The physiology of PCOS

Another one from the archives. This one is an article in the July 2014 JFPRHC explaining the physiology behind PCOS and our treatments, by W. Colin Duncan, clinical fellow and consultant in reproductive medicine at the University of Edinburgh. It’s complicated, but helpful enough that I’ve been keeping it around until I had time to summarise it for this blog.

The production of androgens and oestrogens

Androgens are produced from cholesterol in two places in the body:

  • The thecal cells of the ovarian follicle, under the stimulation of LH
  • The adrenal glands

Normally these produce roughly equivalent amounts of androgen.

There are two other important players we need to know about in this situation:

  • SHBG acts as a ‘sponge’ to mop up any excess circulating androgen. Hence, a deficiency in SHBG will mean that excess androgen levels aren’t dealt with and remain high.
  • Insulin potentiates the action of LH on thecal cells, meaning that the LH will be more effective in inducing androgen production. Higher levels of circulating insulin are therefore associated with higher levels of androgen production.

So, three things that can lead to increased androgen production are:

  • Increased LH
  • Decreased SHBG
  • Increased insulin

Oestrogens are, in their turn, produced from androgens. The androgen –> oestrogen conversion is carried out by aromatase, in the granulosa cell layer of the ovarian follicle, under the stimulation of FSH. (OK, I just gave myself a Cluedo flashback. It was… the FSH, with the aromatase, in the granulosa cells!) Oestrogen then inhibits FSH secretion in a negative feedback loop. (That bit will come in later, so remember it.)

(The thecal cell layer is the outer layer and the granulosa cell layer is the inner layer, but that doesn’t seem terribly important from the point of view of what goes on hormonally.)

So… why is this important?

Because androgen inhibits the growth of larger antral follicles but stimulates the growth of smaller antral follicles in the ovary. Therefore, excess androgen is going to stimulate an ovary to develop multiple small follicles that don’t spontaneously progress – in other words, it’s going to turn it into a polycystic ovary. This has been seen in rhesus monkeys treated with exogenous androgens, and in women with conditions causing pathological increase of endogenous androgens (androgen-secreting tumours and late-onset CAH).

So, the three situations listed above that produce increased androgen production can each contribute to the development of PCOS. In addition, a vicious circle can develop with the increased insulin secretion, as insulin causes weight gain, weight gain causes increased insulin resistance, increased insulin resistance causes increased insulin secretion… and hence causes greater amounts of circulating insulin, which causes weight gain. And weight and SHBG production are inversely related, so all that weight gain causes decrease in circulating SHBG and contributes to the symptoms.

But all this, of course, also gives us our clues about how to solve it. Because, if we can:

  • Decrease LH, and/or
  • Increase SHBG, and/or
  • Decrease circulating insulin…

then we can decrease the amount of androgen in the system, and hence improve the symptoms of PCOS.

How can we do this?

Decreasing LH can be done with the COC (which decreases both gonadotrophins). Note that the progestogens in older COCs tend to be more androgenic in nature, so for obvious reasons the newer ones are preferred. Cyproterone has anti-androgenic effects, so using co-cyprindiol as the COC can be particularly helpful with androgenic symptoms, as it doesn’t just block ovarian androgen production but also inhibits the androgens being produced in the adrenal gland.

(This, of course, doesn’t help when the problem is that pregnancy is desired – that situation is discussed below.)

Increasing SHBG can be done, as you will have gathered from above, by losing weight – which will have the double advantage of improving insulin resistance and thereby also decreasing circulating insulin. This, of course, is why weight reduction is such a key part of improving PCOS symptoms. Note that exercise will improve PCOS symptoms even in the absence of weight loss symptoms.

Decreasing circulating insulin can, as above, be done by losing weight/increasing exercise/eating appropriately. The other way to improve it, of course, is by increasing insulin sensitivity with metformin treatment. Women seem to vary in how they respond to this, but it can improve symptoms of PCOS.

What about fertility?

The goal of fertility treatment, in this case, is to increase FSH levels so that more androgens get converted into oestrogens. As stated above, oestrogens inhibit FSH secretion, so oestrogen antagonists will increase it – and that’s how clomifene works. However, oestrogen is obviously needed for endometrial secretion, as well as to stimulate the mid-cycle surge of LH that triggers ovulation, which is why clomifene is only given at the beginning of the cycle.

If clomifene isn’t effective, the next step is to try injections to raise FSH directly. However, all those partly-developed but arrested follicles in the ovary can still develop if stimulated with FSH, so women with PCOS are at increased risk of ovarian hyperstimulation syndrome as a complication of FSH treatment.

The article also mentioned letrozole, which is an aromatase inhibitor and hence will reduce the amount of initial oestrogen production and raise the FSH level (so… you reduce oestrogen production in order to ultimately increase it? Negative feedback loops can get complicated). It didn’t go into any details about how this might be deployed.

