed by Southern (27.5%, covered by Christchurch and Dunedin Hospitals), Northern (25.4%, covered by Auckland NZRSI and Auckland City Hospital) and Midland (9.8%, covered by the Auckland centres, with a small minority from Waikato Hospital). Of the applications, 3.1% did not have a region associated with them. In the overall data, this distribution shifted to Southern (31.2%), Northern (28.4%), Central (24.7%), Midland (13.7%) and unknown (2.0%). In terms of new SA applications, the Northern and Central regions prescribed proportionally more MF at 57.1% and 53.0%, respectively, while the Midland and Southern regions used more MPHC at 84.2% and 52.8%. By comparison, for overall SA applications, all regions prescribed more MPHC (58.4% in Southern, 56.0% in Northern, 60.9% in Central and 56.1% in Midland).

When the SA application numbers were broken down by age groups, both the total number of new applications and the total number of overall applications were lower than overall new SA provided in Table 2 (148 vs 223 and 710 vs 762, respectively) (Table 4). For new applications, there were three notable peaks in the data: 20–24 years (25.7%), 35–39 years (11.5%) and 75–79 years (3.4%). In contrast, the overall SA applications peaked at 25–29 years (15.1%), 50–54 years (8.2%) and 70–74 years (4.4%). In the new SA group, there was a slightly larger proportion of MPHC used, accounting for 50.7% of applications. This was more pronounced in the overall SA group, where MPHC accounted for 58.9% of applications.

When the SA application numbers were broken down by gender, the total number of new applications was lower than the total new SA number provided in Table 2 (215 vs 223), while the total number of overall applications was accurately reported (762 vs 762) (Table 5). In both the new and overall groups, more SA applications were made for females (68.4% and 58.4%, respectively) compared to males (30.7% and 41.3%) and individuals with unknown gender (0.9% and 0.3%). In both groups, MPHC was prescribed more than MF, with MPHC accounting for 51.2% of new applications and 57.9% of overall applications.

When the SA application numbers were broken down by ethnicity, both the total number of new applications and the total number of overall applications were again lower than that provided for all new SAs in Table 2 (188 vs 223 and 732 vs 762, respectively) (Table 6). In both the new and overall application groups, most applications were made for the NZ European group (78.2% and 78.7%). This was followed by the Māori (8.5% and 13.1%) and Asian (5.3% and 3.8%) groups. The remaining ethnic groups contributed smaller proportions, ranging from 0.4 to 2.1%. In all three main groups, MPHC was prescribed more than MF, with MPHC respectively accounting for 54.4%, 81.3% and 80.0% of new applications for the NZ European, Māori and Asian groups. For overall applications, MPHC accounted for 57.8%, 58.3% and 53.6% of applications in these same groups. The remaining ethnic groups showed more variation in medication use.

Discussion

Between 2021 and 2023 New Zealand conducted 346 MSLTs. Of the 342 tests with diagnostic outcomes, 57 were positive for narcolepsy (16.7%), 106 for IH (31.0%) and 179 for OSA and other diagnoses (52.3%). In these 3 years, Pharmac approved all 762 narcolepsy-associated SA applications, of which 223 were new (i.e., not repeats or renewals). Overall, 52.9% of applications were for MF and 47.1% were for MPHC. Notably, no new, repeat or renewal SA applications have been made for dexamphetamine sulphate since 2021, and clomipramine hydrochloride (a tricyclic antidepressant) is not processed under special authority, meaning Pharmac could not supply data on its use. As such, discussion on both medication types has been omitted.

The global incidence of narcolepsy ranges between 0.3 and 1.3 per 100,000 person-years, while its prevalence sits between 25 and 50 cases per 100,000 people.¹¹ Based on the national diagnostic MSLT data of 2021 to 2023 (Table 1), New Zealand’s estimated incidence (0.36 cases per 100,000 person-years) is comparable to other countries, while our prevalence (21.9 cases per 100,000 people) is slightly lower. These estimates are the first ever calculated for the New Zealand context based on New Zealand’s diagnostic data. However, for two reasons, these measures may be underestimated. Firstly, Pharmac allows physicians to diagnose and prescribe for narcolepsy based on clinical criteria alone, suggesting that some cases may not be represented in the MSLT-based data set. Secondly, the MSLT data used in this study were collected soon after the COVID-19 pandemic. During this time, there were drastic burdens on New Zealand’s healthcare system, reducing access for non-urgent referrals across many disciplines.¹² This likely decreased the number of people receiving MSLT testing. Future work exploring narcolepsy in New Zealand will need to address these factors to ensure a wholly representative sample.

