Primary Care Corner with Geoffrey Modest MD: Take Blood Pressure Meds at Night to Prevent Diabetes?

By Dr. Geoffrey Modest

Two complimentary articles from the same Spanish study database were just released from Diabetologia suggesting that higher night-time blood pressures might increase the risk of new onset diabetes and that giving medications at night might help.

1. In the first study, they found that people with higher night-time blood pressure had increased risk of diabetes (see DOI 10.1007/s00125-015-3748-8).

• Details:
  • 2656 people without diabetes (48% men, mean age 50.6, 13% smokers, 45% obese, 10% OSA (obstructive sleep apnea, defined by apnea/hypopnea index >=10 in in-hospital polysomnography, when there was daytime sleepiness and loud snoring/erratic breathing at night)
  • Baseline and annual 48h ambulatory BP (ABP) were monitored (more frequently with changes in hypertension treatment)
  • Baseline clinic BP 152/87, awake ABP 132/82, asleep ABP 120/69, mean ABP 127/77. (All patients wore wrist actigraph to assess level of physical activity during the ABP monitoring to confirm that they were asleep vs awake).
  • Followed prospectively for 5.9 years
• Results:
  • 190 people developed diabetes (defined as fasting glucose >7 mmol/l=126 mg/dl on at least 2 occasions and at least 3 months apart)
  • Overall there was more diabetes in older people (HR 1.02), those with elevated baseline glucose (HR 3.44), male, if baseline OSA, metabolic syndrome, abdominal obesity (HR 1.03), or chronic kidney disease (CKD, HR 1.53). Those taking at least one BP med at bedtime had lower incidence of diabetes than those taking all meds in AM (HR 0.43) — though no comment on what meds they were taking (see below)
  • The asleep systolic ABP mean was the most significant predictor of developing diabetes, adjusting for age, waist circumference, glucose, CKD, and hypertension treatment
  • The magnitude of association: every 1-SD decrease in asleep systolic ABP mean was associated with a 30% reduction in risk of new-onset diabetes. There was a dose-response curve (step-wise increase in diabetes as the asleep ABP increased). Also, the association held at all levels of baseline ABP, including those who were normotensive (i.e., even if the asleep SBP/DBP were below the current threshold of 120/70)
  • Daytime clinic BP and awake or 48h ABP mean had no predictive value
  • ​Of note, a greater morning or pre-awakening BP surge, was associated with  lower, not higher, risk of diabetes (HR 0.84), though the night-time fall in BP was also protective (HR 0.83)
• So, as with other blogs I have sent out, this study also found clinical correlation (this time, the development of diabetes) being associated with aspects of ABP monitoring and not with office-based measurements.

2. In the second study, they found that those assigned to take their medications at night had lower risk of diabetes (see DOI 10.1007/s00125-015-3749-7)

• Details
  • 2012 hypertensive patients without diabetes (48% men, mean age 53, BMI 29, waist circumference 95cm=37.4 inches, 50% with metabolic syndrome, 15% smokers, 38% obese, 8% with OSA, 22% with CKD, 7% with prior CVD event), randomized to taking all of their BP meds on awakening, vs taking >= 1 med in the entire dose at bedtime. followed 5.9 years
  • Meds: ARB (angiotensin receptor blocker) in 62%, ACE-I (angiotensin converting enzyme inhibitor) in 21%, CCB (calcium channel blocker) in 36%, a-blocker in 21%, b-blocker 21%, and diuretic in 36% — they did not mention which actual med was taken, just the class (except for b-blocker, which was primarily nebivolol)
• Results:
  • 171 people developed diabetes
  • Those taking some meds at bedtime had: lower asleep BP mean, greater sleep-time related BP decline and attenuated prevalence of non-dipping (32% vs 52%, p<0.001) — see below
  • Significantly lower HR of developing diabetes after adjusting for fasting glucose, waist circumference, asleep systolic BP mean, dipping classification, and CKD [adjusted HR 0.43 (0.31-0.61)], noting event rate of 4.8% vs 12.1% for bedtime and morning med-taking respectively (p<0.001)
  • ​Greater benefit was found for bedtime vs awakening treatment for those on ARBs (HR 0.39, p<0.001), ACE-I (HR 0.31, p=0.015), b-blockers (HR 0.35, p=0.021). Not on other BP meds.
  • No difference in adverse events if take all meds in am or some at bedtime
• So, a few points:
  • There is a connection between hypertension and diabetes: hypertension often occurs in the setting of insulin resistance (including activation of renin-angiotensin-aldosterone system or RAAS; elevation of angiotensin II and aldosterone increases hepatic glucose release and decreases insulin sensitivity. And hyperglycemia itself increases aldosterone levels (positive feedback). Studies have found mixed results on whether ACE-I and ARBs decrease the development of diabetes — though, perhaps relevantly, the HOPE trial (Heart Outcomes Prevention Evaluation) which found a 34% decrease in new-onset diabetes, had patients took ramipril at night!! And there are data showing that the RAAS system is more activated during night-time sleep
  • In terms of ABP, there are several studies showing that “non-dippers”, those who do not have the normal decrease in ABP at night, are at increased risk of cardiovascular events, finding more end-organ damage, fatal and non-fatal CVD (in those with hypertension with and without diabetes and even in normotensives). Also, diabetics have a greater likelihood of being non-dippers.
  • The data on ARBs and ACE-I are a bit confusing. This Spanish study suggests that ARBs/ACE-I work better at night in terms of decreasing new-onset of diabetes, and this is independent of their blood pressure lowering effect. The same group previously found in the same cohort that those taking at least one BP med at night also had a 67% decrease in CVD deaths, MIs or strokes (ishcemic and hemorrhagic), with p<0.001, representing 18 vs 55 events (see Chronobiology International 2010; 27:1629).
  • On the other hand, there are studies looking at blood pressure variability, which may be a different marker, finding that meds which are associated with more BP variability (which includes ACE-I/ARBs) are associated with more strokes (see Lancet 2010; 375: 895–905​), and there are studies showing that the duration of BP med effect varies (HCTZ at doses under 50mg has the shortest duration of action, chlorthalidone is much better even at doses of 12.5mg)  — So it would be interesting to know which diuretic was used in the above studies.
  • And there are studies showing that ACE-I are more often associated with strokes, including the ALLHAT trial (JAMA. 2002;288:2981-2997), with the argument posited that ACE-I wear off in the middle of the night, they are not working at the time of the early AM cortisol surge and blood pressure increase, and strokes happen more often in the early AMs. And the added increase in blood pressure variability from ACE-Is, as above, adds to that. So, I’m not sure how to reconcile the differing findings (including the finding in the above Spanish studies that the pre-awakening blood pressure surge seemed to be protective).

How should these new Spanish studies impact our clinical practice??  Hard to draw firm conclusions by these studies alone (and with relatively few events), especially given the perhaps contradictory findings of prior studies about ACE-I in particular. Based on the longer duration of action and least blood pressure variability of CCBs (and perhaps especially with amlodipine), I have been using that medication more over the past several years (and this was reinforced by the NICE guidelines of 2011). And, in the Spanish studies, taking CCBs (?which ones) did not show any difference if taken in the AM or before bed. But, there does not seem to be much downside to giving ACE-I at night, if you plan on prescribing them. The only concern to me is medication adherence, which often is worse with night-time meds, and the likelihood that the night-time meds would mean >once a day dosing. But for patients who are really med adherent, I will start using night-time ACE-I/ARBs.