AHI < 1.0 now, but still tired? Is it UARS, RERA, arousals or FL in your FR peaks?

[2SleepBetta]

This post looks at inspiration time and expiration time, their lengths and proportions of the ventilation cycle time. It looks at how those contribute to the Resmed (RM) flow limit (FL) flag ("FLG") durations and "severity" levels. Sequences of breaths with higher than normal proportions of inspiration  time tend to cause arousals and can explain unrestful low AHI sleep. Beyond harmful arousals, and I have not yet looked into this, lengthened inspiration efforts over long periods of time may have unique associated health detriments, maybe only those of apnea.

In short, anyone who understands--as I did not fully appreciate until recently--the contribution of lengthened inspiratory times (or high duty cycle ratios) to RM FLG signals of flow limitation will find little of interest here. For others who wish to learn more about this, I hope this post will help.

Here my recent posting focus shifts away from tidal volume (TV) and reductions that arise from airflow constrictions (TVd) in our airway. Those may or may not be signaled by the Resmed flow limit flag (FL)--("FLG") hereafter.  Achieving adequate, regular TV inflow (without labored breathing) is most often our no. 1 CPAP therapy goal. But understanding and treating our fragmented sleep--due to arousals from flow limitation--is a close no. 2 goal, if it is not no. 1. In addition to fhe FLG, unrounded tips of our inspiratory wave forms and the relative proportions of our inspiratory and expiratory times are key tells of limited air flow--even in the absence of FLG.

Many times there are no TVd coincident with a FLG.  However, I believe some (many? most? as I hope to learn) of those FLG instances with no TVd can be explained by disproportional  increases in inspiratory work time, the breathing "duty cycle", even if there are few FLG. 

Most all my TV, TVd, generic-flow-limit (fL) and RM FL related posts (errant study notes in truth) are in my thread  http://www.apneaboard.com/forums/Thread-...erstand-FL and in Sheepless' thread, where he and many of us members have posted  http://www.apneaboard.com/forums/Thread-...s-an-index  .

Here's an enlarged reminder to go with my signature zone's disclaimer: my errant posts trying to discover the meanings of "FLG" are a kind of notebook/journal, with warts and all. Mistakes, typos, oversights, misunderstandings and loose ends; all are there with, I hope, some illumination. All faults are mine alone, but I always hope readers will point out blunders, raise questions, and will either post those in my thread (a few have posted questions) or start another thread for their more fruitful and better communicated line of thought. My earlier slow work was started because of my flow limitations and widely shared confusion about meanings and significance of FLG. 

This post revisits an earlier-posted image (as linked below) and will likely be followed over time (soon I intend) with other "second looks" at, and some corrections of, other sleep session comments.   Although I was aware that single breath proportioning (Insp. vs. Exp. time) was factored into the FLG by Resmed, I only recently noticed instances where I had too readily dismissed FLG when TVd did not occur at a FLG. I mistakenly/obliviously assumed, at times, that many large FLG (long duration or high severity, or both) had been caused only by inspiratory wave flattening, "M"-tip and "Chair"-tip waves or some other patent-shrouded wave deformities RM factors into FLG. 

That oversight, I regret to say, is reflected (baked into) in the linked older image I reuse as the core of this post. Some time back I noted in the post or image that FLG (not explained by TVd)  were caused by Resmed (RM) factoring wave shape "flow limit threats" into their FLG. I mistakenly noted only that--that without due regard for the important timing factors RM reflects in FLG, particularly the breathing duty cycle ("dC"). The latter can be determined from OSCAR graphs:  dC = Insp. time/(Insp. time + Exp. time). Though more erratic, a high or growing  I/E ratio is easy to spot quickly:  drag the I-curve above and next to the E-curve (both zoomed in); drag the OSCAR cursor to times where the two curves diverge most widely; note and divide the I and E times (=I/E) at those times. In my experience only a very few narrower gaps would show high duty cycle proportioning.