Beyond this, it was surprising to see that the average number of new SAs approved across 2021–2023 was 3.9 times higher than the average number of new cases diagnosed (Table 2). Logically speaking, there should be minimal differences between these values, as each patient should be treated under a single SA. Four sources for this discrepancy are suggested: 1) there are cases being diagnosed and treated based on clinical criteria alone (as above), 2) there are cases arriving from overseas with existing diagnoses, 3) patients are receiving both MPHC and MF in the same year, and 4) patients who are diagnosed with IH are being treated under the guise of narcolepsy. With the data collected in this study it is not possible to estimate the proportion of cases diagnosed using clinical criteria alone. Similarly, it is difficult to use these data when estimating the number of cases coming and going from New Zealand via migration. As such, points three and four are of primary interest.

For point three, the number of new applications for MF and MPHC were roughly proportional. This could suggest that patients are receiving both medications in the same year. Pharmac requires patients to have trialled MPHC before being prescribed MF.¹³ As MPHC is a second-line medication with little evidence for its use in narcolepsy,^14,15 physicians may be rapidly switching their patients onto MF due to inefficient treatment responses, indicating a redundancy in the SA application process. Other countries often combine MF with other first-line pharmaceuticals to properly manage the symptoms of patients.^14,15 If physicians in New Zealand were switching to MF from MPHC in the absence of access to these other options, hoping for better control, this would suggest a clear need to explore whether new medications should be introduced.

Then, regarding point four, IH and narcolepsy are both central hypersomnolence disorders.¹⁶ IH is not only similar to narcolepsy in clinical presentation, but it shares the same first-line treatment (i.e., MF).¹⁶ However, IH lacks the recognition and treatment avenues afforded to narcolepsy in New Zealand, despite our data revealing it to be the more common of the two (based on MSLT results). That is, there is no option to provide MPHC or MF to patients with IH as SA applications for MPHC and MF are only approved for narcolepsy or attention deficit hyperactivity disorder.¹⁷ Therefore it seems likely that part of the discrepancy between the number of new cases and the number of new applications for SA could come from applications for medications to treat IH being listed as narcolepsy-related. If this were the case, it would be clear that Pharmac should amend the SA criteria to allow funding for IH as well, thereby helping a significantly under-represented cohort of patients and allowing accurate data collection on central hypersomnolence disorders.

However, by assuming this is the case, one must assume the demographic data supplied by Pharmac do not truly represent narcolepsy. Instead, they are likely to represent central hypersomnolence disorders overall. Regardless, inferences for the demographic distribution of narcolepsy can likely be pulled from these data due to the similarities between both conditions. Within this, it should also be noted that the international prevalence of IH has been described as being four times less than that of narcolepsy.¹⁸ Although all IH cases reported in this study are confirmed diagnoses, diagnosed according to clinical and testing criteria from the ICSD-3, the discrepancy between New Zealand and the global population should be queried.

Further difficulties in interpreting the Pharmac data also come from the fact that Pharmac’s privacy policies prevented them from releasing data points with fewer than six values (these were conservatively estimated as one). As such, nearly all of the demographic totals—both new and overall—were underestimated when compared to the reported total for each group. The only exceptions were the overall totals for Health New Zealand – Te Whatu Ora regions and ethnicity, where Pharmac indicated that a single patient could belong to multiple groups. Since these underestimations for each demographic were significantly smaller in the overall group, trends and averages in the medications used to treat patients with central hypersomnolence disorders (and by extension, narcolepsy) are discussed based on these data. This decision was made to minimise the bias caused by unknown values. As a result, several interesting findings emerge.

Firstly, for geographic distribution, most new cases of narcolepsy diagnosed in 2021–2023 were identified by NZRSI and Auckland City Hospital (59.7%). This was followed by WellSleep Wellington (24.6%), Dunedin Hospital (10.5%) and then Christchurch Hospital (5.3%) (Table 3). Yet, in this same period, the Northern and Midland Health New Zealand – Te Whatu Ora regions (covered by the Auckland testing centres) only applied for 42.1% of the overall SAs, while Central (covered by WellSleep) applied for 24.7% and Southern (covered by the Christchurch and Dunedin Hospitals) for 33.2%. This suggests that despite diagnosing the minority of cases by MSLT, the Southern testing centres are applying for a relatively large proportion of narcolepsy-related SAs. Most likely, this would represent the South Island clinicians applying for a higher number of SAs based on clinical criteria alone. A possible reason as to why this might be happening could be the fact that there are only two testing centres on the island despite its large area. There could be a significant geographical barrier reducing access to testing, promoting clinicians to offer clinical diagnoses instead of formal testing. It may also be possible that a small percentage of established cases are being imported to the Southern regions within the student cohort. However, the data presented cannot be used to quantify this variable.

