SCIENCE: Human Sleep Falls Into at Least 16 Distinct Types, Large Data Study Finds.

A sys­tem­at­ic review of sleep data from more than 100,000 peo­ple in the Unit­ed King­dom has revealed 16 dis­tinct ways we snooze. The clus­ters could one day allow experts to bet­ter diag­nose insom­nia and prein­som­nia in individuals.

The research is based on data from smart wrist­bands, used by the Unit­ed King­dom Biobank to deter­mine our pat­terns of wake­ful­ness and sleep. This was done by mea­sur­ing the arm move­ments and twitch­es of par­tic­i­pants over the course of sev­er­al days.

The result of the long-term study was a ‘real-world sleep­scape’ that showed quite the diver­si­ty of sleep cycle patterns.

Over­all, the clus­ters are split into five broad cat­e­gories (1 to 5). Those that fall in the first cat­e­go­ry gen­er­al­ly expe­ri­ence insom­nia with long sleep dura­tion and ‘midawake’, which is when you wake up in the mid­dle of the night.

On the oth­er end of the spec­trum, peo­ple who fall in cat­e­go­ry five sleep all the way through the night with­out tak­ing naps in the day.

Clus­ters 2 to 4 are fur­ther split into a’s and b’s. 2a, for instance, is a clus­ter of those who show irreg­u­lar sleep sched­ules, prob­a­bly as an indi­ca­tion of shift work.

2b, mean­while, is defined by the authors as “frag­ment­ed sleep with short sleep dura­tion” overall.

3a is a clus­ter of those who expe­ri­ence insom­nia with nor­mal sleep dura­tion, and 4a includes those who suf­fer from insom­nia with short sleep duration.

To make things more com­pli­cat­ed, the clus­ters 3b and 4b are fur­ther split into eight sub­cat­e­gories of frag­ment­ed sleep.

3b‑1 is a large clus­ter and includes those who sleep deeply but can­not fall asleep eas­i­ly once they wake up. 3b‑2 includes short-dura­tion sleep­ers who show a mix of short wake-ups and some longer hours spent star­ing at the ceiling.

4b‑1 is anoth­er large clus­ter and fea­tures long sleep­ers. 4b‑2 hous­es all the morn­ing peo­ple and 4b‑6 hous­es all the night people.

4b‑4 and 4b‑5 is defined by the authors as ‘prein­som­nia’. Both clus­ters show nor­mal wake-up dura­tions in the mid­dle of the night, but those in the for­mer clus­ter wake and fall asleep repeat­ed­ly and fre­quent­ly, which sug­gests they strug­gle to main­tain sleep. Clus­ter 4b‑5, mean­while, shows less frag­ment­ed sleep as a whole.

4b‑3 gen­er­al­ly has a short­er 24 hour peri­od­ic sleep-wake cycle that gets out of sync with the usu­al cir­ca­di­an cycle depend­ing on whether some­one is work­ing or on hol­i­day. This is also known as ‘social jet lag’.

The clus­ters iden­ti­fied in the cur­rent research are com­pli­cat­ed and diverse, and they weren’t just put togeth­er for fun. They could very well prove clin­i­cal­ly mean­ing­ful with future research.

Past stud­ies, for instance, sug­gest that insom­nia marked by short sleep dura­tion is linked to impaired neu­rocog­ni­tive func­tion, while insom­nia marked by nor­mal sleep dura­tion is linked to an anx­ious-rumi­na­tive profile.

Data record­ed from peo­ple mov­ing their arms dur­ing sleep could not only help iden­ti­fy those with insom­nia, it could also help reveal those most at risk of devel­op­ing men­tal or phys­i­cal health issues.

To get to that stage, how­ev­er, we need a lot more research. While oth­er forms of brain and mus­cle mea­sure­ments dur­ing sleep are too short-term to iden­ti­fy the nuances of ‘social jet lag’ or ‘morning/evening types’, the clus­ters iden­ti­fied in this research still need to be ver­i­fied further.

The authors hope the “sys­tem­at­ic and unbi­ased clus­ter­ing method” they’ve devel­oped can one day be linked with oth­er per­son­al infor­ma­tion, like present ill­ness, past med­ical his­to­ry, med­ica­tions, edu­ca­tion occu­pa­tions and lifestyle habits to help us bet­ter under­stand the sta­tus of an indi­vid­u­al’s health.