[Extreme digression, out on limb here: My speculation of most recent hours (while gnawing on the FL bone)   is that a FLG reflects a kind of 6-component vector of 2 scalars and 4 severity ranking  components: [TVd, dC, M shape, Chair shape, Flatness shape, Other shape]. The TVd+dC vector resultant (real metrics, as derived from pressure variation) are directly but indistinctly reflected in logged FL severity. The latter four shape discrimination components and their rating may ony be yellow flags for alerting/staging the algorithm. Those rankings may or may not be reflected at all in FLG. That RM patent mystery, again, here.]

The second and third images, attached, take second looks at the linked earlier image, have some new notes added and, more importantly, show at the bottom of the images the role labored breathing (disproportionately long inspiration times) played in causing my long duration and moderately high flow limit flagging. The indicated FLG may also reflect unknowable shape factors (early in FLG build ups?); high duty cycle, IMO, likely accounts for the greater part of FLG not accompanied by TVd.

**********

The stated general purpose of this thread was for us to assemble as many helps here as we can find to address still-unrestful sleep when AHI<1. For me, a mere IFL student (now a VAuto-treated one) the following three paragraphs keep me banging my head against the interesting fL challenges.  

---A. Here is one common repeating inspiratory flow limitation (IFL) phenomena in my (earliest) and others' flow rate (FR) and FL graphs.  Mostly distinct FLG "mountains"   build upward over time, starting at low severity (coming from mainly obscure, not-obvious differing shape effects at first?). Those steps upward progress to higher and higher severities up the mountain which will often be reflected in longer inspiration times (higher duty cycle effects?). Next, a critical high severity is eventually reached when a sudden airway constriction (?) causes a significant fL at a peak FLG severity level. Then our bodies' respiratory drive resolves the limitation somehow (with a micro -arousal?). Next, there is a rapid fall of FLG severity to zero or near it. Usually it is then "Wash and repeat over and over" unless, say, a position change flattens the sleep path ahead. This overall pattern can be seen in the second attachment and I've seen it elsewhere in my and others' FL build ups and collapse. 

---B. As I understand, not every RM user with respiratory effort related arousal (RERA) or IFL will clearly present markers of those conditions in any of their OSCAR graphs except, possibly, in their I-time, their I/E or dC ratio, or inspiratory wave tips. There is sometimes a murky gray area where--aside from disproportionately long I-time RM indicators of difficult inspiration--there are few FLG, few FL mountains and there may or may not be flat (nearly constant fL) high IFL prairies with few or no FLG.  

--C. It seems obvious that CPAP therapy and lifestyle changes to reduce and control apnea must be higher priorities than treatment of low level fL. But at some level the health of those with high fL--"fL" signifying all flow limitation of any kind, including the RM FL but not scoreable apneas-- those with high fL and low AHI suffer equally with the adverse health effects of untreated moderate-to-severely apneic persons. Making matters worse for Ms. or Mr. High IFL,  their respiratory drive masks their IFL (as my attachments show) by rising to meet their metabolic TV demands through higher inspiratory work efforts as shown in  high duty cycle ratios. Others maintain TV with higher, less efficient, respiratory rates. Many or all tend to have more sleep fragmented with arousals. 

Anyone who can shed more light on those three points, or related ones, please, either start a new thread or post your light here.  By all means, hammer me where you see I am amiss.

**********

Finally, I move on to the linked and attached images. Those begin to revisit and correct my earlier posts where some or all FLG were sometimes treated too dismissively as mere indicators of wave shapes the RM devices are "wary" of--i.e, no TVd need be present, as I apparently lapsed into thinking was necessary for a FLG to be thrown.  For those recent times of my earliest dismissal of FLG I plead ignorance (not a defense), having only recently come to learn and deeply appreciate the roles of lengthened inspiratory times or increased duty cycle ratios in our fights against fL . Nevertheless, IMO and AFAIK, RM's may mysteriously mark "offending" wave shapes with FLG if they "offend" enough and are all the RM has. A rebuttable or refutable presumption? (Edit: This disagrees a bit with my later-hour speculation about shape effects on FL above.)

