5D Library › 下丘脑垂体肾上腺轴耦合振荡、皮质醇觉醒反应扁平化与倦怠恢复动力学 · English

第三章 受体把自己数下去

一个由它所约束之物做成的刹车

第一章里那个刹车不是一个固定的部件。它是一群受体分子,而一个被占据过的受体也更容易被处置掉。Oakley 与 Cidlowski 关于受体生物学的综述列出了慢性暴露减少受体数目并改变其信号传导的那些机制,这就是为什么同样的激素浓度在一个已经承受数月负荷的人身上不产生同样的效应 (Oakley and Cidlowski 2013)。McEwen 关于稳态负荷的论述给出临床框架:适应的代价是用适应本身的机件来支付的 (McEwen 1998)。

证据

已发表。 Oakley 与 Cidlowski 综述糖皮质激素受体的生物学,包括持续暴露下受体数目的下调及其对信号传导的后果 (Oakley and Cidlowski 2013)。McEwen 把稳态负荷阐述为适应中介物(其中包括糖皮质激素)所累积的代价 (McEwen 1998)。

把这件事写成算术需要三个速率:一个恒定的合成、一个基础的处置,以及一个与占据成正比的额外处置。稳态密度于是是合成除以总处置。

推导

# Chapter 3. p05: The receptor counts itself down. A glucocorticoid
# receptor that is occupied is also marked for disposal, so a tissue
# bathed in more hormone keeps fewer receptors. Write R for receptor
# density, with a constant synthesis rate, a basal disposal rate and an
# extra disposal proportional to occupancy. At a steady state the
# density is the synthesis rate over the total disposal rate, which
# falls as the hormone rises. The two timescales that matter are the
# minutes of occupancy and the weeks of receptor turnover, and this
# program takes the slow one. Cortisol is in multiples of a normal mean.
SYNTH, BASAL, EXTRA = 1.0, 1.0 / 28.0, 3.0 / 28.0    # per day
KD, HILL = 1.0, 1.0
LEVELS = (0.6, 0.8, 1.0, 1.2, 1.5, 2.0, 2.5, 3.0, 4.0)

def occupancy(c):
    return c ** HILL / (KD ** HILL + c ** HILL)

def steady(c):
    return SYNTH / (BASAL + EXTRA * occupancy(c))

def turnover_days(c):
    return 1.0 / (BASAL + EXTRA * occupancy(c))

if __name__ == "__main__":
    base = steady(1.0)
    print("   basal disposal {:.5f} and occupancy-driven disposal {:.5f}"
          .format(BASAL, EXTRA))
    print("   cortisol   occupancy   receptors   relative   turnover days")
    for c in LEVELS:
        print("   {:<11.1f}{:<12.4f}{:<12.4f}{:<11.4f}{:.1f}"
              .format(c, occupancy(c), steady(c), steady(c) / base,
                      turnover_days(c)))
    print("   at a normal mean the density is {:.4f} and the turnover"
          .format(base))
    print("   time constant is {:.1f} days".format(turnover_days(1.0)))
    print("   doubling the mean leaves {:.4f} of the receptors"
          .format(steady(2.0) / base))
    print("   trebling it leaves {:.4f}".format(steady(3.0) / base))
    lo, hi = 1.0, 20.0
    for _ in range(60):
        mid = 0.5 * (lo + hi)
        if steady(mid) / base > 0.5:
            lo = mid
        else:
            hi = mid
    print("   the cortisol multiple that halves receptor density: {:.4f}"
          .format(0.5 * (lo + hi)))
    print("   but the density cannot fall below the floor set by the")
    print("   occupancy ceiling: at saturation it is {:.4f} of normal"
          .format((SYNTH / (BASAL + EXTRA)) / base))
    print("   so chronic load cannot abolish the receptor, only reduce it")
    print("   feedback strength is proportional to density, so the loop")
    print("   of p01 loses this much of its own restraint:")
    print("   cortisol   feedback left")
    for c in (1.0, 1.5, 2.0, 3.0):
        print("   {:<11.1f}{:.4f}".format(c, steady(c) / base))
    print("   which is why the first thing chronic stress does is raise")
    print("   cortisol further: the brake is made of the same receptor")
#    basal disposal 0.03571 and occupancy-driven disposal 0.10714
#    cortisol   occupancy   receptors   relative   turnover days
#    0.6        0.3750      13.1765     1.1765     13.2
#    0.8        0.4444      12.0000     1.0714     12.0
#    1.0        0.5000      11.2000     1.0000     11.2
#    1.2        0.5455      10.6207     0.9483     10.6
#    1.5        0.6000      10.0000     0.8929     10.0
#    2.0        0.6667      9.3333      0.8333     9.3
#    2.5        0.7143      8.9091      0.7955     8.9
#    3.0        0.7500      8.6154      0.7692     8.6
#    4.0        0.8000      8.2353      0.7353     8.2
#    at a normal mean the density is 11.2000 and the turnover
#    time constant is 11.2 days
#    doubling the mean leaves 0.8333 of the receptors
#    trebling it leaves 0.7692
#    the cortisol multiple that halves receptor density: 20.0000
#    but the density cannot fall below the floor set by the
#    occupancy ceiling: at saturation it is 0.6250 of normal
#    so chronic load cannot abolish the receptor, only reduce it
#    feedback strength is proportional to density, so the loop
#    of p01 loses this much of its own restraint:
#    cortisol   feedback left
#    1.0        1.0000
#    1.5        0.8929
#    2.0        0.8333
#    3.0        0.7692
#    which is why the first thing chronic stress does is raise
#    cortisol further: the brake is made of the same receptor