And don’t forget endometrial protection

Remember that those anovulatory cycles are leaving women exposed to unopposed oestrogen – as well as being a long-term risk, this can cause prolonged heavy bleeding in the short term. Amenorrhoeic women with PCOS need some kind of endometrial protection:

  • The COC obviously does the trick
  • LNG-IUS is an option
  • If women aren’t on either of these, it’s worth inducing three or four withdrawal bleeds a year. This can be done with oral MPA 10 mg twice daily for 7 days or once daily for 10 days; the bleed should occur a few days after stopping treatment. (If the bleed is particularly heavy, try giving the treatment more often so that there’s less chance for the endometrium to build up.) This would, I suppose, be useful in women who don’t want to be using contraception.

Or hirsutism

The COC, particularly one containing cyproterone, may deal with this. If not, look at options such as topical eflornithine or laser therapy. (The article didn’t mention any other options such as spironolactone.)

Posted in Credits 2017, Gynaecology | Leave a comment


I’ve been going through old journals, and found this useful article on urticaria in Prescriber 19.2.15, by Ruth Sabroe.

As well as typical acute urticaria in response to viruses, medication, and who-knows-what, there are a number of what are called physical (or inducible) urticarias:

  • Dermatographism
  • Cholinergic urticaria
  • Delayed pressure urticaria
  • Cold urticaria
  • Aquagenic urticaria
  • Solar urticaria

With the exception of delayed pressure urticaria, the wheals from these types of urticaria will come on within a few minutes of the stimulus and settle within a couple of hours at most (typically 10 – 20 minutes). DPU wheals appear between 30 minutes and 12 hours later and can last for two to three days.

In addition, some of these types of urticarias can present with a spectrum of symptoms; some present just with pruritus and no wheals, while some cause angioedema or, rarely, anaphylaxis.

Dermatographism: The one everyone remembers from medical school textbooks. Linear wheals are produced by scratching the skin or other graded gentle friction.

Cholinergic urticaria: Triggered by a rise in core temperature – can be with heat, exercise, or even with emotion. Tends to come out on the trunk/proximal limbs. Tiny wheals (about 2 mm) with a big surrounding flare that may be confluent.

Delayed pressure urticaria: The pressure in question can be caused by tight clothing, carrying heavy items, or standing on a ladder, and the wheals occur at sites of pressure. Can make it impossible to continue in a manual job.

Cold urticaria: Cold objects, cold food or drink, wind, cold water. Localised wheals. To confuse matters, they may only appear on rewarming.

Aquagenic urticaria: Triggered by water of any temperature. Looks similar to cholinergic urticaria. Note that aquagenic pruritus can be associated with myeloproliferative diseases; check FBC annually.

Solar urticaria: Triggered by sunlight and comes out only at the sites of exposure. Must be differentiated from other photosensitive rashes.

Also note urticarial vasculitis. This can look very similar to DPU, but needs to be distinguished as it is more likely to be associated with underlying disease such as SLE. It needs a skin biopsy and a full vasculitis screen to investigate it, if suspected.

In chronic spontaneous urticaria (no trigger, and lasts >6 weeks), investigation should be FBC, ESR (PV in the case of our local labs), and TSH. Most of the others don’t need much investigation, but do exclude urticarial vasculitis if it’s suspected.


Antihistamines (second-generation) are first-line treatment, as we all know. European guidelines recommend that the second-generation antihistamines can be used at up to quadruple doses for one to four weeks if the ordinary dosages aren’t working. Beware, however, as they can become sedating at those dosages, and some interact with other medicines. Also note that they lower seizure threshold, and be careful about renal or hepatic impairment (excretion method varies between antihistamines, so check). Loratadine or cetirizine are the preferred options if something’s needed in pregnancy (or chlorphenamine).

If a course of prednisolone is being used, it can be used for up to 1 – 2 weeks. Dose wasn’t given. Avoid using it for longer, obviously due to the side effects but also due to the risk of tachyphylaxis.

Other medications include:

  • H2 antihistamines
  • LTIs
  • Ketotifen (a mast cell stabiliser)
  • Low-dose doxepin
  • Immunosuppressives
  • Omalizumab

The article recommends that anything beyond antihistamines and the occasional prednisolone course should be initiated in secondary care, although personally I wouldn’t have a problem with trying ranitidine or montelukast myself while the person was waiting to be seen. Other reasons for referral include:

  • Severe urticaria
  • Testing for physical urticaria
  • Suspected urticarial vasculitis
  • Associated fever or high inflammatory markers
  • Atypical features, such as familial or lifelong disease
  • Pregnancy
  • Diagnostic uncertainty.
Posted in Credits 2017, Dermatology, Immunology | Leave a comment

HPV vaccine

Few points from the latest BMJ article:

The NHS programme currently only covers girls up to 18 years of age. 18 and over aren’t covered, but when I checked the Green Book I did note that it stated it to be ‘reasonable’ to complete a course if a girl had started it prior to 18 and then hit her 18th birthday before completion.

9 – 14-year-olds can take it as a two-dose course, with the doses at least six months and not more than two years apart. A girl who only has the first dose at 15 or older has to have three doses (normally 0, 2, and 6 months) as she doesn’t mount as high an immune response. If she’s had her first dose prior to 15, she can have a two-dose schedule even if the second one is given after 15.