This next linked image has two zooms from its longer-duration top set of graphs. Each newly modified lowest zoom segment, left and right as in attachments, has new comment, newly added I and E curves showing duty cycle values at times. The uppermost image with its two zooms was posted as my August 2021 point of view (POV or then current understanding) as can be found at the lower linked AB page. (Do not be put off by the one-hour time differences that arose from time zone fiddling at different image-snip copying times.) 

http://www.apneaboard.com/forums/attachm...?aid=34974

http://www.apneaboard.com/forums/Thread-...#pid409809

Here is a link to the researchers' paper which studied the relationship between their tested severities of flow limitations (distinct from FL severities) and duty cycle ratio Inspiratory duty cycle responses to flow limitation predict nocturnal hypoventilation | European Respiratory Society (ersjournals.com) . See figure 3. 


Oops! I see the L-R order of the attachments is reversed, but I got burned trying to straighten that out earlier, so will leave it alone.

       

[2SleepBetta]

Main takeaway from this somewhat duplicative post (little is new here, just some improvement and enhancement in the graphic):

If troubled by unrestful low AHI sleep my actionable conclusion from this post recommends this (again): 

---1.   RM users can see on their device screen Sleep Report their Ti and Ti/Te (or I/E average like 1:1.8 which equates to 1/1.8=0.56 or duty cycle=Ti/(Ti+Te)=1/(1+1.8) = 0.36) 

---2. Further, check significantly lengthier periods of wider divergence between the "Insp. Time" and "Exp.Time" curves OSCAR presents. First drag the Insp. Time curve next to and above the Exp. Time curve. At a wider-spread time spot that is of longer duration than most periods drag the cursor from spot to spot there and jot down some time reading pairs from the upper left corners of the two viewing windows where values show and do change as you move the cursor. 

---3.   Check a few ratios by dividing Ti by Te. Those ratios far larger than the I/E=0.56 shown above can (if they do not always) present a problem you may wish to review with your sleep doctor or ApneaBoard experts (among which I am not, though I might offer my thoughts while you await their attention). 

********

This post is mostly an update, for what it's worth: 

It updates the evolving meaning to this novice of the Resmed (RM) FL flags (FLG). It reflects my persistent but naive study of just two individuals' richly disordered breathing in two single sleep sessions. I have used the two examples repeatedly. 

Here I again step away from describing what I think I have learned about the results of sleepers' night long tidal volume losses to look more deeply at their sleep efforts (sleep work)  to inspire, to inhale, sufficiently regular air volumes.  Sleepers who do excessive work to breathe in their air have unrestful, probably fragmented sleep, maybe have or are exacerbating their flow limitation caused co-morbidities like hypertension and A-fib.

Highest level ApneaBoard Administrators, Monitors, Advisors, etc., have pointed to lengthening inspiratory times as revealing and confirming subtle cases of flow limited breathing. Only lately, I began to look at those longer times for help in understanding the seemingly mysterious RM FLG. Those looks were after mistakenly opining that some FLG or FLG severity levels, those not accompanied by tidal volume drops, were mere therapy phantoms, though valuable to the PAP device for its pressure regulation. 

This post focuses on benefits of the "Insp.-Time" and "Exp.-Time" proportions indicated by duty cycle ratio (dC). The dC (less erratic than I/E ratio) largely helps "reconcile" visually large but essentially meaningless apparent severity differences between a FLG and its associated tidal volume (TV).

Overall, my studies, as in the attachment, indicate higher and/or longer duration FLG are driven most by the two most impactul of four (of five, of six?) sleep disordered breathing (SDB) components Resmed CPAP machines factor into a FLG: those two being reductions in tidal volume (TV), which I call "TVd", and our breathing duty cycle ratio, which I call "dC". 

********

Ancillary notes before discussion of the attached graphic, a graphic posted earlier for different, less evidence-informed reasoning:  

---1.   First, some additional "home brew" shorthand I use, not knowing better: "fL" stands for all flow limitation, other than OA and CA whether flagged or not, "Ti" is inspiratory time per breath, "Te" is expiratory time per breath, "Ttot" is Ti+Te, "RR" is respiratory rate, and TVb is the sleeper's assumed (estimated), night-long, "normal" average tidal volume base .

