Trends for prevalence and incidence of resistant hypertension: population based cohort study in the UK 1995-2015BMJ 2017; 358 doi: https://doi.org/10.1136/bmj.j3984 (Published 22 September 2017) Cite this as: BMJ 2017;358:j3984
All rapid responses
Manage uncontrolled hypertension as ‘treatment failure’ before labelling patients with resistant hypertension
We read with great interest the study on resistant hypertension by Sinnott et al., 2017. Uncontrolled hypertension (≥140/90 mm Hg) despite adherence to a prescribed drug regimen of three or more concurrent antihypertensive drugs which include a diuretic; or a four anti-hypertensive agents, non-diuretic based-regimen, is considered ‘resistant hypertension’.
Classical diagnostic reasoning suggests that ‘treatment failure’ should be considered if a diagnosis is made(e.g. hypertension) but the patient fails to respond to treatment (i.e. antihypertensive medication). We therefore recommend that the cohort of patients identified by Sinnott et al., 2017  should in fact be considered to have treatment failure. The management of treatment failure requires a systematic approach. Consider that perhaps:
1. The treatment is not reaching the circulation (concordance issues, malabsorption) or is ineffective (counterfeit medication)
2. The choice of treatment was suboptimal for a given patient (e.g. thiazide diuretic in a young Caucasian?)
3. The diagnosis is incorrect (pseudohypertension, faulty measurement, white coat hypertension etc.)
4. The patient has true resistant hypertension
We suspect that the true incidence of resistant hypertension is far lower than that suggested by Sinnott et al., 2017 . This is because compliance with antihypertensive medication is generally very poor. One meta-analysis reported that nearly 85% of patients with uncontrolled hypertension were not taking their medications. Sinnott et al., 2017  sought to detect compliance by assessing the collection of repeat prescriptions. However a prescription refill does not confirm that the patient is actually taking their medications. Even inspection of patient’s medications and pill counting at each clinic visit may not always reflect adherence. Inquiring about the side effects of medications, such as cough with ACE inhibitors or ankle swelling with calcium channel blockers can be helpful. However, these side-effects may in some instances be a consequence of the ‘nocebo’ effect.
Whilst measurement of serum levels of medications can provide definitive evidence of adherence these tests are rarely available outside the context of clinical trials. Clinical signs such as bradycardia (beta-blockers) or hypokalaemia (diuretics) are among the best tools to gauge adherence with antihypertensive medications that are currently available to jobbing clinicians.
In conclusion, we reiterate the importance of assessing for objective evidence of adherence and excluding other possible causes of treatment failure before labelling patients with resistant hypertension.
Rajkumar Rajendram,[1,2] Azra Mahmud 
 Consultant in Internal Medicine, Department of Medicine, King Abdulaziz Medical City, Riyadh, Ministry of the National Guard Health Affairs, Saudi Arabia
 Chairman, Medication Utilization & Process Evaluation Subcommittee, Medication Safety Program, Central Region, Ministry of the National Guard Health Affairs, Saudi Arabia
 Consultant in Adult Cardiology, Department of Cardiac Sciences, King Abdul Aziz Cardiac Center, King Abdulaziz Medical City, Riyadh, Ministry of the National Guard Health Affairs, Saudi Arabia
1 Sinnott SJ, Smeeth L, Williamson E, Douglas IJ. Trends for prevalence and incidence of resistant hypertension: population based cohort study in the UK 1995-2015. BMJ. 2017;358:j3984.
2 Mancia G, Fagard R, Narkiewicz K, et al. Task Force for the management of arterial hypertension of the European Society of Hypertension Task Force for the management of arterial hypertension of the European Society of Cardiology. 2013 ESH/ESC Guidelines for the Management of Arterial Hypertension. Blood Press. 2013;358:193-278.
3 Abegaz TM, Shehab A, Gebreyohannes EA, Bhagavathula AS, Elnour AA. Nonadherence to antihypertensive drugs: A systematic review and meta-analysis. Medicine (Baltimore). 2017;96:e5641.