Efficacy data are, as yet, based on immune responses to HPV. There are as yet no outcome data (not surprisingly – it’ll take a long time to get this data in). The vaccine is therefore not a substitute for screening as an adult (and obviously not a substitute for safe sex practices).

The vaccines contain only virus-like particles, without viral DNA.



Posted in Vaccines | Leave a comment

When Q waves are a normal variant

It seems I’ve been overdiagnosing past MIs on ECGs. A cardiologist has now sent me the helpful information that isolated Q waves in III are a variant of normal. Old inferior MI can only be diagnosed if there are Q waves in II and aVF in addition to III. Good to know!

Posted in Cardiovascular, Credits 2017 | Leave a comment


Some interesting bits and bobs in the latest BMJ:

The three most effective NSAIDs for hip and knee osteoarthritis appear to be, in descending order, diclofenac, etoricoxib, and rofecoxib. Now, of course, that’s of limited use, given the high risk of cardiac events from both the first and third of those; but I can see it being useful in patients with a short life expectancy who are desperate for some relief from their pain.

Falls prediction scores don’t work (Minerva). In a study in A&E in Australia, the Falls Risk for Older Persons Community Setting Screening Tool and the Two-Item Screening Tool (I so want to find a way to reverse the last two words on that second one) were little better than a coin toss in predicting falls over the next six months.

It’s possible to develop abdominal pseudohernia after thoracic shingles (Minerva again) causing a lateral bulge in the abdomen due to thoracic nerve root paralysis. In the case given, this resolved spontaneously three months later, though we weren’t given figures on the likelihood of this. (The likelihood of it happening in the first place, if you’re interested, is 0.7%.)

Posted in BMJ, Credits 2017, Elderly Medicine, Medication, NSAIDs | Leave a comment

Reflux in infants

(BMJ 2017;357:j1802, supplemented by checking the original guidance)

Not to be referred to as GORD unless the symptoms are interfering with the baby’s quality of life or causing complications (poor weight gain, difficulty sleeping, recurrent chest infections). Otherwise, it’s physiological GOR, minus the D.

Red flag symptoms

  • Frequent projectile vomiting – forceful enough to stain a wall or land across a room. This is associated with pyloric stenosis, especially if starting from third or fourth week in a typically hungry infant who’s failing to gain weight.
  • Bilious vomiting – can indicate intestinal obstruction and requires urgent surgical referral.
  • Blood in vomit, unless swallowed blood from nipple crack etc.
  • Vomiting unrelated to feeding – can be secondary to RICP and hence a sign of NAI.
  • Bulging fontanelle – also sign of RICP
  • Constipation/loose stools – might indicate CMP or lactose intolerance. (This is included on the list of symptoms requiring paediatric referral, which surprises me; surely a trial of a hydrolysed formula is the best first step?)
  • Blood in stool – CMP-induced enterocolitis (again, not sure why requires paeds referral rather than trial of hydrolysed formula)
  • Onset after 6/12 – unlikely to be reflux, may be infection
  • Persisting after 1 year – suggests alternative diagnosis
  • Systemic symptoms (may be infection or RICP)


Usually conservative.

  • For bottle-fed babies, try the following:
  1. Reduce volumes by about 20% but increase frequency to maintain appropriate total daily amount of milk.
  2. If that isn’t successful after 2/52, offer a trial of thickened formula (Aptamil anti-reflux, Enfamil anti-reflux, SMA Staydown).
  • For breastfed babies, advise seeing lactation expert/community midwife.
  • Try keeping babies in an upright position for the first hour or so after feeding, except when sleeping.
  • Always remember to ask how parents are coping.

If these measures fail, consider a trial of alginate. If that doesn’t work, and if the child seems distressed or has poor growth or choking symptoms as well as the vomiting, try a two-week trial of either H2 antihistamine or proton pump inhibitor. If that still doesn’t work, stop the medication and refer to secondary care.

Don’t use prokinetics without specialist advice.

Posted in BMJ, Credits 2017, Paediatrics | Leave a comment

Chondrodermatis nodularis helicis

I just checked some photos and confirmed my suspicions that this is what one of my patients has, so am making some notes on the advice:

  • Main thing is to avoid pressure on the ear during sleep. Cutting a hole in the pillow can help but sounds a bit drastic. Using a piece of foam rubber or a bath sponge, cutting a hole in that, and holding it to the head with a headband, sounds more feasible. Purchasing a silicone splint from a dermatology organisation (not sure where; maybe BAD?) can also be done but is obviously more expensive.
  • Protect from cold and wind; wear a warm hat when outdoors. (Luckily not an issue at the time of year that I’m writing this.)
  • Ulceration can be treated with petroleum jelly or with antibiotic ointment under a light dressing. In view of antibiotic resistance issues, I prefer the thought of the former.
  • Steroid injections, cryotherapy, and collagen injections are all options, although obviously not ones I could try.
  • GTN ointment 2% bd can be used in severe cases.
  • Surgical excision is an option, but there’s a 10 – 30% recurrence rate.

(DermnetNZ site, plus emedicine)

Posted in Credits 2017, Dermatology, Skin lesions | Leave a comment