 ---2.   A FLG can have two (up to four?) other contributing factors for RM detection of and handling of inspiratory wave shape irregularities, any wave tip lacks of roundedness: M-tips, Chair-tips, flattenings, other deformations. I assume RM detection of these alert it to airflow conditions which may either deteriorate (or improve) and require a timely PAP pressure adjustment up or down. It seems we can ignore some wave tip deformities which reflect abnormal airway conditions other than significant TVd and/or greater inspiratory labor (higher dC ).

---3.   My usual rough (Excel) data analysis and graphing and (Publisher) publishing skills are inflicted on you in the small print of stacked, cobbled together, individual pieces of work done at different times, often with unnoticed typos, even earlier blunders.  One help for them: In Windows 10 there is a magnifier one can call up by hitting the Windows key and "+" sign (close it by hitting "-").

********

My earlier study, work, and discussion focused mainly on TVd vs FL meaning. It is largely in later posts to this thread http://www.apneaboard.com/forums/Thread-...#pid409245  and, is also along with others' discussion of FL in this thread http://www.apneaboard.com/forums/Thread-...#pid389057 .) Members cathyf, Amirkas and many others have informative threads centered on flow limitations, too. 

The attachment below:

Most work is presentation of breath-by-breath graphs of 25 hz RM data, and my depictions of and derivations from that data: FR (my shapes agree with OSCAR graphs) and derived, TV, TVd, TVb, Ti, Te and dC. 

Along the second axis down the dark green is the FR curve and the light green shows calculated breath by breath TV. TV shortfall below the 0.4 L assumed TVb baseline is depicted by red bars descending from the top axis of the image. Red-colored small fuzz above and below that upper axis reflects normal small variations in TV; the more significant red TVd descend lower. Large rises of red above the upper axis show extra-large TV greater than 0.4 L and are mostly the result of large recovery breaths. Yellow-color depicts FL duration and severity and it turns to orange/rust color wherever yellow overlaps red TVd.

Along the third axis down, the green color shows Ti, red color Te, and blue color dC. Notice that the green Ti with few but important exceptions, is smoothest of all curves. I believe it shows the sleeper exerted nearly constant, night-long longer inspiratory efforts to take in each breath and thereby maintain supply of needed TV despite high fL. When breath sufficed with less effort, expiration was more "leisurely" and more red-color y-extent (Te) showed and was reflected in a lower dC=Ti/(Ti+Te). One caveat, however: crowded together and overlapped 2-pixel wide traces along a night-long time axis can obscure important detail which our great tool OSCAR can zoom in on and reveal.

Where (lowest) green traces obscure red traces the dC is highest. Where the y-extent of the raggedy band of red (Te) is wider and widest the dC is lower and lowest, TVd and FL are also lowest. At times of those extremities, look at larger y-extent of bands of light green (TV); they are narrower at high Ti and dC and are wider and less grassy at lower Ti and dC.

The rectangle- enclosed items f, i, and k, particularly k, are puzzling departures from other patterns. Irregular f and i might be because of transitions, possibly position effects. Item k baffles me: TV is high, not grassy, dC is low to moderate yet FL duration and it long high severity may be greatest of the session. My intention is to look at the Ti wave tips to see how those compare to those elsewhere.

Notice how slanted lines agree and disagree among themselves and how the two largest and most pronounced two FL up trends--of three largest FL--in FL are matched with, met with increased dC.

The three yellow lines through the curves along the third axis mark the following. A RM patent includes a kind of rule of thumb that dC near 0.30 and 0.40 tend to indicate sleepers' sleep or wake states, respectively. The upper most line was intended to represent the research finding that approximately 0.52 (+/-SEM) was the most severe Upper Airway Obstruction finding for severely flow limited breathing dosages given its 26 test subjects. Visually, the illustrated sleep dC looks to be about 0.46 which the study characterized as mild UAO. The study "Inspiratory duty cycle responses to flow limitation predict hypoventilation" can be found online.

The study points out how men and women differ in handling flow limitations with their natural dC and RR responses to them. Interestingly, in non-REM sleep TV/Ti and MV (using our abbreviations) was 40% lower for BMI and age matched women vs men, not to mention the lesser dead space losses of females. Variations among individuals for most measures deviated wider than expected.