在常态均值下,这些单位里的密度是 11.2000,而周转时间常数是 11.2 天。把均值翻倍留下 0.8333 的受体,翻三倍留下 0.7692。这个下调是真的,而且是温和的,这是本章作出的第一处更正:慢性负荷并不会把受体剥光。存在一个下限,因为占据本身也会饱和,而在饱和占据下密度仍是常态的 0.6250。

这个下限对第四章的论证要紧。如果受体可以被压到零,刹车的削弱本身就能解释任意幅度的皮质醇上升。它不能,所以随后的那场坍塌必须由别的东西来交代。

下调确实解释的是第一阶段的方向。反馈强度与密度成正比,所以在均值 1.5 时这个环还剩 0.8929 的约束,在 2.0 时还剩 0.8333。一个把皮质醇抬高的负荷,同时削弱了本该把它带回去的那个刹车;于是慢性压力做的第一件事,就是把皮质醇抬得比负荷本身所能抬到的更高。

全书其余部分围着转的那个数

恢复以同一个时间常数、朝相反方向进行,而这个常数是本书里后果最大的一个数。

推导

# Chapter 3. p06: How long the receptor takes to come back. Once the
# load is removed and the hormone returns to its normal mean, receptor
# density climbs back towards its own steady state as a first-order
# process whose time constant is one over the total disposal rate. That
# time constant is the number this chapter exists to compute, because
# every popular claim about recovering from chronic stress is a claim
# about it. The starting density is the steady state under the chronic
# load; the target is the steady state at a normal mean.
SYNTH, BASAL, EXTRA = 1.0, 1.0 / 28.0, 3.0 / 28.0
KD, HILL = 1.0, 1.0
CHRONIC = 2.5                     # multiples of a normal mean
DAYS = (2, 7, 14, 21, 30, 45, 60, 90, 120)

def occupancy(c):
    return c ** HILL / (KD ** HILL + c ** HILL)

def steady(c):
    return SYNTH / (BASAL + EXTRA * occupancy(c))

def tau(c):
    return 1.0 / (BASAL + EXTRA * occupancy(c))

def density(t, start, target, tc):
    return target + (start - target) * 2.718281828459045 ** (-t / tc)