Competing interests: No competing interests
Re: Trends for prevalence and incidence of resistant hypertension: population based cohort study in the UK 1995-2015
we very much welcome this important contribution to the study of resistant hypertension. This is one of the first general practice record studies published in this area, as opposed to those derived from trial or observational cohorts, which greatly enhances the generalisability of the reported findings. This is an impressive study, with a nationally representative sample of 2.2 million patients followed over 10 years. The suggested prevalence of 6.5%, which is significantly lower than that which has been reported elsewhere (see Achelrod, Wenzel, & Frey, 2015), resonates clinically and suggests that this condition is being over-diagnosed. With regard to pseudo-resistance, the examination of adherence through prescribing records, albeit a limited measurement approach, is most welcome; however, consideration of white coat hypertension and maximal dosing are still overlooked. These are likely to reduce prevalence even further. A study of individual patient records, which we are currently undertaking, may be helpful in this regard.
Peter Hayes 1, Monica Casey 1, Gerry Molloy 2, Hannah Durand 2, Andrew W Murphy 1, 3.
1Discipline of General Practice, School of Medicine, National University of Ireland, Galway, Ireland
2 School of Psychology, NUI Galway, Ireland
3 HRB Primary Care Clinical Trials Network, Ireland
Achelrod, D., Wenzel, U., & Frey, S. (2014). Systematic review and meta-analysis of the prevalence of resistant hypertension in treated hypertensive populations. American Journal of Hypertension, 28(3), 355-361. https://doi.org/10.1093/ajh/hpu151
Competing interests: No competing interests
Uncontrolled hypertension “resistant” to three, concurrent, antihypertensive drugs is an alarming diagnosis with significant, subsequent cardiovascular morbidity (1). Treating the condition as a syndrome may suit contemporary pharmacological approaches but, in many cases, there is an historical, renal arteriolar injury that may better define the condition. In 1937, Moritz & Oldt reported arteriolar stenosis in kidney, spleen, pancreas and adrenal in hypertensive patients of both sexes, across the age range with an increasing incidence with advancing age (2). The aetiology of that injury had not been elucidated though their central question was; is this injury a cause or a consequence of hypertension ?
We find that injuries to vasomotor nerves in renal (and other visceral) arterioles leads to release of cytokines, regeneration of injured nerves with concurrent hyperplasia of the denervated-vessel walls (3). Injured arterioles express purinergic, P2X3 ”stretch” receptors, and, are susceptible to hyaline and fibrinoid changes. Salt, sugar, stress, smoking, and straining at stool may all contribute to hypertension associated with these vasomotor injuries (4). Subsequent narrowing of renal arterioles also plays into the development of “resistant” hypertension through classical, Goldblatt mechanisms associated with relative ischaemia. The arteriolar injury may, therefore, be both a “cause and consequence” of hypertension producing a downward spiral of deteriorating consequences that ends as the syndrome of resistant hypertension. Preventing the initial injuries by amending lifestyle choices to avoid peaks of high blood pressure that damage injured, renal arterioles, may further reduce the rate of resistant hypertension ?
(1) Sinnott SJ, Smeeth L, Williamson E, Douglas IJ.
Trends for prevalence and incidence of resistant hypertension: population based cohort study in the UK 1995-2015.
(2) Moritz A, Oldt M
Arteriolar Sclerosis in Hypertensive and Non-Hypertensive Individuals.
Am J Pathol. 1937 Sep;13(5):679-728.
(3) Wu XQ, Xu XX, Zhou ZY, Quinn MJ.
Potential mechanisms in pregnant and non-pregnant hypertension.
BJOG, 2017 (in press).
(4) Wu XQ, Cai YY, Xia WT, Quinn MJ.
The aetiology of preeclampsia, 1945-53.
BJOG 2016; 123(13):2130.
Figure: Features of injured renal arterioles and nerves in hypertension is at http://www.preeclampsiaexplained.com/publications/letter-to-bmj-october-...
[A] Narrowed, renal arterioles with medial hyperplasia, and, irregular layers of perivascular hyalinisation (“fried egg” appearance) created by raised intravascular pressures causing traumatic injuries to renal vasomotor nerves (x100, hematoxylin and eosin).
[B] Regenerating, injured vasomotor nerves in the perivascular layers that also regenerate in the arteriolar wall (x200, anti-S100).
[C] Expression of purinergic P2X3 “stretch” receptors (brown) in endothelial and perivascular layers in narrowed renal arterioles (x100, anti-P2X3) that correspond with the layers of hyaline change in Fig. 1A.
[D] Collateral sprouting of injured, renal nerve bundles with loss of individual nerve fibres within the bundle, adjacent to narrowed arterioles that are pathognomonic for traumatic injury, which in this case, was associated with prolonged and persistent, physical efforts during defaecation (x100, anti-S100).
Competing interests: No competing interests