The respiratory rate curve is the usual OSCAR presentation of 0.5 hz data. It shows a bit lower but nearly constant RR after a couple hours as well as periods of most disturbed breathing.

   

[sheepless]

I'm responding just so you know I'm following along and looking forward to your eureka moment!

[2SleepBetta]

In my previous post I noted that the graphic item designated by the letter "k" was puzzling and it should be followed up. This is that follow up. 

Item "k" was puzzling because there was a long series of high FL severities which were not reflected in TVd, TV values were among the highest of the night. Further, at that graph scale it was not clear that high or varying dC could account for the high FL. I mentioned that possibility, but also thought deformed wave tips might account for the FL. Wrong.

I believe the image below explains the high FL. It shows a combination of both high dC and large variations of that dC. Whether one or the other or both dC factors could cause such high FL flagging, I have no idea, but those in combination with  prominent cardio ballistic artifact (CBA) effect  are the only plausible explanation for the persistent relatively high FL severities.

CBA is  a problem Resmed devices and patents address, best they can (quite well), and it complicates the kind of novice work I've been doing to understand the FL flag. CBA have tended to obscure and confuse all timing issues when prominent--possibly causing (IMO) much of the sharp variations in TV, Insp. and Exp. Time and RR we see in OSCAR portrayals of device data. Further, CBA effect on inspiratory wave tips no doubt causes some of the M-tips the Resmed devices are sensitive to.

All said, when FL indications are high as wave tips are round, I'll be looking carefully at the inspiration and expiration times and their variations. As I type this post, I don't recall if those times (for Resmed devices) are what are called high rate data (sampled at 25 hz) or low rate data (0.5 hz).

See item "k" here:

attachment.php (1632×1056) (apneaboard.com)

   

[2SleepBetta]

Square on topic, here is the most comprehensive survey of research about arousals and the role of arousals in sleep fragmentation and its health and function impacts. It is likely there are later such reports than this one from 2007, however, this paper is too rich with information to neglect before searching out later reports.

Link where a PDF file can be downloaded https://jcsm.aasm.org/doi/10.5664/jcsm.26815

   

This section was one among several which caught my attention. The gold standard sleep lab device for detecting arousals is the EEG and its interpretation. This excerpt holds some promise we PAP user might sufficiently approximate EEG value by means more accesible to us.

"7.2 Measures of autonomic arousals

"A number of studies have explored the utility of measuring autonomic events, usually blood pressure or heart rate changes, as
a way to quantify arousals during sleep. This work is based on the observation that changes in measures of autonomic function,
such as increases in arterial blood pressure, occur in response to acoustic stimuli12 and at the conclusion of obstructive respiratory
events.82 Studies of autonomic arousals have generally addressed one of the following 3 areas: 1) the utility of using autonomic 
arousals as a screen for sleep disordered breathing, 2) use of autonomic arousals as a more easily scored proxy for EEG arousals, or 
3) autonomic arousals as a predictor of daytime sleepiness as compared to EEG arousals. 

"The first of these questions is beyond the scope of this paper and will not be addressed. The other 2
questions are related to the goals of this paper, and relevant studies are reviewed below. Twelve studies,
including 6 level III stud ies,32,83,12,36,82,81 5 level IV studies,84,85,66,69,68 and 1 level V86 study
are summarized in Table 6 (which can be accessed on the web at http://www.aasmnet.org).


"7.2.1 Autonomic arousals and daytime sleepiness

"Autonomic arousals occur with obstructive respiratory events and with snoring, even in the absence of EEG arousals.82,86 
Autonomic arousals have also been detected in normal sleepers in response to experimental stimuli when EEG arousals did 
not occur.85,83 In contrast, EEG arousals were always accompanied by an autonomic arousal in those studies. These results 
suggest that autonomic arousals have a lower threshold and are more sensitive to perturbations in the CNS during sleep. 
Therefore, it is theoretically possible that these events may be a more accurate indicator than EEG arousals of CNS changes 
sufficient to cause daytime impairment"[.]