if __name__ == "__main__":
    start = steady(CHRONIC)
    target = steady(1.0)
    tc = tau(1.0)
    print("   chronic load at {:.1f} times normal leaves {:.4f} of the"
          .format(CHRONIC, start / target))
    print("   receptors; recovery runs at a time constant of {:.2f} days"
          .format(tc))
    print("   day    density   share of the gap closed   relative to normal")
    for d in DAYS:
        r = density(float(d), start, target, tc)
        print("   {:<7d}{:<10.4f}{:<26.4f}{:.4f}"
              .format(d, r, (r - start) / (target - start), r / target))
    for frac in (0.50, 0.80, 0.90, 0.95):
        lo, hi = 0.0, 1000.0
        for _ in range(60):
            mid = 0.5 * (lo + hi)
            r = density(mid, start, target, tc)
            if (r - start) / (target - start) < frac:
                lo = mid
            else:
                hi = mid
        print("   days to close {:.2f} of the gap: {:.1f}"
              .format(frac, 0.5 * (lo + hi)))
    print("   the time constant is the whole answer, and it is set by")
    print("   receptor turnover, not by how tired anyone feels")
    print("   turnover under load is faster, at {:.2f} days, which is"
          .format(tau(CHRONIC)))
    print("   why the fall is quicker than the recovery: {:.2f} times"
          .format(tc / tau(CHRONIC)))
    print("   two days of rest close {:.4f} of the gap"
          .format((density(2.0, start, target, tc) - start)
                  / (target - start)))
    print("   and fourteen days close {:.4f}"
          .format((density(14.0, start, target, tc) - start)
                  / (target - start)))
    print("   so the popular claim that a long weekend resets the axis")
    print("   is wrong by more than an order of magnitude in time")
#    chronic load at 2.5 times normal leaves 0.7955 of the
#    receptors; recovery runs at a time constant of 11.20 days
#    day    density   share of the gap closed   relative to normal
#    2      9.2837    0.1635                    0.8289
#    7      9.9738    0.4647                    0.8905
#    14     10.5436   0.7135                    0.9414
#    21     10.8487   0.8466                    0.9686
#    30     11.0427   0.9313                    0.9860
#    45     11.1588   0.9820                    0.9963
#    60     11.1892   0.9953                    0.9990
#    90     11.1993   0.9997                    0.9999
#    120    11.1999   1.0000                    1.0000
#    days to close 0.50 of the gap: 7.8
#    days to close 0.80 of the gap: 18.0
#    days to close 0.90 of the gap: 25.8
#    days to close 0.95 of the gap: 33.6
#    the time constant is the whole answer, and it is set by
#    receptor turnover, not by how tired anyone feels
#    turnover under load is faster, at 8.91 days, which is
#    why the fall is quicker than the recovery: 1.26 times
#    two days of rest close 0.1635 of the gap
#    and fourteen days close 0.7135
#    so the popular claim that a long weekend resets the axis
#    is wrong by more than an order of magnitude in time

从常态 2.5 倍的慢性负荷出发(留下 0.7955 的受体),恢复的时间常数是 11.20 天。闭合一半的缺口要 7.8 天,闭合 0.90 要 25.8 天,闭合 0.95 要 33.6 天。两天休息闭合缺口的 0.1635,十四天闭合 0.7135。负荷之下的周转更快,为 8.91 天,比恢复快 1.26 倍,所以下落比爬回来更快。

受体密度对恢复天数,图中标出时间常数与 0.90 这一点(p06)。

注

那个流行数字的来处。 几乎每一份关于从慢性压力中恢复的说法都会引用「二十一到四十五天」这个区间,它不是一个时间常数,也不该被当作时间常数用。在这个模型里,它是恢复比例从第 21 天的 0.8466 走到第 45 天的 0.9820 的那一段,把它称作恢复的尾声是合理的,但它底下的时间常数是 11.20 天。这个区分之所以要紧,是因为时间常数能告诉你两天买到什么,而一个恢复窗口不能:两天买到 0.1635,这就是为什么那个周末感觉上什么也不是,而它确实什么也不是。

假说 · 尚未验证

作者的诠释(尚未验证): 决定从倦怠中恢复需要多久的是受体恢复的时间常数,而不是先前负荷的严重程度;所以轻症与重症的恢复时间应当大致相同,而恢复的量不同。什么会证明它错: 如果在一个经重复测量追踪整个恢复过程的队列里,达到某一固定恢复比例的时间随初始严重程度缩放,而不是大致恒定,那么单一时间常数的模型就不够用。

小结

受体在被占据时被处置得更快,所以慢性负荷会减少它的密度,但只减到饱和占据下常态的 0.6250 这个下限。反馈强度随密度下降,这就是为什么一个负荷先把皮质醇抬得比负荷本身所能抬到的更高。恢复的时间常数是 11.20 天,所以闭合 0.90 的缺口要 25.8 天,0.95 要 33.6 天,而两天只闭合 0.1635。第四章把这件事与肾上腺耦合起来,找出那三个阶段。

参考文献

McEwen, Bruce S. 1998. “Protective and Damaging Effects of Stress Mediators.” The New England Journal of Medicine 338 (3): 171 to 179.
Oakley, Robert H., and John A. Cidlowski. 2013. “The Biology of the Glucocorticoid Receptor: New Signaling Mechanisms in Health and Disease.” The Journal of Allergy and Clinical Immunology 132 (5): 1033 to 1044.

Fan, Y. (2026). HPA-Axis Coupled Oscillations, Cortisol Awakening Response Flattening and Burnout Recovery Kinetics: Glucocorticoid Receptor Downregulation, Ultradian Pulsatility and Allostatic Load Reset [下丘脑垂体肾上腺轴耦合振荡、皮质醇觉醒反应扁平化与倦怠恢复动力学] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090252 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).