Age-wise, three peaks were seen in the overall SA data (Table 4). The first and largest was seen in the 25-to-29-year-old cohort. Its distribution was relatively wide, spanning from 20 to 39 years (all values above 10.0%). The second largest was seen at 50–54 years, and the smallest was seen at 75–79. The first two peaks seem to mirror what is seen internationally. That is, previous research has noted that the incidence of narcolepsy peaks between 14.7 and 18.1 years and again at 35 years.^18,19 However, due to a diagnostic lag of up to 13 years, many patients do not start treatment until years after.²⁰ Therefore, it makes sense that New Zealand prescriptions are peaking in these decades. Comparatively, the third peak is unexpected. As narcolepsy becomes increasingly recognised, we would expect the peak age of diagnosis/prescription to converge on the ages of onset. Given the fact we see peaks in the 60+ cohort, there may be some level of diagnostic catch-up occurring as knowledge of narcolepsy increases among healthcare professionals—consistent with data showing that some patients still report diagnostic lags of 16–28 years.²¹ Beyond this, as a point of interest, the use of MPHC over MF appears to increase with age. In this regard, MF only became funded in New Zealand from 2004.²² This pattern may therefore reflect the long-term, continued use of MPHC by 50-to-70-year-old patients who were initiated on this medication and preferred not to switch to MF over time.

Then, regarding gender equity, our data indicate there may be gender-related biases in diagnosis and/or prescribing (Table 5). At the time of data collection, New Zealand’s population was 99.3% cisgender and 0.7% transgender or non-binary.¹⁸ Of the cisgender population, 49.3% were male, while 50.3% were female.¹⁸ However, the percentage of females receiving SAs overall was high (58.4%) when compared to males (41.3%). The higher number of SAs approved for women is consistent with some international data showing that women are diagnosed with narcolepsy more than men—although such findings vary between studies, and the risk of narcolepsy among both genders has been shown to be equivocal with variations in symptom profiles and diagnostic access causing biases in the scarce, gender-related findings.²⁴ By comparison, the 0.3% patients with unknown gender may closely represent the number of gender diverse individuals within New Zealand’s population—especially when factoring in the conservative estimation of one for values less than six.

Finally, New Zealand has historically struggled to meet the health needs of its minority populations, especially for the Indigenous Māori.²⁵ Historically, Māori have been less likely to seek healthcare and less likely to receive appropriate management when they do. This has resulted in long-standing health outcome disparities, especially when compared to NZ Europeans.²⁵ These disparities have been a major public health focus in New Zealand for several years now. The percentage of overall SAs approved for NZ Europeans (78.7%) was high compared to the percentage of people identifying as NZ European (67.8%) but low in Māori (13.1%) compared to those identifying as Māori (17.8%) (Table 6).²⁶ It should be considered that the number of Māori accessing diagnosis and treatment for narcolepsy may be under-representative of the group’s needs. This is especially true when considering that Māori suffer from more frequent and severe viral and bacterial infections when compared to NZ Europeans,²⁷ which is argued to be a trigger for narcolepsy.²⁸ Māori also tend to live in more rural communities with lower access to healthcare,²⁹ potentially compounding any disparities. However, it is also known that globally, different ethnicities have different propensities for narcolepsy.¹¹ Therefore, New Zealand’s prescription rates could still be ethnically representative. More research is needed to confirm.

Limitations and future directions

Overall, this study has several limitations that need to be considered when interpreting its results. Firstly, the case numbers presented may be incomplete as some individuals with narcolepsy may not have sought diagnosis or may have been diagnosed based on clinical criteria alone. Secondly, regarding the SA applications, it is likely that the data presented actually represent central hypersomnolence disorders, reducing the specificity for narcolepsy. This is further complicated by the privacy policies of Pharmac—a system that naturally causes underestimations in the data for rarer disorders. To overcome these limitations, it is recommended that future research focusses on prospective diagnostic data collection encompassing both objective and clinical cases. Access to diagnostic testing services needs to be improved through Health New Zealand – Te Whatu Ora sleep services. It is also recommended that Pharmac develop and fund separate SAs for the treatment of IH—allowing proper reporting of both disorders, guiding future management.

Conclusion

This study provides the first estimates of narcolepsy incidence and prevalence in New Zealand and describes potential disparities in the medications used to treat it. Changes in policy and future prospective data collection are recommended.