[My emphasis added]

[2SleepBetta]

This post revisits my quest to understand unrestful low AHI sleep using, for a time in 2021, the Dreem 2 EEG-like device. The attachment is intended to illustrate what correlation the reader can see there was, or was not, between flow rate disturbances and the D2 sleep stage indication. Two of attachment OSCAR graphs are most important: flow rate and sleep stage and, I hope, those two can  be discerned without eye strain.

A member prompted this post by asking what my experience was with the Dreem 2 shown in my post sidebar. My thoughts and comments now may not be the same as I might have had 18 months ago when the structural part of the Dreem 2, the headband, parted, and I have let the D2 lie--not that impressed enough with its helpfulness, as noted in other posts, to fix it ahead of other task priorities.

My main goal in buying D2 was to see, in OSCAR graphics, how well and consistently it would associate my variously disturbed flow rate wave forms, patterns and sequences with changes of sleep stage from any stage of sleep to a lighter stage or wakefulness. By my lights, the D2 was useless for that so I could not suggest its use here for that purpose. If I recall correctly, literature about the D2 noted its weakness in distinguishing light sleep from being awake--and my experience often confirmed that when I was definitely awake.  

It is possible my waveform peculiarities, and/or lack of capacity to visually correlate wave patterns with sleep stage changes, underlie my dissatisfaction. However, I do believe the D2 would be somewhat useful to monitor effectiveness of sleep hygiene changes for those who are not chronically troubled with unscored mini-arousals like RERA or have UARS. 

Graphics in the attachment were gleaned from many (15-20?) sessions, were dated sequentially--but with sleep sesssion gaps--within an old OSCAR/D2 profile; that to revisit this topic. Focus was on  sessions with most disturbed wave patterns (possible arousals) in relation to sleep stage and stage changes. Graphics are very small, so as to show many examples of worst wave forms vs stage change or lack of stage change. The graphic depicts top to bottom: flow rate, flow limitation, sleep stage and tidal volume. Stacking exceptions: a couple graphics show pulse rate above SPO2 inserted into the scheme, just below flow rate.

Draw such conclusions as you may. I made no effort to be selective other than to show the more outstanding waveform-bursts/series of the session or  to show there was no stage change from the disturbance. The choices are simply representative of that approach.

The bottom right graphic: starting sleep is LIGHT, sleep deepens to DEEP and then stages change to LIGHT, REM and WAKE (-4, -3, -2, and -1, respectively, in the OSCAR frame). Look at those first to calibrate understanding stages shown in  other graphics.

Three check marks indicate graphics I consider illustrations of coughs. Not to say  there has to be arousal or stage change from coughing, but, by my light, it seems strange that the D2 did not signal any stage change. Is thayt due to its scoring window size, possibly 2 minutes comprised of four 30 second positive detections?  All this about D2 resolution limits is far above my pay grade.

   

[EddyDee]

Very interesting, thanks! Might help if the panels are numbered.

Row 2 panels 1, 3, 5 seems to show progressively shallower breathing followed by an irregular pattern - cough, arousal, whatever. What that might indicate specifically is hard (for me at least) to say.

It would be useful to see if regular sleep breathing resumed immediately thereafter. For example, is row 1 panel 3 simply the infamous SWJ?

IIUC the newer ResMed firmware versions include RERA detection. Presumably that is based on this kind of analysis?

[2SleepBetta]

In addition to numbering the images, as you mentioned, EddyDee, it would have been good to date stamp them. After seeing your post I attempted to revisit the items you designated and respond to your comments, but was unsuccessful locating the instances shown, within reasonable time, upon scanning my final month of D2 usage.

My focus in that attachment--my main purposes for buying the D2--were to show changes or absences of changes in sleep stage at times of large flow rate disturbances and increases and to learn to detect arousals. I hoped to make some kind of catalogue of associations. But I saw little if any correlation. 

You raise good points others, here, could respond to much better than I.

[2SleepBetta]

My two aims as I request others' thoughts here: 

1. Illustrate a possibly meaningful difference between use of an AirCurve10 Vauto in Vauto mode vs Vauto in CPAP mode. (Besides my fraught first guess at a CPAP pressure setting for the VAuto there is this somewhat off point matter; I believe—may misremember—that AB experts have pointed out that the Vauto algorithm and its PS deliver better FL therapy than that of the  AirSense Autoset with EPR....even with Vauto PS setting"="EPR setting as ilimited up to 3.) 

2. Request others' suggestions of how my Vauto's CPAP and VPAP settings might be changed to minimize arousals which I believe are likely reducing my REM sleep.
Rusty on whatever I have learned about the topics IFL, FL, UARS, and SDB, I looked last night for information to help me find better CPAP or VPAP settings which might correct my near absence of dream recollections—those REM stage indicators– after waking. 

Regrettably I've lost track of a thread, read last night, discussing Lefty Lanky's negative position regarding VPAP and the issue of arousals caused by pressure changes. That discussion is what motivated me to experimentally switch my VAUTO to CPAP mode for sleep last session.

The upper half of the attachment shows my long-time typical sleep and settings for VAUTO mode. and the lower half shows my CPAP mode results last night—er, this morning. 

I switched to the VAUTO several years ago because of my AirSense Autoset FL like and worse than that of the AirCurve 10 Vauto (set to CPAP mode) in the lower graphic.

My SWAG last night was to try, as I did, 12 cm pressure and EPR 2. Got no dreams, no surprise, but lotsa FL as in the past with the AutoSet. Can't say sleep was any different. Can't say pressure bothered me; in VAUTO mode it has ranged 10-14 cm for about the last 4 months with Max IPAP 17 and PS 4. Am thinking of trying 12.5 cm tonight, same EPR 2. 

A number of AB members have previously drawn attention to PubMed's PPMC4688581 regarding flow limits.
At the linked page there are rich AB results dealing with FL from my Google search of AB:

PMC4688581 site:apneaboard.com - Google Search

[2SleepBetta]

The motivation for this and the previous post was to try cpap mode, after long using vauto mode, with a goal to see how I could increase REM sleep. The trials of cpap as in my previous and this post show so much more flow limitation than in my use of vauto mode. It's hard to believe there is so little difference in restfulness of sleep. This new observation and evidence puzzles me.

This post follows up my previous one which should be reviewed to help interpret both posts. The uppermost Flow graph here differs little from the nature of the similar graph in my previous post. It is presented mainly because the attachment below depicts a sleep session with AHI=0, high density of FL, and my same usual restful sleep as I've long had from vauto mode of my Vauto. This does not seem to make sense. Can it possibly be that cpap mode is picking up cardio ballistic dips in tips of my inhalation wave?
 
The attachment below presents, again, differences in OSCAR's portrayals of my Vauto's flow limits (FL) in vauto vs cpap modes. FL are much more dense in cpap mode, yet in that mode a few nights now, I've surprisingly perceived little, if any, sleep quality difference from sleep in vauto mode.

This attachment: the upper pair of graphed data show Flow and Flow Limits (FL) for cpap mode, in the IPAP 14 max, EPR=3 sleep session.  
Again, in vauto mode (previous attachment) the sleep quality was typical (good) and virtually all my sleep session vauto pressures are in the range 10-14 cm water column with PS 4 and max pressure is, if I remember, 18 cm.

The lower graphic depicting flow, TV and FL, has finally convinced me that the "M" tips at max flow are from cardio-ballistic effect, do not account much for FL flagging. Those are present in most all my flow peaks. It seems those may be interpreted by the imperfect algorithm as airflow limitation at some confounding points and may contribute some to FL flagging.

The rough graphic correlation of the slightly lagging TV minima with each series of dipped flow rate tips is evident, but not striking. I have seen that before and may have presented graphics showing that and shown, also, that there is some degree of correlation of tip dipping, TV minima, and FL flagging.  Of course, variation in the areas under each inhalation wave, wave densities and the aggregate total area of a series of waves determine variations in TV.  

Long past bedtime and rummy, please excuse my more than usual prolix, redundancies and injection of my own therapy—the latter contrary to stated intent of this thread. Good intention was to raise these questions and get input because the matter does fall within the thread topic and only lately and lightly touches on my attempt to get more REM sleep.